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).

Notice: Only Gecko based browsers prior to FF4 support the multipart/mixed "server push" method used by this log reader to auto-update. Since you do not appear to use such a browser, this page will simply show the current log, and not automatically update.

#rockbox log for 2009-09-10

00:00:07rasherwell, you can do them, but you'd have to insert a blank line after it to get the effect you want
00:00:07pixelmarasher: this is *your* WPS and I could use the same argument against yours
00:00:10*domonoky1 suggest just normal inline comments with # (linefeed not eaten if it isnt a full comment line) and some other token to continue the line, like "\"
00:00:31domonoky1this also prevent current wps from breaking.
00:00:33rasherpixelma: You're going to claim widecabbie.wps is "too simple"?
00:01:13 Join truthtaco_ [0] (n=truthtac@adsl-74-12-230.aby.bellsouth.net)
00:01:17JdGordon|domonoky1: that would break current wps.. but yes thats my suggestion also..
00:01:23JdGordon|i tinhk
00:01:33gevaertsexcept for the \ :)
00:01:38rasherThis wouldn't exactly be the first time we have broken wpses, even in non-obvious ways
00:01:41JdGordon|% makes more sense for us
00:01:47pixelmarasher: no, but it's only one example
00:02:00rasherpixelma: You only need one example to disprove a negative.
00:02:11domonoky1JdGordon|: why would it break ? Are inline comments expected to eat newlines at moment ?
00:02:42JdGordon|yes
00:02:43rasherpixelma: you claimed, basically, that inline comments are useless and can't be used in real wpses. I've shown you a real wps where they could easily be used. Case closed.
00:02:46*domonoky1 thinks % strange, it look like a token without identifier..
00:02:47 Quit p3tur ("Zzzz")
00:03:33gevaertsdomonoky1: that's exactly why it's the best character for this
00:04:03rasherIt's already part of the WPS "language"
00:04:08rasherMakes sense to me
00:04:13JdGordon|it *is* the wps language
00:04:20 Part zzy
00:04:23gevaertsAll other characters would need some way of escaping them
00:04:39rasherOooh, very good point
00:04:59 Quit archivator (Read error: 104 (Connection reset by peer))
00:05:21 Join archivator [0] (n=Delyan@77.70.28.57)
00:06:06pixelmarasher: I just couldn't understand why you want to use them as for me they aren't easier to read but I didn't say they were "useless"
00:06:32pixelmaoh well
00:06:44domonoky1do inlined comments end at a continued lines ?
00:07:01JdGordon|?
00:07:24domonoky1ie if you break up a long line, and add comments to each section.
00:07:30rasherpixelma: Then I fail to understand your point, unless it's just "I wouldn't use inline comments"
00:07:44JdGordon|domonoky1: "onetwo#comment\nthree" is "onetwothree" as far as the tokeniser is concerned
00:07:52domonoky1%t0.1%?bc<%xdea|>; #first subline %
00:08:12rashershouldn't that be
00:08:18rasher%t0.1%?bc<%xdea|>; %#first subline %
00:08:24*rasher shrugs
00:08:37JdGordon|no the % before the # makes it a literal #
00:08:46rasher%t0.1%?bc<%xdea|>; % #first subline %
00:08:51rashergar
00:08:53rasherwithout the last &
00:09:00rasherMaybe I should just stop typing.
00:09:12 Join soap [50] (n=soap@rockbox/staff/soap)
00:09:32domonoky1but the current parser wont see the ending % and leave a newline there.
00:09:42 Join zzy [0] (n=opera@c-71-202-247-23.hsd1.ca.comcast.net)
00:09:49*rasher maintains that the current parser is pants
00:09:58 Part zzy
00:10:12pixelmarasher: yes that was - which is why I didn't understand why the current behaviour needs changing (especially as I saw it as a downside having to know now that # after some WPS code is different to # on its own line)
00:10:18domonoky1maybe we should just introduce a new wps lanugage :-)
00:10:30rasherpixelma: Because someone else would like to use newline comments.
00:10:38soapLUA!
00:10:43soap;)
00:10:45 Join webguest73 [0] (n=18ec7c6f@gateway/web/cgi-irc/labb.contactor.se/x-cekzcmuuymftwtse)
00:10:55rasherpixelma: you keep claiming it's different. I say that the current behaviour is magic
00:11:05JdGordon|magic! haha
00:11:09domonoky1yes, LUA wps, with selfcreated aninmations :-)
00:11:15pixelmathat's the programmer's point of view
00:11:19JdGordon|soap: na, xml!
00:11:36 Join Rob2222 [0] (n=Miranda@p4FDCDF3A.dip.t-dialin.net)
00:11:37JdGordon|pixelma: no, the programmer says its well defined.... its the user that thinks its magic
00:11:44domonoky1rml = rockbox markup language :-)
00:11:46rasherWhat JdGordon| said.
00:11:56rasherSaying "oh that's just consistent" is the programmer's view
00:12:27pixelmabut how comments work in other languages is not the non-coder
00:12:33JdGordon|ok, so go back to domonoky1's suggestion... inline comments wont eat the \n if it ends with # maybe?
00:12:39pixelmaknowledge
00:12:41webguest73The current e200v1 build seems to be broken. Attempting to play from the file menu freezes. Going back to daily build 20090908 fixes it
00:12:48 Quit webguest73 (Client Quit)
00:13:13JdGordon|6 thumbs waaay up for webguest73~
00:13:30rasherpixelma: It's not even related to how it works in other languages. Eating \n is just magic unless you think the programmer way
00:14:06 Quit niekie (Read error: 104 (Connection reset by peer))
00:14:19 Join niekie [0] (n=niek@dreamworld.bergnetworks.com)
00:15:41 Quit Lynx_ (" HydraIRC -> http://www.hydrairc.com <- Po-ta-to, boil em, mash em, stick em in a stew.")
00:15:45 Quit bertrik ("De groeten")
00:16:10 Join MG_Man [0] (n=MGMan@11.208.101.97.cfl.res.rr.com)
00:16:12pixelmaI don't think I'm thinking the coder's way... I just think it's easier to just say "# always eats the newline" than "lines starting with # will ignore the newline, if you put the # after your WPS code the newline will not be ignored"
00:16:18 Quit parafin (Read error: 60 (Operation timed out))
00:16:32rasherpixelma: Just the fact that you're using the word "newline" shows you are thinking coder's way
00:16:50rasherThe ordinary user doesn't know of the newline at all
00:17:12rasherHe just sees "this is a comment that should get ignored"
00:17:17 Quit TruthTaco (Read error: 113 (No route to host))
00:17:28domonoky1users think like: i put a comment somwhere and it should just work :-)
00:17:50gevaertsApparently, there are just three themes on the theme site that use inline comments at all, with two of them in a way where it makes no difference whether the EOL is eaten
00:17:56*JdGordon| doesnt understand why people tink themers are regular users...
00:18:45rasherJdGordon|: I don't think themers necessarily understand the finer points of newlines
00:18:50pixelmaso, just because I have a certain word for it in a basically foreign language (because I don't know a different one) means I'm thinking the coder way? How else would I speak about it here?
00:18:58rasherProving this hypothesis shouldn't be hard
00:19:35rasherpixelma: I mean that the whole concept of the newline being a separate entity is something that most people don't think of
00:19:46gevaertsAnd the third looks a bit weird. '%ar%# %pp/%pe\n%ar%pt'
00:19:52domonoky1can we just stop the silly discussion, and return to making a sane wps language without breaking too much current wps ?
00:20:02pixelmagevaerts: that's not a comment
00:20:11pixelmathat's a displayed #
00:20:13gevaertspixelma: ah indeed. Well spotted
00:20:21 Join parafin [0] (i=parafin@paraf.in)
00:20:52 Quit petur ("Zzzzz")
00:21:04 Quit Rob2222 (Remote closed the connection)
00:21:05gevaertsOK. So my conclusion is that exactly zero themes on the theme site depend on how inline comment end of lines are handled, so if we want, we can change their behaviour
00:21:07JdGordon|domonoky1: I dont tihnk the laguage is insane...
00:21:23 Join Strife89 [0] (n=Strife89@adsl-154-25-9.mcn.bellsouth.net)
00:21:25JdGordon|it should be consistant though
00:21:37pixelmadomonoky1: what is "works" in regards to inline comments or comments on their own lines, what would the "user" expect?
00:22:04 Join stripwax_ [0] (n=Miranda@87-194-34-169.bethere.co.uk)
00:22:09rasherJdGordon|: It *is* consistant. I'll give pixelma that much. But it shouldn't be when it breaks a use-case
00:22:09domonoky1nothing unexpected, the comment shouldnt change anything displayed.
00:22:14***Saving seen data "./dancer.seen"
00:22:23*gevaerts is now in favour of never eating line endings unless there is an "eat-line-ending" character present
00:22:40*domonoky1 is with gevaerts
00:22:42rashergevaerts: Line ends in #?
00:22:47pixelmagevaerts: then you change all the half-broken WPSs
00:22:59gevaertspixelma: there are none of them on the theme site
00:23:17domonoky1pixelma: there are no wps witch rely on this feature at moment..
00:23:34domonoky1so fixing them is easy :-)
00:23:35pixelmahuh, am I misunderstanding again?
00:23:35JdGordon|gevaerts: I agree... but opposite
00:23:44JdGordon|always eat unless there is a # at the end
00:24:23gevaertspixelma: no. I'm confused...
00:24:26pixelmaI don't believe no-one used "comments on their own lines"
00:24:40 Quit saratoga (orwell.freenode.net irc.freenode.net)
00:24:40NSplitorwell.freenode.net irc.freenode.net
00:24:48domonoky1pixelma: no one wants to change full comment lines.
00:24:58*gevaerts does some new analysis
00:25:18pixelmathat's how I understood gevaerts' statement
00:25:37gevaertsplease ignore my recent statements for now. I'm working on a revised edition
00:25:54 Join kugel [0] (n=kugel@rockbox/developer/kugel)
00:25:57*domonoky1 thinks comment should act as their werent there. so full comments eat newlines, inline comments dont. add a new token to split long lines.
00:26:38*kugel thinks that a) comments should always eat \n and a special char would invert that, or b) comments never eat \n and a special char would invert that
00:27:05*kugel prefers b)
00:27:27JdGordon|domonoky1: thats what the patch did, but I dont like the 2nd part anymore... the reason to split the line should be to add a comment so we should just keep it as is
00:27:28*domonoky1 thinks b) breaks current wps
00:27:29kugelI can't make any sense of it that comments have different behavior depending on where they are
00:27:34pixelmaa is the current situation (without the inverting) and would need less changes
00:27:51pixelmaand what domonoky1 said
00:27:54rasherThen you get the "when do I need to add the X character to the end of my comments?" issue
00:28:02kugelso?
00:28:20domonoky1the behavior of comments should be to not do anything, eating newlines or not is just a implementation detail.
00:28:21rasherIt wouldn't be necessary if the language was sane :|
00:28:28kugelI don't think most themers never heard of the "newline problem".
