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 2012-10-30

00:01:12 Join ender1 [0] (
00:05:59 Join factor [0] (
00:09:42 Join webguest93 [0] (
00:10:37 Quit webguest93 (Client Quit)
00:22:35 Quit Joakim (Quit: Joakim)
00:22:47 Quit guymann (Quit: brbbrbrb)
00:25:48 Join guymann [0] (~c@unaffiliated/guymann)
00:26:50 Quit Wardo (Read error: Connection reset by peer)
00:27:52 Quit lebellium (Quit: ChatZilla 0.9.89 [Firefox 17.0/20121023124120])
00:32:50 Quit mirak_ (Quit: Ex-Chat)
00:36:06tjb0607so I'm having some issues with ID3 tags/vorbis comments to include album art with easytag
00:36:31tjb0607basically, I do everything right and everything shows up right, but when I try it on Rockbox it doesn't show up
00:39:02tjb0607all the other albums I download from Amazon have ID3 artwork working perfectly
00:40:48amayerbluebrother^: do you know php?
00:43:08amayerscorche`: could you approve this since you already tested it on your server?
00:43:37amayeror "review" it. not sure what the appropriate term is
00:51:09 Quit ender1 (Quit: The best way to accelerate a Macintosh is at 9.8m/sec┬▓)
00:54:00 Quit KiwiCAM_ (Remote host closed the connection)
00:57:41 Quit Totalled (Quit: PETTAN PETTAN,´╗┐ TSURUPETTAN!)
00:58:53 Join KiwiCam [0] (~quassel@
01:00:48KiwiCamtjb0607: Have you tried Puddletag?
01:06:07tjb0607everything shows up right in there too
01:07:19***Saving seen data "./dancer.seen"
01:12:39 Join Gallomimia [0] (
01:30:44 Join Totalled [0] (
01:45:07amayerbluebrother^: ping
01:57:11 Quit Totalled (Ping timeout: 252 seconds)
02:00:59 Join Totalled [0] (
02:11:38amayerAlexP: do you know php to approve something for the themesite?
02:12:32amayerTorne: ping
02:44:28 Quit guymann (Ping timeout: 264 seconds)
02:45:17 Join guymann [0] (~c@unaffiliated/guymann)
02:53:37 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
03:07:22***Saving seen data "./dancer.seen"
03:28:32 Quit dfkt (Quit: -= SysReset 2.55=- Sic gorgiamus allos subjectatos nunc.)
03:41:19 Quit Gallomimia (Quit: Gallomimia)
03:48:22 Join Scr0mple [0] (~Simon@
03:52:25 Quit Scromple (Ping timeout: 272 seconds)
04:02:45 Quit PathosesNorms (Ping timeout: 244 seconds)
04:25:39 Join TheSphinX^ [0] (
04:25:55 Join [Saint] [0] (~quassel@rockbox/user/saint)
04:27:00 Quit TheSphinX_ (Read error: Operation timed out)
04:37:56 Nick [Saint] is now known as [Saint_] (~quassel@rockbox/user/saint)
04:38:02 Nick [Saint_] is now known as [Saint__] (~quassel@rockbox/user/saint)
04:38:10 Nick [Saint__] is now known as [Saint___] (~quassel@rockbox/user/saint)
04:38:17 Nick [Saint___] is now known as [Sinner] (~quassel@rockbox/user/saint)
04:38:17DBUGEnqueued KICK [Sinner]
04:38:21 Nick [Sinner] is now known as [Sinner_] (~quassel@rockbox/user/saint)
04:38:21DBUGEnqueued KICK [Sinner_]
04:38:57 Nick [Sinner_] is now known as [Sinner__] (~quassel@rockbox/user/saint)
04:38:57DBUGEnqueued KICK [Sinner__]
04:38:57***Alert Mode level 1
04:39:16 Nick [Sinner__] is now known as [Sinner___] (~quassel@rockbox/user/saint)
04:39:16DBUGEnqueued KICK [Sinner___]
04:39:16***Alert Mode level 2
04:39:23 Nick [Sinner___] is now known as sinner (~quassel@rockbox/user/saint)
04:39:23DBUGEnqueued KICK sinner
04:39:23***Alert Mode level 3
04:39:33 Nick sinner is now known as [Saint] (~quassel@rockbox/user/saint)
04:39:33DBUGEnqueued KICK [Saint]
04:39:33***Alert Mode level 4
04:48:50 Quit amiconn (Disconnected by services)
04:48:50 Join amiconn_ [0] (amiconn@rockbox/developer/amiconn)
04:48:51 Quit pixelma (Disconnected by services)
04:48:51 Join pixelma_ [0] (pixelma@rockbox/staff/pixelma)
04:48:53 Nick amiconn_ is now known as amiconn (amiconn@rockbox/developer/amiconn)
04:48:53 Nick pixelma_ is now known as pixelma (pixelma@rockbox/staff/pixelma)
04:49:34***Alert Mode OFF
04:58:05amayer[Saint]: i forget if i asked you. do you know php(enough to review it)?
04:59:19JdGordonamayer: what are your plans for the theme site? if you just want to do fixups then i think we should go about getting you commit access
04:59:43JdGordonif its more then you really should just fork it, put up your finished site and them switch
04:59:48amayerJdGordon: im game for that
05:00:07amayeri am confused about the forking tho
05:00:16amayerfor now its just minor things here and there
05:00:34amayerim not like rebuilding the way it works just making it more efficent
05:00:58amayerim guessing a fork is just making a new branch in git?
