Previous day | Jump to hour: 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 | Next day

Seconds: Show Hide | Joins: Show Hide | View raw
Font: Serif Sans-Serif Monospace | Size: Small Medium Large

Click in the nick column to highlight everything a person has said.
The Logo icon identifies that the person is a core developer (has commit access).

#rockbox log for 2011-02-09

00:00:04LloreanI just feel if there were objections before, at least a *strong* attempt should've been made to contact the objectors.
00:00:13gevaertsthomasjfox: done
00:00:15kugelthomasjfox: is sunday a special date?
00:00:16LloreanI objected to the initial commit of autoresume too.
00:00:29thomasjfoxkugel: See the ML :o)
00:00:47kugelah, the freeze?
00:00:51thomasjfoxyes :)
00:01:31 Quit mlt- (Quit: CGI:IRC)
00:01:58thomasjfoxgevaerts: thanks
00:03:36JdGordon1BORING.... Llorean, you arnet here 24/7 anymore, you arnt in -community much anymore either.... how is anyone supposed to know the difference between activly leaving and just just being silent
00:03:57JdGordon1you shouldnt have disspapeared while an arguent was happening which you were involved in
00:04:11LloreanJdGordon1: I'm sorry I had oral surgery
00:04:21sideralLlorean: Several objections of various people have been addressed, including complexity of configuration, wording of the config options, and lang-related stuff. Here's part of the IRC discussion I mentioned:
00:04:22LloreanIn the future I'll schedule my two cubic centimeter cysts around our discussions.
00:04:46JdGordon1thank you
00:04:50JdGordon1that would be appreciated
00:05:24kugelLlorean: we couldn't know that, for us you were just silent/gone inactive
00:06:15Lloreankugel: Yes, but the fact that there wasn't *any* further discussion on the ML (the initial venue of discussion) doesn't help at all.
00:07:31*Llorean learns from the linked discussion that the comma separated list will match "pod" to the podcasts folder, and thinks it's terribly easy to get false positives.
00:07:50gevaertsLlorean: /pod/ won't
00:08:54Lloreangevaerts: My concern is genre tags, which don't typically contain characters that make it easy to exclude folders.
00:09:10 Quit bertrik (Read error: Operation timed out)
00:09:16gevaertsAh, right. That trick won't work there, true
00:09:22*gevaerts is too tired to think straight
00:09:26Lloreangevaerts: For example, the genre "voice" isn't uncommon for things I might want to resume, but "voice" is a word that could easily show up in album names, and thus folders
00:09:47LloreanIn fact I wouldn't be surprised if my player already has content that would be fouled up by this.
00:10:06JdGordon1are *you* going to use the feature at all?
00:10:43LloreanJdGordon1: No, clearly not, that's why I spent so long trying to discuss how we could meet both his needs and my needs with it. The fact that I have desires and needs for the features doesn't in any way indicate I would use it, of course.
00:11:44LloreanIf I weren't going to use it, I'd have either said "I object to the bloat of it" if I thought nobody would use it, or "as long as there's an off switch, I don't care."
00:13:30sideralLlorean: There's no reason we cannot add support for your use case as well. I just wasn't willing to give up my major use case for yours. Sorry.
00:13:52***Saving seen data "./dancer.seen"
00:14:18Lloreansideral: The majority of your use case was addressed just fine, it was just manual instead of automated...
00:14:24LloreanThat's not giving up your use case.
00:14:46sideralLlorean: I didn't say majority of use case, but major (as in most important) use case
00:15:17kugelLlorean: we are not stopping you from implementing the context menu item and deactivating autoresume
00:15:24Lloreansideral: And the most important thing to me is that it's not automatic.
00:15:53Lloreankugel: You know I cannot realistically remove an existing feature days or weeks down the line.
00:15:54sideralLlorean: Fine, turn it off then. Then let's add a context menu for manual resume operation
00:16:11kugelLlorean: by deactivating I meant in the settings, not in the code
00:16:13Lloreansideral: Then you've got overlapping features.
00:16:42Lloreansideral: I didn't see any sort of significant evidence of changed opinions in that link. Mostly you talking with gevaerts and kugel.
00:16:43sideralLlorean: I disagree. You get two ways (manual+automatic) to get at the same feature
00:16:58 Quit thomasjfox (Remote host closed the connection)
00:17:21Lloreansideral: And most of Rockbox is not automated.
00:17:34JdGordon1thats nonesense
00:17:51JdGordon1just about everything IS atomated
00:18:00LloreanJdGordon1: Rockbox doesn't have any automated system for turning on or off *any* other setting I think
00:18:22LloreanJdGordon1: How do I automate whether or not a folder is shuffled when I play it? Or whether or not to use EQ? Or...?
00:18:23JdGordon1that's not what you said, and thats incorrect anyway
00:18:46LloreanJdGordon1: If "use autoresume automatically" is automation, then *not* toggling settings automatically qualifies as parts of the system that aren't automated.
00:19:57LloreanOr, rather than saying "toggling settings automatically" we could say "automatically selecting whether to use settings during playback of specific files"
00:20:13LloreanSince, strictly speaking, the setting isn't actually *changed* just applied or not
00:20:18kugelI actually thought a shuffle.<mode> file would be nice to automatically shuffle folders
00:20:27 Quit GeekSh4dow (Quit: The cake is a lie !)
00:20:48 Quit Jerom (Quit: Leaving.)
00:20:59*JdGordon1 points out that any sane argument would remind everyone that having folder/playlist based .cfg files is a strict nodo
00:21:00Lloreankugel: If we were to do that, wouldn't a .canresume file make sense for marking folders (and thus all their subfolders) resumable, rather than a comma separated list?
00:21:05LloreanIt'd be very similar to database.ignore
00:21:07JdGordon1well, folder anyway
00:21:27LloreanJdGordon1: Yes, so there's a strict nodo for folder based application of *most* settings, but not this one.
00:21:27sideralLlorean: I do not agree that you have sufficiently demonstrated that the feature "does not fit" into Rockbox.
00:21:34kugelJdGordon1: shuffling without toggling the setting
00:21:50kugelno setting should be toggled by something else then the user IMO
00:21:53Lloreansideral: So express for me some similar features, since JdGordon1 very nicely expressed a nodo that's very similar to this feature you've just committed.
00:22:27JdGordon1"just" being a week ago?
00:22:28sideralYou said any setting was being toggled automatically, which is clearly not the case.
00:23:19LloreanJdGordon1: I have it as 20:31 today
00:23:33Lloreansideral: No, I clarified that the setting is being *applied* automatically.
00:23:44*JdGordon1 has absolutly no idea what those commits acheive
00:23:57 Quit liar (Read error: No route to host)
00:24:02JdGordon1sideral: I hope you created a task to add it to the manual? :D
00:24:05LloreanJdGordon1: Not the auto-resume, but the ones that add settings for using a comma separated list of search terms to determine which files to auto-resume on and which not to.
00:24:22sideralJdGordon: that's on my todo list
00:24:42sideralLlorean: No, that's not what the setting does.
00:24:57 Join liar [0] (
00:25:15JdGordon1Llorean: well, the nodo i mentinoed was because of the file check, not necessarily for the autosetting changing
00:25:17Lloreansideral: "Tracks can be selected based on their their file location or genre tag (comma-separated list of filename / genre substrings)"
00:25:33LloreanJdGordon1: Really? Most objections I'd heard to it was because of the setting changing.
00:25:44 Quit evilnick_B (Quit: Page closed)
00:25:45LloreanJdGordon1: It's not like playlists and playback don't regularly open files...
00:25:58sideralit toggles whether automatic track change triggers resume for the next track depending on a user-defined set of tracks
00:26:11kugelJdGordon1: why is that a nodo?
00:26:11LloreanYes, whether the setting applies based on automation.
00:26:14sideralthe settings are all static.
00:26:28sideralThis is nitpicking
00:26:42Lloreansideral: So it's not automation?
00:27:19 Quit LambdaCalculus37 (Quit: Clunk)
00:27:40sideralSure it is automation. But it's not the settings that are dynamic. It's the behavior, as configured by a (static) user setting.
00:28:11LloreanYes, exactly as would be by static files that determined which settings are applied to a given folder (which as JdGordon1 just pointed out, is more or less a NoDo)
00:29:12LloreanIt has even been proposed that these files not be able to permanently change settings, but only change them until that folder is left, or playback is stopped, being a similar "only for this session/selection" thing as yours, and still rejected
00:29:24LloreanSo I'd ask why this kind of automation is okay for autoresume, but not for any other setting.
00:30:33sideralI don't care about that nodo, and see no significant overlap with autoresume. Nothing of this is risking the integrity of the software or user experience.
00:30:55Lloreansideral: So you're choosing to ignore an existing nodo, or do you mean you think you can explain how it's not parallel?
00:31:07kugelLlorean: I honestly can't see your point
00:31:31Lloreankugel: You think there's no parallel between "use automation to determine whether or not to apply autoresume" and "use automation to determine whether or not to apply other settings"?
00:31:34sideralI didn't understand the similarity in the first place.
00:31:36LloreanOr you don't see why that's valid?
00:32:06Lloreansideral: Your feature uses automation to determine whether or not to apply a feature (autoresume), but the option to use automation to determine whether or not to apply *any* setting has been rejected.
00:32:10JdGordon1well the 2nd is a given if the first is true
00:32:15sideralThe principle you state has no benefit as is.
00:32:19Lloreankugel: How can you think you can judge if it's valid or not, if you don't understand it?
00:32:39kugelwho said I don't understand it?
00:32:44Lloreansideral: What principal.
00:33:06kugelthe user clearly made a decission on what to apply automation on. what's wrong with that?
00:33:20Lloreankugel: Well if you don't see it the way I do, in terms of existing as a parallel, your judgment is on something I'm not saying, so it doesn't bear any meaning in rejecting my claim.
00:33:31Lloreankugel: The user clearly makes a decision what to apply automation on in the NoDo case too...
00:33:37LloreanIt's just as explicit
00:33:39LloreanMoreso actually
00:33:41kugelwhat nodo?
00:33:47Lloreankugel: The one JdGordon1 referenced.
00:33:57kugelit's a nodo for him
00:34:09kugelwe only have one official nodo list
00:34:27 Join Keripo [0] (
00:34:48kugelas I said, I actually would like a feature that shuffled the selected folder if it has a shuffle.<mode> file in it
00:34:52sideralLlorean: It doesn't make sense to state that filtering settings by some definition of a subset of files is bad per se. There has to be a case-by-case analysis.
00:35:04sideralClearly podcasts != audiobooks != music albums
00:35:07kugel(for the folder only, without changing the setting that is)
00:35:17Lloreankugel: It's also been rejected many, MANY times in the past, though. It may not be a nodo, but the same reasoning for rejecting it applies to this.
00:35:36kugelif it's not on the nodo list, it's not a nodo. period
00:35:44kugelwe discussed that on the devcon
00:35:48Lloreansideral: Or you could not automate it, so that the filtering isn't necessary. You haven't addressed how yours doesnt parallel an often rejected feature idea.
00:36:17kugeldatabase.ignore has been in for years now
00:36:18Lloreankugel: Yes, so it's not a nodo. That doesn't change that the reasoning for rejecting it also applies to this.
00:36:24 Quit elcan (Read error: Operation timed out)
00:36:39sideralLlorean: I frankly don't care whether my feature _does not_ parallel another often rejected feature idea. Sorry.
00:37:10Lloreansideral: Well, if the other feature was rejected, and yours parallels, why shouldn't yours be rejected?
00:37:17kugelalso prevent clipping, track gain if album gain isn't available too
00:37:43kugelthere's a number of setting/files that toggle some automation. I don't see anything wrong with this
00:37:52 Join elcan [0] (
00:37:55Lloreankugel: That's part of the replaygain feature standard though.
00:37:59sideralI contend there's no parallel. And my personal opinion is that you cannot generalize from negative examples and create principles out of them.
00:38:08LloreanThat's like saying "bass boost is automation because it automatically determines which frequencies to boost"
00:38:21Lloreansideral: Explain how there's not a parallel then?
00:38:47JdGordon1loud users dont get to dictate feature inclusions
00:38:55sideralLlorean: No I won't, because it doesn't serve any purpose as I just tried to explain
00:38:56Lloreansideral: If the same *reason* applies to both features, because of the parallel, you can generalize.
00:39:02 Join froggyman_ [0] (~seth@
00:39:14Lloreansideral: If we say "we don't want that because we shouldn't automate that sort of behaviour" and your patch *automates that sort of behaviour*...
00:39:21 Quit froggyman (Quit: Ex-Chat)
00:39:39LloreanThen it applies as a valid reason to reject yours, and you can explain how yours doesn't automate that sort of behaviour to counter it, rather than saying "it doesn't apply because I say it doesn't"
00:40:13sideralLlorean: I don't buy the principle in the first place. You have to explain why you don't want something, and then we see whether or not that applies to the new feature.
00:40:23kugelLlorean: are you too blind to see that we have loads of automation in rockbox?
00:40:37LloreanJdGordon1: But several people objecting to it can, which was previously the case, and no attempt was seriously made to determine if that situation had changed before it went in, despite past discussions about *exactly* such a situation and how it should be resolved.
00:41:07Lloreankugel: Explain to me a similar case of automation that chooses whether or not to apply a setting based on the location of a file?
00:41:34JdGordon1kugel: even ignoreing the whole file -> buffer-> codec -> output system, there is tons of user configurable automation going on!
00:41:39Lloreansideral: I don't want it because it automates the application of a setting based on an automated procedure, a reason that has been deemed valid for rejecting features in the past.
00:41:47kugelWhy should we reject something because it adds a setting that configures on what to apply automation?
00:42:11kugelespecially if we have implemented that scheme in a lot of places already?
00:42:18Lloreankugel: So tell me some, like I asked.
00:42:35sideralLlorean: I think that both Saint and pixelma stopped their opposition.
00:42:57kugelsideral: everyone did
00:43:08Lloreansideral: I didn't see that at what you linked me to, at least.
00:43:27sideralLlorean: That may have happened in other conversations.
00:44:01sideralpixelma and Saint helped me streamline the configuration.
00:44:05Lloreankugel: Database.ignore is completely unable to affect playback. I will give you that in the general case it's similar, but not to the degree that folder-based configs and folder-based autoresume is.
00:44:30AlexPI did too, on the understanding that I'm going to keep annoying sideral until he lets me add tracks as and when needed to an otherwise empty database :)
00:44:33Lloreansideral: I would have helped you streamline the configuration while still objecting, in the case that even if I didn't like it, if it went in despite my objections I'd still want it the best possible
00:44:38Lloreansideral: "helped me" doesn't mean 'agreed with me'
00:45:13kugelLlorean: honestly, to me it looks like you want to aggrandise yourself because we didn't wait for you to recover from your surgery (keep in mind we didn't even know about). I seriously have a hard time to see your point, because your arguments are plain wrong. But I'm out of this stupid discussion for today
00:45:17LloreanI'm not saying they don't agree with you
00:45:24LloreanJust saying, that's an erroneous conclusion to jump to.
00:45:50Lloreankugel: Basically "I don't have any features other than database.ignore, which doesn't affect playback like the rejected ones did?"
00:45:52sideralLlorean: Yes, I took some educated guesses :)
00:45:57Lloreansideral: Educated by what?
00:46:46Lloreansideral: In a situation like this, the effort should be made for explicit agreement. Not "I haven't heard from them in a while, so it's good enough" like apparently was done with me. We had a whole discussion in the past about contentions features where there was even talk about how many had to agree more than objectors.
00:46:48sideralSimply by my conversations with folks here.
00:46:58Llorean"I assume" really isn't good enough for that.
00:47:13JdGordon1nothing short of signed letter is
00:47:48LloreanWell, for example, it should be assumed that I still object until I say otherwise. If the feature is good enough, even with that assumption it'd be possible to get enough explicit agreements.
00:48:06LloreanThere should never be a case of implied change of opinion.
00:48:16sideralLlorean: I'm afraid we all make assumptions all the time. Also, I accept the reality that there's no way everyone will always agree with me.