00:28:31 Quit niekie (Read error: 104 (Connection reset by peer))
00:28:34 Join niekie_ [0] (i=quasselc@dreamworld.bergnetworks.com)
00:28:37kugeland if not, we can make them sensible for it
00:29:15kugelI rather think most wps coders are familiar with other script/markup or even real languages
00:29:26pixelmaforcing every third to update his WPS isn't nice
00:29:36 Join PaulJam [0] (n=Paule@p54BEE436.dip.t-dialin.net)
00:29:38kugellooking at abi, half of the themes were done by 2 people where one can code, and the other can photoshop
00:30:01Zagorisn't it better to add a "line continuation tag" to allow line-splitting instead of fidgeting with the comments?
00:30:05kugelpixelma: didn't gevaerts just conclude that 0 themes on the theme site need to change?
00:30:09pixelma(the "third" is just a guess but I think a lot use full line comments)
00:30:17*rasher is wary of using abi as an example of anything
00:30:18Zagorsort of like you split #defines with \ in cpp
00:30:20 Join Strife1989 [0] (n=Strife89@adsl-154-11-147.mcn.bellsouth.net)
00:30:22gevaertskugel: I concluded that everything I said was suspicious
00:30:34JdGordon|Zagor: I did but, but dont like it anymore
00:30:38*domonoky1 is with zagor..
00:30:43pixelmakugel: he was confused and needed exami...
00:30:45*kugel is too
00:30:48ZagorJdGordon|: why?
00:31:06kugelI quite like "\"
00:31:23JdGordon|because the reason to split lines should be to add a comment, or make it easier to read... using the comment token is a good way to make themers think about writing something there
00:31:23rasherkugel: there's no good reason to use \ rather than %. And a good reason not to.
00:31:38*kugel isn't against % either
00:31:49pixelmaZagor: we already have that (by the means that # always eats newlines)
00:31:54gevaertsOK. No theme uses partial-line comments in a way that cares about whether or not they eat the \n, so changing those not to eat them (and leaving full-line comments unchanged) would not break anything
00:32:11Zagormixing tags is asking for trouble. don't add "features" to the comment tag
00:32:13*JdGordon| slaps gevaerts
00:32:17JdGordon|comments SHOULD eat the \n
00:32:21domonoky1JdGordon| but what would happen if you have just two lines of wps code with a inline comment at the end of the first.. they would get joined if we eat the newline there.
00:32:26kugelso we're basically free to change stuff? (ignoring other theme repos)
00:32:30rasherJdGordon|: not partial-line comments
00:32:40ZagorJdGordon|: possibly, but comments should not be the mechanism for line splitting
00:32:44rasherkugel: Weren't we always?
00:32:57JdGordon|Zagor: I can accept that
00:32:58rasherZagor: That's the current situation.
00:33:02gevaertsJdGordon|: I don't care what they should do. I'm trying to evaluate the cost of *all* options
00:33:07kugelrasher: 2-3 people are apparently caring about not braking themes here
00:33:10JdGordon|so how about always eat unless % is the last char of the line?
00:33:10Zagorrasher: that it is, or isn't?
00:33:16kugels/themes/0 themes/
00:33:21 Quit matsl (Read error: 60 (Operation timed out))
00:33:21rasherZagor: It is. You can use # to split lines.
00:33:34Zagorrasher: yes, but can you use anything else?
00:33:58kugelZagor: that's my opinion too. comments shouldn't be making up line splitting or not
00:33:59 Quit Riku (Nick collision from services.)
00:34:06rasherkugel: If breaking a few themes (or many) is the cost of fixing a bug (I still regard this as a bug), I don't have any issue with that.
00:34:12pixelmayou wouldn't need to and I think it's not the question
00:34:21kugelimo comments should behave independantly of their position, and an *explicit* char should handle \n
00:34:27 Join Lss__ [0] (n=Lss@cm204.delta92.maxonline.com.sg)
00:34:27domonoky1JdGordon|: i think that then you would % in the normal case (just indipendend lines with comments at the end) and the special case wont need one.
00:34:39gevaertskugel: an explicit char handles \n. It's \n
00:34:57pixelmathose are two chars
00:34:59bughunter2fixed my bricked e200v2 :)
00:34:59kugelgaah
00:35:05bughunter2lesson learned: do not spam the root directory with 40 thousand files
00:35:06bughunter2:D
00:35:18Zagorbughunter2: haha
00:35:28bughunter2rockbox handled it well btw, it just said "Dir Buffer is Full"
00:35:30bughunter2the OF failed
00:35:31bughunter2hard
00:35:31bughunter2:P
00:35:52bughunter2using recovery mode + winhex allowed me to remove all the entries
00:36:00pixelmakugel: I'm definitely *against* changing the behaviour of full line comments
00:36:01gevaertsChanging the way full-line comments behave will break nearly all themes
00:36:18bughunter2TheSeven: ^ :D
00:36:38*domonoky1 is still with: comments on full lines act as they werent there (ie thy eat the newline), inline comments act as they werent there (ie they dont eat newlines) and add a new char for line joining :-)
00:36:50kugelpixelma, gevaerts: I'm ok with that
00:37:05gevaertskugel: it breaks them in a way that cannot be automatically detected
00:37:07*gevaerts is with domonoky1
00:37:09Zagordomonoky1: I agree
00:37:10TheSevenbughunter2: it just crashed?
00:37:16bughunter2no I fixed it ;)
00:37:31bughunter2the OF hung when rebooting
00:37:37bughunter2instead of saying "Not enough free space to create database"
00:37:38bughunter2hehe
00:37:43gevaertskugel: so you're basically saying we should just purge the theme site
00:37:54JdGordon|sounds like a good idea!
00:37:55kugelgevaerts: read my 2 suggestions
00:37:56bughunter2bought a set of micro screw drivers to fix the player (I anticipated this happening, luckily)
00:38:19TheSevenbughunter2: you need to OPEN that thing to fix a fs-level issue!?
00:38:23JdGordon|domonoky1: I can live with that.... but then you have to ask, why not allow comments to be closed with a # also then?
00:38:25domonoky1if we really want to make such a incompatilbe change, we should take the chance and make a new wps language :-)
00:38:29*kugel isn't with domonoky1, but that topic doesn't touch me enough anyway
00:38:31bughunter2TheSeven: yes, RB doesn't have a native USB driver for this player yet
00:38:32JdGordon|which seems to make more sense
00:38:32 Quit stripwax (Read error: 110 (Connection timed out))
00:38:34*TheSeven is glad he has an ipod...
00:38:36bughunter2so that's why I needed the OF
00:38:38bughunter2but the OF failed
00:38:39bughunter2:P
00:38:39pixelmagevaerts: nah, I think he means "I'm fine with that" that he's fine with not changing the behaviour of full-line comments
00:38:52ZagorJdGordon|: sure you can close it, but it doesn't make a difference in the newline issue
00:38:57bughunter2TheSeven: anything but iPod imo :), I like this player a lot :)
00:38:58gevaertskugel: I'm trying to find them again in this mess :\
00:38:58pixelmamaybe not
00:39:02 Quit Strife1989 ("If you hold a Unix shell to your ear, you can hear the C.")
00:39:03domonoky1JdGordon|: that would be a fine extension too. (but only close the comment, no action on line joining or alike)
00:39:14bughunter2I bricked (and fixed) it myself, after all :)
00:39:17kugel"kugel thinks that a) comments should always eat \n and a special char would invert that, or b) comments never eat \n and a special char would invert that"
00:39:23gevaertsthanks
00:39:23JdGordon|#% is going to look totally fucked thoguh!
00:39:36TheSevenbughunter2: I won't buy a player where you need to void warranty in order to fix such a thing
00:39:53bughunter2TheSeven: they can't know whether the warranty was voided
00:40:06pixelmaJdGordon|: would that be a new inline comment marker (sorry I missed the first mention of this)
00:40:11bughunter2anyway, DRM is disabled after installing RB, for yet unknown reasons, so by that they might be able to detect it, though
00:40:23TheSevenwell, opening it is also always another bricking risk
00:40:24bughunter2TheSeven: btw, it might have happened with your ipod as well
00:40:28gevaertskugel: as far as I'm concerned, (b) is out because of breaking everything. (a) would work, but I think it's ugly
00:40:31bughunter2depends :)
00:40:52JdGordon|pixelma: no that would be something like this line..... "%ia #some comment but dont end the line #%"!
00:40:53TheSevenbughunter2: there is no way we know of putting an ipod in such a borked state
00:41:00bughunter2I'm sure I could help :D
00:41:20bughunter2making drivers go nuts is dangerous ;)
00:41:23domonoky1kugel: this would also mean comments act different on where they are placed regaring the produced wps..
00:41:28ZagorJdGordon|: sure, but why add the complexity? just add the line-continue tag before the comment
00:41:37TheSevenyou would at least to either mess up the FTL entirely in a manner that locks up FTL code (which I don't htink is possible) or you need to damage the NOR flash contents while keeping the checksum intact
00:42:11gevaertsZagor: won't that complicate the parser
00:42:11TheSevennothing else can put an ipod into a non-recoverable (via key combos + usb) state
00:42:11gevaerts?
00:42:14JdGordon|Zagor: thats even more dangerous.... "%# this is a comment" would get displayed!
00:42:27Zagorcan we use %\ ?
00:42:28TheSevenbughunter2: we should move this discussion to -community, though, i think
00:42:49bughunter2done
00:42:51JdGordon|Zagor: % is our \.... %\ is a literal \
00:42:53 Join kugel_ [0] (n=kugel@e178089197.adsl.alicedsl.de)
00:42:55 Quit kugel (Nick collision from services.)
00:42:58Zagorah right
00:43:05 Nick kugel_ is now known as kugel (n=kugel@e178089197.adsl.alicedsl.de)
00:43:35gevaertsJdGordon|: huh? Why is %\ a literal \? \ doesn't need to be escaped, does it?
00:43:52pixelmajust wanted to ask the same
00:43:55JdGordon|thats the example
00:44:05JdGordon|but yeah, \ is a regular char
00:44:11rasher%\ should be an error, right?
00:44:19gevaertsso %\ is available for special things
00:44:20JdGordon|yeah
00:44:41kugeldomonoky1: how would they act different?
00:44:48domonoky1so we could use "... %\ # comment" to have a joined lined with a comment at the end.
00:45:13gevaertsI'd just go for %\ being a "joined-line comment" then
00:45:33JdGordon|ah, from a code POV it will be difficult to work if the join-line token is not right before the \n
00:45:37gevaertsor rather a non-joined-line comment, so # keeps its current meaning
00:45:46 Quit Strife89 (Read error: 110 (Connection timed out))
00:45:47Zagorgevaerts: no. separate line-joining and comments. even if it technically could act as it. but don't document that!
00:45:47domonoky1kugel: in your ways it would modifiy the linejoining depending on comment position.
00:46:01JdGordon|also, control stuff should be 1 char... not two
00:46:33kugeldomonoky1: I still don't understand
00:46:53Zagorwe only have a single comment char: # the %\ tag is strictly for line joining. don't drag our themers into this confused debate :)
00:47:00 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
00:47:37gevaertsZagor: ok. That keeps the problem of compatibility though
00:47:44JdGordon|I think % by itself is a better line joiner than \ and definetly %\
00:47:57JdGordon|but it wont work right before comments :(
00:47:57gevaertsWhich you solve by using a non-join character
00:47:58domonoky1kugel: a) would mean you have to place a newline-eating char at full comments, but not on inline comments. b) the otherway round.