05:01:16 Join Rower85 [0] (
05:01:31JdGordonexactly what you've done for the gerrit review... thats a fork
05:01:53JdGordonbut i mean setting it up on a server you own and then taking it as far as you want and seeing what it looks like before merging it back
05:01:53amayer*totaly confused*
05:01:55amayerthat is all in the master branch
05:02:36amayerJdGordon: i already do that. not a public server but i have a dedicated box(ok its a laptop) just for the theme site
05:03:29amayerso how do i go about getting commit access?
05:03:59JdGordonI've sent an email to the secret brotherhood of faceless men
05:04:50amayerahhh... do i get to wear a cloak after i get commit access?
05:04:52amayerand program in the dark with toruches burning in the background
05:06:27JdGordonyou dont *have* to... but its recommended
05:07:07amayer*arm pump* YES!!!!
05:07:26***Saving seen data "./dancer.seen"
05:10:25CtcpIgnored 6 channel CTCP requests in 1 minute and 16 seconds at the last flood
05:10:25*scorche` hands out the cloaks
05:13:16 Quit sentriz (Quit: Leaving)
05:15:06 Join Galois [0] (
05:15:46 Nick scorche` is now known as scorche (~scorche@rockbox/administrator/scorche)
05:28:40 Quit guymann (Quit: brb irssi bork again)
05:29:43 Join guymann [0] (~c@unaffiliated/guymann)
05:32:08 Quit [7] (Disconnected by services)
05:32:16 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:07:41 Part amayer
06:15:46 Join mortalis [0] (
06:33:08 Join sentriz [0] (~Senan@
06:35:26 Quit Totalled (Quit: PETTAN PETTAN,´╗┐ TSURUPETTAN!)
06:35:41 Join Totalled [0] (
06:40:14 Quit alexbobp (Ping timeout: 240 seconds)
06:41:11 Join alexbobp [0] (
07:03:59 Quit sentriz (Ping timeout: 252 seconds)
07:07:27***Saving seen data "./dancer.seen"
07:23:07 Join stoffel [0] (
07:26:11 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
07:29:12 Quit scorche (Disconnected by services)
07:29:16 Join scorche` [0] (~scorche@rockbox/administrator/scorche)
07:32:10 Join xvat [0] (
07:32:23 Join akaWolf [0] (~akaWolf@unaffiliated/akawolf)
07:33:46xvatFor the Sansa Clip+ chipset, what is the highest capacity microsd card it can handle?
07:35:12*[Saint] wonders what the chipset has to do with this
07:35:35[Saint]64GB, anyway (works with Rockbox, idea about the OF, never used it)
07:35:46 Join derk0pf [0] (
07:43:47 Join Keripo [0] (
07:54:14 Quit Keripo (Ping timeout: 240 seconds)
07:58:27 Quit advcomp2019__ (Read error: Connection reset by peer)
07:58:50 Join advcomp2019__ [0] (
07:58:50 Quit advcomp2019__ (Changing host)
07:58:50 Join advcomp2019__ [0] (~advcomp20@unaffiliated/advcomp2019)
07:59:00amiconn[Saint]: Technically nothing, but due to the weirdness in the SD standard that specifies the filesystem to use, SDXC won't work out of the box (due to being ExFAT formatted)
07:59:22 Join Keripo [0] (
08:06:05 Quit bertrik (Ping timeout: 264 seconds)
08:06:27 Quit Keripo (Read error: Connection reset by peer)
08:11:20 Join Zagor [0] (
08:11:20 Quit Zagor (Changing host)
08:11:20 Join Zagor [242] (~bjst@rockbox/developer/Zagor)
08:19:24 Join Keripo [0] (
08:20:21 Quit [Saint] (Quit: Quit)
08:30:42 Join LinusN [0] (
08:32:08 Join Joakim [0] (
08:33:00 Join ender1 [0] (
08:35:44 Join [Saint] [0] (~quassel@rockbox/user/saint)
08:47:09 Quit Joakim (Quit: Joakim)
08:49:43 Quit ender` (Ping timeout: 246 seconds)
08:50:10 Join ender` [0] (~ender1@2a01:260:4094:1:42:42:42:42)
09:07:28***Saving seen data "./dancer.seen"
09:18:10 Join petur [0] (~petur@rockbox/developer/petur)
09:25:34 Part LinusN
09:30:14 Join pamaury [0] (
09:30:14 Quit pamaury (Changing host)
09:30:14 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
09:37:59 Quit pamaury (Ping timeout: 260 seconds)
10:00:14 Quit derk0pf (Ping timeout: 240 seconds)
10:01:07 Join derk0pf [0] (
10:14:06 Join pamaury [0] (
10:14:07 Quit pamaury (Changing host)
10:14:07 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
10:26:38 Join Joakim [0] (
10:31:22 Quit tchan (Ping timeout: 255 seconds)
10:41:41*[Saint] wonders what to do with the metadata in cabbie for targets that can display 6-7-8+ lines of text (+2-3-4 lines of text in the "no album art" case)
10:43:08[Saint]In the changes I have in my tree at the moment, I've fixed up the whole 'always expect there to be a next track' thing, and that leaves a whole lot of room for additional data (if it's wanted) on some targets.
10:51:06 Quit tjb0607 (Ping timeout: 265 seconds)
10:55:19[Saint]JdGordon: you around?
10:56:53 Join einhirn [0] (~Miranda@2001:638:605:4:1c8e:15bb:86b0:8d30)
10:59:30 Join tjb0607 [0] (~tjb0607@
11:02:37 Join PathosesNorms [0] (
11:02:38 Join LinusN [0] (
11:07:29***Saving seen data "./dancer.seen"
11:13:55JdGordon[Saint]: ?