00:48:19JdGordon1and in that vein it should be assumed that sideral is going to keep ignoreing your nonesense objections
00:48:25LloreanEven if you can't get in contact with everyone, changes could easily change the opinion of enough people that can give an explicit response.
00:48:50LloreanJdGordon1: Just because you don't agree with them doesn't mean they're nonsense. Please, try to stay civil. No value-assertions are needed.
00:49:29Lloreansideral: Yes, assumptions are often made. That doesn't mean that you should use them for this sort of decision making, to the contrary of what was basically eventually agreed upon as the policy for this sort of conflict resolution.
00:49:38LloreanThe fact that no real attempt was made to follow that is kind of a problem.
00:49:51LloreanThat means anyone can say "I assume they agree with me, so I'll commit"
00:50:15LloreanHow much would have hurt to wait a short period to actually ask the people who'd objected and give them a day or two to respond?
00:52:52sideralLlorean: I'm sorry for the loss of communication.
00:52:53sideralBut I look at it this way: There's no way your objections can be cleared up −− they are based on invalid (in my view) principles. Yet, I was given commit access, meaning I was trusted with making the decision to commit at the right time.
00:53:15Lloreansideral: And I too have commit access. I could remove your commit if I felt it was appropriate, and then we'd be stuck again.
00:53:52JdGordon1and there would be a massive shitstorm if you did that, and you know it
00:53:58Llorean"Have commit access" doesn't mean "should override opinions before settling the issue"
00:54:08LloreanJdGordon1: Last time I reverted a change I didn't like, I believe it stuck.
00:54:23sideralLlorean: Yes you could. I actually invite you to do that if you think it is right. That would mean our disagreement would have to worked out by others pretty quickly.
00:54:27JdGordon1what, 2 years ago?
00:54:39AlexPI can't remember what it was, but I seem to remember others agreed then
00:55:00LloreanJdGordon1: I'm not sure it'd be the exact one you think it would.
00:55:11LloreanIt would at least bring discussion of this to the forefront.
00:55:35LloreanI know when I previously objected to this issue, I was told by some others that they agreed with my objections but didn't speak up because I'd already said them.
00:55:56LloreanWhich is why I feel discussion should've been settled on the issue before the commit.
00:57:33Lloreansideral: My principals may be invalid in your view, but distinguishing autoresume from bookmarks is invalid in mine since, fundamentally, there's no reason they shouldn't be the same feature, and working to separate them like this just makes it a more significant change to users later if the separation is resolved.
00:57:55LloreanNot to mention if the automation *is* rejected, and later removed.
00:58:08saratogaLlorean: honestly its pretty difficult to care about discussions with people who really have no interest in actually working on things
00:58:19LloreanOr, if the fact that it's too easy for false positives (a question that wasn't even addressed) means that the behavior needs to change.
00:58:49saratogaso i think its maybe unrealistic for you to expect people to care so much about what you think in situations like this
00:58:58Lloreansaratoga: And yet people were happy to criticize me all the time on the things *I* worked on in the past. So, you know what, I think their behaviour has validated that I'm allowed to voice my opinion on what gets worked on in the areas I contribute less to.
00:59:14LloreanMy opinion doesn't have to be listened to, and nothing has to be done about it.
00:59:21sideralLlorean: Yes, there's potential for false positives, but I think we arrived at a good compromise between more configurability (which is what I wanted) and simplicity (one of your goals)
00:59:22LloreanBut never suggest I shouldn't voice it as I like.
00:59:44saratogavoice all you like
01:00:01Lloreansideral: I don't think it's a good compromise at all. It should only do explicit, full matches. So no file paths unless there's a leading /, no substrings.
01:00:23Lloreansaratoga: Then what was your point? Obviously I know that I'm not guaranteed to have people do what I say or anything.
01:00:50Lloreansaratoga: If I felt I could dictate, I wouldn't give reasons and attempt to explain the grounds upon which I object.
01:01:23Lloreansideral: This still gives as much configurability as you could want, but means almost no false positives (unless your genre happens to exactly match a file path)
01:02:37sideralLlorean: I stand by my assessment that from a user perspective, autoresume has to be different from bookmarks because the point of autoresume is that _is has no UI_. From an implementation perspective, I'd be fine with sharing a common backend with bookmarks as JdGordon has proposed.
01:02:52linuxstbkugel: Here's a first attempt at changing configure to split out the applications into separate targets - any comments before I post it to flyspray (so thomasjfox can test) ?">
01:03:11linuxstbkugel: (or let me know if you don't have time/interest in looking now, and I'll just put it on flyspray)
01:03:14JdGordon1Llorean: ok, good, so you understand that not everyone has to care what you say, you've now said your peice so accept that it isnt being accepted and be quiet
01:03:17sideralLlorean: Yes, we can work on making the configuration more precise. I'm open on that
01:03:26LloreanJdGordon1: See, there you are telling me not to voice my opinions.
01:03:37Lloreansideral: So why not change it to explicit matching? Why wasn't that used in the first place?
01:04:05JdGordon1Llorean: you've been stating your opinions for over an hour... this is circular, the rest of us are bored
01:04:08sideralLlorean: Because no one suggested it despite extensive review?
01:04:53Lloreansideral: It seems pretty obvious too me. Too bad I was unconscious. That's exactly the sort of thing I'd have suggested while saying "it shouldn't go in, but if it does, it really should at least..."
01:05:32LloreanJdGordon1: Then don't pay attention, and/or don't respond. Any time you offer a counter-argument if I think that either part of my opinion has been misunderstood, or misrepresented, or just missed, I will continue to respond. That's how it works.
01:05:41JdGordon1linuxstb: good idea, you should make the menu alphabetical though (android then sdl)
01:05:43 Quit toffe82 (Read error: Connection reset by peer)
01:05:58Lloreansideral: Any chance it's relatively trivial to fix? the sooner the better on behaviour changes so users don't get used to it.
01:06:39linuxstbJdGordon1: And when the next target is added? I'm just following the convention of listing them in the order they were added.
01:06:48sideralLlorean: Glad you're being constructive! That could be enhance pretty quickly. Could you add a tracker comment outlining your idea?
01:07:10Lloreansideral: Does it really need outlining? Match only whole, exact values rather than substrings?
01:07:26JdGordon1linuxstb: sure, but because its a moderate change you may as well make it neat to begin with...
01:08:27saratogawe need to get GSOC project ideas on the wiki
01:08:33sideralI may want to think about this some more, and logging it in the tracker makes sure it doesn't get lost. For example, I think we need some substring capability to capture both "podcast" and "PODCASTS" to make the default more useful.
01:08:40Lloreansaratoga: Are we even possibly accepted this year?
01:08:42saratogait would be great if people could copy their proposals from last year over to this year's page
01:09:02Lloreansideral: Well, I think it shouldn't be case sensitive. Exact matching of letters, not character values.
01:09:18LloreanSince FAT32 isn't a case sensitive filesystem, and I don't think that ID3 tags are generally considered to be.
01:09:37Lloreansaratoga: I seem to remember last year there was significant thought we wouldn't be acceptable this year.
01:09:42JdGordon1linuxstb: while you're playing, can you add a set of common LCD sizes instead of having to do width/ehight manually?
01:09:42sideralLlorean: that, but also mind the "s" in "podcast" vs "podcasts"
01:10:01Lloreansideral: I think it's not too hard to type "podcast, podcasts, /podcasts"
01:10:17 Quit dfkt (Quit: -= SysReset 2.53=- Sic gorgiamus allos subjectatos nunc.)
01:10:21LloreanJust make sure the default contains both.
01:11:04sideralLlorean: You're probably right. But please do log a comment −− I'm too tired and probably won't recall this completely tomorrow :)
01:11:16LloreanThere's no reason not to update the default to take into account new limitations.
01:11:19LloreanAnd I've posted to the tracker
01:11:35LloreanAnd, just to be explicit - my being helpful doesn't mean I agree with the way the feature currently works. :-P
01:11:49saratogai have no idea
01:11:57sideralLlorean: I wanted to say one more thing publicly:
01:12:07Lloreansaratoga: Do we still have contacts?
01:12:31saratogamaybe ask Bagder or AlexP?
01:12:32pixelmakugel: there was a time were I suddenly didn't get flyspray notification for no apparent reason (still subscribed to the same tasks and email address was correct, it suddenly fixed itself. I also heard that from someone else, I believe it was funman. I objeced (only here) to saratoga's initial commit btw. but got no-one to back it up. I'd just given up opposition though because I was tired discussing it and since I don't listen to podcasts
01:13:20Lloreanpixelma: It looks like I legitimately was not subscribed, so in this case it's probably my fault. I thought I had, but I'm not willing to say that it chose to forget that I had entirely.
01:13:25pixelmait would still be nice if some issues were addressed and seeing about sorting this out before a feature inclusion would be nice
01:14:01pixelmaas I stated about a week ago - we usually don't toss features once included
01:14:20sideralLlorean: To make up for my first impression of you during our conversion, I've listened to that FLOSS podcast you did, and have to say: That was really cool, and I think you're quite a sympathetic person. :)
01:14:36sideralGood night
01:14:43Lloreansideral: I don't come across nearly as well in text, unfortunately, as I do in person. Ask anyone who met me at DevCons past.
01:14:50LloreanSorry we seem to be butting heads on things.
01:15:00kugellinuxstb: looks ok
01:15:18linuxstbJdGordon1: I'ld rather do that afterwards. Maybe it would be better as a sub-menu - i.e. choose the lcd size from a list instead of specifying width/height. There is also the question about whether a different lcd size means a different target_id (which IIUC is just used for voice/lang files identification)
01:15:58JdGordon1linuxstb: uh, yeah,, thats what I meant :)
01:24:35 Quit sideral (Ping timeout: 240 seconds)
01:31:01 Quit kugel (Remote host closed the connection)
01:44:39 Quit mudd1 (Ping timeout: 260 seconds)
01:47:25 Quit pamaury (Remote host closed the connection)
01:48:03 Quit factor (Read error: Operation timed out)
01:55:38 Quit linuxguy3 (Ping timeout: 272 seconds)
01:56:53 Join linuxguy3 [0] (
02:05:23 Quit Dexpid (Remote host closed the connection)
02:11:28 Quit Stephen__ (Quit: Leaving)
02:13:53***Saving seen data "./dancer.seen"
02:18:52[Saint][12:42] <sideral> Llorean: I think that both Saint and pixelma stopped their opposition. <−− fwiw...I didn't stop my opposition, I simply decided my voice was too small.
02:19:22[Saint]and, weighed up how much it actually meant to me...I don't *have* to use it...right?
02:29:02 Join factor [0] (~factor@
02:35:49 Join milk [0] (
02:56:32 Quit saratoga (Quit: Page closed)
03:23:49 Quit robin0800 (Quit: Leaving)
03:26:55 Quit qurvel (Ping timeout: 276 seconds)
03:27:00 Join robin0800 [0] (
03:33:52 Quit robin0800 (Quit: Leaving)
03:34:07 Join robin0800 [0] (
03:37:17 Quit liar (Ping timeout: 255 seconds)
03:42:38 Quit robin0800 (Quit: Leaving)
03:44:34 Quit [Saint] (Ping timeout: 240 seconds)
03:45:11 Join [Saint] [0] (S_a_i_n_t@
03:46:52 Join enthdegree [0] (
04:05:34 Quit Keripo (Ping timeout: 255 seconds)
04:07:17 Join godeater_ [0] (3a9337c7@rockbox/staff/GodEater)
04:08:15 Quit amiconn (Disconnected by services)
04:08:15 Join amiconn_ [0] (quassel@rockbox/developer/amiconn)
04:08:33 Nick amiconn_ is now known as amiconn (quassel@rockbox/developer/amiconn)
04:08:46 Join Keripo [0] (
04:09:21 Quit pixelma (Disconnected by services)
04:09:23 Join pixelma_ [0] (quassel@rockbox/staff/pixelma)
04:09:25 Nick pixelma_ is now known as pixelma (quassel@rockbox/staff/pixelma)
04:10:41 Quit Barahir (Read error: Operation timed out)
04:10:52 Quit TheSeven (Ping timeout: 245 seconds)
04:13:56***Saving seen data "./dancer.seen"
04:14:28 Join Barahir [0] (
04:14:42 Quit Kitr88 (Read error: Connection reset by peer)
04:14:58 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
04:20:21 Quit godeater_ (Quit: Page closed)
04:22:57 Join Kitar|st [0] (
04:25:29 Join DerPapst1 [0] (
04:28:01 Quit DerPapst (Ping timeout: 250 seconds)
04:44:52 Nick froggyman_ is now known as froggyman (~seth@
04:45:08 Quit froggyman (Changing host)
04:45:08 Join froggyman [0] (~seth@unaffiliated/froggyman)
04:50:52 Join Dexpid [0] (
04:53:05 Quit DerPapst1 (Quit: Leaving.)
04:54:19 Quit Dexpid (Client Quit)
04:59:46[Saint]Hmmm...what information would one like to see present (when no next track is available) on a wps to fill a big gap, considering that track name, artist, album, genre, year, bitrate, codec, and samplerate are already displayed?
05:00:46[Saint]in the (reasonably infrequent) event there's no next track present, there's a bit of a hole in my WPS if there is no album art (another reasonably unlike occurence).
05:01:18 Quit Keripo (Quit: Leaving.)