00:47:58 Quit ender` (" We are surprised at our own versatility in being able to fail in so many different ways. -- Samuel McChord Crothers")
00:48:08 Quit thegeek_ (Read error: 54 (Connection reset by peer))
00:48:17Zagorgevaerts: what compatibility?
00:48:21rasherIf the end result is (what we "agree" is) more sane, then I don't see breaking compatibility as a big problem.
00:48:30pixelmadomonoky1: no. that's not what he meant (I think)
00:48:50kugeldomonoky1: depends on what you want (you want different things if you use different comment styles). the comments would behave the same
00:49:14gevaertsZagor: currently all comments join lines implicitely. Changing that is tricky, because it's impossible to look at a file and decide which behaviour it wants, so you can't autoconvert
00:49:39Zagorgevaerts: you're thinking in code. stop it! :-)
00:49:55*domonoky1 just wants that comments dont change anything, and i also dont want to use a different notation beween full comments and inline comments.
00:50:16*gevaerts doesn't mind wps changes that break themes in a way that they stop loading, but he does mind changes that break themes by subtly changing their effect
00:50:17Zagorgevaerts: I still want full-line comments to be ignored completely, just like domonoky1 said
00:50:30Zagorjust don't think of it as line joining
00:50:49gevaertsZagor: ok. So you want domonoky1's proposal plus a continuation character for partial-line comments?
00:50:57Zagoryes
00:51:02pixelmaZagor: the problem is if you change # to not eat the newline even for full-line comments, every WPS that uses full-line comments will have displaced lines
00:51:10domonoky1ignoring a full-line comment completly doesnt count as line-joining from the users-perspective.
00:51:27Zagorpixelma: I just said I *do* want full-line comments to be completely ignored (the whole line).
00:51:53gevaertsI agree that this is the most sensible thing to do
00:51:57ZagorI refuse to mix comments and line joining, conceptually. it will only end in massive confusion
00:52:14pixelmadomonoky1: might be important to mention in this discussion here - proof: I misunderstood Zagor
00:52:48*domonoky1 thinks there is many missunderstandments here, and we all want the same :-)
00:53:22pixelmaZagor: but you have to if you make differences between full-line and inline comments
00:53:28*gevaerts wants to get rid of wps files and get javascript+css instead :)
00:53:50*domonoky1 slaps gevaerts
00:54:21gevaertspixelma: not really. You strip away the comment, and if there's something left on the line the line counts, if there's nothing left, the line is fully ignored
00:54:33 Join thegeek [0] (n=nnscript@s243b.studby.ntnu.no)
00:54:35Zagorpixelma: not really. all comments are deleted by the parser. if there is anything left on the line after that, it is treated as a line.
00:55:15pixelmaand what about blank lines on purpose?
00:55:23JdGordon|gevaerts: but thats what we are arguong about.. what ends the comment? the \n of the char before \n?
00:55:31Zagorpixelma: " # this line is intentionally blank"
00:55:31JdGordon|ideally comments should be endable
00:55:54 Quit domonoky1 (Read error: 104 (Connection reset by peer))
00:56:05Zagorpixelma: you have a point there. we don't ignore empty lines today...
00:56:22ZagorJdGordon|: why?
00:58:06NHeal(timeout) orwell.freenode.net irc.freenode.net
00:58:29pixelmawhich is on purpose to position text without putting everything into an own viewport (and they didn't always exist). I guess it's still used sometimes if it is easier and you'll get the effect you wanted anyways
00:59:21JdGordon|Zagor: sorry.. why what?
00:59:23gevaertsit's definitely used
00:59:44 Nick truthtaco_ is now known as TruthTaco (n=truthtac@adsl-74-12-230.aby.bellsouth.net)
00:59:51ZagorJdGordon|: why should comments be endable? do you really want to put tags *after* a comment?
01:00
01:00:25JdGordon|no, but thats exactly how you would have to use the joiner tag if you wanted a comment also (unless you want to make the parser more complicated)
01:00:47Zagorhuh? why not %\ # comment
01:00:55kugelgah
01:01:01*JdGordon| agrees with kugel :p
01:01:08Zagorwhat's wrong?
01:01:10kugelcan't we just have c-style, or even c++ style comments?
01:01:17 Part TruthTaco ("Leaving")
01:01:27JdGordon|OK... how about ## doesnt eat the line # does though!
01:01:37pixelmakugel: how does that solve the newline issue?
01:01:59gevaertsJdGordon|: 25 themes from the theme site have '##'
01:02:00kugelc-style need to be closed explicitely
01:02:15JdGordon|shuush kugel
01:02:23Zagorwhy do you want to require %\ to be last on the line?
01:02:35JdGordon|purely from a code P.O.V
01:02:43 Quit stripwax_ (Read error: 110 (Connection timed out))
01:02:46JdGordon|the parser is dead simple
01:02:56ZagorJdGordon|: %\ = ignore rest of line, including newline
01:02:58*gevaerts thinks that continuation lines should be done by starting the continuing line with a backspace character
01:02:59JdGordon|I also dont agree that %\ is a good joiner
01:03:19ZagorI have no opinion about which tag to use
01:03:38JdGordon|gevaerts: that will be a total mess to implement
01:04:05gevaertsJdGordon|: yes, but that's nicely balanced by also being a pain for the theme authors :)
01:04:17JdGordon|EXCELLANT!
01:05:27JdGordon|there is some logic to using ^ to say "join onto the previous line" though
01:05:40gevaertsDo people actually want continuation lines combined with end of line comments?
01:06:00JdGordon|thats the best time to use them
01:06:31Zagorgevaerts: I though that was the big point of splitting lines?
01:07:30 Quit KBH ()
01:07:43gevaertsit's the big point of splitting lines, sure (except in some of my experimental wpses that have 500-character lines...), but splitting lines are not the only reason for end-of-line comments
01:08:31Zagorof course
01:08:49Zagorhence my insistance that we don't mix the two functions
01:10:06JdGordon|you know what... screw it... we should use % to tell it to ignore the \n, and just hope themers are smart enough to put a space between it and #
01:10:22PaulJamkugel: did you notice that your custom SB patch is broken with the current SVN revision?
01:10:38pixelmaif you transfer kugel's c reference to WPS - you could see the /* as the # and you would only need a */ equivalent for someone who wants inline comments, no?
01:11:06kugelPaulJam: yes, I think it's this bug actually: http://www.rockbox.org/tracker/task/10596
01:11:11 Join HBK [0] (n=hbk@rrcs-97-77-51-170.sw.biz.rr.com)
01:11:24JdGordon|pixelma: please dont entertain that idea!
01:11:30gevaertsJdGordon|: so '<whatever valid wps code> % #comment' for continuation?
01:11:41JdGordon|yeah
01:11:43rasherpixelma: it's not a good comparison, since blank lines in C don't have any significance
01:12:02kugelwe could also say the current situation isn't bad, and tell everyone to insert a blank line who wants inline comments which doesn't eat newlines?
01:12:18JdGordon|NO!
01:12:38gevaertsJdGordon|: that would only break for new themes if people aren't careful, and I'd expect people to look at the wps screen before uploading, so I think that's fine
01:12:44pixelmawell, just thought it would be an easier solution - maybe I shouldn't have drawn the analogy
01:13:17JdGordon|we could do %* blaaaa *% though :p
01:14:17JdGordon|I tinhk that is actually more understandable than an extra line joiner tag
01:14:46JdGordon|nice and easy to implement also
01:14:49Zagorhow does that do away with the line joiner tag?
01:14:56*gevaerts thinks that JdGordon|'s '% #' proposal is fine. It doesn't fully meet Zagor's wish for total independence of comments and continuation handling, but it seems impossible to meet that while keeping compatibility
01:15:28JdGordon|no more ambiguity over what comments do # is the rest of the line, %* *% is everything between them
01:15:32JdGordon|it also allows you to comment out blocks
01:15:37*kugel thought compatibility isn't an issue?
01:16:07ZagorJdGordon|: so you're commenting out the newlines? hmm...
01:16:46JdGordon|you're commenting out everything....
01:16:50kugelJdGordon|: that's c-style!
01:16:53JdGordon|between and including the tags
01:16:56*JdGordon| knows :(
01:16:58JdGordon|it works though
01:17:03kugelwhat I said
01:17:05 Quit bubsy (Read error: 60 (Operation timed out))
01:17:43JdGordon|it could even just be a single % on either end... but that might be harder to follow
01:17:55 Quit archivator (Read error: 104 (Connection reset by peer))
01:18:45kugelblock comments would be nice for debugging
01:19:31ZagorI don't quite see it as a convenient replacement for a line joiner
01:19:59 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
01:20:06*kugel votes for the empty line line-joiner :p
01:24:12kugelJdGordon: did my commit break non-backdrop themes?
01:24:29JdGordon|your whowho did what now?
01:24:45kugel:?
01:24:54JdGordon|which commit?
01:25:07kugelr22662
01:25:44JdGordon|I dont tihnk so?
01:26:51kugelmy statusbar patch worked with r22657
01:27:15kugelI'm thinking PaulJam's problem is because sbs don't have backdrops
01:27:54 Quit faemir ("Leaving")
01:28:53kugelall commits inbetween are mine :p
01:29:00JdGordon|oh good... your bug then :)
01:29:07JdGordon|I cant imagine why that would have broken it though
01:30:59kugelI'm wondering why data->num_tokens is incremented once more since r22618, and why it's incremented after setting the current token to WPS_NO_TOKEN
01:32:55kugelalso (for the album art problem), didn't the token always return "C"? Why is the C showing now, but didn't before?
01:34:39 Quit PaulJam (".")
01:34:48JdGordon|I'm not entirely sure...
01:35:12JdGordon|that first bit I have to look at... but im at work still
01:36:29kugelwhat does %x|a|filename|x|y| do?
01:36:44kugeldancingpuffdou has this, for the backdrop apparently
01:36:48 Quit Zagor ("Clint excited")
01:36:53pixelmastatically load a bitmap
01:37:05kugelnot %xl
01:37:30pixelmawhich would be dynamically loaded bitmaps...
01:37:41kugelwhat's the difference?
01:37:45 Quit Thundercloud (Remote closed the connection)
01:37:52 Part toffe82
01:38:04kugelCustomWPS doesn't tell anything about it
01:38:04JdGordon|no l means its always displayed
01:38:40pixelmathe %x bitmap is always shown whereas you can control whether the %xl bitmap is shown or not with the%xd tag
01:39:39*kugel didn't connect *loading* statically or dynamically with that
01:40:04pixelmaand it's often used in monochrome WPSs instead of backdrops (DancePuffDuo is basically a monochrome WPS, works on greyscale and colour too of course
01:40:13kugelah, but I just see CustomWPS does tell about it, I'm apparently blind
01:40:54pixelmawell you load it in a way of "do something with it now and be done with it" or "load it but let's use it later"
01:41:43JdGordon|so what did we decide re the comments?
01:43:22pixelmaI see the album art problem now with my WPS that worked fine before (some early 22600 rev) -> blank space at parts of the screen (as you said at the left corner of the screen) just updated when there are other bitmaps or dynamically updating WPS tags in that area
01:43:23kugelhrm, the test wps at fs#10596 doesn't parse here
01:43:47pixelmakugel: my WPS has no backdrop so that's not the problem
01:44:14kugeldoes it have viewports?