11:16:52JdGordonget me tomorrow
11:34:18 Quit einhirn (Ping timeout: 260 seconds)
11:34:54 Quit Provel_ (Read error: Connection reset by peer)
11:35:19 Join Provel_ [0] (
11:43:26kugel[Saint]: I wouldnt put too much information
11:43:54kugelalso don't change too much in one go, incremental changes are better
11:44:23[Saint]most of it is purely cosmetic code changes
11:44:48[Saint]and the rest is just making the various cabbies more similar with each other
11:45:24[Saint]there's some weird, completely unnecessary, differentiation between some of the cabbie ports that I'm ironing out.
11:46:55 Quit bootinfdsds (Read error: Connection reset by peer)
11:47:15[Saint]ahhhhhhhh, bum.
11:47:19[Saint]I missed Jd.
11:54:13[Saint]Anyway, yeah. I haven't actually changed the layout of any of the cabbies at all. I've just unified the metadata between the ports, and made the different ports code as similar to each other as possible.
11:54:40[Saint]and updated old syntax to new syntax in a few places.
11:55:32[Saint]But,'s pretty much all "behind the scenes" cosmetic stuff that I hope will make cabbie a better teaching tool.
12:02:13 Join wodz [0] (
12:02:50pixelmaI always wanted to have the c200 version use a 10 pixel font instead of 11 so that on the WPS there could be one more row and hence having space for the playing time info line all the other ports have. Since using my own WPS on the c200 though, motivation wasn't very high
12:03:29wodzamiconn: ping
12:06:54[Saint]pixelma: does a suitable 10px font exist?
12:07:34[Saint]I have made a set of anti-aliased fonts from GNU Unifont that I was considering useng.
12:07:38pixelmait's currently using 11-Sazanami-Mincho and we also have a 10-pix version of that
12:08:23[Saint]although the aa fonts are large (and GNU Unifont is ugly) I think that the default theme that Rockbox ships with should "just work" for any language.
12:08:42[Saint]I haven't made the change yet, but I was going to bring it up on the ml
12:08:44 Quit bluebrother^ (Read error: Operation timed out)
12:09:04 Quit fs-bluebot (Ping timeout: 260 seconds)
12:09:04xvat[Saint]: the OF? I was asking what the Sansa Clip+ supported in capacity for sd memory, but inserted chipset to also alert the answerer that I am interested if there are theoretical limits in the hardware that are different to that of the rockbox firmware
12:09:07pixelmahard to make it work for the smaller screens though
12:09:09xvat64gb soudns good
12:09:24 Quit factor (Read error: Connection reset by peer)
12:09:47[Saint]well, there's no limit to the Rockbox firmware.
12:10:08[Saint]it is entirely limited by the hardware.
12:10:28 Join fs-bluebot [0] (
12:10:45Tornethe limit for SD cards is 2TB
12:10:49[Saint]I have no idea if the OF can support a 64GB card or not, so dual-boot might get interesting...but I know that Rockbox can.
12:11:21Torneif you somehow get hold of a 2TB card and it doesnt' work in rockbox that's a bug and if you let us know we can probably fix it :p
12:11:21xvatOF = original firmware
12:11:38xvatI haven't thought of using more than one firmware, i was set on using only rockbox
12:11:40[Saint]Torne: Bwahahaha ;)
12:11:43xvati don't see why one would do otherwise
12:12:22 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
12:12:50Tornethere isn't really a hardware limit, the controller does whatever we tell it to.
12:13:02[Saint]Subscription/paid content, is one that I can think of.
12:13:12Torneuntil someone actually makes a kind of SD card that isn't physically/electrically backward compatible it will work if we implement the code right :)
12:13:15[Saint](wrt: reasons to use the OF)
12:14:08[Saint]Torne: well, sorry...I wasn't very clear, but by "hardware" I was also including the card itself.
12:14:34 Join T44 [0] (
12:14:39[Saint]that was non-obvious to third parties though, apologies.
12:14:51xvat[Saint]: ah, afaik Sansa Clip+ has none of that
12:14:53Tornethat doesn't make sense, though
12:15:29Tornethere's no actual reason why you couldn't make an SD card that was a trillion terabytes, or whatever, and still have it be compatible with existing reader hardware :p
12:15:38Torneyou'd need to invent a pretty entertaining variant on the SD interface spec..
12:15:42Torneand write some funky drivers :p
12:15:53Torneand you might not be very happy with the performance.
12:17:03wodzwho cares about performance when handing trillion terabytes SD :-)
12:18:02xvatwhy not develop an open source digital audio player with built-in asic/fpga codecs for lame/vorbis/etc?
12:18:24 Quit Topy (Ping timeout: 260 seconds)
12:18:35Tornexvat: largely pointless
12:18:41wodzxvat: because its silly?
12:18:49Tornemost current codecs are very easy to decode on modern CPU arches
12:18:58wodzwhy not use off the shelf modern SoC?
12:19:04Torneusing dedicated decoder hardware is very oldfashioned and way more expensive
12:19:13Torneand much less flexible when new codecs show up
12:20:35xvatTorne: the rockbox developers do not feel limited by what hardware manufacturers put on the market?
12:20:39[Saint]see also the various open source music player projects that have dies a slow and expensive death in our wiki
12:21:18Tornexvat: not raelly, no
12:21:31Tornewhat's limiting about them?
12:21:53Tornesomething made in huge volume by a consumer manufacturer is going to be cheaper and better-built than anything we can arrange to ahve made
12:22:19Tornethere isn't really anything rockbox can do that current players have a problem with
12:22:23Torneso, what's the point? :)
12:22:32[Saint]It's just way cheaper to work with existing hardware, and most of the time these players have more than enough "guts" to play damn near anything (even foolish audiophile quality audio).