05:03:47 Join Rob2223 [0] (
05:04:49 Join Keripo [0] (
05:08:22 Quit Rob2222 (Ping timeout: 276 seconds)
05:20:16 Join Dexpid [0] (
05:24:24 Quit zu (Ping timeout: 276 seconds)
05:24:53 Quit Keripo (*.net *.split)
05:24:53 Quit [Saint] (*.net *.split)
05:24:53 Quit Galois (*.net *.split)
05:24:53 Quit Battousai (*.net *.split)
05:24:54 Quit CIA-7 (*.net *.split)
05:24:54 Quit yosafbridge (*.net *.split)
05:24:54 Quit logiclost (*.net *.split)
05:24:54 Quit cjcopi (*.net *.split)
05:24:54 Quit enthdegree (*.net *.split)
05:24:55 Quit factor (*.net *.split)
05:24:55 Quit froggyman (*.net *.split)
05:24:55 Quit JdGordon1 (*.net *.split)
05:24:55 Quit Horscht (*.net *.split)
05:24:55 Quit jae (*.net *.split)
05:24:55 Quit eGen_ (*.net *.split)
05:24:55 Quit Strife89 (*.net *.split)
05:24:55 Quit jfc (*.net *.split)
05:24:55 Quit DrDnar (*.net *.split)
05:24:55 Quit AlexP (*.net *.split)
05:24:55 Quit kkit|sh (*.net *.split)
05:24:55 Quit dionoea (*.net *.split)
05:24:55 Quit pikytcus (*.net *.split)
05:24:55 Quit scorche|sh (*.net *.split)
05:24:55 Quit MagusG (*.net *.split)
05:24:55 Quit ChanServ (*.net *.split)
05:24:56 Quit Quazgaa (*.net *.split)
05:24:56 Quit ehntoo (*.net *.split)
05:24:56 Quit guymann (*.net *.split)
05:24:56 Quit balintx (*.net *.split)
05:24:56 Quit feisar- (*.net *.split)
05:24:56 Quit Rob2223 (*.net *.split)
05:24:56 Quit amiconn (*.net *.split)
05:24:56 Quit bluebrother (*.net *.split)
05:24:57 Quit n17ikh (*.net *.split)
05:24:57 Quit Dhraakellian (*.net *.split)
05:24:57 Quit Farthen (*.net *.split)
05:24:57 Quit linuxguy3 (*.net *.split)
05:24:57 Quit plux (*.net *.split)
05:24:57 Quit Dreamxtreme (*.net *.split)
05:24:58 Quit panni_ (*.net *.split)
05:24:58 Quit jepler (*.net *.split)
05:24:58 Quit Kohlrabi (*.net *.split)
05:24:58 Quit amee2k (*.net *.split)
05:24:58 Quit ps-auxw (*.net *.split)
05:24:58 Quit bzed (*.net *.split)
05:24:59 Quit powell14ski_ (*.net *.split)
05:24:59 Quit ranmachan (*.net *.split)
05:24:59 Quit Elfish (*.net *.split)
05:24:59 Quit milk (*.net *.split)
05:24:59 Quit mystica555_ (*.net *.split)
05:24:59 Quit Sochiro (*.net *.split)
05:25:00 Quit preglow (*.net *.split)
05:25:00 Quit crwl (*.net *.split)
05:25:00 Quit Loto (*.net *.split)
05:25:00 Quit gevaerts (*.net *.split)
05:25:00 Quit TBCOOL (*.net *.split)
05:25:00 Quit YPSY (*.net *.split)
05:25:41 Join zu [0] (
05:25:41 Join Keripo [0] (
05:25:41 Join Rob2223 [0] (
05:25:41 Join amiconn [0] (quassel@rockbox/developer/amiconn)
05:25:41 Join enthdegree [0] (
05:25:41 Join [Saint] [0] (S_a_i_n_t@
05:25:41 Join milk [0] (
05:25:41 Join factor [0] (~factor@
05:25:41 Join linuxguy3 [0] (
05:25:41 Join froggyman [0] (~seth@unaffiliated/froggyman)
05:25:41 Join JdGordon1 [0] (
05:25:41 Join plux [0] (
05:25:41 Join Dreamxtreme [0] (~Dre@
05:25:41 Join Horscht [0] (~Horscht@xbmc/user/horscht)
05:25:41 Join jae [0] (
05:25:41 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
05:25:41 Join panni_ [0] (
05:25:41 Join eGen_ [0] (
05:25:41 Join mystica555_ [0] (
05:25:41 Join Strife89 [0] (~Strife89@
05:25:41 Join jfc [0] (
05:25:41 Join Quazgaa [0] (
05:25:41 Join Galois [0] (
05:25:41 Join Battousai [0] (~bryan@gentoo/developer/battousai)
05:25:41 Join Sochiro [0] (~Sochiro@
05:25:41 Join cjcopi [0] (~craig@
05:25:41 Join n17ikh [0] (
05:25:41 Join DrDnar [0] (
05:25:41 Join ehntoo [0] (
05:25:41 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
05:25:41 Join guymann [0] (~charles@
05:25:41 Join AlexP [0] (~alex@rockbox/staff/AlexP)
05:25:41 Join kkit|sh [0] (
05:25:41 Join Dhraakellian [0] (
05:25:41 Join bzed [0] (
05:25:41 Join balintx [0] (
05:25:41 Join jepler [0] (~jepler@emc/developer/pdpc.professional.jepler)
05:25:41 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
05:25:41 Join yosafbridge [0] (
05:25:41 Join preglow [0] (
05:25:41 Join crwl [0] (
05:25:41 Join Loto [0] (~nfs@xbmc/user/Loto)
05:25:41 Join CIA-7 [0] (~CIA@
05:25:41 Join feisar- [0] (
05:25:41 Join Kohlrabi [0] (
05:25:41 Join Elfish [0] (amba@2a01:4f8:100:90a1:abc:abc:abc:abc)
05:25:41 Join amee2k [0] (
05:25:41 Join pikytcus [0] (
05:25:41 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
05:25:41 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
05:25:41 Join MagusG [0] (
05:25:41 Join ChanServ [0] (ChanServ@services.)
05:25:41 Join Farthen [0] (
05:25:41 Join ranmachan [0] (
05:25:41 Join powell14ski_ [0] (
05:25:41 Join logiclost [0] (
05:25:41 Join TBCOOL [0] (
05:25:41 Join YPSY [0] (
05:25:41Mode"#rockbox +o ChanServ " by
05:25:44 Quit Barahir (*.net *.split)
05:25:44 Quit Llorean (*.net *.split)
05:25:44 Quit advcomp2019 (*.net *.split)
05:25:44 Quit ack` (*.net *.split)
05:25:44 Quit parafin (*.net *.split)
05:25:44 Quit ThomasAH (*.net *.split)
05:27:26 Quit MethoS- (Remote host closed the connection)
05:27:38 Quit markun (Ping timeout: 240 seconds)
05:27:51 Quit Dexpid (Read error: Connection reset by peer)
05:28:49 Join markun [0] (~markun@
05:28:49 Quit markun (Changing host)
05:28:49 Join markun [0] (~markun@rockbox/developer/markun)
05:29:25 Join Barahir [0] (
05:29:25 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
05:29:25 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
05:29:25 Join ack` [0] (
05:29:25 Join parafin [0] (
05:29:25 Join ThomasAH [0] (
05:29:29 Quit panni_ (Quit: ( :: NoNameScript 3.81 :: ))
05:33:28 Join DanielleKayser [0] (
05:33:50DanielleKayserrequesting approval of wiki account DanielleKayser please
05:37:29[Saint]one sec.
05:38:50 Quit Horscht (Quit: Verlassend)
05:38:56 Quit DanielleKayser (Quit: CGI:IRC (Ping timeout))
05:39:42[Saint] DanielleKayser: if you come back...done, don't break our wiki! ;)
06:03:37 Join Klevi [0] (
06:03:50KleviHey guys! Question for everyone
06:04:48KleviI have an OpenPandora >> << if I were to build rockbox simply to be an Mp3/Video player for it, how should I go about doing that?
06:08:17froggymanThe easiest way would be to set up the build environment on your pandora, and compile the simulator for the M-robe 500 as I'm pretty sure it uses the same screen size as the pandora
06:08:56KleviOh! Didnt see you were here as well, froggyman >.o
06:10:00Kleviand alright, I should be more than capable of tackling this myself then now that I know the target build I need. Thanks. =) Will report back how it works out
06:11:23froggymanWell the M robe 500's screen resolution is actually 640x480... you might be able to change the screen size for the sim somewhere in one of the source files... not sure where though
06:13:59***Saving seen data "./dancer.seen"
06:16:17[Saint]why could it not run the SDL app?
06:16:28[Saint]then you could feed it whatever screen size you wanted.
06:17:28froggyman[Saint], Isn't the sim the same as the SDL app?
06:17:50KleviWhats different?
06:18:02JdGordon1if openpandora provides a nice posix api then just build the sdl app
06:22:22froggyman[Saint], JdGordon1, is there a wiki page for the SDL app?
06:22:53[Saint]not that I know's fantastically poorly doccumented.
06:23:18froggymanhaha... alright
06:40:24 Join ZhangNing [0] (~ZhangNing@
06:51:25 Quit Klevi (Quit: Leaving)
07:23:35 Join Choicefresh [0] (~Liam@unaffiliated/choicefresh)
07:23:48 Quit Barahir (Ping timeout: 245 seconds)
07:24:35 Join Barahir [0] (
07:25:00ChoicefreshHey, is there any work currently being done on accesory compatibility? (i.e. if i upgrade from the stable release to the current build, will it work with more dock speakers?)
07:30:19 Quit JdGordon1 (Quit: leaving)
07:32:43Choicefreshor, in other words, are there release notes for the "current build"?
07:32:53Choicefreshas in advantages/disadvantages
07:47:32 Quit Judas_PhD (Quit: This is a quitting message)
07:57:17 Join kugel [0] (~kugel@
07:57:19 Quit kugel (Changing host)
07:57:19 Join kugel [0] (~kugel@rockbox/developer/kugel)
08:07:26 Quit Choicefresh (Quit: Leaving.)
08:09:34 Join JdGordy [0] (
08:14:03***Saving seen data "./dancer.seen"
08:19:54 Join sideral [0] (~sideral@
08:19:55 Quit sideral (Changing host)
08:19:55 Join sideral [0] (~sideral@rockbox/developer/sideral)
08:20:07 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
08:22:13 Join bertrik [0] (
08:22:14 Quit bertrik (Changing host)
08:22:14 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
08:23:49 Quit Zambezi (Changing host)
08:23:49 Join Zambezi [0] (Zulu@unaffiliated/zambezi)
08:28:21 Join B4gder [0] (~danielx@rockbox/developer/bagder)
08:29:10 Quit rasher (Ping timeout: 240 seconds)
08:33:48sideralpixelma, Saint: Thanks for clarifying your position re autoresume. I had not intended to misrepresent what you've said. Of course I'm sorry you still don't like the feature as it is now; no feature ever is for everyone, I guess. But I'm impressed that, despite this, you've helped to improve it. Thanks! :)
08:34:45 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
08:38:56 Quit JdGordy (Ping timeout: 255 seconds)
08:39:51 Quit mc2739 (Ping timeout: 240 seconds)
08:40:25[Saint]As much as I don't necessarily think it's the right way to go, if it's the way we're going I still want it to be the best it can be.
08:40:32[Saint]Even if I don't use it.
08:40:37 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
08:40:50 Quit bertrik (Ping timeout: 240 seconds)
08:42:41sideralMuch appreciated!
08:46:00 Quit mc2739 (Ping timeout: 246 seconds)
08:46:20ZagorFYI: the Rockbox Steering Board has been asked to look into this issue
08:48:15 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
08:50:34 Quit ZhangNing (Remote host closed the connection)
08:50:51 Join ZhangNing [0] (~ZhangNing@
08:51:19 Join ender` [0] (
08:54:04 Join pamaury [0] (
08:54:05 Quit pamaury (Changing host)
08:54:05 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
08:54:59 Join mudd1 [0] (
08:57:00 Join efyx [0] (
08:57:46 Join tchan1 [0] (
08:59:05 Join petur [0] (d408b802@rockbox/developer/petur)
08:59:31 Join rasher [0] (
08:59:32 Quit rasher (Changing host)
08:59:32 Join rasher [0] (~rasher@rockbox/developer/rasher)
08:59:38 Quit tchan (Ping timeout: 240 seconds)
09:06:46sideralThanks for the heads up, Zagor. I welcome this move −− although I'm sorry to be the reason for its summoning. :(
09:07:16 Join liar [0] (
09:10:34 Join LinusN [0] (~linus@rockbox/developer/LinusN)
09:15:38kugelfwiw, I'm fine with autoresume as of now (in what it does/offers and the related settings; I would still welcome if it would be unified with bookmarks internally)
09:16:22 Join n1s [0] (~n1s@rockbox/developer/n1s)
09:18:12 Join jr [0] (
09:18:38JdGordon|Zagor: is there any procedure (publix or private) as to how it will be handled?
09:19:47ZagorJdGordon|: not many. the summoning is confidential, and the board is free to decide however it wants. including to decide not to decide.
09:19:50 Quit kugel (Ping timeout: 240 seconds)
09:21:42 Quit sideral (Ping timeout: 246 seconds)
09:24:55*JdGordon| is curious as to what the request wording was (by whom can probably be guessed and that is less interesting)
09:25:06JdGordon|but I understand that that is unlikely to happen
09:28:47Zagoryes, all mails sent to the rsb are strictly confidential
09:35:10 Join sideral [0] (~sideral@rockbox/developer/sideral)
09:37:00 Quit liar (Ping timeout: 255 seconds)
09:38:04 Join liar [0] (
09:43:39sideralZagor: Is this the first time ever the board was summoned?
09:44:41CIA-7New commit by jethead71 (r29257): buffering: Fix a case that could allow widx to fully wrap to ridx and overflow the ringbuffer.
09:46:32 Quit jr (Remote host closed the connection)
09:49:30CIA-7r29257 build result: All green
09:49:50 Quit GeekShadow (Ping timeout: 276 seconds)
10:00:07 Quit DrDnar (Read error: Connection reset by peer)
10:05:27 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
10:14:06***Saving seen data "./dancer.seen"
10:19:43 Quit mc2739 (Read error: Connection reset by peer)
10:19:54 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
10:24:30 Join Highlander [0] (
10:30:13CIA-7New commit by jethead71 (r29258): Buffering should align itself and not rely on buffering_reset parameters when storage alignment matters so that wrapped reads maintain alignment.
10:30:30 Join Judas_PhD [0] (
10:34:16CIA-7r29258 build result: All green
10:41:19 Quit ZhangNing (Remote host closed the connection)
10:43:38 Join einhirn [0] (
10:43:46 Quit Keripo (Quit: Leaving.)
10:45:00 Join DerPapst [0] (
10:45:35 Join Keripo [0] (
10:54:06 Quit Keripo (Quit: Leaving.)
11:03:31 Quit GeekShadow (Ping timeout: 240 seconds)
11:04:18 Join vnl [0] (
11:05:04 Quit vnl (Client Quit)
11:09:18 Join qurvel [0] (
11:09:52 Join mt [0] (~mt@
11:13:32 Quit mt (Changing host)
11:13:32 Join mt [0] (~mt@rockbox/developer/mt)
11:16:49 Quit Highlander (Quit: Quitte)
11:27:26 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
11:30:54[Saint]sideral:'s not the first time it's been considered ;)
11:33:19 Join dfkt [0] (dfkt@unaffiliated/dfkt)
11:37:25 Quit GeekShadow (Quit: The cake is a lie !)
11:42:21 Join kevku [0] (~kevku@2001:7d0:0:f9af:215:c5ff:fe7e:794d)
11:51:15 Join Raptors [0] (
11:51:24Raptorsis there a way to queue a song to be played nexT?
11:56:11[Saint]bring up the context menu on the track, choose Playlist, then Queue Next (or similar).
11:56:37[Saint]the manual describes it quite well ;)
11:57:07RaptorsThere is no queue next in manual
11:57:34[Saint]are you looking in the menu, or the context menu?
11:58:08[Saint]there's two options there, "yes" doesn't vut it.
11:58:25RaptorsI'm looking in the context menu
11:58:42[Saint]is a track playing?
11:59:18[Saint]I have no idea where you're looking then, as I'm looking at it right now.
11:59:49RaptorsMaybe the sanza fuse doesn't have it :/
11:59:54[Saint]if a track is playing, bring up the context menu on another track, select Playlist, then Queue Next
12:00:07[Saint]and, yes...I'm pretty sure it does ;)
12:01:22RaptorsIt's not there...
12:01:53[Saint]It doesn't even matter if you launch playback from the database of the file browser...there just needs to be a track playing currently so there's a playlist for it to be added to.
12:02:11[Saint]afaik it should be available on all targets.
12:02:34gevaertsIf it's not there, you're not looking in the right place.
12:02:40RaptorsIt's not launched from database
12:02:45gevaertsWhere did you start the context menu from?
12:02:57RaptorsFrom track playing screen
12:03:14[Saint]well, there you go.
12:03:28RaptorsI tried from the playlist too :|
12:03:29[Saint]you need to bring up the context menu on the track you wish to add.
12:03:52[Saint]from either the database, or the file browser.
12:04:00RaptorsI see it now
12:04:11RaptorsIt can't be in the playlist
12:05:10[Saint]if you're in the playlist, you can always use "move"
12:05:28[Saint]then you can just find the track you want and change its position manually.
12:08:08RaptorsRockbox is so much better then the stock firmware
12:14:09***Saving seen data "./dancer.seen"
12:18:15 Quit n1s (Ping timeout: 255 seconds)
12:35:14 Quit Rob2223 (Quit: Rob2223)
12:46:15 Join Rob2222 [0] (
12:48:29 Quit Rob2222 (Read error: Connection reset by peer)
12:48:39 Join Rob2222 [0] (
13:10:43 Join user890104 [0] (
13:11:09 Join thomasjfox [0] (
13:12:00*thomasjfox summons zagor
13:12:37*Zagor appears from the shadows
13:12:56Zagoryou rang?
13:13:17thomasjfoxyes. My wiki registration email got stuck somewhere. Could you check the logs?
13:13:43thomasjfoxI was some days ago
13:13:59 Quit CIA-7 (Read error: Operation timed out)
13:14:12thomasjfoxAnd elevated access rights in FS would be nice, too. Then I can close my own tickets :)
13:15:32Zagorwhat is your wiki name?
13:16:54Zagorthere is no such wiki account. try registering again.