01:44:21pixelmalots
01:45:10 Quit froggyman (Read error: 104 (Connection reset by peer))
01:45:36kugelthat wps isn't invalid, is it?
01:45:45 Quit dfkt ("-= SysReset 2.53=- Ph'nglui mglw'nafh Cthulhu R'lyeh wgah'nagl fhtagn.")
01:45:48kugelit really doesn't parse successfully on my e200 sim
01:45:52pixelmamine?
01:46:17pixelmaplease be carefull with "it" and "that" ;)
01:46:23kugelthe one at fs#10596
01:47:48kugelit seems a wps starting with a vp doesn't work
01:48:37kugeloh, nevermind
01:48:48kugelthat wps works fine here :S
01:57:02kugelmy custom statusbar patch also still works :/
01:59:07 Quit JdGordon| ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
02:00
02:00:51kugelpixelma: you mean you see the "C" now?
02:01:04kugelwould be interesting which commit introduced that
02:03:45 Quit DerPapst ("Leaving.")
02:14:38 Join JdGordon_ [0] (i=ad7fa80e@gateway/web/freenode/x-prefxzgxvqmrxomd)
02:15:19JdGordon_it should be obvious which commit made C show
02:15:22 Quit stripwax (Read error: 54 (Connection reset by peer))
02:15:41JdGordon_the one I changed when drawing AA was moved to after images
02:17:08kugelJdGordon_: have you actually been able to repro the crash at FS #10596?
02:17:44 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
02:17:50JdGordon_I havnt tried.. those 2 fixes where literally while i was havcing breakfast
02:18:38 Join AndyI [0] (n=pasha_in@212.14.205.32)
02:18:41JdGordon_you cant repro?
02:18:54 Quit robin0800 (Remote closed the connection)
02:19:00kugelno
02:20:21 Quit bughunter2 ("Leaving.")
02:21:09 Quit mcuelenaere ()
02:21:16 Quit Klowner (Read error: 110 (Connection timed out))
02:22:15***Saving seen data "./dancer.seen"
02:23:20*JdGordon_ thinks about displaying the whole line of a broken wps with a fancy arow pointing to the problem in the debug output
02:24:54kugelhrm, UnhelpfulAA seems to crash
02:25:11JdGordon_going by its name, that is unsurprising
02:25:51kugeldid we freeze already, btw?
02:26:34JdGordon_I was wondering the same
02:26:44*JdGordon_ is in no rush to send out the freeze email :p
02:29:38kugelJdGordon_: is the skin buffer not free'd properly?
02:30:14JdGordon_what gives you that idea?
02:30:18kugelstrange
02:30:36kugelunder gdb, the skin buffer overflowed after changing wpses a few times
02:31:05 Quit stripwax ("http://miranda-im.org")
02:31:30 Quit AndyIL (Read error: 110 (Connection timed out))
02:31:50kugelit also shows 2 "Skin buffer usage: XX" lines per wps change
02:32:05JdGordon_that shouildnt be possible...
02:32:19JdGordon_which target?
02:32:30kugele200 sim
02:32:38JdGordon_does the skin load?
02:32:40kugelI didn't get the skin buffer overflow again yet though
02:32:46JdGordon_or is the 2nd one the default?
02:33:20JdGordon_OH yeah! it probably isnt freeded after a failed load
02:34:46kugelsomething is really strange here
02:34:59kugelit crashes at another position now if I go to wps from theme settings
02:35:17JdGordon_svn? or your patched?
02:35:20kugelsvn
02:36:15kugelhttp://pastie.org/611805
02:36:40JdGordon_img == 0
02:37:14JdGordon_pastebin the wps?
02:38:22kugelJdGordon_: http://pastie.org/611808
02:38:36 Join Strife89 [0] (n=michael@adsl-154-11-147.mcn.bellsouth.net)
02:38:49kugelthat doesn't even have images
02:39:40JdGordon_yeah
02:39:41JdGordon_wtf?
02:40:16JdGordon_something is not right
02:41:34JdGordon_back in 15
02:42:19kugelif (data->tokens[i].type == WPS_TOKEN_IMAGE_PRELOAD_DISPLAY) shouldn't be true if there's no images
02:44:35 Quit JdGordon_ (Ping timeout: 180 seconds)
02:50:46JdGordoncorrect
02:53:35 Join fdinel [0] (n=Miranda@modemcable204.232-203-24.mc.videotron.ca)
02:54:15kugelJdGordon: anyway, I added a if (img == NULL) check, the other crash remains
02:54:31JdGordonthat shouldnt happen so the check is bad
02:54:35JdGordonit hides a real issue
02:54:44kugelyes..
02:54:56kugelI'm saying it wouldn't even fix the other problem
02:55:30kugelthe other crash is more fishy anyway. get_token_value returns a strange address (gdb just says "Address out of bounds")
02:59:47kugelit crashes on any token, apparently, I commented all out -> no crash, commented all but one out -> crash, commented all but (another) one out -> crash too
03:00
03:06:26kugelis there an easy way to get from a number to the enum memberß
03:06:37JdGordonnope
03:15:23 Quit thegeek (Read error: 104 (Connection reset by peer))
03:23:26kugelJdGordon: getting more strange: if I comment everything but 1 line out, it crashes. if I actually delete everything but that line, it runs
03:24:50kugelI wonder why my breakpoint is hit 4 times even though I have only 1 token
03:24:58kugelbefore it crashes
03:26:01 Quit MG_Man (Read error: 104 (Connection reset by peer))
03:26:15JdGordonfuck
03:27:28kugelhaha
03:28:02JdGordonits only that wps though yeah?
03:28:04kugelI have to correct myself. It only does't crash with everthing but 1 line removed if I leave an empty line at the very beginning
03:28:25JdGordonAAAAAAAAAAAAAHHHHHHHHHHHHHHHHHHHHH
03:29:21kugelhttp://pastie.org/611837
03:29:28kugelit might be a hint :p
03:30:08JdGordoncurr_line->curr_subline->last_token_idx = wps_data->num_tokens - 1;
03:30:19JdGordonlast_idx is -1 !
03:30:35JdGordonwhich is probaly why that wasnt there to start with :)
03:30:41*JdGordon curses this sub/.line handling
03:30:42kugelhow?
03:30:51JdGordonnum_tokens at that point is still 0
03:31:44kugelhow does the empty line make a difference there?
03:37:06JdGordonwhere the heck is tokens[-1] land?
03:37:14JdGordonit still gets done
03:37:21JdGordonactually... well...
03:37:29JdGordonit should still start at 0 and get no further
03:39:18kugelah, a newline char increases num_tokens
03:40:22kugelso, starting with a vp is a bad idea
03:41:25kugelI'm thinking num_tokens should start with 1. if even an empty line increases it
03:42:22kugelit doesn't crash if I do that
03:43:05JdGordonthat will cause other problems
03:43:10kugelor, it could also be changed to increase num_tokens before parsing
03:43:57kugelgoing into vp parsing with num_tokens == 0 doesn't seem totally correct to me
03:44:10 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
03:45:22kugelquick fix seems to cap last_token_idx to 0 in viewport parsing?
03:50:43JdGordonyeah
03:50:51JdGordonsomething like that
03:50:55 Join Canabama [0] (n=andrew@207.157.10.2)
03:51:18JdGordonI think we need to start adding every token to the array, regardless of it its used or not...
03:51:22JdGordonneed to think about it
03:51:27JdGordonheading out now
03:51:48 Part Canabama
03:53:07 Join Dhraakellian [0] (n=ntryon@cpe-72-226-197-191.rochester.res.rr.com)
03:54:18DhraakellianI have an artist that shows up in the database without its album or tracks
03:54:29 Nick advcomp2019_ is now known as advcomp2019 (n=advcomp2@unaffiliated/advcomp2019)
03:54:34DhraakellianI think there may have been an error when I restarted after updating the database
03:55:36Dhraakellianso when I go to export modifications before doing a reinitialize, I get a data abort at 30040958
03:55:58DhraakellianSansa FuzeV1, r22667
04:00
04:04:58 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
04:06:12 Quit TheSeven (Nick collision from services.)
04:06:27 Join The_Seven [0] (n=theseven@dslb-084-056-169-196.pools.arcor-ip.net)
04:06:35 Nick The_Seven is now known as TheSeven (n=theseven@dslb-084-056-169-196.pools.arcor-ip.net)
04:07:30*kugel doesn't feel good at this fix
04:08:06 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
04:11:55 Join froggyman [0] (n=chatzill@pool-72-69-75-180.chi01.dsl-w.verizon.net)
04:14:47 Quit froggyman (Client Quit)
04:20:05 Quit Strife89 (Read error: 110 (Connection timed out))
04:20:10 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
04:21:48 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
04:22:19***Saving seen data "./dancer.seen"
04:22:41 Nick fxb is now known as fxb__ (n=felixbru@h1252615.stratoserver.net)
04:23:51 Quit liar (Remote closed the connection)
04:30:20 Quit HellDragon (Read error: 104 (Connection reset by peer))
04:30:32 Join HellDragon [0] (i=jd@modemcable178.248-201-24.mc.videotron.ca)
04:37:01 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
04:39:14kugelJdGordon: Why is num_tokens right for the index if the viewport is the first thing to parse, but off by one anytime after (which was the reason for the -1)
04:43:23 Quit Rondom (Nick collision from services.)
04:43:33 Join Rondom [0] (n=Rondom@dslb-084-057-162-243.pools.arcor-ip.net)
04:46:08 Quit kugel (Remote closed the connection)
05:00
05:24:47 Join Klowner [0] (n=klown@71-215-18-224.dvnp.qwest.net)
05:27:37 Quit panni_ ("( www.nnscript.de :: NoNameScript 3.81 :: www.XLhost.de )")
05:36:02 Join JohnTeddy [0] (i=unstable@tor/regular/sid)
05:36:40JohnTeddyIf I have playlists on rhythmbox, is it easy to make a playlist on rockbox the same?
05:42:12n1srockbox uses m3u playlists, so if rythmbox can create those it should be pretty easy
05:45:11JohnTeddyn1s: yes, It can save m3u playlists. but the directories for my playlist is stuff on my laptop hard disk, I don't have playlists for the directory structure on my rockbox
05:45:49JohnTeddyI could mimick it though, and make symbolic links. Does rockbox fs support symbolic links?
05:46:01JohnTeddySo do like /home/johnteddy/Music/.
05:46:06n1sno, rockbox uses FAT
05:46:25JohnTeddyAre there shortcuts of any kind?
05:46:53JohnTeddyotherwise everytime I have to play my music, I have two directories to click into. /home/johntedy
05:47:48 Quit Horscht ("Verlassend")
05:48:00n1sthere are shortcuts of one kind, implemented using some plugin but IIRC they don't do what you want, an m3u can use relative paths though so if you manage to get rythmbox to save it like that and uses the same dir structure on your DAP it shoukd work i think
05:48:27JohnTeddyactually whatever, it's easy to modify
05:48:41JohnTeddygedit can just change all '/home/johnteddy' to '', in one second
05:48:45JohnTeddyand save the file
05:48:50n1ssure
05:51:32JohnTeddyI just put the m3u anywhere?
05:51:41JohnTeddyI can make like a 'playlists' directory or something?