12:22:34wodzfor open hardware DAP project there are two limiting factors 1) cost 2) mounting technology will most users can't handle
12:23:13[Saint]s/play damn near anything/play damn near anything in real time/
12:23:18[Saint](missed an important bit)
12:23:22Torneaudio processing doesn't really cost any more processing power now than it did a long time ago, but CPUs are way, way more powerful and efficient
12:23:32*[Saint] nods
12:24:48 Quit stoffel (Ping timeout: 252 seconds)
12:26:24xvatfor US people
12:26:43wodzAs a side note - In front of me there is R4600 CPU ripped off INDY SGI workstation. It has about the same fixed-point performance as ATJ213x SoC.
12:26:46Tornethe whole Clip family is super awesome. :)
12:26:54Torne*so* tiny and cheap
12:27:26 Join factor [0] (
12:27:28copperI bought a refurbished Clip+ recently - its screen went very dim a few days after delivery
12:28:07xvatbut it seems they break Torne
12:28:13xvatshould i buy a soldering kit just for this
12:28:23Tornenone of mine are broken
12:28:33Tornenever had a problem
12:28:35xvathow do you take care of them
12:28:41Tornenot at all
12:28:46Tornethrow them in the bottom of my bag or wahtever.
12:28:54Tornethey're solid lumps of plastic
12:29:02Torneit's about as indestructible as anything else i own
12:29:13xvatthe earphone jack seems to be a weakspot
12:29:15xvatand FM tuning
12:29:24xvat"I purchased a silicone skin protector and put on a screen film to protect against scratching and thought I was good to go for a long time."
12:29:36Tornenot sure what FM tuning has to do with anything..
12:29:43Tornethe jack is fine in mine
12:29:56Tornei don't leave the headphones plugged in when i throw it in somewhere
12:30:07Torneso, there's rarely any forces applied to the socket.
12:30:13Tornei can see that breaking it maybe?
12:30:47Tornewhen i'm using it nothing is likely to happen that pushes on the headphone plug enough to do any damage
12:30:56Torneand when i'm not using it nothing is plugged in, so the socket can't really get broken
12:31:47xvatTorne: So what is your favored of the Sansa Clip family
12:32:00*funman pushes people to #rockbox-community
12:32:06wodzusually headphone socket gets broken on insertion of the jack
12:34:14wodzmortalis: ping
12:34:22mortaliswodz: pong
12:34:23 Quit Keripo (Quit: Leaving.)
12:35:31wodzmortalis: I was thinking about cache problem and I see one race condition in previous code. If interrupt fires during cache invalidation it may go nuts
12:35:46wodzmaybe thats the reason of the problems
12:36:52 Join Nux|Vomica [0] (
12:37:54Tornethere's a way for interrupts to happen during our cache invalidation?
12:37:57Tornethat would be bad, yes :)
12:38:49mortaliswodz: I have the same thought after I saw g339
12:38:51fs-bluebotGerrit review #339 at : Fix FS #12391 : Memory corruption on PP502x after commit_discard_idcache() by Boris Gjenero (changes/39/339/1)
12:38:57wodzTorne: in rk27xx cache controller is peripheral like anything else.
12:39:12wodzmortalis: heh, exactly the same here
12:40:04Tornewodz: right, then you need to disable interrupts while you wait for it, or else prove that no interrupt handler can ever touch the cachelines you care about :p
12:40:43wodzor disable caching
12:41:09Tornedepends on the semantics of disabling the cache
12:41:27wodzI mean disable cache; invalidate; enable cache
12:41:46Torneright, but that doesn't actually work on all cache implementations :)
12:42:08Torneit's possible you still read from cache while it's disabled
12:42:14Tornejust, no linefill operations take place
12:42:19Torneor other weird semantics :/
12:42:39Tornemost arches are not designed to ever have their cache turned off again for anything other than preparation for a reset/chainload of different binary/etc
12:42:46Torneso sometimes the behaviour is weird
12:43:09wodzSo you mean that it is actually safer to disable interrupts in wait loop of invalidate?
12:43:23Torneyes, that's what i'd suggest
12:43:40Tornealso, even if disabling the cache works for an invalidate, it doesn't for a clean
12:44:24wodzrk27xx cache doesn't distinguish this two - it has one commit and invalidate opcode AFAIK
12:44:37Tornethen you absolutely can't disable the cache to do it
12:44:44Nux|VomicaHi, i have a question about the compatibility of Rockbox with external media players over usb; In detail, its my Denon AV-Receiver, that has an USB Port where music can be played from USB Devices. It runs well with standard thumb drives, but for some reason cannot detect my IPod Mini 2g with Rockbox on it. Is there any way I could enable this? (The Denon also has a direct ipod control feature, but since I dont use the IPod defa
12:44:44Nux|Vomicault firmware, this feature is useless anyway.)
12:44:47Tornebecause anything that touches a dirty cacheline while it was disabled may read the stale data from ram
12:44:51Torneinstead of the current data from teh cache :)
12:44:54Torneso, yeah, disable interrupts.
12:45:12wodzTorne: ok, thanks for clarification
12:45:24wodzmortalis: could you test this on hifiman?
12:45:39TorneNux|Vomica: this usually happens because it recognises that it's an ipod by the USB vendor/product ID and then tries to communicate with it using the ipod accessory protocol, or to read the itunes database, or something else we don't undertand
12:45:40Nux|VomicaI read a bit about it, this might come from the "unformatted" 39.22MB part at the beginning of my IPods drive; or it might be that the denon is just not able to detect partitions other than the first one.
12:45:50TorneNux|Vomica: that;s unlikely.
12:46:10Torneyou will probably find if you patch rockbox to claim a different USB vendor/product that's not an ipod it will just work
12:46:18Tornebecause it'll treat it like any old usb device
12:46:24mortaliswodz: sure
12:46:45Tornewe've had this come up a bunch of times before; someone should really work out a nice way to solve this with an option :/
12:47:04Tornebut we don't have a usb id of our own to use.