13:17:02thomasjfoxok, hang on a sec
13:17:14 Quit user890104 ()
13:17:21thomasjfoxMaybe it got deleted as it wasn't confirmed after XX hours
13:17:25 Join CIA-70 [0] (~CIA@
13:17:38gevaertsthomasjfox: your wiki name should be your real name
13:18:02Zagorflyspray account upgraded
13:18:02thomasjfoxthen it was "Thomas Jarosch" - I'm a bit vague on the details
13:18:40Zagorok, that account is present
13:19:46Zagorthough no page is created. the wiki sure seems flaky these days.
13:20:11gevaertsZagor: that seems to happen a lot these days
13:20:43thomasjfoxZagor: I can't seem to reset the password for this account
13:21:03thomasjfox"" -> Access denied / The "Main/System" web does not exist
13:21:10[Saint]Zagor: that ties in with what happened to MysteryKeeper
13:21:14 Join n1s [0] (
13:21:15Zagorthomasjfox: I think wiki is confused since you have no account page
13:21:15 Quit n1s (Changing host)
13:21:15 Join n1s [0] (~n1s@rockbox/developer/n1s)
13:21:25[Saint]making the page manually and adding the default form "just works"
13:21:31thomasjfoxI guess the account is not confirmed yet
13:22:31gevaertsI'm not sure if the account page is that critical. I've added people without one to WikiUsersGroup and they could edit things
13:23:12[Saint]i thought the page *was* the account
13:23:25[Saint]and only passes and acces rights were stored.
13:24:40Zagorthomasjfox: you should be getting a reset mail now
13:25:09[Saint]as soon as I made MysteryKeeper's (I forget the real name) page, his login automagically worked...I thought the page/account was quite tightly tied.
13:25:24[Saint]happy to be wrong though ;)
13:26:05Zagor[Saint]: iirc I had to manually fix his password
13:26:52 Join casainho [0] (
13:29:16linuxstbZagor: Could you add RaaA to flyspray as a player type?
13:29:35thomasjfoxZagor: Password has been reset. I tried to edit something like the AndroidPort page. Is it normal that I don't have access to that?
13:29:44B4gderperhaps even specific raaa versions?
13:29:55thomasjfoxZagor: Same for the main wiki page, though I guess I don't have access to that either
13:30:33Zagorthomasjfox: you haven't been given access yet. doing it now.
13:30:46thomasjfoxAh sorry, the error message cleary states that :)
13:31:35Zagortry now
13:32:54thomasjfoxZagor: I get the edit box now. Thanks
13:33:49ZagorI added Android and Maemo to the player types
13:35:26linuxstbZagor: An overall "RaaA" type would be useful as well though IMO - I think a lot of tasks will be generic (my first RaaA patch is...)
13:35:57Zagoryes, but most work is generic to a lot of players.
13:36:10thomasjfoxlinuxstb: Or "Generic RaaA" to make it more precise?
13:36:32Zagorwe don't have "generic swcodec" or some such either
13:36:46Zagoractually we do ;*)
13:37:10*Zagor 's argument fails miserably
13:37:29gevaertsI'd say RaaA is swcodec :)
13:39:22 Quit kevku (Ping timeout: 260 seconds)
13:39:37 Quit thomasjfox (Remote host closed the connection)
13:39:52 Join thomasjfox [0] (
13:41:39 Join kevku [0] (~kevku@2001:7d0:0:f9af:215:c5ff:fe7e:794d)
13:42:03 Quit sasquatch (Ping timeout: 240 seconds)
13:43:09*Zagor restarts apache
13:46:15sideralZagor: While you are at it, could you help me setting up a Wiki account as well? I'm not sure in which state my registration / edit privs are.
13:46:35Torneanyone can do that
13:46:43Torneif you have an account and you can edit pages, that's it
13:46:49Tornethere are no speciak wiki privileges for developers
13:47:04Torneeveryone who's been granted edit permission at all can do anything
13:47:29Torne(and if you haven't then just tell us your username and nayone can add you)
13:47:38sideralJust tried reregistering / logging in, but it said "Access check on Main.MichaelHohmuth failed. Action "CHANGE": access not allowed on web"
13:48:09sideralAlso, didn't receive any registration email...
13:48:33Zagorapparently nobody does
13:48:46TorneYour user page doesn't exist
13:48:49Torneso I don't think you are registered
13:48:59Zagorhe exists in the .htpasswd file
13:48:59sideralShall I try again?
13:49:08Torneoh, maybe it gets created on first login?
13:49:18Zagorno, there's something wrong with the wiki
13:49:25Torneah :)
13:50:55sideralTorne: the wiki home page actually says to ask for write permission here, so I thought I needed that in addition to the account
13:51:04Zagorsideral: yes, you do
13:51:11Torneyou do, but anyone who already has permission can do it
13:51:20Torne..assuming the wiki is working properly ;)
13:51:39Tornethe write permission includes permission to write to the list of people who have write permission ;)
13:51:48sideralI see
13:52:23Zagorsideral: I sent you a new password mail and added you to the permission group
13:53:08 Join afk [0] (~Dre@
13:53:12 Quit thomasjfox (Quit: Back to day job)
13:55:21sideralThanks Zagor. The email message hasn't arrived yet. Standing by...
13:55:37 Quit Dreamxtreme (Ping timeout: 240 seconds)
13:57:34 Join TheLemonMan [0] (
14:01:50 Quit mystica555 (Read error: Connection reset by peer)
14:01:50 Quit mystica555_ (Read error: Connection reset by peer)
14:03:07 Quit Topy (Read error: Connection reset by peer)
14:03:07 Join T44 [0] (
14:04:08 Join mystica555 [0] (
14:04:09 Quit antil33t (Read error: Connection reset by peer)
14:04:18 Join antil33t [0] (
14:08:43 Join sasquatch [0] (~username@
14:13:00 Join mystica555_ [0] (
14:13:17 Join Dexpid [0] (
14:13:57sideralZagor: I got the email, and could log in. Thanks!
14:14:13***Saving seen data "./dancer.seen"
14:15:39sideralI have no user page, is that a problem?
14:23:16gevaertsWell, it's not a *big* problem
14:23:40gevaertsYour user page is where you store certain settings though, but I suspect most people don't use that
14:24:08Zagoryou might want to create one, to avoid anyone else doing it :)
14:24:16gevaertsAh, that too :)
14:24:38n1sgevaerts: there's one crucial setting, the one that turns off the broken editor
14:24:48gevaertsn1s: yes, that's the main one
14:25:11sideralAh, OK. I guess I should copy n1s' user page then. :)
14:25:37Zagorjust make sure you change the ALLOWTOPICCHANGE line!
14:26:10Zagoris the wysiwyg editor active? I thought I disabled it globally
14:26:29n1ssideral: afair gevaerts is the one that added that setting to my page after i broke the UserGroup(?) page locking every non admin out from editing
14:26:30gevaertsZagor: not changing that one saved me when I was playing with disabling the wysiwyg editor back when it was new :)
14:26:49gevaertsSomeone else was able to fix my settings then
14:27:17gevaertsZagor: maybe you did. Having my own settings makes that invisible :)
14:32:34sideralHmm, I told it to create a child of TWikiUsers, but the edit dialog says I'm ion Main
14:32:56 Join Topy [0] (
14:33:01Zagoryes, the wiki structure is flat. the "child" notion is only a tag.
14:33:19sideralah, I see.
14:36:03 Quit T44 (Ping timeout: 240 seconds)
14:37:32sideralDone. Thanks for your help!
14:39:06 Quit afk (Ping timeout: 276 seconds)
14:45:21 Join Dreamxtreme [0] (~Dre@
14:48:56 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
14:50:48 Quit sideral (Quit: Leaving.)
14:59:11 Quit mudd1 (Ping timeout: 240 seconds)
15:01:14 Join panni_ [0] (
15:04:59 Quit antil33t (Read error: Connection reset by peer)
15:05:08 Join antil33t [0] (
15:06:04 Join T44 [0] ( was accessed by 2245 unique IP addresses yesterday
15:08:32Tornethat seems like a lot
15:08:42n1sB4gder: any idea of how many completed downloads?
15:09:02B4gderwe also have manuals there
15:09:16 Quit Topy (Ping timeout: 240 seconds)
15:10:27B4gder3335 HEAD requests
15:10:34B4gderwhich are rbutil
15:12:06B4gder2556 requests for build-info
15:12:23B4gderalso rbutil querying for info
15:13:16B4gder6661 downloads, roughly
15:13:47[Saint]could we drop a file to check for a new install compared to an update?
15:13:54[Saint](with rbutil)
15:14:06[Saint]statistics would be *great*
15:14:58B4gder2125 of the downloads were rbutil
15:16:54B4gder422 were =)
15:17:14[Saint]haha! is the by far most downloaded file, 1697 downloads
15:18:04 Join manotroll [0] (
15:18:24B4gderall this is only during 24 hours
15:19:25 Quit JdGordon| (Ping timeout: 246 seconds) 560 downloads (!)
15:19:58 Join webguest20 [0] (
15:20:06[Saint]what the!?!
15:20:11webguest20not possible to recognize the rockbox playlist from WMP12?
15:20:22*[Saint] expected the #1 to be an iTarget...surely 103 downloads
15:20:40gevaertsB4gder: unique IPs? Plausible user agents?
15:21:17B4gderright now I just filter out GETs and remove most non-files, so I don't see 206 responses
15:21:41B4gder(206 being used for partial responses)
15:21:45 Quit webguest20 (Client Quit)
15:21:49 Quit antil33t ()
15:21:58gevaertsI mean, some of those might be spiders, which are not very interesting
15:22:07 Join manotroll-brasil [0] (
15:22:10manotroll-brasilnot possible to recognize the rockbox playlist from WMP12? because it does not recognize the list and do not know how to solve this
15:22:28B4gderthis is just some quick grepping and sorting because I'm curious
15:22:29 Join mudd1 [0] (
15:22:39B4gderI'll get the proper stats later on using a better stats script
15:22:55[Saint]if it's not an .m3U playlist, then no...I doubt RB can read it.
15:23:04 Quit manotroll (Quit: CGI:IRC (Ping timeout))
15:23:08[Saint]manotroll-brasil: ^
15:23:43[Saint]if it *is* an .m3u, but has specific paths reffering to files on your PC, it will also fail.
15:23:48manotroll-brasiland how to create one. m3u or convert to this format?
15:24:19[Saint]it's a let less painful to just create the playlist on the diveice
15:24:32[Saint]since you'll be playing files present on it anyway.
15:24:43 Quit cjcopi (Ping timeout: 264 seconds)
15:24:53*[Saint] tries again.
15:25:08[Saint]"it's a lot easier to just create your playlists with Rockbox"
15:25:54Torneer, it's okay if the playlist refers to files on the PC as long as those files are in the same layout on the player somewhere
15:26:08TorneIt will strip drive letters/directory names off the front of the tracks until the path matches
15:26:32Torneso if the playlist refers to e:\stuff\files\music\foo.mp3 then that will work if the file on the player is called \music\foo.mp3
15:27:03Torneso as long as you have all your music under a single folder on both devices, and it makes m3u format playlists, it should work okay
15:27:09manotroll-brasiland separate it by artist and with the list I can choose which song to listen and not have to throw everything into a single folder that he does not accept the playlist from wmp12/11/10 since he and a program that open this support did not win
15:28:42 Join komputes [0] (~komputes@ubuntu/member/komputes)
15:29:33manotroll-brasilbut it is not possible to convert? and where to put that differently than does the rockbox
15:29:50Tornewhat format are the playlists in now?
15:29:57Tornewhat is the file extension?
15:31:56manotroll-brasilthe problem here in Brazil that have special characters like "~, ç,^,~, and others that rockbox does not recognize
15:32:33TorneRockbox does recognise all characters.
15:32:47Torneyour filenames or tags are probably encoded wrong :)
15:33:04[Saint]If your font doesn't support those characters, they won't be displayed.
15:33:21[Saint]or incorrect codepage.
15:35:35manotroll-brasil8 Unicode (UTF-8) and widely used but has variables but this is not possible to read the list of WMP12? the important thing was to read this type of list
15:38:09 Join MethoS- [0] (~clemens@
15:38:34[Saint]it seems WMP12 uses it's own XML format, WPL...this is not supported.
15:39:15manotroll-brasiluse from when the support was experimental sansa c200v2 thought it would be the solution to the problem of not getting the full 8gb card (only recognizes 4gb) but from the look and use it only 4gb
15:39:29 Nick tchan1 is now known as tchan (
15:39:34 Quit tchan (Changing host)
15:39:34 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
15:41:06manotroll-brasilbut it also creates m3u but not recognized and
15:41:46[Saint]as explained earlier, if it creates an m3u playlist it will work if the paths match
15:41:53[Saint](at least in part)
15:42:14[Saint][03:26] <Torne> It will strip drive letters/directory names off the front of the tracks until the path matches
15:42:16[Saint][03:27] <Torne> so if the playlist refers to e:\stuff\files\music\foo.mp3 then that will work if the file on the player is called \music\foo.mp3
15:42:16[Saint][03:27] <Torne> so as long as you have all your music under a single folder on both devices, and it makes m3u format playlists, it should work okay
15:43:07Torneyou're talking about several different problems and it's not clear what you mean with *any* of them really, i'm afraid your english is not very clear
15:46:34 Join sideral [0] (~sideral@
15:46:34 Quit sideral (Changing host)
15:46:34 Join sideral [0] (~sideral@rockbox/developer/sideral)
15:49:02manotroll-brasilI use a translator when I create a list for him but the songs are scattered sample in the System of a Down has the folders Mezmerize, Steal This Album, System Of A Down Toxicity and looks like this so I can choose which to listen or to use the playlist to play them all in an order defined by min the buffer error when I try to make a playlist with he or she does not touch anything even with the correct path can not make a video with the emulator
15:50:34Lloreanmanotroll-brasil: Try using short, simple sentences. Say one thing at a time. It often helps.
15:50:55LloreanAlso, if you get errors, say what the exact error message is
15:52:46manotroll-brasilafter the end of sync with wmp 12 I'll try to make another playlist and come back here to see which will give error and inform
15:57:44 Quit manotroll-brasil (Quit: CGI:IRC)
15:59:52 Quit sideral (Quit: Leaving.)
16:05:44 Quit B4gder (Read error: Operation timed out)
16:06:06 Part Zagor
16:06:10 Part LinusN
16:06:13 Join B4gder [0] (~danielx@2a00:1a28:1200:9::2)
16:06:13 Quit B4gder (Changing host)
16:06:13 Join B4gder [0] (~danielx@rockbox/developer/bagder)
16:06:14 Quit B4gder (Remote host closed the connection)
16:06:28 Quit Dexpid (Quit: Computer has gone to sleep.)
16:06:46 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
16:06:50 Join B4gder [0] (~danielx@rockbox/developer/bagder)
16:13:51 Quit mt (Ping timeout: 240 seconds)
16:14:14***Saving seen data "./dancer.seen"
16:15:02 Quit Dreamxtreme (Read error: Connection reset by peer)
16:16:47 Join LinusN [0] (~linus@rockbox/developer/LinusN)
16:18:11 Quit TheLemonMan (Quit: free(me))
16:19:01 Quit eGen_ (Read error: Connection reset by peer)
16:20:20 Part LinusN
16:20:28 Join L-Strife89 [0] (
16:22:52 Join eGen_ [0] (
16:22:52 Part Zagor
16:22:52 Join Zagor [0] (~bjst@rockbox/developer/Zagor)
16:26:01 Quit n1s (Quit: Lämnar)
16:39:50 Join LinusN [0] (~linus@rockbox/developer/LinusN)
16:45:38*TheSeven is running his ipod classic at 288MHz
16:46:25*[Saint] smells smoke.
16:47:19 Join toffe82 [0] (
16:48:30 Join Keripo [0] (
16:49:25TheSeven[Saint]: 318MHz now, still running smoothly
16:49:53[Saint]"Do not release the genie"
16:50:05 Join Dexpid [0] (
16:50:31enthdegreeWhy is FM recording so low quality?
16:51:06enthdegreeis it because you can't write to the disk fast enough?
16:52:22TheSeven[Saint]: at 348MHz it started calculating garbage
16:52:42[Saint]reason that thought when compared to recording fron any other input...