05:52:47n1syes and no :) it depends on how it is created... if the paths are relative it has to be in the right place, if they are absolute (which i guess yours are) you can put it anywhere
05:53:32 Quit Overand_ (Client Quit)
05:53:35 Join Overand [0] (i=overand@crappy.domain.name)
06:00
06:06:52 Join Andy [0] (n=461682be@gateway/web/cgi-irc/labb.contactor.se/x-aqbmhuesjavkljxw)
06:08:25Andycan anybody tell me where a full list of supported players are?
06:08:53scorchewww.rockbox.org
06:09:28Andydamn, is that it?
06:10:20AndyI thought the zen vision M would definatly be supported
06:10:37scorchewell, it isnt ;)
06:10:57scorchethere is a port in development, but it is not near up to the usable level
06:11:17Andymaybe I'm mistaken but isn't it the most popular next to the ipod?
06:11:37Andyin fact it's the only one I've people actually using besides an ipod
06:11:43Andyseen*
06:11:45scorchei seriously doubt that, but popularity does not dictate which devices work with rockbox
06:16:53 Quit Andy ("CGI:IRC (Ping timeout)")
06:19:51 Join _Tristan [0] (i=tristan@i.dont.want.to.die.virgin.net.in)
06:22:24***Saving seen data "./dancer.seen"
06:22:29 Join webguest58 [0] (n=4b3e7fdc@gateway/web/cgi-irc/labb.contactor.se/x-upzygppxqqmpspbg)
06:24:05 Quit fdinel ("Miranda IM! Smaller, Faster, Easier. http://miranda-im.org")
06:29:21TheSevenscorche: it's rather popularity within rockbox developers that dictates it, right? :-)
06:29:52 Quit webguest58 ("CGI:IRC (EOF)")
06:32:28 Quit Tristan46 (Read error: 110 (Connection timed out))
06:35:17 Join LinusN [0] (n=linus@rockbox/developer/LinusN)
06:39:16 Join shoe_ [0] (i=4b3e7fdc@gateway/web/freenode/x-ggcyvrgnzvzzimfv)
06:49:51 Nick niekie_ is now known as niekie (i=quasselc@dreamworld.bergnetworks.com)
06:51:07 Quit gevaerts (Nick collision from services.)
06:51:19 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
06:56:13shoe_Hi I'm having trouble installing the bootloader on a Gigabeat S using the precompiled sendfirm.exe. It reports success but the player does not reboot as it should. I can provide more details if anyone can help.
07:00
07:18:42pixelmaJdGordon (kugel): I don't see a "C" now, but I preload album art at a position of x=1 and y=1 but don't show it in a viewport that's in the upper right corner of the screen. This worked without any glitches before, now if album art should be shown, parts of the screen get blanked ou. It looks like an area the size of my album art and position x=1, y=1 of the screen is affected, not everything vanishes though - dynamic bitmaps are still drawn (a
07:18:42pixelma satic one is affected), everything of the textual info (partially from tags) vanishes except the time info (current playtime and length of the track)
07:19:29pixelmafirst "don't" shouldn't have been there... also, I guess I can provide a screenshot if you like
07:24:44 Quit shoe_ ("Page closed")
07:27:36 Quit n1s (Read error: 145 (Connection timed out))
07:45:35 Join stoffel [0] (n=quassel@p57B4E3B4.dip.t-dialin.net)
08:00
08:09:34 Join flydutch [0] (n=flydutch@host214-146-dynamic.15-87-r.retail.telecomitalia.it)
08:10:46 Quit antil33t ()
08:17:47 Part safetydan ("Leaving.")
08:21:20 Quit z35 (Read error: 110 (Connection timed out))
08:22:27***Saving seen data "./dancer.seen"
08:24:39 Join antil33t [0] (n=Mudkips@119.224.12.185)
08:25:04 Join Zagor [242] (n=bjorn@rockbox/developer/Zagor)
08:32:13 Join ender` [0] (i=krneki@foo.eternallybored.org)
08:37:07 Join bertrik [0] (n=bertrik@87.211.49.117)
08:41:54 Join GodEater__ [0] (n=9372e2b6@rockbox/staff/GodEater)
08:46:22 Join petur [50] (n=petur@rockbox/developer/petur)
08:53:07 Join decayedcell_ [0] (n=decayed_@60-241-92-53.static.tpgi.com.au)
09:00
09:01:10 Join pamaury [0] (n=pamaury@140.77.26.69)
09:01:34pamaurypcc1: hello
09:06:02 Quit fyrestorm ("lamers envy me like they envy bill g -- main boot xp, just the way it should be!")
09:07:04 Quit bertrik ("De groeten")
09:11:21 Join daggett [0] (n=denis@lab.openpattern.org)
09:21:17 Join PaulJam [0] (n=Paule@p54BEEBAB.dip.t-dialin.net)
09:21:21 Join Thundercloud [0] (i=thunderc@persistence.flat.devzero.co.uk)
09:24:47 Join DerPapst [0] (n=DerPapst@91-64-221-175-dynip.superkabel.de)
09:32:20 Quit pamaury ("exit(*(int *)0 / 0);")
09:43:59 Join n1s [0] (n=n1s@rockbox/developer/n1s)
09:47:17 Quit Thundercloud (Remote closed the connection)
09:55:14 Join pamaury [0] (n=pamaury@140.77.26.69)
09:56:44 Join robin0800 [0] (n=robin080@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
09:58:54robin0800JdGordon: problem in album art display this conditional %?C< displays album art if true think only %C should do this?
10:00
10:01:10pcc1pamaury: hello
10:01:26pamauryhello
10:01:41pamauryI've wommited some thing about object properties on github
10:02:42pcc1yes, I saw that
10:03:00pamauryI've still not checked date work correctly because mtpfs does not retrieve date and times. PErhaps there is anoter mtpfs implementation that does it
10:05:35pcc1so mtpfs uses the date/times in ObjectInfo? or not at all?
10:05:49pcc1(I've been testing with gphoto2)
10:07:37pamauryNot at all as far as I could see it. Basically, all dates are set to 1970, ... It's possible that it is a bug in my implementation and that the objets properties regarding date/time are buggy but as I'm sure they work for size, I would be strange
10:14:07pcc1hm, I was testing my SendObjectInfo implementation and got a message on screen: "*PANIC* Stkov usb (0)". anyone familiar with this?
10:15:14 Quit HBK (Read error: 104 (Connection reset by peer))
10:21:00pamauryis it a kernel message or a libmtp one ?
10:21:18pamauryAh you mean on your device ?
10:22:01pamauryPerhaps you send empty messages, I read somewhere that some usb hardware didn't like them
10:22:29***Saving seen data "./dancer.seen"
10:23:28pcc1yes, on the device
10:24:11pcc1some googling indicated stkov stood for stack overflow
10:34:16pcc1hm, I may have made a silly mistake
10:39:54 Quit PaulJam (Nick collision from services.)
10:40:01 Join PaulJam_ [0] (n=Paule@p54BEEBAB.dip.t-dialin.net)
10:49:32 Join liar [0] (n=liar@83.175.83.185)
11:00
11:08:35 Quit Llorean ("Leaving.")
11:09:09 Join Llorean [0] (n=DarkkOne@adsl-99-4-146-40.dsl.hstntx.sbcglobal.net)
11:09:47 Join DarkDefender [0] (n=rob@78.69.30.229)
11:18:08 Quit stoffel (Remote closed the connection)
11:20:26 Part decayedcell_
11:20:55 Join decayedcell_ [0] (n=decayed_@60-241-92-53.static.tpgi.com.au)
11:21:08 Part decayedcell_
11:43:35 Quit tarbo_ (Read error: 60 (Operation timed out))
11:43:46 Join Zarggg_ [0] (n=zarggg@65-78-69-194.c3-0.eas-ubr6.atw-eas.pa.cable.rcn.com)
11:45:47 Quit ehntoo (Read error: 60 (Operation timed out))
11:45:57 Join ehntoo [0] (n=ehntoo@adsl-99-156-196-216.dsl.applwi.sbcglobal.net)
11:47:35 Quit aevin (Remote closed the connection)
11:48:27 Quit jfc (Read error: 60 (Operation timed out))
11:49:20 Join jfc [0] (n=john@dpc6682208002.direcpc.com)
11:50:47 Join aevin [0] (i=eivindsy@unaffiliated/aevin)
11:53:49 Join tarbo [0] (n=me@unaffiliated/tarbo)
11:54:58 Join matsl [0] (n=matsl@dhcp126.contactor.se)
12:00
12:02:14 Quit Zarggg (Read error: 110 (Connection timed out))
12:07:03 Quit pamaury (Remote closed the connection)
12:19:12 Quit gevaerts (Nick collision from services.)
12:19:22 Join gevaerts [0] (n=fg@rockbox/developer/gevaerts)
12:22:32***Saving seen data "./dancer.seen"
12:27:34 Quit shodanX (Read error: 110 (Connection timed out))
12:44:24 Quit robin0800 (Read error: 110 (Connection timed out))
12:44:57 Quit PaulJam_ (".")
12:45:14 Join PaulJam [0] (n=Paule@p54BEEBAB.dip.t-dialin.net)
12:46:45LloreanGodEater__: Off of the statement in -community, I actually wonder why "autorock" is still a define, and not simply an option in the "startup screen" choices.
12:47:04GodEater__good point
12:47:07LloreanI seem to recall it's not normally defined so as to skip looking for that file and slowing bootup, but now that there's a start-screen choice, it kinda makes sense to have it as an option, and drop the define.
12:47:18GodEater__indeed
12:47:37GodEater__I'd use my new found commit powers to change it, only I have of course no access to svn during the day, and I'll have forgotten by tea time ;)
12:47:40LloreanEspecially now that we have one plugin that could realistically be a useful startup
12:54:13 Join webguest [0] (n=464605b9@gateway/web/cgi-irc/labb.contactor.se/x-mkdwobdndsatibjm)
12:56:32 Quit webguest (Client Quit)
12:58:47AlexPLlorean: Rockbox1D? :)
12:58:51 Join jfc^2 [0] (n=john@dpc6682208002.direcpc.com)
12:58:54AlexPer, blox
13:00
13:02:25LloreanI guess it would also need a context menu option for .rock files to "set as autorock" or something
13:02:33 Join FOAD_ [0] (n=dok@dinah.blub.net)
13:05:44GodEater__not particularly hard to implement though I wouldn't have thought
13:10:55 Join bubsy [0] (n=Bubsy@94.139.72.137)
13:11:50LloreanI can't imagine it would be. It's basically the same functionality as setting the default recording folder should be I think - store a path somewhere.
13:14:28 Quit jfc (Read error: 110 (Connection timed out))
13:14:52 Quit FOAD (Read error: 110 (Connection timed out))
13:14:52 Nick FOAD_ is now known as FOAD (n=dok@dinah.blub.net)
13:16:41amiconnLlorean: It's not entirely the same. Autorock should allow to set it to *nothing*, skipping the check altogether
13:19:47Torneamiconn: you don't actually need that if it's a start screen choice though, no?
13:19:52Torneyou can just set the start screen back to main menu
13:20:07amiconnHmm, yes
13:20:09Tornethere's no technical reason to need to clear the autorock setting
13:20:14Tornethough aesthetically might be nice ;)
13:20:19*amiconn has it normally set to 'file browser' though
13:20:53*GodEater__ has it set to WPS
13:21:13LloreanIs there a way to unset the recording folder / set it to root?