12:47:47wodzhas anyone get in touch with openmoko to get vids:pids for our own?
12:47:49Tornewodz: hah, in fact yes, this means you can never disable the cache without having first disabled interrupts and cleaned the whole thing anyway :)
12:47:58Tornewodz: so yeah.
12:48:19Tornei don't think anyone has, no
12:48:33wodzI guess this would be the first step
12:48:42wodzsetting bit is easy
12:50:51 Quit zoktar (Ping timeout: 245 seconds)
12:51:11 Quit wodz (Quit: Leaving)
12:52:14 Quit Nux|Vomica (Ping timeout: 260 seconds)
12:59:52 Quit petur (Quit: *plop*)
13:01:12 Join Nux|Vomica [0] (
13:06:18 Join wodz [0] (
13:07:32***Saving seen data "./dancer.seen"
13:07:43mortaliswodz: works on 801
13:08:50 Quit Nux|Vomica (Ping timeout: 256 seconds)
13:09:27wodzmortalis: I believe Torne in this area so we should probably commit this
13:10:26Tornei assume you're invalidating the entire cache, here, btw?
13:10:43Torneok, just checking
13:11:04 Join Nux|Vomica [0] (
13:11:11wodzDon't we need to disable interrupts for invalidate_range also?
13:11:32Torneit depends how that's implemented
13:11:43Torneis the underlying cache op "invalidate by address"?
13:11:54Torneand you just call it once for each cache line in the range
13:12:25Torneif you can prove that interrupt handlers will never touch any of the cache lines you are invalidating, then it's fine
13:12:34Torneotherwise, you need to turn off interrupts there too
13:13:26Tornewill look later, need to go
13:17:02 Quit Nux|Vomica (Ping timeout: 276 seconds)
13:22:07 Join dfkt [0] (dfkt@unaffiliated/dfkt)
13:22:48 Join zoktar [0] (~zoktar@unaffiliated/zoktar)
13:30:29 Join stoffel [0] (
13:35:49 Quit derk0pf ()
13:37:26 Join michaelr [0] (
13:39:08michaelrOkay, so I did what you guys told me to do, changed the english.lang file, changed "Database" to "Library" and when I tested it on the simulator, it still said "Database" on the root menu.
13:42:02wodzdid you rebuilt simulator or just changed the file?
13:42:49michaelrI downloaded a new one. I also did "../tools/configure" and then make and then make zip and used that zip
13:44:35wodzjust checking: 1) cloned repo 2) changed lang file 3) ../tools/configure && make && make zip ?
13:45:17michaelrone sec
13:46:47michaelrWhat's cloning the repo? And I ran ../tools/configure, make, and make zip all as separate commands, right after one another.
13:47:20wodzthats basically the same as && it
13:47:47wodzso how did you get the source other then trough cloning our repo?
13:48:57michaelrI gib, I believe but that was a day ago, I didn't download a new one, should I?
13:50:45wodz"michaelr> I downloaded a new one." <- Please first describe your problem consistently then we can (probably) help you
13:50:47 Join amayer [0] (
13:51:04 Quit Joakim (Ping timeout: 244 seconds)
13:52:21michaelrSorry, are we talking about the source code? I used git clone to get the source a few days ago, now I'm editing it. I changed english.lang to change "Database" to "Library" on the root menu, but it didn't change.
13:54:28wodzdescribe the steps you did *in actual order*, then pastebin your change
13:55:39michaelrI edited language file, compiled the source with "make", then used "make zip" to create ~/rockbox/build/, when I used ".rockbox" on the simulator the changed to the language file did nothing.
13:56:35wodzshow your change
13:57:52michaelrwhat I changed in the language file?
13:59:55 Join yxo [0] (59a984ae@gateway/web/freenode/ip.
14:00:29wodzI just checked - it works - edit apps/lang/*.lang and rebuild
14:00:40michaelr<phrase> id: LANG_TAGCACHE desc: in the main menu and the settings menu user: core <soruce> *: "Library" </source> <dest> *: "Library" </dest> <voice> *: "Library"</voice> </phrase>
14:00:48michaelrThen I must have done something wrong...
14:00:54wodzaaa, you changed source entry
14:00:58wodzyou can't do that
14:01:07wodzchange dest and voice
14:01:18yxohi. could anyone helpme to unbrick my clip plus? i do all like in a manual, but player doesn't start.
14:01:23michaelrOh. The guy yesterday told me changing all of them won't hurt.
14:01:34*wodz checking
14:01:52michaelrYesterday around this time.
14:02:24wodzthat also works
14:04:16michaelrSo it wasn't my editing, I just did something silly wrong?
14:09:06michaelrAh. Okay, thanks. I'll try it here in a sec.
14:09:50 Join derk0pf [0] (
14:13:09 Quit derk0pf (Read error: Operation timed out)
14:18:46 Join tchan [0] (
14:18:47 Quit tchan (Changing host)
14:18:47 Join tchan [0] (~tchan@lunar-linux/developer/tchan)
14:22:48 Quit wodz (Quit: Leaving)
14:28:53Torneamayer: i can take a look at your php?
14:30:32michaelrI don't understand, it's not changing it in the simulator...
14:31:30amayerTorne: i looked at the first php change i submitted and noticed you were one of the people that approved it and i was wondering if you could review these:
14:31:44Torneamayer: yeah, i will have a look now.
14:31:53Tornei am not a php guru and i don't know a lot about the themesite#
14:31:59Tornebut i can probably help
14:32:44michaelrWhat do you need in PHP?