16:52:48[Saint]enthdegree: ^
16:52:48TheSeven342MHz seems to be stable
16:52:55DexpidIs rockbox written in just C?
16:53:22TheSevenDexpid: the core, yes
16:53:32TheSeventhere are a few LUA plugins as well
16:53:34Lloreanenthdegree: What player?
16:53:43Lloreanenthdegree: Some players have very limited recording capabilities in general.
16:53:48enthdegreeClip plus
16:53:51TheSevenand the PC-side tools are of course partially perl or shell scripts as well
16:54:21TheSeven158% factory clock... nice :)
16:54:31Lloreanenthdegree: Now, what do you mean by "low quality"? What are your settings, recorded filetype, and what's wrong with the outputted file?
16:54:35TheSeven(at factory Vcore)
16:54:48enthdegreeTheSeven: What is the slowest you can clock it?
16:55:05[Saint]0Mhz idle? ;)
16:55:14TheSevenenthdegree: that depends on the peripherals that need to work
16:55:28TheSevenI think USB starts to cock up at <50MHz somewhere
16:55:39 Quit sasquatch (Quit: WeeChat 0.3.2)
16:55:48enthdegreeLlorean: The outputted file sounds more grainy than what I hear from the radio interface
16:56:02TheSevenalso the LCD interface speed scales linarly and is at ~25fps at 216MHz
16:56:03enthdegreeI will pull up my settings...
16:56:51[Saint]enthdegree: have you compared Fm recording to recording from line in?
16:57:03[Saint]your statement sounded to me as though it was worse.
16:57:45enthdegreeI have not, I will do that.
16:58:08TheSevenenthdegree: is there some kind of buzzing or ticking involved?
16:58:10 Quit bluebrother (Ping timeout: 255 seconds)
16:58:46LloreanDoes the ClipV2 record at full sample rate?
16:58:49LloreanEr Clip+
16:59:04LloreanI seem to recall some of those players had restrictions in sample rate or bit rate (14 bit instead of 16?)
16:59:09LloreanEr bit depth
16:59:44LloreanBy "those devices" though I just mean "sandisks that I can no longer keep straight"
16:59:59 Join bluebrother [0] (
16:59:59 Quit bluebrother (Changing host)
16:59:59 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
17:01:14TheSeventhe UI is running at ~3fps @ 13.5Mhz
17:01:26TheSevenUSB works down to 27MHz apparently
17:01:49enthdegreerun mpegplayer
17:01:50[Saint]bet the UI feels nice and responsive ;)
17:03:34 Quit Keripo (Quit: Leaving.)
17:03:34 Join sideral [0] (~sideral@
17:03:34 Quit sideral (Changing host)
17:03:34 Join sideral [0] (~sideral@rockbox/developer/sideral)
17:03:34 Part LinusN
17:03:34timccchows the battery life on the classic now?
17:04:24TheSeventhe HDD is doing something, but failed to load rockbox.ipod @13.5MHz
17:04:24 Part Zagor
17:04:38 Quit B4gder (Remote host closed the connection)
17:05:24TheSeventimccc: i have no idea
17:05:41TheSevenbut if i port undervolting and clock scaling, it should be better at least
17:06:11 Quit Judas_PhD (Quit: This is a quitting message)
17:06:12 Quit logbot (Ping timeout: 240 seconds)
17:06:12***ERROR: (Closing Link: (Ping timeout: 240 seconds)) from
17:06:12***Saving seen data "./dancer.seen"
17:06:14***Started Dancer V4.16
17:06:14***Connected to on port 6667
17:06:14***Logfile for #rockbox started
17:06:14Mode"logbot :+i" by logbot
17:06:18***Server message 501: 'logbot :Unknown MODE flag'
17:06:18 Join logbot [0] (
17:06:18 Join sideral [0] (~sideral@rockbox/developer/sideral)
17:06:18 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
17:06:18 Join Dexpid [0] (
17:06:18 Join toffe82 [0] (
17:06:18 Join eGen_ [0] (
17:06:18 Join L-Strife89 [0] (
17:06:18 Join MethoS- [0] (~clemens@
17:06:18 Join komputes [0] (~komputes@ubuntu/member/komputes)
17:06:18 Join mudd1 [0] (
17:06:18 Join T44 [0] (
17:06:18 Join panni_ [0] (
17:06:18 Join evilnick_B [0] (0c140464@rockbox/staff/evilnick)
17:06:18 Join mystica555_ [0] (
17:06:18 Join mystica555 [0] (
17:06:18 Join kevku [0] (~kevku@2001:7d0:0:f9af:215:c5ff:fe7e:794d)
17:06:18 Join casainho [0] (
17:06:18 Join CIA-70 [0] (~CIA@
17:06:18 Join Rob2222 [0] (
17:06:18 Join Raptors [0] (
17:06:18 Join dfkt [0] (dfkt@unaffiliated/dfkt)
17:06:18 Join qurvel [0] (
17:06:18 Join DerPapst [0] (
17:06:18 Join einhirn [0] (
17:06:18 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
17:06:18 Join liar [0] (
17:06:18 Join rasher [0] (~rasher@rockbox/developer/rasher)
17:06:18 Join petur [0] (d408b802@rockbox/developer/petur)
17:06:18 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
17:06:18 Join efyx [0] (
17:06:18 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
17:06:18 Join ender` [0] (
17:06:18 Join Barahir [0] (
17:06:18 Join ThomasAH [0] (
17:06:18 Join parafin [0] (
17:06:18 Join ack` [0] (
17:06:18 Join advcomp2019 [0] (~advcomp20@unaffiliated/advcomp2019)
17:06:18 Join Llorean [0] (~DarkkOne@rockbox/user/Llorean)
17:06:18 Join markun [0] (~markun@rockbox/developer/markun)
17:06:18 Join YPSY [0] (
17:06:18 Join TBCOOL [0] (
17:06:18 Join logiclost [0] (
17:06:18 Join powell14ski_ [0] (
17:06:18 Join ranmachan [0] (
17:06:18 Join Farthen [0] (
17:06:18 Join @ChanServ [0] (ChanServ@services.)
17:06:18 Join MagusG [0] (
17:06:18 Join gevaerts [0] (~fg@rockbox/developer/gevaerts)
17:06:18 Join scorche|sh [0] (~scorche@rockbox/administrator/scorche)
17:06:18 Join pikytcus [0] (
17:06:18 Join amee2k [0] (
17:06:18 Join Elfish [0] (amba@2a01:4f8:100:90a1:abc:abc:abc:abc)
17:06:18 Join Kohlrabi [0] (
17:06:18 Join feisar- [0] (
17:06:18 Join Loto [0] (~nfs@xbmc/user/Loto)
17:06:18 Join crwl [0] (
17:06:18 Join preglow [0] (
17:06:18 Join yosafbridge [0] (
17:06:18 Join dionoea [0] (~dionoea@videolan/developer/dionoea)
17:06:18 Join jepler [0] (~jepler@emc/developer/pdpc.professional.jepler)
17:06:18 Join balintx [0] (
17:06:18 Join bzed [0] (
17:06:18 Join Dhraakellian [0] (
17:06:18 Join kkit|sh [0] (
17:06:18 Join AlexP [0] (~alex@rockbox/staff/AlexP)
17:06:18 Join guymann [0] (~charles@
17:06:18 Join ps-auxw [0] (~arneb@2001:470:c807:0:1532:4e5f:2ad3:4123)
17:06:18 Join ehntoo [0] (
17:06:18 Join n17ikh [0] (
17:06:18 Join Sochiro [0] (~Sochiro@
17:06:18 Join Battousai [0] (~bryan@gentoo/developer/battousai)
17:06:18 Join Galois [0] (
17:06:18 Join Quazgaa [0] (
17:06:18 Join jfc [0] (
17:06:18 Join Strife89 [0] (~Strife89@
17:06:18 Join jae [0] (
17:06:18 Join plux [0] (
17:06:18 Join froggyman [0] (~seth@unaffiliated/froggyman)
17:06:18 Join linuxguy3 [0] (
17:06:18 Join factor [0] (~factor@
17:06:18 Join milk [0] (
17:06:18 Join [Saint] [0] (S_a_i_n_t@
17:06:18 Join enthdegree [0] (
17:06:18 Join amiconn [0] (quassel@rockbox/developer/amiconn)
17:06:18 Join zu [0] (
17:06:18 Join Kitar|st [0] (
17:06:18 Join TheSeven [0] (~TheSeven@rockbox/developer/TheSeven)
17:06:18 Join pixelma [0] (quassel@rockbox/staff/pixelma)
17:06:18 Join elcan [0] (
17:06:18 Join mordocai_afk [0] (~mordocai@
17:06:18 Join shai [0] (
17:06:18 Join krazykit [0] (
17:06:18 Join timccc [0] (~timccc@
17:06:18 Join Guinness [0] (
17:06:18 Join Slasheri [0] (miipekk@rockbox/developer/Slasheri)
17:06:18 Join alexbobP [0] (
17:06:18 Join Xerion [0] (
17:06:18 Join jhMikeS [0] (~jethead71@rockbox/developer/jhMikeS)
17:06:18 Join [fred] [0] (
17:06:18 Join literal [0] (
17:06:18 Join simonrvn [0] (
17:06:18 Join w0m [0] (~wom@
17:06:18 Join jordan` [0] (
17:06:18 Join scorche [0] (~scorche@rockbox/administrator/scorche)
17:06:18 Join Zambezi [0] (Zulu@unaffiliated/zambezi)
17:06:18 Join iq [0] (~iq@unaffiliated/iq)
17:06:18 Join linuxstb [0] (~linuxstb@rockbox/developer/linuxstb)
17:06:18 Join thegeek_ [0] (
17:06:18 Join Bagder [0] (~daniel@rockbox/developer/bagder)
17:06:18 Join niekie [0] (quasselcor@CAcert/Assurer/niekie)
17:06:18 Join sinthetek [0] (~sinthetek@unaffiliated/sinthetek)
17:06:18 Join blast007 [0] (~blast007@bzflag/developer/Blast)
17:06:18 Join GodEater [0] (~bibble@rockbox/staff/GodEater)
17:06:18 Join sbhsu [0] (~a6530466@
17:06:18 Join vedos [0] (
17:06:18 Join soap [0] (~soap@rockbox/staff/soap)
17:06:18 Join aevin [0] (eivindsy@unaffiliated/aevin)
17:06:18 Join merbanan [0] (
17:06:18 Join FOAD [0] (~dok@
17:06:18 Join simabeis [0] (
17:06:18 Join knittl [0] (~knittl@unaffiliated/knittl)
17:06:18 Join miceh [0] (
17:06:18 Join ved [0] (
17:06:18 Join maraz [0] (
17:06:18 Join pjm0616 [0] (~user@
17:06:18 Join Hadaka [0] (
17:06:18 Join Rondom [0] (
17:06:18 Join Utchy [0] (
17:06:18 Join tmzt [0] (~tmzt@
17:06:18 Join Torne [0] (~torne@rockbox/developer/Torne)
17:06:18 Join Unhelpful [0] (~quassel@rockbox/developer/Unhelpful)
17:09:56 Join Judas_PhD [0] (
17:11:11 Quit Judas_PhD (Client Quit)
17:12:35 Quit sideral (Remote host closed the connection)
17:12:49 Join sideral [0] (~sideral@
17:12:49 Quit sideral (Changing host)
17:12:49 Join sideral [0] (~sideral@rockbox/developer/sideral)
17:13:46 Quit liar (Ping timeout: 255 seconds)
17:14:40 Quit L-Strife89 (Quit: Returning to the other building.)
17:17:27TheSevenwas that gui boosting patch committed?
17:17:42TheSeveni think it will be a prerequisite for ipod classic cpu clock scaling
17:17:49TheSevenat <200MHz the LCD is lagging badly
17:18:13[Saint]It would be good to get it in, but it needs work.
17:18:20 Quit shai (Read error: Connection reset by peer)
17:18:22TheSevenwhat's wrong with it?
17:18:27[Saint]it currently messes up keyclick spectacularly.
17:18:47 Join shai [0] (
17:18:56[Saint]it uses key events to decide when to boost.
17:19:03 Join liar [0] (
17:19:11[Saint]and...that is playing havoc with keyclick.
17:19:29[Saint]piezo and SW
17:19:58 Join Keripo [0] (
17:22:21TheSevenhm, the PLLs don't like to lock at low frequencies
17:23:36pamauryis there a way to get the virtual address of the TTB on arm ?
17:23:58 Quit Dexpid (Remote host closed the connection)
17:24:47Tornevirtual address?
17:24:55Torneit's wherever you mapped it :)
17:25:26TheSevenpamaury: what are you planning to do? some kind of reverse page table lookup?
17:26:46pamauryI have no control over the virtual space, it's setuped by the WinCE bootloader, it seems the physical address I provide to the LCD controller doesn't match the virtual address I'm thinking about. So I would like to hva e alook at the translation table to understand :)
17:27:24pamauryOf course, there is the solution of setting up my own virtual space but...
17:27:48TorneThere's no way, anyway.
17:28:01TorneYou can get the physical address from TTBR in cp15
17:28:12pamauryyes but that's not really helpful
17:28:36Torneno, I know.
17:28:37Tornebut that's it.
17:28:45TheSevenpamaury: the first thing i'd do when playing with such a platform would be to kill the MMU :)
17:28:48Torneyou are expected to have mapped the pagetables yourself ;)
17:29:38TheSevendamn, messed up some calculations above
17:29:47TheSeven96MHz seems to be the lowest speed that USB works at
17:29:56pamauryTheSeven: the difficult point is that each time I try to touch the MMU, I can't make the things work :) And I don't have JTAG or something similar
17:29:58 Join n1s [0] (
17:29:59 Quit n1s (Changing host)
17:29:59 Join n1s [0] (~n1s@rockbox/developer/n1s)
17:30:33 Join domonoky [0] (~Domonoky@rockbox/developer/domonoky)
17:30:33 Join t0rc [0] (~t0rc@unaffiliated/t0rc/x-5233201)
17:30:42pamauryAnyway, I'll continue to try until I suceed
17:30:43 Join LinusN [0] (~linus@rockbox/developer/LinusN)
17:32:06TheSevenpamaury: are you relying on any bootloader services or something?
17:32:21TheSevenif you fully take over control, fixing up the MMU shouldn't be hard
17:33:07pamauryThere is bootloader, loading WinCE, but I don't fully understand how does it setup the virtual space. And currently, I can't get rid of WinCE code (it's using some weird format), so I overwrite parts of it
17:33:33pamauryIt's just that I don't know much about ARM :) So I write buggy code ;)
17:33:44*TheSeven suggests to reverse the bootloader until you can get rid of all vendor code :)
17:33:50pamauryI can't
17:33:54pamauryI don't have access to it
17:34:01TheSevencan't you dump it?
17:35:40 Quit [Saint] (Quit: I'm only going to Heaven if it feels like Hell, I'm only going to Heaven if it tastes like caramel...)
17:35:40TheSevenwhat means of output do you currently have?
17:35:42pamauryUsing some wince tools I found on the net, I dumped something which perhaps contains the bootloader, it's already a second stage bootloader. And the trick is the second seems to do calls to the first for all the hardware parts. And I don't have the first
17:35:59pamaurySo I don't understand anything :)
17:36:07 Join sasquatch [0] (~username@
17:36:19TheSeventhen you can try to impersonate the second stage :)
17:40:36pamauryI can try to kill the MMU, setup a new stack and pray that gcc generated position independent code for the rest. Or else I'll have to copy it :-/
17:41:17 Join benedikt93 [0] (~benedikt9@unaffiliated/benedikt93)
17:41:59 Join [Saint] [0] (S_a_i_n_t@
17:43:59TheSevenpamaury: PIC doesn't help you here
17:44:14TheSevenin the worst case your code is scattered all around the memory
17:44:57pamauryyes but I'm optimistic and think it's contiguous :)
17:45:06pamauryso how do I disable the MMU ?