13:22:17AlexPDelete the directory :)
13:22:42GodEater__stkov
13:22:46AlexPLlorean: http://download.rockbox.org/manual/rockbox-h100/rockbox-buildch10.html#x13-15400010.9
13:23:24 Quit bubsy ("I'll be back somewhere in time...")
13:26:52 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
13:39:36 Quit PaulJam (Nick collision from services.)
13:39:45 Join PaulJam_ [0] (n=Paule@p54BEEBAB.dip.t-dialin.net)
13:47:32 Quit PaulJam_ (Nick collision from services.)
13:47:34 Join PaulJam__ [0] (n=Paule@p54BEF705.dip.t-dialin.net)
13:50:16 Quit ch4os (Remote closed the connection)
13:52:43 Quit DarkDefender (Remote closed the connection)
13:52:49 Quit bughunter2 (Read error: 104 (Connection reset by peer))
13:53:21 Join bughunter2 [0] (n=bughunte@ip4da4427e.direct-adsl.nl)
13:56:26 Quit bughunter2 (Client Quit)
13:58:36 Join dfkt [0] (i=dfkt@unaffiliated/dfkt)
13:59:52 Join DerPapst1 [0] (n=DerPapst@91-64-221-175-dynip.superkabel.de)
14:00
14:03:59 Join bubsy [0] (n=bubsy@94.139.72.137)
14:11:10 Quit DerPapst (Read error: 110 (Connection timed out))
14:11:12 Join shodanX [0] (n=shodanX@jazz.informatik.uni-erlangen.de)
14:13:55 Join gregzx [0] (n=chatzill@dsv220.neoplus.adsl.tpnet.pl)
14:22:35***Saving seen data "./dancer.seen"
14:27:34 Part LinusN
14:27:52 Quit bubsy ("I'll be back somewhere in time")
14:27:55 Join bubsy [0] (n=bubsy@94.139.72.137)
14:28:00 Quit linuxstb (Read error: 113 (No route to host))
14:30:24 Join LambdaCalculus37 [0] (i=44a0430d@rockbox/staff/LambdaCalculus37)
14:33:52 Join z35 [0] (n=z35@ool-45701ce5.dyn.optonline.net)
14:34:03 Join bubsy_ [0] (n=bubsy@94.139.72.137)
14:34:22 Quit bubsy (Read error: 54 (Connection reset by peer))
14:38:36 Join robin0800 [0] (n=robin080@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
14:44:25 Join Strife89 [0] (n=michael@168.16.238.225)
14:54:22 Quit bubsy_ (Read error: 54 (Connection reset by peer))
14:54:30 Join bubsy__ [0] (n=bubsy@94.139.72.137)
14:55:09 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
14:56:38 Join barrywardell [0] (n=barrywar@rockbox/developer/barrywardell)
14:59:49 Quit bubsy__ ("I'll be back somewhere in time")
14:59:52 Join bubsy [0] (n=bubsy@94.139.72.137)
15:00
15:17:22 Join TopyMobile_ [0] (n=topy@chello084010205095.chello.pl)
15:19:05 Quit TopyMobile (Read error: 104 (Connection reset by peer))
15:20:22 Nick fxb__ is now known as fxb (n=felixbru@h1252615.stratoserver.net)
15:24:41 Quit Strife89 ("They say that if you hold a Linux shell to your ear, you can hear the C.")
15:32:00 Join decayedcell_ [0] (n=decayed_@60-241-92-53.static.tpgi.com.au)
15:34:59CIA-43New commit by 03zagor (r22671): Committing current state of buildserver scripts.
15:42:19 Quit BulldogPO (Read error: 110 (Connection timed out))
15:51:17 Join ch4os [0] (n=ch4os@gentoo/user/ch4os)
15:53:59 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-fqsifjurrujthjoh)
15:57:16 Quit robin0800 (Remote closed the connection)
16:00
16:02:30 Quit bubsy ("I'll be back somewhere in time")
16:02:34 Join bubsy [0] (n=bubsy@94.139.72.137)
16:04:13 Part decayedcell_
16:07:36 Join kugel [0] (n=kugel@rockbox/developer/kugel)
16:07:45 Join JohannesSM64 [0] (n=Johannes@cm-84.208.207.140.getinternet.no)
16:22:39***Saving seen data "./dancer.seen"
16:28:04 Join chandoo [0] (n=chandoo@ool-4353b978.dyn.optonline.net)
16:33:37 Join toffe82 [0] (n=chatzill@12.169.218.14)
16:33:46 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
16:45:04 Quit matsl (Read error: 110 (Connection timed out))
16:45:59 Quit evilnick (Ping timeout: 180 seconds)
16:49:27 Nick jfc^2 is now known as jfc (n=john@dpc6682208002.direcpc.com)
16:53:04 Join archivator [0] (n=Delyan@77.70.28.57)
16:53:31 Join robin0800 [0] (n=robin080@cpc3-brig8-0-0-cust436.brig.cable.ntl.com)
17:00
17:00:16 Quit petur ("*plop*")
17:04:14 Quit killan_ (Read error: 104 (Connection reset by peer))
17:04:35 Quit kugel (Nick collision from services.)
17:04:39 Join kugel [0] (n=kugel@e178066030.adsl.alicedsl.de)
17:05:46 Join evilnick [0] (i=0c140464@gateway/web/freenode/x-umhublhkbzlqtccq)
17:07:18 Quit Zagor ("Don't panic")
17:11:06 Join xavieran__ [0] (n=xavieran@ppp118-208-209-164.lns10.mel6.internode.on.net)
17:13:51 Quit Lss__ (Read error: 104 (Connection reset by peer))
17:18:24 Join thegeek [0] (n=nnscript@129.241.139.243)
17:19:33 Quit GodEater__ ("CGI:IRC")
17:20:06 Join GodEater__ [0] (n=9372e2c1@rockbox/staff/GodEater)
17:20:36 Quit BlakeJohnson86 (Remote closed the connection)
17:22:44 Join explore [0] (n=msparker@173.57.115.183)
17:24:18 Quit xavieran_ (Read error: 105 (No buffer space available))
17:24:20 Quit ehntoo (Read error: 105 (No buffer space available))
17:24:37 Quit kugel (Nick collision from services.)
17:24:41 Join kugel [0] (n=kugel@e178077144.adsl.alicedsl.de)
17:26:25 Join mt [0] (n=Miranda@rockbox/developer/mt)
17:29:29 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
17:36:59 Join daurn| [0] (n=daurnima@freenode/staff/daurnimator)
17:37:48 Nick fxb is now known as fxb__ (n=felixbru@h1252615.stratoserver.net)
17:41:27 Quit bughunter2 (Read error: 104 (Connection reset by peer))
17:46:49 Quit kugel (Remote closed the connection)
17:47:05 Join DarkDefender [0] (n=rob@78-69-30-229-no36.tbcn.telia.com)
17:48:10 Quit GodEater__ ("CGI:IRC")
17:50:04 Join Shlee [0] (n=Shlee@115-166-22-228.ip.adam.com.au)
17:50:35 Join DerPapst [0] (n=DerPapst@91-64-221-175-dynip.superkabel.de)
17:52:52 Quit daurn (Read error: 110 (Connection timed out))
17:54:32 Join kugel [0] (n=kugel@rockbox/developer/kugel)
17:59:40 Quit JohannesSM64 (Read error: 54 (Connection reset by peer))
17:59:53 Join JohannesSM64 [0] (n=Johannes@cm-84.208.207.140.getinternet.no)
18:00
18:01:05 Quit DerPapst1 (Read error: 110 (Connection timed out))
18:04:37 Quit evilnick (Ping timeout: 180 seconds)
18:05:27 Join HBK [0] (n=hbk@rrcs-97-77-51-170.sw.biz.rr.com)
18:18:17 Join domonoky [0] (n=Domonoky@rockbox/developer/domonoky)
18:18:27 Quit archivator ("Leaving")
18:18:31 Join fyrestorm [0] (n=nnscript@24.90.84.240)
18:19:10 Join thegeek_ [0] (n=nnscript@s243b.studby.ntnu.no)
18:22:43***Saving seen data "./dancer.seen"
18:28:42 Quit HBK (Read error: 104 (Connection reset by peer))
18:29:16 Join archivator [0] (n=archivat@77.70.28.57)
18:29:35 Join moos [0] (i=mostafa@85.171.100.120)
18:29:55 Join HBK [0] (n=hbk@rrcs-97-77-51-170.sw.biz.rr.com)
18:34:52 Quit antil33t (Read error: 131 (Connection reset by peer))
18:35:06 Join antil33t [0] (n=Mudkips@119.224.12.185)
18:36:35JdGordon(07:39:15 PM) kugel: JdGordon: Why is num_tokens right for the index if the viewport is the first thing to parse, but off by one anytime after (which was the reason for the -1) <−−- because %V is not put into the tokens array so num_tokens isnt incremented... later on when %V happens there are toens there and we need to tell the previous line to finish on the (i thought) current toekn
18:38:23 Quit thegeek (Read error: 105 (No buffer space available))
18:40:33 Join merbanan [0] (n=banan@c-83-233-172-245.cust.bredband2.com)
18:46:04 Quit JohannesSM64 (Read error: 54 (Connection reset by peer))
18:46:41 Join JohannesSM64 [0] (n=Johannes@cm-84.208.207.140.getinternet.no)
18:49:43kugelJdGordon: you're saying the "-" is done before the "?:" ?
18:50:21 Quit fyrestorm (Read error: 131 (Connection reset by peer))
18:50:22 Join fyre^OS [0] (n=nnscript@cpe-24-90-84-240.nyc.res.rr.com)
18:50:39 Quit fyre^OS (Read error: 104 (Connection reset by peer))
18:50:54 Join fyrestorm [0] (n=nnscript@cpe-24-90-84-240.nyc.res.rr.com)
18:51:06 Join panni_ [0] (i=hannes@ip-95-222-19-21.unitymediagroup.de)
18:51:37 Quit bubsy ("I'll be back somewhere in time")
18:52:10kugelamiconn: ping
18:53:42robin0800kugel: amended what I wrote in the bug I meant %?C<t10.... does this which I think should not
18:54:15kugeldepends on what the rest of the line is...
18:54:26 Quit archivator (Remote closed the connection)
18:55:25robin0800kugel: There is ni %C in it or any where else in the wps and I have album art displayed
18:55:54JohnTeddyIs it possible to have my ipod charging, and playing music at the same time? with rockbox
18:58:01domonokyJohnTeddy: sure, you have to press the right button while you inster usb. see manual for the correct button.
18:58:28robin0800kugel: This is the wps http://paste.ubuntu.com/268664/
18:59:03 Join Grahack [0] (n=chri@ip-222.net-82-216-222.rev.numericable.fr)
18:59:05pixelmakugel, JdGordon: did anyone of you see my description (answer to this night's question)? That was with rev. 22668 (downloaded from the site) on my c200
18:59:10JohnTeddydomonoky: Is it possible to have this activity by default?
18:59:30kugelpixelma: I saw that, but I didn't understand it
18:59:32JohnTeddyI have my ipod plugged into my car 24/7. I want when I start my car, it turns on and starts playing music basically.