14:32:52amayerthe database class was already tested on the live themesite(as the description says)
14:33:20amayermichaelr: i just need people to review and approve my code so we can commit it to the themesite
14:34:55michaelrI'll check it out.
14:35:59amayermichaelr: see links ^
14:37:18michaelrThe code viewer skips a bunch of lines...
14:38:02amayermichaelr: it only shows you diff(aka the lines i change/added)
14:38:11Torneit shows you whatever you tell it to show you :)
14:38:20Tornepreferences -> context
14:39:17amayerTorne: what is the criteria for reviewing?
14:39:19amayerdo you just have to have an account or is that a privliage that has to be granted by the powers-that-be
14:39:28Torneanyone can review changes
14:39:42Tornebut only committers can give a score of +2 or -2, which allows/blocks the commit
14:39:42amayerok. ive never tried reviewing anyones code
14:39:45Torneeveryone else's scores are advisory only
14:40:04amayeroh ok
14:40:52Torneanywya i'm looking now but may not respond soon, rsi flaring up
14:43:40michaelrwhy not use empty instead of == ""
14:43:53 Part xvat
14:45:49amayermichaelr: which file what line?
14:47:25michaelrdb.class.php line 208
14:47:51michaelrAnd 216
14:48:25 Quit mortalis (Quit: Leaving)
14:50:38 Quit stoffel (Ping timeout: 260 seconds)
14:51:21amayeri dont fully understand the question. and i didnt make changes to those lines anyway.
14:51:23amayer(i assume you are talking about the right side(new file)
14:52:23michaelrYeah, okay, nevermind then. It just does the same thing.
14:53:31amayerTorne: did you leave a message with your review?
14:53:34amayeri cant find any an am confused why you gave it a -1
14:54:09amayernvm. i think i found the comments
14:54:51Tornethe code is pretty bad in general already; bad db usage :/
14:55:40amayeri was talking about the theme class. but i found your comments.
14:55:50Torneso probably you arent really making it worse from that pov
14:56:26Torneif you actually want to work on the therme site you may want to look at fixing the db stuff though.
14:56:40amayeras far as the db class goes i was just updating it so it would work on my server... i dont think i want to rewrite the whole thing right now(maybe in the near future)
14:57:01Tornenot the db class
14:57:14Tornethe entire approach to queries
14:57:20Tornesprintf is a terrible way to do it
14:58:17amayeri was just using the same stuff that was already being used(trying to follow the same style)
14:58:20amayeri kinda get what it does but ive never used that sprintf function before
14:58:43Tornethe specific function isn't important
14:58:49Torneusing string manupulation to build queries is just wrong
14:59:02Torneregardless of how you manipulate strings
15:01:43Tornemichaelr: thanks for looking, but "works on target" is something you would only set if you have actually tested the code in question
15:03:23Torne"target" refers to a rockbox target, which is not entirely appropraite as a review category for the themesite :)
15:03:30Tornejust for reference
15:04:44amayerTorne: i will make the changes that you commented inline and resubmit over my lunch break(about 2 hours from now)
15:05:14Tornebut seriously, if you actually want to work on the themesite code in general that would be great
15:05:31Tornebut you need to approach it from the perspective that what we have now is kinda actively bad and probably has all kinds of sql injection holes in it
15:05:36Torneand understand how to avoid this :)
15:06:37amayeri would like to work on the rockbox code. i have alot of ideas i would like to try and implement. ive never worked on something like this and it seems like alot of code to read before jumping in.
15:07:18amayeri dont mind working on the theme site and updating it to be more secure.
15:07:36***Saving seen data "./dancer.seen"
15:08:08Tornei'm not trying to discourage you
15:08:26Tornei also realise it's difficult because few people in the project know enough or care enough about php or the themesite to comment on your changes
15:08:31Tornei don't really know much about how the themesite code works
15:08:51Torneand i avoid php wher epossible, so can't exactly tell you the most current best practises for how to do this stuff
15:09:07Tornebut i know what things are very bad :p
15:16:19 Quit ender1 (Ping timeout: 244 seconds)
15:17:49amayerTorne: well im glad you took the time to review my code(even if it was a bad review) i helps me alot.
15:17:52amayerand i believe i will continue to work on the theme site.
15:18:10Tornesure, the -1 is largely directed at things like spelling
15:18:22Tornei'm not expecting you to actually replace all teh db code with parameterised queries
15:18:29Tornejust, do the best you reasonably can :)
15:18:48Tornebut if you want a *next* project then replacing all the db code with parameterised queries would be pretty rad
15:19:15Tornebtw, "rsi" == "RSI" == my hands hurt and i can't type much :p
15:19:38amayeri noticed... apperently i have a mental block when it comes to spelling "newest"
15:20:11amayerin my defence if you look at the time i posted it and apply it to EST it was like 1 in the morning on a work night that i commited them
15:20:57Torneyou don't have to defend yourself
15:21:11Tornereview comments are constructive criticism :)
15:24:03amayerTorne: can you give me an example of a parameterised query?
15:26:31amayerdo you just mean escape all user input that goes into the query?
15:26:34amayeri googled "parameterised query php" and i didnt find much
15:28:23 Quit michaelr (Quit: CGI:IRC (EOF))
15:28:36 Join wodz [0] (
15:29:22 Join ender1 [0] (
15:30:14wodzamayer: He probably means use user input to build up thoroughly checked parameter list to pass to the function instead of including parts of user input in query
15:30:33wodzpretty basic approach wrt to sql injection prevention
15:31:22amayerare you talking about Prepared statments?
15:32:23wodzI didn't look at the code of quetions - I am talking in general
15:33:23wodzalso I can't remember if I coded something over 10 lines of code in php
15:34:25amayerI know what an sql injection is. i just never heard of parameterised query.