17:45:24TheSevenso you basically need to allocate a contiguous region of pyhsical memory that doesn't contain any of your code, do a 1:1 mapping for that, copy the code there, jump there, and then kill the MMU
17:45:43LinusNwoo, i managed to repair my tblcf! :-)
17:46:12*TheSeven 's ipod classic is idling at 23.6mA now
17:46:41TorneTheSeven: which you can't do if you don't know where the pagetables are
17:46:44TheSevenif I disable USB i can probably get way lower
17:46:50TorneTheSeven: since you can't make mappings
17:46:56pamauryI can't, I must disable the mmu first
17:47:21TorneThere's no way to disable the MMU without the hardware basically immediately falling over, anyway. :)
17:47:22TheSevenTorne: ah, right, the TTB base address is a physical address ./
17:47:44TheSevenTorne: yeah, at least if you don't know the current setup
17:47:47Tornecan you write to whatever's at physical address zero?
17:47:55Tornei.e. is SDRAM there?
17:48:10Torneif so you can probably do it by installing your own data/prefetch abort handlers
17:48:32Tornethen assuming that the machine will take an external abort after you disable the mmu
17:48:34 Quit [Saint] (Ping timeout: 250 seconds)
17:48:35TheSevenTorne: I like that devious idea :)
17:48:45Torneexcept quite possibly it won't, it depends where you are in the address space ;)
17:49:00Torneyou need to jump to a virtual address which you know doesn't map to anything as a physical address ;)
17:49:20pamaurywhat I know is that the running code is mapped at 0x8c000000. SRAM is normally at 0x30000000, registers seems to be identify mapped
17:49:37Torneyeah you're probably screwed
17:49:46pamaurywait, I think there is a way to know the physical address
17:49:53TheSevenTorne: how does icache behave if the MMU is being disabled?
17:49:57TorneTheSeven: inconsistently
17:49:59pamaurylet me dig into the wince nightmarish format
17:50:08 Join [Saint] [0] (S_a_i_n_t@
17:50:10TheSevenso one can hope that it still executes a few "old" instructions?
17:50:15TorneIt will
17:50:21TorneBut you don't know how many
17:50:33TheSevenbut that will be enough to trigger an undefined instruction abort :)
17:50:46TorneBut that's only any use if the hardware has writable memory at 00000000 or FFFF0000
17:50:58Tornesince otherwise the undef abort will probably go to a rom vector
17:51:00TheSevenif it doesn't you're screwed anyway
17:51:41pamauryI could do a conservative assumption: 0x8c000000 maps to 0x30000000 so I just disable the MMU and the link the code to 0x30000000
17:52:05TheSevendammit, USB is unstable in these conditions
17:52:07TorneThe problem ist hat disbaling the MMU is not atomic
17:52:17TorneThe next instruction is not guaranteed to execute witht he MMU disabled :)
17:52:25TorneThis is why you normally need an identity mapping
17:52:25pamaurybut if you put nops ?
17:52:26TheSevenif i switch it off i get down to 12.2mA drawn from the bus, so maybe like 10mA when running on battery
17:52:35Tornepamaury: the problem is you have to put exactly the right number of instructions
17:52:37pamauryah ok, I understand
17:52:42Torneif you branch too early you die, if you branch too late you die
17:52:49Torneand the right number may not be constant :)
17:52:53Tornesince it depends on the icache
17:53:15TorneIt's probably possibel to figure out a way to do this that's reasonably reliable but it won't be particularly easy.
17:53:28TorneIt's likely that the whole of SDRAM is mapped as a single block
17:53:35TorneI'd suggest just dumping the whole of where it would logically be
17:53:45Torneand looking for things that appear to be PDEs in there
17:53:54Tornesince it's probably not a complicated mapping
17:53:56 Quit Keripo (Quit: Leaving.)
17:53:58TheSevenpamaury: can you write to virtual 0x30000000?
17:54:14pamaurydidn't try
17:54:52TheSevenwhat means of output do you have currently?
17:54:57TheSevencan you download big amounts of data?
17:55:25pamauryI can write to the LCD
17:56:02pamauryI just let windows setup the memory and then take control
17:56:36pamaurythere is another option: reverse engineer WinCE mmu code
17:58:28TheSevenhm, 12.5-17mA depending on the cpu load @96MHz, Vcore=0.95V
17:59:33 Quit petur (Quit: Page closed)
17:59:47TheSevenlocking the hold switch increases the current by ~3.7mA
17:59:56TheSevenwhat the hell
18:02:01TheSevener, that was 24MHz actually :P
18:03:20pamauryhaha, I've come up with a nastu idea
18:03:43*TheSeven wonders if he should just commit the voltages that work for him and see if people complain
18:04:29pamauryI should have think about it earlier: I can ask WinCE to identify map the SDRAM !
18:07:01TheSevenoh damn, there's another thing that's holding back clock scaling:
18:07:07 Quit t0rc (Quit: Give someone code, help them with one project. Teach someone to code, help them rule the world.)
18:07:10TheSeventhe timers seem to be derived from the cpu clock :/
18:07:44pamauryhaha, you can slow down the worl which such a device :)
18:10:06TheSevenhm, i can actually change the PLL division factors without the device crashing, even if the CPU clock is coming from that PLL
18:11:04TheSevenUSB works at 51MHz, is unstable at 48MHz and unusable below that
18:11:22sideralTheSeven: Out of interest: How are you measuring these currents −− with an on-chip meter + ADC or with an external meter hooked up to the battery?
18:11:34TheSevenexternal meter on the USB bus
18:11:44TheSeveni haven't figured out how to use the internal one yet
18:12:01TheSevenand on the nano2g the internal one seems to have a ~7mA offset, so that isn't very usable either
18:12:55 Quit kevku (Ping timeout: 272 seconds)
18:13:15sideralTheSeven: Wouldn't you only be able to measure the charge current on the USB bus?
18:13:39TheSevenif the battery is fully charged i get reasonable values
18:13:58sideralah, interesting, thanks
18:13:58TheSeven(the charger is only acting as an LDO regulator in that case, and eating like 1mA by itself)
18:14:29TheSevenseems like the ATA UDMA timings are derived from the CPU clock as well :/
18:14:31 Part LinusN
18:17:42 Quit evilnick_B (Ping timeout: 245 seconds)
18:29:27 Join Keripo [0] (
18:37:42 Quit DerPapst (Quit: Leaving.)
18:38:04 Join Dreamxtreme [0] (~Dre@
18:38:05 Join GeekShadow [0] (
18:38:06 Quit GeekShadow (Changing host)
18:38:06 Join GeekShadow [0] (~Antoine@reactos/tester/GeekShadow)
18:38:18 Join Jerom [0] (~jerome@
18:43:00 Join dfkt|x [0] (~dfkt@unaffiliated/dfkt)
18:43:21 Quit Galois (Quit: Leaving)
18:47:35 Quit sideral (Quit: Leaving.)
18:54:44 Quit dfkt|x (Remote host closed the connection)
19:02:17 Join Galois [0] (
19:06:16***Saving seen data "./dancer.seen"
19:07:53 Join datagutt [0] (~datagutt@unaffiliated/datagutt)
19:15:33 Join mlt- [0] (
19:17:57linuxstbIs anyone with an Android build environment able test ?
19:19:54 Join nieuwbie [0] (
19:20:28nieuwbieHello, I have a problem with rockbox toolchain.
19:21:38 Quit mlt- (Quit: CGI:IRC)
19:22:08 Quit shai (Read error: Connection reset by peer)
19:22:35nieuwbieI tried to compile hello world from freemyipod on emCore, but I've got a *PANIC*.
19:25:06AlexPlinuxstb: Just see if it builds?
19:25:07 Join MaynardWaters [0] (
19:26:11 Join {phoenix} [0] (
19:26:24linuxstbAlexP: What would be useful is to create two new build directories - run the current configure (but nothing else) in one, and then apply my patch and run the patched configure in the other build directory. Then do a diff on the two directories and see if the generated autoconf.h and Makefiles are the same.
19:26:41AlexPlinuxstb: OK, no problem
19:27:16nieuwbieI spoke about it with TheSeven, and he analysed the binary and suggested that relocs from .text to .data/.bss are apparently relative to the start of .data/.bss instead of .text so it seems like linking step problem or maybe some inportablility in the linker script.
19:27:42TheSevener, does this really belong here?
19:27:49TheSeventhe very same toolchain links it correctly for me
19:28:29TheSevenand it isn't really "the rockbox toolchain" but rather the gnu toolchain built by with some rockbox-specific patches applied
19:28:54nieuwbieWell my mistake then.
19:28:57AlexPIs this relevent to Rockbox?
19:29:42 Join BHSPitMonkey [0] (~stephen@unaffiliated/bhspitmonkey)
19:32:21 Join kugel [0] (~kugel@rockbox/developer/kugel)
19:33:43nieuwbieTheSeven: So the problem should be relative to Machine or compilation, right?
19:34:33kugelI would like to commit FS #11216 before the freeze. is anyone opposed?
19:35:28 Join DerPapst [0] (
19:36:28gevaertskugel: I can't comment on the patch itself, but I think that since we're not frozen yet, that should be OK
19:36:38AlexPkugel: No comment on the technical side obviously, and I think embedded album art is a bit silly, but sure, why not - it'll help people out
19:37:09 Quit powell14ski_ (Ping timeout: 272 seconds)
19:37:21kugelI also find it silly, but a fellow student bugged me that rb doesn't show AA on his desire ;)
19:38:09*TheSeven wonders where that code grabs its memory from
19:38:31TheSevendoes this mean a 150KB ramsize hit, or is it doing something more clever?
19:38:39kugelit reads the mp3 as jpeg, with an appropriate offset and max size; since for id3v2 the jpeg file is embedded as blob
19:38:57TheSevenso directly from the MP3 file to the framebuffer?
19:39:00 Join powell14ski_ [0] (
19:39:13kugelto the audiobuffer, just like normal albumart
19:39:30TheSevenok, so no memory overhead (except binsize of course) involved?
19:39:58 Join TheLemonMan [0] (
19:39:59gevaertskugel: hm, wasn't there this issue with escaping some byte sequences?
19:40:15kugelTheSeven: yes
19:40:51kugelgevaerts: that's for desync'd pictures, which are simply unsupported yet
19:41:03kugelI haven't seen one in reality so far
19:41:18 Join Horscht [0] (
19:41:18 Quit Horscht (Changing host)
19:41:18 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:41:23gevaertskugel: what's the impact if there is one?
19:41:38kugela desync one?
19:42:00 Part nieuwbie ("ERC Version 5.3 (IRC client for Emacs)")
19:42:41kugelthe metadata rejects to parse this then (unsync'ing is supported in our parser, but more tricky for large data such as album art)
19:43:46gevaertsIf it doesn't crash and audio and other tags are still handled correctly (or as correctly as before), I think that's fine
19:44:41linuxstbAlexP: Thanks. I forgot APP_TYPE...
19:45:05pixelmawasn't there some buffering guy who had objections about the patch in general (at some point), I didn't follow though
19:45:13pixelmakugel: ^
19:45:48kugelthat was when the patch still needed some spare buffers which is not the case anymore
19:46:32kugelI and several people had objections that the initial patch used the codec buffer for storing the decoded image
19:46:51linuxstbkugel: Your patch is just for mp3/id3v2?
19:47:35pamaurydo I mean to do something special after writing an entry of the TTB ?
19:48:09linuxstbkugel: Any idea about other formats of album art in mp3s, e.g. could a PNG be there, and would your code try to parse it as a jpeg?
19:48:37kugelit checks the mime type so png is rejected
19:49:42kugeland IIRC the jpeg decoder checks for the presence of jpeg magic as well
19:51:07 Quit GeekShadow (Ping timeout: 276 seconds)
19:52:13linuxstbkugel: Well, it obviously needs extending to other file formats, so IMO the question is whether we want to put an incomplete feature in 3.8, or commit it after 3.8 and hope other formats are implemented by 3.9...
19:53:56gevaertslinuxstb: isn't that similar to not releasing 3.0 because we didn't have wma pro yet?
19:55:14 Quit qurvel (Ping timeout: 246 seconds)
19:58:40 Join saratoga [0] (9803c6dd@gateway/web/freenode/ip.
19:58:44kugellinuxstb: I wouldn't say it's incomplete
19:59:08Lloreangevaerts: It'd be like not releasing 3.0 because the "artist" tag only works for Mp3 and not any other metadata...
19:59:40linuxstbgevaerts: I'm not sure about that example, but I'm sure we've released incomplete features before... Another concern could be that this is dealing with core playback code, so bugs would be serious.
19:59:49LloreanI don't think it's particularly bad to commit it right after the release branch.
19:59:54saratogawhere is the jpeg decoder currently located? in the main binary or plugin?
19:59:57kugelid3v2 does advice against desync metadata, it's ok to not support it IMO
20:00:10LloreanJust in case there are unforseen bugs in it, too, as well as it being one of those features where it may not be easily clear to users when it will or won't work and why it's not working.
20:00:22LloreanIt has a high "is this a bug?" "no it just doesn't do that yet" probability
20:00:26kugelfor other formats it's not as simple to implement (and I have little interest to do it currently) as they not always binary blobs
20:01:24kugelwe also don't support other types of jpeg in our decode, that hasn't been a problem so far
20:02:08linuxstbkugel: I'm just playing devils advocate a little to see what others think - I'm not strongly against committing it, as it's a long-requested feature and mp3 is probably still the most-used format... I'm also not expecting you to implement other formats - commit mp3 and others will no doubt add others over time.
20:02:16kugelthis patch catches the vast majority of cases as I see it
20:03:09LloreanIsn't album-art-in-metadata incredibly common in AAC as well?
20:03:13LloreanWell, MP4 in general
20:03:25gevaertsLlorean: right, that's a bit closer
20:03:32kugelLlorean: no idea
20:03:54LloreanAlso, to a lesser extent, FLAC I think, where a lot of whole-album CUE+FLAC seems to be not too uncommon.
20:03:58linuxstbThe one track I bought from itunes had embedded AA - I've no idea if they all do...
20:04:11LloreanBut really MP4 is I think the one that needs to be addressed if the goal is "vast" majority
20:04:44LloreanI'd say MP3 is probably the majority, but not by as wide a margin as it could be now that iTunes is DRM-free in many cases.
20:04:47kugelthat's not the goal
20:04:59saratogaconsidering our MP4 parser is barely functional at this point, it may be premature to worry about album art in it
20:05:04Lloreankugel: You were claiming you felt it catches the vast majority of cases, I'm just saying I'm not certain that's true.
20:05:55LloreanI don't think I've ever acquired an MP3 with embedded album art, but I know that many MP4 files I've come across have had it.
20:06:00kugelI think AAC more common among (new) ipod users, but not so much for all other people
20:06:10linuxstbFLAC at least seems to be just a binary blob -
20:06:44LloreanI'm just saying I'd like to suggest it wait for after 3.8, both so that it has a longer bug shakedown period, and in the hope that *someone* implements it for more of the major formats.
20:07:05LloreanI don't really see that as a majorly negative thing, to wait a short time.
20:07:56linuxstbChecking my mp4 file - a hexdump shows that the jpeg data appears to just be a binary blob.
20:08:04LloreanIf 3.9 comes along and the situation has improved, at least there's been time for a manual entry to have been written, us see if it caused confusion about what's supported, clarifications to be made, etc.
20:08:12LloreanEr *hasn't* improved
20:08:38kugelI don't think it's very risky as the unsupported cases are pretty much catched. and there's still more than two weeks left
20:08:41*Llorean is okay with only jpeg being supported by the way, PNG or BMP in metadata seems somewhat insane.
20:08:59saratogadoes png album art work as separate files right now?
20:09:21LloreanI just saw PNG asked about above, so included it there, is all
20:09:44 Quit liar (Ping timeout: 255 seconds)
20:10:41LloreanI just don't think we'd say "only MP3 gets album art from files" is good enough for a release, so I'm not sure why it's a good idea to do it for embedded when we can wait a couple weeks, then hopefully some interest might grow before the next release and we can say "many common embedded album arts will be displayed"
20:10:58LloreanIs there some urgency to this going in I'm missing?
20:11:30gevaertsLlorean: why wouldn't we accept only mp3 files for a release?