19:00
19:00:11robin0800JohnTeddy: Car mode
19:00:27JohnTeddyaww, you guys rock. literally
19:00:28JohnTeddythanks
19:00:47pixelmaI'm not sure that works as expected on an Ipod but you can try
19:01:05 Join bubsy [0] (n=bubsy@94.139.72.137)
19:04:23 Quit Shlee ()
19:05:39JdGordonrobin0800: try this http://pastebin.com/m723b8cdd
19:07:07 Join saratoga [0] (i=9803c6dd@gateway/web/freenode/x-nkptzyqeccsxioxx)
19:07:46kugelgcc 4.4.1 + binutils 2.20 seems successful so far, doing a few test_codec runs rightnow
19:08:12 Join stoffel [0] (n=quassel@p57B4CEA1.dip.t-dialin.net)
19:08:19kugelJdGordon: "JdGordon: you're saying the "-" is done before the "?:" ?"
19:08:40 Quit panni_ (Connection reset by peer)
19:08:42JdGordonyeah, just checked the precedance table and i tihnk thats correct
19:08:53 Join panni__ [0] (i=hannes@ip-95-222-19-21.unitymediagroup.de)
19:08:56kugelthe one on wikipedia?
19:09:01JdGordonno
19:09:19JdGordonhttp://docs.roxen.com/pike/7.0/tutorial/expressions/operator_tables.xml
19:09:32 Join Johannes[SM64] [0] (n=Johannes@cm-84.208.207.140.getinternet.no)
19:09:56 Quit Johannes[SM64] (Read error: 54 (Connection reset by peer))
19:10:25 Quit Sajber^ (Read error: 104 (Connection reset by peer))
19:11:29 Join bertrik [0] (n=bertrik@ip117-49-211-87.adsl2.static.versatel.nl)
19:19:37 Quit linuxstb (Read error: 113 (No route to host))
19:22:32 Join Horscht [0] (n=Horscht2@xbmc/user/horscht)
19:23:46PaulJam__kugel: btw, i inserted a blank line at the top of the .sbs file and now it works again with a current revision (no freeze).
19:29:21 Quit daggett ("Ex-Chat")
19:31:40 Join bughunter2 [0] (n=bughunte@unaffiliated/bughunter2)
19:35:00pixelmakugel: http://imagebin.org/63345 (I darkened the light blue a bit to make it better visible): album art is preloaded with %Cl|2|1|40|40| at the start of the file with other bitmap definitions but only shown in a conditional viewport starting at x=89, y=1. With current SVN an area gets blanked out that has the dimensions of the album art at the position x=2, y=1 of the screen. That is a bit hidden by the fact that dynamic info/bitmaps/progressbar
19:35:00pixelma is getting drawn on top of it again
19:36:26kugelinteresting
19:36:55kugelsaratoga: ping
19:37:15 Join linuxstb [0] (n=linuxstb@rockbox/developer/linuxstb)
19:37:20saratogakugel: pong
19:37:27pixelmathe "i" bitmap is a statically loaded one so vanishes too
19:38:08kugelsaratoga: I'm running a build with gcc 4.4.1 (release) and binutils 2.20 (snapshot). Everything seems to run fine, and codecs seem a bit faster.
19:38:36saratogaany test_codec results?
19:38:42saratogaand AMS?
19:38:58kugelI'm doing tests on my e200 right now
19:39:41 Quit barrywardell (Read error: 145 (Connection timed out))
19:39:57kugelI haven't done the tests on a normal build (apparently my FS just broke), but comparing to CodecPerformanceComparison all my tested codecs seem to be faster by about 1MHz
19:40:19 Join rvvs89 [0] (n=ivo@pdpc/supporter/base/rvvs89)
19:45:06 Quit explore ("leaving")
19:47:47 Join Lss [0] (n=Lss@cm204.delta92.maxonline.com.sg)
19:58:12 Quit PaulJam__ (".")
20:00
20:04:45kugelsaratoga: hrm, only mp3 is slighly faster
20:05:01kugelaac/ogg is slightly slower
20:05:07kugelflac is noticeably slower
20:07:56 Join bluebrother [0] (n=dom@rockbox/developer/bluebrother)
20:08:02 Join petur [0] (n=peter@rockbox/developer/petur)
20:10:22 Join Thundercloud [0] (i=thunderc@persistence.flat.devzero.co.uk)
20:17:39 Join gregzx [0] (n=chatzill@dsv220.neoplus.adsl.tpnet.pl)
20:18:36 Join matsl [0] (n=matsl@1-1-4-2a.mal.sth.bostream.se)
20:21:38 Quit flydutch ("/* empty */")
20:22:46***Saving seen data "./dancer.seen"
20:25:34 Join gregzx_ [0] (n=chatzill@cmn229.neoplus.adsl.tpnet.pl)
20:28:15 Quit gregzx (Nick collision from services.)
20:28:20 Nick gregzx_ is now known as gregzx (n=chatzill@cmn229.neoplus.adsl.tpnet.pl)
20:30:37amiconnkugel: pong
20:32:16kugelamiconn: I just had the (small?) problem that under my gcc 4.4.1 setup (see -dev ml) doom was linking gcc-support from the plugin lib. that resulted in 2 memcpy versions (doom has it's own wrapper for rb->memcpy). That doesn't happen under 4.0.3
20:32:33kugelremoving the wrapper from doom obviously solved the problem
20:33:18kugelthat was the only thing I needed to change in the source for gcc 4.4.1
20:33:18kugelI only build for arm
20:33:39amiconnWell if that actually *works*, it should be done in svn too imo
20:36:18kugelI didn't run doom, nor applied that change to SVN doom for building with 4.0.3, but the wrapper is the same
20:37:04kugelI was just wondering how that happened, why doom links that under 4.4.1
20:37:48amiconnThat's probably due to linking order
20:38:12amiconnI think this wrapper in doom is an oversight from before the "automatic" wrappers
20:38:47 Nick fxb__ is now known as fxb (n=felixbru@h1252615.stratoserver.net)
20:39:55 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
20:42:16 Join JdGordroid [0] (n=jdg@m4b0436d0.tmodns.net)
20:45:47 Join explore [0] (n=msparker@pool-173-57-115-183.dllstx.fios.verizon.net)
20:47:13 Quit explore (Client Quit)
20:47:28 Join explore [0] (n=msparker@pool-173-57-115-183.dllstx.fios.verizon.net)
20:49:46 Quit pixelma (Nick collision from services.)
20:49:48 Join pixelma_ [0] (i=quassel@rockbox/staff/pixelma)
20:50:08 Nick pixelma_ is now known as pixelma (i=quassel@rockbox/staff/pixelma)
20:50:58 Quit amiconn (Nick collision from services.)
20:51:00 Join amiconn_ [0] (i=quassel@rockbox/developer/amiconn)
20:51:20 Nick amiconn_ is now known as amiconn (i=quassel@rockbox/developer/amiconn)
20:54:43 Join stripwax_ [0] (n=Miranda@87-194-34-169.bethere.co.uk)
21:00
21:01:48 Join drakonik [0] (n=moot@72.40.159.202)
21:02:08drakonikI have a question about how Rockbox's database scanning works.
21:02:25drakonikWhen the db scanner is working, it finds a new file, and adds it to some kind of queue that gets processed on the next bootup.
21:03:23drakonikOr rather, on shutdown or completion of the ENTIRE scan, it adds the files to the commit queue.
21:03:53drakonikI want to change it so that each file gets added to the queue after it's found. What kind of slowdown would I see? Would it be a horrible horrible idea?
21:06:12bertrikkugel, nice job on the gcc comparison
21:07:01bertrikdo newer gcc's still need to be patched like our current dev setup?
21:10:10 Join gregzx [0] (n=chatzill@cmn229.neoplus.adsl.tpnet.pl)
21:15:00kugelI think so
21:16:38robin0800JdGordon: cant get that pastbin to patch isit right?
21:17:33 Quit JdGordroid (Remote closed the connection)
21:25:44robin0800JdGordon: i've got "patch: **** Only garbage was found in the patch input"
21:26:33Slasheridrakonik: what do you mean with that? each file _is_ added to the queue, or actually a temp file after it's found
21:27:00kugelrobin0800: that's because JdGordon can't learn to not post patches at pastebin.com
21:27:47drakonikSlasheri: I'm just inferring what I can. See, for whatever reason, the scanner crashes my ipod at random points, and NO files are added. So I'm assuming that the files are added to the db-commit-queue when the scan is completed or on sane shutdown of the ipod.
21:28:11robin0800kugel: Was it for me anyway?
21:28:23drakonikSo, I want to change the scan process to add a file to the commit queue as each file is succesfully scanned.
21:28:34 Quit TopyMobile_ (Success)
21:28:47drakonikThat way, even if it crashes, on next bootup, it will be able to commit the files it DID manage to scan to the DB.
21:29:10Slasheridrakonik: ah, actually the temp file will be appended after each file is found
21:29:32 Quit kugel (Remote closed the connection)
21:29:32Slasheribut that file probably gets corrupted until it's closed properly (when the entire scan finishes)
21:29:36drakonikHm.
21:29:40drakonikOkay.
21:29:46drakonikI understand now.
21:30:19drakonikSo I just need to figure out how to safely close (and presumably reopen) the file for each individual file scan.
21:30:19drakonik=\
21:30:34Slasherianyway, you might try to look at the end of the file with a hexedit to get a clue
21:30:46drakonikProbably won't be very fast, but it'll work.
21:30:49drakonikWhere's the file?
21:30:53gevaertsdrakonik: it sounds much more sensible to me to try to find out why it crashes
21:30:59drakonikgevaerts: I have tried.
21:31:05drakonikExtensively.
21:31:10 Join JdGordroid [0] (n=jdg@m4b0436d0.tmodns.net)
21:31:20drakonikThere's just nothing conclusive.
21:31:37Slasheridrakonik: rockbox/database_tmp.tcd
21:31:46drakonikHm.
21:31:50drakonikNo such file.
21:31:52Slasheriit's a binary file
21:33:02drakonikLet's see if I can MAKE it by initializing my DB
21:33:22Slasheriyou have to look for it immediately after the crash. otherwise that file will vanish
21:33:27drakonikright
21:33:35drakonikworkin' on it
21:33:58Slasheriand by using the bootloader disk mode (before rockbox boots up and removes the file)
21:34:16drakonikAha
21:34:39drakonikIs there an upperlimit on how many files the scanner can handle, maybe?
21:35:03Slasherinot any absolute limit
21:35:28drakonikI've used the simulator, and when I tried to scan my DB, in the db scan debug screen, it mentioned something about too many items or something. tbh, it's been maybe six months since I actually did it so I can't remember precisely what it said.
21:35:33Slashericrashes during database scan are most likely caused by a bad file (or bad tags)
21:35:42drakonikSee, I thought that
21:35:48drakonikBut I eliminated that.
21:36:05drakonikUnfortunately.
21:36:05Slasherihow did you eliminate that?
21:36:25drakonikBreaking down my music collection. Taking half of it out of the scan, seeing if it still crashed,
21:36:46drakonikAnd halving each half
21:37:04Slasheriand finally having just a single file and still a crash?
21:37:44drakonikThat was the problem. I used the db scanner debug window, and made note of what file caused the crash. And it's completely random.
21:37:46Slasherithen i would suggest checking the file system for errors
21:37:50drakonikI have.