15:34:28amayerive seen prepared statments. and heard that doing it that way is more secure.
15:36:28 Join lebellium [0] (
15:38:08wodzamayer: thats strange since 'sql parameter query' returns example on the first place in google
15:39:33amayerwodz: peoples google results arent always the same
15:39:51amayeri googled parameter query php and didnt get much
15:39:54amayerwhen i google sql parameter query
15:40:01amayerit gives a little better results
15:40:16Torneprepared statements are usually parameterised, but you don't necessarily have to actually prepare statements to parameterise queries.
15:40:29 Join iPodMan5 [0] (
15:40:30Tornedepend son the db api
15:40:52wodzanyway my understanding of the term paramterised was correct :-)
15:40:56Torneescaping user input is also just more string manipulation
15:41:00Tornedon't bother with that either
15:41:04iPodMan5Which would save more battery: sleep mode or shutting RockBox down?
15:41:05Tornei mean, escaping is better than not escaping
15:41:13TorneiPodMan5: we don't have a sleep mode
15:41:18Torneso, shutting rockbox down :)
15:41:32amayeri think i could update db class to use prepared statments pretty easy
15:41:41Tornegreat :)
15:42:01amayerprobably not today. but as Torne said "*next* project"
15:42:04iPodMan5So locking my iPod when no music is playing and having the screen go off isn't sleep mode?
15:42:12Torneno, that's just the screen being off
15:42:17Torneeverything else is still running exactly the same
15:42:38Tornethe power consumption of an ipod that's not playing audio and has the screen off is pretty low
15:42:47Tornebut, when it's off it's off, so yaknow
15:43:03Torneoff always consumes less :)
15:44:59 Join Topy [0] (~Topy44@
15:45:03 Join lebellium_ [0] (
15:45:13Torneamayer: i would also suggest that you read up on things like sql injection prevention/etc in general :)
15:45:29Tornesince if you aren't already familiar with this stuff there are probably other things you should be aware of before web pgoramming too
15:45:33Tornethuogh, er
15:45:40Tornea lot of material on this topic is really bad and wrong
15:45:46Torneespecially material associated with php :)
15:45:53 Quit Clear_runway (Ping timeout: 276 seconds)
15:45:55Tornenot entirely sure what i would recommend specifically :p
15:46:03 Quit lebellium (Ping timeout: 260 seconds)
15:46:12 Nick lebellium_ is now known as lebellium (
15:46:28amayerive found some things. and i understand the concept behind sql injection. so im confident i can filter out the good stuff
15:46:42 Quit T44 (Read error: Connection reset by peer)
15:47:43wodzfiltering user input has the drawback that you don't know how exactly query will be build. It is easy to miss some corner case
15:48:09wodzwith parametrization you gain full control over queries
15:48:11Torneyeah. you should not need to filter user input, generally
15:48:30amayeri ment filter the articles i read(nothing to do with the code)
15:48:36Tornethe only escaping/quoting you actually need to do in a webapp is escaping any user data you *output* as part of a page
15:48:45Torneand html escaping is easy to do and very predictable
15:49:00Tornesince it's largely just "replace < with &lt;"
15:50:53 Quit AlexP (Ping timeout: 240 seconds)
15:52:03 Quit tjb0607 (Ping timeout: 248 seconds)
15:53:51copperIn System -> Rockbox Info, there's a "buffer" value, is that the size of the buffer that gets filled at the start of playback?
15:54:08Torneit's the amunt of memory available to buffer playback data in general
15:54:22Tornewe don't wait until it's full before playback starts
15:54:33Tornebut we buffer as much as will fit (if there's enough in the playlist)
15:56:07 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
15:57:03 Quit iPodMan5 (Quit: CGI:IRC)
16:09:23 Join AlexP [0] (~alex@rockbox/staff/AlexP)
16:09:44 Part LinusN
16:11:31 Join tjb0607 [0] (~tjb0607@
16:16:15 Join AlexP_ [0] (~alex@rockbox/staff/AlexP)
16:16:58 Quit AlexP (Ping timeout: 256 seconds)
16:17:34 Join sentriz [0] (~Senan@
16:23:38 Quit wodz (Ping timeout: 260 seconds)
16:33:46 Quit LambdaCalculus37 (Quit: LambdaCalculus37)
16:35:41 Quit pamaury (Quit: No Ping reply in 180 seconds.)
16:36:05 Join pamaury [0] (
16:36:06 Quit pamaury (Changing host)
16:36:06 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
16:42:53 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
16:44:43 Quit amayer (Remote host closed the connection)
16:46:15 Quit pamaury (Ping timeout: 246 seconds)
16:46:24 Join pamaury [0] (
16:46:24 Quit pamaury (Changing host)
16:46:24 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
16:47:15 Quit soap (Quit: soap)
16:47:59 Join wodz [0] (
16:48:36 Join amayer [0] (
16:51:56 Join stoffel [0] (
16:54:16 Join mortalis [0] (~mortalis@
16:55:01 Join y4n [0] (~y4n@unaffiliated/y4ndexx)
16:55:49 Quit Zagor (Quit: Clint excited)
17:07:40***Saving seen data "./dancer.seen"
17:09:20 Quit sentriz (Ping timeout: 252 seconds)
17:15:08 Quit LambdaCalculus37 (Quit: LambdaCalculus37)
17:19:21 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
17:22:08 Quit LambdaCalculus37 (Client Quit)
17:25:19 Join soap [0] (
17:25:19 Quit soap (Changing host)
17:25:19 Join soap [0] (~soap@rockbox/staff/soap)
17:40:00 Quit pamaury (Ping timeout: 265 seconds)
17:41:16soapI haven't done a careful study of how many spam account are "repeat customers". I'm guessing most are single-use hit-and-run accounts. But...