20:12:13kugelI would like to see it in 3.8, it hopefully increases the chance for adoption to other formats sooner
20:12:16Lloreangevaerts: Is there a good reason it shouldn't work for the other formats, if the data is already available to us in a format we can recognize?
20:12:34Lloreankugel: How does having it in 3.8 make it more likely to be improved than having it in the daily build?
20:13:05kugelmany people are using releases and not current builds
20:13:08gevaertsLlorean: no, but is there a good reason to deny it to mp3 users because there are other formats?
20:13:15linuxstbEven though I first mentioned it, I don't really see a problem with "Support for embedded album art - currently only MP3 files"
20:13:23Lloreankugel: That doesn't explain how it increases the likelihood of new development.
20:13:35Lloreangevaerts: We're not. It'll be in the daily build at the time of 3.8 release.
20:13:53gevaertsLlorean: we're trying to get regular users to use stable releases
20:13:56Lloreangevaerts: I'm not suggesting any delay to it beyond the release, and it'll be in 3.9 even if it doesn't improve.
20:14:06 Join petur [0] (~petur@rockbox/developer/petur)
20:14:11kugelanyway, as nobody is strongly opposed I'll put it in
20:14:22*Llorean likes being "nobody"
20:14:25gevaertsI think that the potential of (serious) bugs is the only possible issue, and I think it's up to the indiviual developer to estimate that
20:14:33kugelLlorean: are you strongly opposed now?
20:14:38Lloreankugel: I'm opposed.
20:14:40LloreanI stated that.
20:14:42kugelyou weren't some minutes ago
20:14:42saratogai don't really see the need to encourage regular users to use any particular build
20:14:56Lloreankugel: I've been saying I don't think it should go in until after 3.8 is branched this whole time.
20:15:24LloreanI just think there's no harm in delaying it until 3.8
20:15:26pixelmakugel: no, I meant your patch and a discussion from last week or so
20:15:27 Join markun_ [0] (~markun@rockbox/developer/markun)
20:15:50LloreanYou can go ahead and commit it, I just don't appreciate the idea that nobody opposed it.
20:16:03gevaertsLlorean: he said "strongly opposed"
20:16:08gevaertsThat's not the same as "opposed"
20:16:20Lloreangevaerts: "strongly" is so thoroughly subjective that it's not reasonably useful here.
20:16:44Lloreangevaerts: "strongly enough to object to it being committed", which I have.
20:16:51kugelpixelma: then that was probably Unhelpful about desync files
20:17:11kugelthe patch as of now doesn't use additional buffers and leverages the existing album art system
20:17:24 Join liar [0] (
20:17:25Lloreangevaerts: The alternative would be "weakly enough to say 'I don't like it, but I don't stand in the way of delaying it'"
20:17:33 Quit markun (Ping timeout: 240 seconds)
20:17:40Lloreankugel: How hard would it be to adapt it for use in other formats where the jpeg data is a binary blob?
20:17:52 Join sideral [0] (~sideral@
20:17:52 Quit sideral (Changing host)
20:17:52 Join sideral [0] (~sideral@rockbox/developer/sideral)
20:17:57kugelI don't know, I haven't looked into other formats
20:18:20kugelexcept ogg vorbis where I found *very little* information about it
20:18:29LloreanWhat I mean to ask is, is what you've put in place relatively flexible/expandable, or deeply ingrained into ID3 in some manner?
20:19:13kugelit's expandable, but the parser needs to implement it
20:21:55 Join bertrik [0] (
20:21:55 Quit bertrik (Changing host)
20:21:55 Join bertrik [0] (~bertrik@rockbox/developer/bertrik)
20:22:44AlexPI don't see the issue with just saying currently only MP3 files as linuxstb said
20:22:49LloreanI guess there's no strong reason not to include it now. I'm against it because I feel there's no harm in waiting, and I'd really like us to be able to advance it in the changelog for a release as "most common embedded album art" rather than just "ID3 embedded album art"
20:23:02Lloreanannounce it, rather
20:23:30LloreanI'd dislike just as much if we only supported the "title" tag in MP3 but not Ogg, MP4 or others.
20:23:35 Quit casainho (Remote host closed the connection)
20:24:32 Join mlt- [0] (
20:24:50LloreanBut I don't understand the metadata system, so I don't understand why we can decode the data from one format, but others might have it the same way and we can't extract it for decoding?
20:24:58*linuxstb spots that $cmdline is broken with his RaaA configure patch, as they all have the model name "application"...
20:25:09kugelthe patch is already 9 month old
20:25:35mlt-I'm still trying to build simulator under MS Windows. Can someone take a look what "tools/genlang" may not like among options?
20:25:47*gevaerts doesn't really like the idea of always waiting until a feature is as perfect as it's likely to ever get before committing
20:26:11kugelLlorean: I also don't know the metadata system very well, the id3v2 parser isn't my work actually. I only fixed the buffering side
20:26:23linuxstbLlorean: It's not a matter of "can't", it's just that it hasn't been implemented in all metadata parsers.
20:26:25pixelmamlt "under Windows" how?
20:26:29Lloreangevaerts: As I said, I only suggest waiting until after the 3.8 branch, so it gets better shakedown before being in a release.
20:26:42Lloreangevaerts: I even said explicitly, if it still hasn't been expanded by 3.9, it goes in anyway
20:27:02Lloreangevaerts: Just a *small* delay in the hopes that having it in SVN will prompt the "rounding out" of the feature before it's in a release.
20:27:04mlt-pixelma: I mean if someone is familiar with that genlang tool
20:27:10kugelmlt-: ah, that's where I stopped my last adventure about compiling under windows
20:27:13Lloreangevaerts: Since many things get less attention as patches.
20:27:29LloreanSo I'm not proposing any prevention of the commit, at all
20:27:33LloreanJust timing of it.
20:27:45pixelmamlt-: and I meant to ask which build environment you use ;)
20:28:08mlt-kugel: last time we figured out that command line is too long for windows. So I used @response file for gcc to shorten it. But I'm stuck with this genlang now :(
20:28:12gevaertsLlorean: waiting a bit is indeed an option, but my (rather strong) opinion is that unless we're in a freeze (and we're not) it's up to the committer to decide if the feature is likely to be stable in time for release
20:28:16kugelI question the existence of the freeze (or it's length) if new features can't go in a couple of days before it
20:29:03AlexPI agree - otherwise we might as well start the freeze now
20:29:07mlt-pixelma: I'm using MinGW gcc 4.50 with make 3.81. I'm running Windows XP SP3 32 bit
20:30:16AlexPHowever, should anyone try and commit a new feature after Sunday I'll hit then with a sharpened stick :)
20:30:23Lloreangevaerts: It is up to the committer, and I don't dispute that. I'm just offering my opinion that in the case of a feature like this, were I the committer I'd certainly wait since I think it'll provide a better all around user experience that way.
20:31:38LloreanAlexP, kugel, gevaerts: I've never liked the idea of "squeeze a feature in shortly before the freeze." I think the feature freeze is a "no features period" but the more likely the feature is to have bugs, the further before the freeze someone should be thinking "do I really want to commit this now with a release coming up?" since we've seen releases delayed for a *long* time by new features too late before them (longer than our freeze period).
20:31:47LloreanBut in the case of this one, I'm not saying it's buggy or even that I think there will be bugs.
20:31:53LloreanI just think it will be better over all if it waits.
20:32:05kugelwaiting even more is always an option
20:32:37AlexPI think it makes no difference when between freezes something gets added as long as the committer thinks it'll be OK by the next release. It is the same if it is three months before or three days before
20:32:56*mlt- was surprised that genlang turned to be a perl script...
20:33:11kugelbut commit early, commit often also and I prefer that (that shouldn't imply the embedded AA patch needs any post-commits, not that I'm aware of anyway)
20:33:35AlexPBut of course, whould this turn out to be broken I'll beat kugel with a shoe :)
20:33:40LloreanCommit early is good. It's sad that it's been in the tracker for 9 months, and this conversation would've been completely different from my side if this was taking place even 2 or 3 weeks ago.
20:33:57gevaertskugel: speaking of AA... Do you still have plans for the *other* AA? :)
20:34:04LloreanI just feel it's very close to the release now, and there's not significant harm in waiting, and multiple advantages (from my perspective) in doing so.
20:34:21kugelgevaerts: I want to have a look in my semester break
20:34:23LloreanIf the branch is Sunday, that's waiting 4 days for the commit
20:34:32AlexPfreeze is sunday
20:34:38AlexPbranch a week after that
20:34:47LloreanAh, well slightly longer then.
20:34:53AlexPbut I think it'd be nice for 3.8
20:34:59AlexPanyway, it is up to kugel
20:35:15LloreanIt is up to Kugel. I just really would like to see it for at least MP4 when we announce it.
20:35:51LloreanI wouldn't be entirely certain that more of our users don't have that than MP3 embedded considering how many iPod users we seem to have.
20:36:16 Quit liar (Quit: Leaving)
20:38:22MaynardWatersDoes rockbox, allow you to do some type of 5 star rating which can be written to the id3 tags? during playback?
20:39:31AlexPThere is rating, but it gets stored in the Rockbox database
20:39:57 Join leavittx [0] (~lev@
20:40:49MaynardWatersI am no expert on id3 tags, but v1, v2 and v2.3.... at least one of those has data for some type of 5 star rating, I believe, ...
20:40:59LloreanWe don't write to your music files at all
20:40:59bertrikI'd rather not have rockbox change the audio files it plays back
20:41:45LloreanSomeone could write a plugin that would read ratings in the database and write them out to files or something, but it may be best left to that sort of manually triggered operation.
20:41:45MaynardWatersI can understand that, avoids having to deal with the stock firmware issues, I am guessing
20:44:03 Join thomasjfox [0] (
20:49:07 Join p3tur [0] (~petur@rockbox/developer/petur)
20:50:48*mlt- found out that perl from MinGW badly treats -s option that causes troubles for tool/genlang
20:52:40 Join liar [0] (
20:52:44 Quit Keripo (Quit: Leaving.)
20:53:29thomasjfoxlinuxstb: ping
20:54:52linuxstbthomasjfox: Hi
20:55:02thomasjfoxJust gave your patch a test run
20:55:17thomasjfoxOh well, it doesn't compile anymore
20:55:33thomasjfox"No rule to make target `/home/maemo/MyProjects/rockbox/BUILD/sim_tasks.h', needed by `/home/maemo/MyProjects/rockbox/BUILD/apps/main.o'"
20:56:11thomasjfoxalso maemo4cc needs SDL, glib and libosso, too.
20:56:19linuxstbCan you compare the autoconf.h and Makefile created by my patch with the current version of configure?
20:56:34thomasjfoxmake reconf dies with: "Please select a supported target platform!"
20:56:38linuxstbI've made a few more changes this evening and am about to upload a new version of the patch.
20:56:46linuxstbYes - make reconf is something I've just fixed (I hope)
20:57:11thomasjfoxI also noticed maemo4cc doesn't have -O2
20:57:54linuxstbBut its a bit of a hack at the moment because RaaA is using "application" for the modelname for all RaaA targets.
20:58:52thomasjfoxregarding the autoconf.h compare, I think it's better to wait for the updated version of your patch
20:59:33linuxstbOK, I've just fixed maemo4cc I think.
20:59:36kugellinuxstb: IIRC modelname is only interesting for the target tree
20:59:59linuxstbkugel: grep tells me it's also used in
21:00:19kugelfor creating rockbox-info.txt?
21:00:55kugelthomasjfox: I'm going to commit the codec buffer patch
21:01:12kugelucontext needs to wait a bit longer I'm afraid
21:01:12thomasjfoxkugel: already christmas?
21:01:42thomasjfoxthat's fine
21:01:59thomasjfoxI'll try to look into the shutdown() sequenze right now
21:02:21 Join Keripo [0] (
21:03:19linuxstbthomasjfox: New version of patch added.
21:05:17thomasjfoxlinuxstb: Just an idea: Can't we have a unified maemocc like before and pass the desired version number as $1?
21:06:05linuxstbthomasjfox: I was just thinking about that. Or maybe a "maemocc_common" function?
21:06:19***Saving seen data "./dancer.seen"
21:06:32thomasjfoxmaemocc would do ;)
21:06:50thomasjfoxIf the function ever would get too big, we could still split it later on
21:07:48 Quit datagutt (Quit: kthxbai)
21:10:33thomasjfoxlinuxstb: still fails at sim_tasks.h. I'll try to compare autoconf.h and Makefile
21:12:05*TheSeven starts to understand the s5l8702 clocking
21:12:20TheSevenand there's bad news :/
21:13:07TheSevenseems like the AHB:APB ratio must be at least 2:1
21:13:10TheSeveneven at low base clocks
21:13:22CIA-70New commit by kugel (r29259): Embedded album art support in MP3/ID3v2 tags. ...
21:13:39TheSeventhat's really nasty as it means that we can't go below 108MHz without affecting timers etc.
21:13:55TheSeven(APB running at 54MHz)
21:14:15 Quit {phoenix} (Remote host closed the connection)
21:14:16thomasjfoxlinuxstb: autoconf.h is identical, Makefile diff is here:
21:14:36 Quit mystica555_ (Read error: Connection reset by peer)
21:14:46n1sTheSeven: how much difference does lower freq make? a lot of new chips seem to not be much affected by the freq but rather how much they sleep
21:15:08thomasjfoxlinuxstb: Wild guess: It's the missing APP_TYPE=sdl-app
21:15:15TheSevenwhat does seem to matter is the AHB frequency
21:15:40linuxstbthomasjfox: Hmm, but it's not _the_ sdl-app...
21:16:04kugelannoying "suspicious change request" warning on the wiki...
21:16:05thomasjfoxlinuxstb: Still uses various parts from it (kernel, button stuff etc.)
21:16:24TheSevenn1s: and i still have the suspicion that rockbox isn't sleeping very well
21:16:40linuxstbthomasjfox: Yes, I'll add that. I'll also revert back to maemocc and add the version as a parameter.
21:16:40TheSevenit reproducibly sucks more current when it's idle than when it's playing music, on multiple targets
21:17:00thomasjfoxlinuxstb: And install scratchbox ;)
21:17:10kugelthomasjfox: app_type shouldn't be sdl-app for maemo I think
21:17:28kugelnasty thing is that you don't have logical or in Makefiles
21:18:13n1sTheSeven: ah, didn't know that but i meant more when the core sleeps for a tcik because it has nothing to do during playback doesn't cost much
21:18:19CIA-70r29259 build result: 329 errors, 20 warnings (kugel committed)
21:18:28linuxstbthomasjfox: Yes, I should do that... ;) Plus the Android stuff.
21:18:56thomasjfoxkugel: Nice commit :o)
21:19:39 Quit leavittx (Ping timeout: 260 seconds)
21:20:24TheSevenwhich one was it this time? charcell? greyscale? hwcodec? :P
21:20:29gevaertsyes :)
21:20:31AlexPkugel: How does pictureflow deal with embedded? Still just ignores it I guess?
21:21:00kugelgood question
21:21:05kugelI assume it does
21:22:56CIA-70New commit by kugel (r29260): Fix red. Not all targets have album art support.
21:23:11TheSevenkugel: so basically charcell? :P
21:23:20gevaertsTheSeven: and monochrome
21:24:36thomasjfoxkugel: Congrats on embedded album art support. I was really missing it!
21:25:05kugelah good, so we have at least one developer using it ;)
21:26:47pixelmathe commit message is missing "in core"
21:27:01kugeloh sorry
21:27:28CIA-70New commit by kugel (r29261): Disable buffering codecs (and code generally) on RaaA. ...
21:27:40CIA-70r29260 build result: 54 errors, 20 warnings (kugel committed)
21:32:38CIA-70r29261 build result: 54 errors, 20 warnings (kugel committed)
21:32:54TheSevenwell, at least no *new* errors :P
21:34:14 Quit milk (Ping timeout: 260 seconds)
21:35:41 Quit factor (Ping timeout: 255 seconds)
21:37:51 Quit enthdegree (Quit: leaving)
21:37:51 Join milk [0] (
21:38:37*mlt- has successfully built working rockboxui.exe under MS Windows
21:38:41CIA-70New commit by kugel (r29262): Fix remaining reds.