21:38:06Slasheriokay, that's bad
21:38:10drakonikYeah
21:38:18drakonikSomeone sugested that it was a hardware problem.
21:38:20Slasherimight be something wrong with you hardware, cpu, memory and so on
21:38:22drakonikAnd I understand why.
21:38:48drakonikBut I'm inclined to believe that if there were hardware issues, my ipod wouldn't work flawlessly, the way it has until I started trying to use rockbox.
21:38:50Slasheriindeed, i would thought a hardware problem also if the crashing is completely random
21:39:49drakonikkay, database_tmp.tcd is 0kb
21:40:01drakonikI realy don't think I even need to bother opening it.
21:40:44pixelmawhat type of files are those and how are they tagged?
21:40:49drakonikmp3s
21:41:14 Join bmbl [0] (n=Miranda@unaffiliated/bmbl)
21:41:18drakonikAnd...normally? I dunno, how many ways are there to tag an mp3?
21:41:22 Join archivator [0] (n=archivat@77.70.28.57)
21:41:33drakonikWait. What's nvram.bin?
21:41:50drakonikI don't think I remember seeing that a moment ago.
21:42:00archivatorHey all, Is there an alternate keymap for the simulator that doesn't use the numpad?
21:42:45domonokynvram.bin is just a file to store some settings, nothing todo with database.
21:42:46 Part Grahack
21:42:49pixelmawell, e.g. Rockbox only supports only the "native" tagging format which is ID3 for mp3, some PC programs allow Ape tags too
21:42:52drakonikOkay.
21:42:57drakonikAh.
21:43:03drakonikNope. IDv3 all the way
21:43:15domonokythe 0kb database_tmp file is probably because the file wasnt closed correctly at the crash.
21:43:22drakonikRight.
21:43:29drakonikI gathered that much.
21:43:49drakonikI guess all I can do is get into the source and see what I can see.
21:43:57 Join froggyman [0] (n=chatzill@pool-72-69-75-180.chi01.dsl-w.verizon.net)
21:44:21 Join kugel [0] (i=kugel@rockbox/developer/kugel)
21:44:25domonokyyes, a try with opening/closing the temp file beween scanned files, might help to find out more.
21:45:12drakonikOff the top of your head, does anyone know what file/function I should look for? I can find it fine on my own, I just wanna save some time
21:45:36 Quit JohannesSM64 ("Leaving")
21:46:28 Quit bmbl (Client Quit)
21:46:34 Join KBH [0] (n=hbk@rrcs-97-77-51-170.sw.biz.rr.com)
21:49:47archivatorby the way, does anyone else use an alternate keymap? I'm using opti.kbd but I seem to have found a bug - the first time the virtual keyboard is displayed, everything's fine. The second time, however, it's only 3 rows (as opposed to 5) and the line spacing is all messed up. Am I doing something wrong or is it a bug?
21:50:51n1sarchivator: sounds like a bug, i don't think it's a very commonly used feature
21:51:08domonokydrakonik: apps/tagcache.c/h
21:51:20drakonikThank you
21:51:56 Quit stripwax_ ("http://miranda-im.org")
21:52:41 Quit n1s ("Lämnar")
21:53:09drakonikText flowchart. I like it.
21:53:16domonokylooks add_tagcache() is the recursive function which adds the entrys to the tempfile. also enabling the logging at the top of this function might help to find out where it stops.
21:54:05drakonikurgh
21:54:14drakonikmy C is way rustier than it has any business being
21:54:46drakonikI guess line 1660 is the function definition?
21:55:35domonokyyes
21:55:40drakonikky, I got it
21:55:42drakonikThanks.
21:57:20 Join safetydan [0] (n=deverton@rockbox/developer/safetydan)
21:59:10 Quit LambdaCalculus37 ()
21:59:13archivatorWhere can I find the keymaps for the simulator?
22:00
22:00:23domonokyuisimulator\sdl\button.c
22:00:46*domonoky opens beer :-)
22:01:49archivatorThe resident file-finder strikes again :)
22:02:25 Join killan [0] (n=nnscript@c-0efa70d5.06-397-67626721.cust.bredbandsbolaget.se)
22:02:27 Quit liar (Read error: 113 (No route to host))
22:04:48drakonikHm.
22:05:26 Quit HBK (Read error: 110 (Connection timed out))
22:07:25bertrikbughunter2, could you try another patch?
22:07:55bughunter2bertrik: roger that
22:08:01 Join AsaelReiter [0] (n=d44c6ebe@gateway/web/cgi-irc/labb.contactor.se/x-xjvtjttelexjwxme)
22:08:25bertrikbughunter2, can you handle a source code patch or do you need a binary?
22:08:43bughunter2atm. a binary would be best (sorry for the inconvenience)
22:09:33AsaelReiterbetrik: will a test on Clip help you?
22:09:51bertrikAsaelReiter, no
22:10:11drakonikOkay. I'm not seeing anything that resembles a "close temp db file" in this function. =\
22:10:33drakonikI see where I should put it. But I don't know where the file is being opened or where it's being closed.
22:10:49domonokydrakonik: you wont find one there. its done at the place where this function is called.
22:10:56drakonikI am looking there.
22:11:13drakonikI admit, my C-fu is pretty weak, so I might just be missing it.
22:11:20drakonikaround line 4170
22:11:38domonokydrakonik: but you could also try to enable the logging (ifdef for sim at moment) to find out more.
22:11:48drakonikAlright.
22:11:49 Quit Thundercloud (Read error: 54 (Connection reset by peer))
22:12:35drakonikWait. This just write the file being scanned, doesn't it?
22:13:05domonokydrakonik: yes, so you can findout where it fails.
22:13:15drakonikAh
22:13:21domonokytagcache_build() contains the opening and closing of the temp file.
22:13:22drakonikYeah.
22:13:32drakonikThis is pretty useless. I've used that logger before.
22:13:44drakonikAnd it never stops at a consistent file.
22:14:21 Quit stoffel (Read error: 104 (Connection reset by peer))
22:14:27domonokyif it stops, does it completly crash, or what does exactly happen ?
22:14:38bertrikbughunter2, this is the binary http://filebin.ca/amkevz/rockbox.sansa
22:14:40drakonikMy ipod screen flashes white and then powers down completely.
22:15:14bughunter2bertrik: what should I look out for?
22:15:36drakonikInteresting, but unrelated, is the fact that when it reboots, evne if the battery indicator showed 100% charge before the crash, after the crash, it invariably shows 0 charge and very quickly shuts itself down.
22:16:07domonokydrakonik: and it also crashes with only one file ?
22:16:40drakonikI never quite got it down to that point.
22:16:51bughunter2bertrik: anyway, it runs...
22:17:01drakonikBecause of that battery issue, I'd have to charge it for an hour or two after a crash.
22:18:26*domonoky wonders why the battery would be emtpy from scanning only a few files. sounds strange.
22:18:33 Join TopyMobile_ [0] (n=topy@chello084010205095.chello.pl)
22:20:20bughunter2bertrik: I've just checked out the sources in a VM, compiling them now with gcc 3.3.5
22:22:09JohnTeddyHow can I skip through lists more quickly?
22:22:36 Quit killan (Read error: 104 (Connection reset by peer))
22:22:49***Saving seen data "./dancer.seen"
22:22:56 Join Thundercloud [0] (i=thunderc@persistence.flat.devzero.co.uk)
22:23:59 Quit JdGordroid (Remote closed the connection)
22:24:07Slasheridrakonik: hmm, maybe the battery is finished? the db scanning is very hard for the battery
22:24:24drakonikI really doubt it
22:24:29drakonikIt's got fine battery life
22:24:46 Quit AsaelReiter ("CGI:IRC")
22:24:46drakonikBut it's possible
22:24:48drakonikI dunno man.
22:27:18drakonikSo if I understand this code...tagcache_build() is called on bootup or when a scan is completed. And I guess the argument is where to find the database_tcd.tmp file...
22:28:35domonokyno tagcache_build scans for files, the parameter is the startdir..
22:28:38Slasheridrakonik: the argument tells from what directory to start scanning
22:28:40drakonikAha
22:28:52drakonikJust shows how bad my C is.
22:28:54drakonikAlrighty.
22:29:13domonokyit may also commit the changes. but that depends on different things..
22:29:38Slasheritagcache_build will create the temp file, and when it's proper time, also commit that file into a database
22:30:06domonokyits tagcache_build() -> check_dir() -> add_tagcache() for the scanning process.
22:30:17drakonikGot it.
22:31:15bughunter2bertrik: still there?
22:35:53drakonikWelp,this is pretty much beyond my ken.
22:41:47 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
22:42:33 Join stoffel [0] (n=quassel@p57B4CEA1.dip.t-dialin.net)
22:47:10 Quit gregzx ("ChatZilla 0.9.85 [Firefox 3.5.2/20090729225027]")
22:48:05bughunter2bertrik: what should I test? :)
22:48:23bertrikjust see if the buttons/wheel still work
22:48:27bughunter2they do
22:49:13bughunter2btw, when I said scrolling really really fast didn't work too well, that's when I put the player on my desk and gave the wheel a quick spin
22:49:22bughunter2with normal use it's fine I guess
22:54:17 Join killan [0] (n=nnscript@c-0efa70d5.06-397-67626721.cust.bredbandsbolaget.se)
22:59:15bughunter2bertrik: if you have anything else, I might help testing tomorrow :)
22:59:16bughunter2night all
22:59:32bertrikthanks so far
22:59:38bughunter2np :)
22:59:53 Quit bughunter2 ("Leaving.")
23:00
23:00:49 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
23:05:06 Quit panni__ (Client Quit)
23:05:50 Join panni_ [0] (i=hannes@ip-95-222-19-21.unitymediagroup.de)
23:08:47 Join liar [0] (n=liar@83.175.83.185)
23:11:09 Quit merbanan (Read error: 110 (Connection timed out))
23:11:35 Join BlakeJohnson86 [0] (n=bjohnson@c-24-118-162-123.hsd1.mn.comcast.net)
23:15:17 Join Zagor [242] (n=bjst@46.35.227.87.static.tab.siw.siwnet.net)
23:16:30 Quit Eosphere46 (Read error: 104 (Connection reset by peer))
23:18:27 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
23:20:31 Quit archivator ("Leaving")
23:23:38 Quit bluebrother ("leaving")
23:26:26 Quit stoffel (Remote closed the connection)
23:26:56 Join FOAD_ [0] (n=dok@dinah.blub.net)
23:32:00 Quit bertrik ("De groeten")
23:32:15 Quit Eosphere46 (Read error: 54 (Connection reset by peer))
23:35:23 Join Eosphere46 [0] (i=Eosphere@74.197.221.60)
23:36:32 Quit stripwax ("http://miranda-im.org")
23:36:58 Join stripwax [0] (n=Miranda@87-194-34-169.bethere.co.uk)
23:40:13 Quit Eosphere46 ("It's a dud! It's a dud! It's a du...")
23:42:15 Quit FOAD (Read error: 110 (Connection timed out))
23:42:16 Nick FOAD_ is now known as FOAD (n=dok@dinah.blub.net)
23:44:32 Quit stripwax ("http://miranda-im.org")
23:45:38 Quit petur ("Zzzzz")
23:58:20 Quit domonoky (Read error: 104 (Connection reset by peer))

Previous day | Next day