21:38:52kugelmlt-: great!
21:39:20mlt-can someone add me to wiki users group so I can add few words on my experience as out-of-the-box build is ultimately broken?
21:40:12kugelmlt-: what's your username?
21:40:39mlt-I tried to register with wiki name MikhailTitov
21:41:02mlt-I'm not sure where to enter it though)
21:41:58thomasjfoxmlt-: You'll need it once you click on "Edit"
21:42:49kugelmlt-: done, promise not spam :)
21:43:34CIA-70r29262 build result: 56 errors, 0 warnings (kugel committed)
21:46:50mlt-Thanks! I don't have time right now, but I'll explain probably later today what needs to be changed for UI simulator under MS Windows.
21:47:23 Quit mlt- (Quit: CGI:IRC)
21:48:48TheSevenkugel: are you color blind? you fixed yellows, not reds :P
21:49:26kugelwrong. I fixed all reds and introduced other ones ;)
21:49:49kugelthe numbers are misleading :)
21:51:15CIA-70New commit by kugel (r29263): Hopefully all green now
21:55:21CIA-70r29263 build result: All green
21:55:43 Join shai [0] (
21:56:13dfktkugel: \m/
21:57:37 Quit benedikt93 (Quit: Bye ;))
21:57:41kugelthomasjfox: did I forget anything to commit (except ucontext)?
22:03:36TheSevenstupid hold switch
22:04:00*TheSeven has never seen a player before whose hold switch has been that hard to read out
22:04:12TheSevenit's still glitching even with a majority detect across 5 samples!
22:04:56thomasjfoxkugel: All in :)
22:05:48*TheSeven wonders why plugins are data aborting
22:05:55TheSevendid someone forget to bump the api?
22:12:02n1sTheSeven: a full remake would fix it if that's the problem
22:13:07TheSevenyeah, i'm running with an old set of plugins
22:13:13TheSevenbut that still shouldn't happen
22:13:49n1syes, true, i just meant that if that fixed it it would indicate that api breakage was indeed the problem
22:15:08n1ssometimes things that don't directly touch the plugin api structs still break the binary compatibility though
22:15:33n1slike changing a struct somewhere
22:16:52 Join factor [0] (~factor@
22:18:02 Quit efyx (Quit: Quitte)
22:19:12 Join wodz [0] (
22:20:20wodzLeaked SDK for rockchip rk27xx is the most buggy piece of code I have ever seen!
22:20:36kugelpixelma, AlexP: can you check this patch for the manual?
22:23:30AlexPkugel: It was already there, but line 13 we should use \disk{} instead of disk I think
22:24:00AlexPAnd line 14 picture not pictures
22:24:21AlexPActually, again already there, but I think image(s) is better than picture(s) here
22:24:43AlexPout of interest, what does unsynchronized mean in this context?
22:25:08AlexPbut yeah, looks fine :)
22:25:32kugelyou can fix the bits you mentioned and commit if you like
22:25:43AlexPOK, will do
22:25:50 Quit factor (Ping timeout: 276 seconds)
22:26:36kugelunsynchronized means that it has addtional zero bits to split up 11111111111 sequences which is some special mpeg marker
22:27:05AlexPAnd did you try pictureflow? We should mention here if embedded doesn't work in pictureflow
22:27:29kugelI'll have a look at pictureflow
22:28:12 Quit Keripo (Quit: Leaving.)
22:30:39pixelmaI find the embedded jpeg in \fname a bit odd, but I guess it only looks weird in code, not in the compiled manual
22:31:03 Join factor [0] (~factor@
22:32:40bluebrotherhmm, we have embedded album art support now?
22:33:21bluebrotherand why put the "embedded jpeg" in \fname? I don't see a reason for not putting it as plain text
22:33:37AlexPpixelma: I'm just looking, but I'm guessing it'll fit better
22:34:12AlexPyeah, no need
22:35:11 Quit n1s (Quit: Ex-Chat)
22:35:21linuxstbShouldn't it be JPEG?
22:37:46TheSevenn1s: updating fixed the plugins problem, so someone indeed didn't bump the API
22:37:50 Join stripwax [0] (
22:38:11TheSeven=> we should maybe make sure bumped it will get bumped at least once before the release
22:39:36pixelmaAlexP, linuxstb, kugel: isn't this the list of which type of files/pictures are supported and which take priority?
22:39:44CIA-70New commit by theseven (r29264): iPod Classic: This time really fix the hold switch. Read it out through the power manager, and cache the result for 100 milliseconds because the power ...
22:39:52AlexPpixelma: yes
22:40:08AlexPwell, it was filename and then priority
22:40:14pixelmaand does an embedded file really take priority over external images?
22:40:17AlexPSo we don't actually need the JPEG I think
22:40:19AlexPpixelma: yes
22:40:37AlexP(see kugel's commit message) :)
22:41:15AlexPI think this makes sense - this way you can have e.g. an album folder.jpg but override t for a specific track
22:41:44linuxstbDoesn't using "trackname.jpg" do that?
22:42:13pixelmathat should also work
22:42:20TheSevenhm, one can once again hard lock rockbox by plugging/unplugging usb too fast
22:42:24AlexPbut I mean for people who use embedded albumart
22:42:24kugelAlexP: pictureflow is tricky, as it prefers album art over track art
22:42:55AlexPkugel: Does it do anything with embedded at the moment?
22:43:32AlexPI guess for that look for an album art on disk then if you don't find it use the first track art you find for an album
22:43:38TheSevenhm, rockbox playing at 55mA...
22:44:19CIA-70r29264 build result: All green
22:44:22kugelAlexP: that's how it works
22:44:39kugelso for pictureflow I need to look for embedded at last
22:45:32AlexPkugel: But are you going to now, or shall I just put a note in saying pictureflow doesn't support embedded yet?
22:45:43kugelgive me some minutes
22:45:45TheSevenso about 10mA improvement...
22:45:56AlexPkugel: OK :)
22:46:04CIA-70New commit by theseven (r29265): iPod Classic: Enable boosting by switching the CPU between 1x and 2x AHB clock
22:47:58TheSeventhere is definitely something wrong
22:48:34TheSeventhat kind of boosting barely saves anything within rockbox, but 6mA within emcore
22:49:20linuxstbthomasjfox: New (and hopefully final) version of patch uploaded...
22:50:26CIA-70r29265 build result: All green
22:50:40thomasjfoxlinuxstb: ok, checking
22:50:51TheSevenplaying/not playing doesn't make a huge difference, while in emcore i can even see if the clickwheel is being touched on the multimeter
22:51:24TheSevenalso, after starting and stopping playback, rockbox consumes more power than before
22:52:34thomasjfoxlinuxstb: I think the "-atleast-version=4" won't work with the version == 5 check, no?
22:54:45linuxstbthomasjfox: I'm not sure what you mean. The intention is that if $1==5 and version 4 is detected, then it's an error.
22:55:44thomasjfoxlinuxstb: sorry, my fault: it's inside if / elseif. This will work
22:56:07thomasjfoxlinuxstb: I thought it just for version >=4 and then complains if version == 5
22:57:11thomasjfoxcompile output looks correct, let's diff the Makefile again
22:57:40TheSevenplaying flac or ape -c4000 doesn't make a difference power consumption-wise, that just *can't be right*!
22:59:54 Quit TheLemonMan (Quit: free(me))
22:59:57TheSevenis there a way to see how much time rockbox is sleeping?
23:00:18TheSeven(this should maybe be added to the buffering debug screen)
23:00:27 Join JdGordon| [0] (
23:00:27 Quit JdGordon| (Changing host)
23:00:27 Join JdGordon| [0] (~jonno@rockbox/developer/JdGordon)
23:01:23thomasjfoxlinuxstb: Looking good. I think I found an unrelated issue (maybe with maemocc): If you run "make reconf", the LD_OPTS append themselves every time you run it
23:01:32thomasjfoxlinuxstb: Gonna check maemo 4 now
23:02:13 Join Keripo [0] (
23:03:10thomasjfoxlinuxstb: "make reconf" successfully complains "ERROR: Maemo 5 SDK required." on maemo 4
23:05:25thomasjfoxlinuxstb: argh, maemo4 doesn't have pkgconfig for SDL. I'm wondering how this worked before...
23:06:14gevaertswodz: I got the package on monday. Thanks!
23:06:22***Saving seen data "./dancer.seen"
23:07:54linuxstbthomasjfox: Looking at simcc(), it just sets LDOPTS to "-lm" - I'm guessing maemocc should do the same (i.e. not include $LDOPTS)
23:08:11linuxstbthomasjfox: But I'll leave you to fix that after I commit this patch.
23:08:17thomasjfoxlinuxstb: ok
23:08:43thomasjfoxlinuxstb: I think we need to hack sdl-config support back in for maemo 4
23:08:58kugelAlexP: add a note, it's more tricky than I thought
23:09:06thomasjfoxafter you commit the patch. Guess this broke when I switched from simcc() to maemocc()
23:09:15AlexPkugel: OK, will do
23:09:39linuxstbthomasjfox: OK, so you're happy for me to commit v3?
23:10:44thomasjfoxlinuxstb: Yes. Except for the part that it breaks the "debian/rules" stuff. Could you prepare a fix for that?
23:10:52 Quit p3tur (Quit: here today, gone tomorrow)
23:11:43thomasjfoxlinuxstb: We should be able to get the maemo version in there too as the maemo-version-dev package is a build requirement
23:12:29kugelAlexP: it'd mean making the db and metadata subsystems album art aware (they are not now) for !HAVE_ALBUMART, just for pictureflow
23:12:39CIA-70New commit by alex (r29266): Manual: We now support embedded JPEG album art in ID3v2 tags.
23:12:46thomasjfoxkugel: Could you take a look at FS #11924 v3 if it's ok for android? (the paths etc.)
23:13:23AlexPkugel: I use neither embedded AA nor pictureflow so fine by me :)
23:15:08 Quit liar (Ping timeout: 255 seconds)
23:15:10linuxstbthomasjfox: Hmm, I'm not sure what to do in debian/rules. We now have two different applications being built (−−target=nokian8xx and −−target=nokian900) - how should it know what the user is building? Checking for the maemo SDK version seems a bit of a hack...
23:15:15wodzgevaerts: good - does it work?
23:15:22gevaertswodz: I haven't tried yet
23:15:56wodzgevaerts: First simple test - connect it to usb and look if it gets detected
23:16:34thomasjfoxlinuxstb: As scratchbox sets up symlinks for everything (libraries, gcc) depending on the active target, detecting the maemo-version-dev version is fine
23:16:48CIA-70r29266 build result: All green
23:17:01kugellinuxstb: looks good for android (Makefile: target id is different, and platform=A is missing from the reconf line;autoconf.h: identical)
23:17:59linuxstbkugel: OK, thanks for looking.
23:19:04linuxstbthomasjfox: I am thinking more generally - how would we, for example, extend it to handle a Debian SDL app?
23:19:46thomasjfoxlinuxstb: Then we need distro specific code in the rules file anyway. The maemo distro version would still do auto detection.
23:19:57 Quit factor (Read error: Connection reset by peer)
23:20:10gevaertslinuxstb: debian is actually easy. We could use the v3 source format, which implies being able to provide a new debian/ directory in a separate tar file
23:21:08linuxstbthomasjfox: Would you be able/willing to fix it?
23:21:11gevaertsIt's if there are more systems that use older dpkg versions that we have problems
23:22:36thomasjfoxlinuxstb: This is way beyond my debian packaging skills...
23:23:09linuxstbthomasjfox: And mine... I can't even test things at the moment.
23:23:24linuxstbBut debian/rules is just a Makefile IIUC...
23:23:44thomasjfoxThe rules file: ye
23:24:30thomasjfoxLook at the "control" file. It will be a bit tricky to get the dependencies right for multi distro
23:26:18 Join liar [0] (
23:26:39linuxstbAren't the dependencies already wrong? i.e. before my changes to configure?
23:27:22thomasjfoxlinuxstb: They are correct for maemo
23:27:30 Quit wodz (Quit: Leaving)
23:28:22linuxstbBut doesn't it say you need "maemo <= 5" for both Meamo4 and Maemo5 builds?
23:30:18 Quit petur (Quit: Leaving)
23:32:22thomasjfoxlinuxstb: There was something about this. Can't remember the details right now. I could try with "maemo < 4" some other time
23:32:54linuxstbthomasjfox: But I mean that the dependencies depend on what version of Maemo you want to build for... i.e. they are not fixed, even for maemo.
23:33:52thomasjfoxlinuxstb: Yes, that is true. I just said it will get more tricky to add normal debian systems or other debian based distros into that equation
23:34:55kugelthomasjfox: do you still plan to work on pulseaudio backend?
23:35:34thomasjfoxkugel: Yes, though with low priority. I want to get the usability right first
23:35:56thomasjfoxkugel: proof of concept change to fix the shutdown stuff for SDL/maemo:
23:36:26 Join factor [0] (~factor@
23:37:57thomasjfoxkugel: Is there a master event loop on android?
23:38:25thomasjfoxkugel: The event loop needs to keep running until shutdown_hw() is finally called
23:38:45thomasjfoxkugel: Would that be easily doable?
23:39:12kugelnot sure about "easily" but yea
23:40:50 Quit komputes (Remote host closed the connection)
23:41:15TheSeveni now have proof that rockbox never sleeps!
23:42:29kugelthomasjfox: any idea what's the difference between shutdown_hw() and sys_poweroff()?
23:42:38thomasjfoxkugel: Yes :)
23:43:03thomasjfoxkugel: shutdown_hw() does the actual shutdown. On native platform it calls poweroff() at the end
23:43:07*TheSeven wants to see that fixed before 3.8!
23:43:26kugelTheSeven: can you elaborate?
23:43:46thomasjfoxkugel: sys_powerroff() sets a timeout and broadcasts the SYS_POWEROFF event
23:44:06kugelTheSeven: are you saying core_sleep() (in thread-arm.c) doesn't work?
23:44:15TheSeveni'm saying that it doesn't seem to be called
23:44:29TheSeveni just added some statistics, and they show all-zero
23:44:31kugelit's called by switch_thread()
23:44:42TheSevenalso the power consumption measurements don't show any traces of sleeping
23:44:51kugelTheSeven: it has a bunch of #ifdef, is your target in that list?
23:44:55TheSevenyes, it is
23:45:07TheSevenand the #warning below doesn't come up (and did before i added it to the ifdef)
23:45:23sideralJdGordon, n1s: Could you comment on the lang-related question in ?
23:45:40kugelI'm pretty sure rockbox is sleeping, I noticed a *huge* difference in CPU usage on the android port
23:45:58kugel100% load w/o proper core_sleep(), 0% with core_sleep()
23:47:06kugelTheSeven: are you on armv6?
23:47:30TheSevenarmv4 and armv5e
23:47:57kugelstrange that core_sleep() isn't called. it definitely is on android
23:48:35TheSeveni just made it panic, and well, sometimes it is called
23:48:41TheSevenbut there is something wrong somewhere
23:48:48TheSevenit seems to be called 90 times per second
23:48:54TheSeven(when idle)
23:48:54 Quit factor (Read error: Connection reset by peer)
23:49:10kugelthat sounds about right
23:49:33TheSevenyeah, but the time inside is near-zero
23:49:55TheSeven(or there is a huge bug in my measuring...)
23:50:13kugelis some interrupt/tick task hogging the cpu?
23:51:06kugelit goes like timer_int->all tick tasks->wakeup scheduler->run all runnable threads->core_sleep()->wait for interrupt->timer_int->...
23:52:18 Quit Keripo (Quit: Leaving.)
23:53:40 Join webguest95 [0] (
23:53:41 Quit krazykit (Ping timeout: 240 seconds)
23:53:46 Join factor [0] (~factor@
23:54:30kugelon idle it should be almost entirely be in core_sleep(). if it isn't one of the other parts of this chain is hogging
23:54:42 Join mlt- [0] (
23:58:08 Quit JdGordon| (Ping timeout: 265 seconds)

Previous day | Next day