17:41:38soapWe need to do something about the fact Rockbox Experts have the power to delete spam messages, but not ban the spammer.
17:42:36soapI'm not a big fan of the fact that "Removed Posts" is populated heavily with obvious spam accounts which have had their post removed, but not their posting privileges.
17:54:51 Join bertrik [0] (~quassel@rockbox/developer/bertrik)
17:55:31 Quit stoffel (Ping timeout: 255 seconds)
18:13:21 Join kevku [0] (
18:31:48mortalisTorne, wodz: Speaking about invalidating cache range. I guess there is no guarantee that interrupt handlers doesn't touch cache lines we are invalidating, right?
18:46:19 Join derk0pf [0] (
18:50:23Tornein general, no
18:50:26Torneit depends what you are invalidating
18:50:30Torneand why.
18:50:52TorneFor a lot of cases in an OS it's possible to know that a given cacheline cannot be touched by any code right now
18:51:09Tornebecause, say, you know that the only references to that range of memory are in your local callstack where you are doing the invalidation
18:51:21Torneand this can help you do cache operations without blocking other things from running
18:51:27Tornebut in the general case, you can't always prove it
18:51:48yxopls help to unbrick clip plus. i write OF in recovery mode. i read dump. but clip don't start when i try to turn it on
18:52:32 Join saratoga [0] (123e0cfc@gateway/web/freenode/ip.
18:52:37saratogawhat did you do to brick it?
18:53:10yxoit was rockboxed
18:53:24yxoi turn it off. dry
18:53:44yxothen i try to turn it on and OF don't start
18:53:47Tornethe "unbrick" process is to recover from the firmware being corrupted
18:53:56Tornephysical damage is basically never going to do anything to the firmware
18:54:01Torneso, i expect your hardare is just broken
18:54:15yxorockbox startted but i get meaasge that loading rockbox.sansa bad checksum
18:54:57bertrikhave you tried starting it with the left/previous button pressed (should boot into the OF)
18:55:19yxono i go to recovery mode, i can write firmware i can read firmware from play it identical. but player do not start
18:55:55yxoi think that hardware is oko
18:56:27yxoi think that it's a logical only trouble
18:56:44Torneif it doesn't boot after writing a new firmware image to it in recovery mode and you're sure you did it correctly then no, the hardware isn't ok :)
18:56:51Tornepretty much by definition
18:56:58saratogawasn't there some trick where you could zero a few bytes when you write the firmware to get it to boot?
18:57:34yxobut when i go next time to recovery mode and read dump - dump is ok.
18:57:46 Join bertrik_ [0] (~quassel@rockbox/developer/bertrik)
18:57:47 Quit bertrik (Read error: Connection reset by peer)
18:58:02 Quit tjb0607 (Ping timeout: 256 seconds)
18:58:17 Quit wodz (Ping timeout: 276 seconds)
18:59:15 Join tjb0607 [0] (~tjb0607@
19:00:14yxosaratoga i don't understand your mesage. pls write simply. i not very wel understand english
19:01:04yxoi think that partition o player corrupted and player can't reinstal uploaded firmware
19:01:35yxobut i don't know how to repair corrupted partition
19:01:45 Join LambdaCalculus37 [0] (~LambdaCal@rockbox/staff/LambdaCalculus37)
19:01:45yxoin recovery mode
19:02:18saratogai think sometimes just flashing the .bin file wasn't enough
19:02:32yxoi think it to now =)
19:02:37saratogasometimes parts of the device's memory needed to be set to zero too
19:03:17yxomay be i need dong "dd if=/dev/zero of=/dev/clip"?
19:03:55yxoand than "dd if=/clppa.bin of=/dev/clip"?
19:05:54bertrik_"recovery mode" on the AMS players is not very well known
19:06:51 Join pamaury [0] (
19:06:52 Quit pamaury (Changing host)
19:06:52 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
19:07:09yxoif i write to raw disk any data - may it occure to unbrickable state?
19:07:41***Saving seen data "./dancer.seen"
19:07:46yxoor can i experimenting and write anything that i want?
19:08:21 Join mystica555 [0] (
19:10:05 Nick bertrik_ is now known as bertrik (~quassel@rockbox/developer/bertrik)
19:11:35saratogayxo: in that mode i think you can't break the player more
19:11:40 Join |akaWolf| [0] (~akaWolf@
19:12:04saratogai remember posts on the forums about this, but searching i can't find anything
19:14:06yxoi read that for unbricking fuze require more that simply write OF firmware
19:14:26yxowas needed additional info
19:14:26 Quit akaWolf (Ping timeout: 260 seconds)
19:15:20yxowhere i can read how to rockbox booting step by step
19:16:31yxomay be need place clppa.bin to first byte on memaory... or something else
19:16:36 Quit |akaWolf| (Ping timeout: 256 seconds)
19:16:43yxofor proper start booting process
19:17:30funmanyeah clppa.bin
19:18:23 Join wodz [0] (
19:18:25mortalisTorne: Even if you know that the only references to that range of memory are in the local callstack, how could you know that some other variable or code used during invalidation are not mapped to the same cache line?
19:18:28yxoif i write dd command it start write byte by byte from 0 byte memory?
19:29:40 Join Horscht [0] (
19:29:41 Quit Horscht (Changing host)
19:29:41 Join Horscht [0] (~Horscht@xbmc/user/horscht)
19:31:13 Join yxo1 [0] (
19:36:17 Join yxoo [0] (59a984ae@gateway/web/freenode/ip.
19:36:27 Quit yxo1 (Quit: CGI:IRC)
19:36:42amayerTorne: i hate to do this to you but can you re-review my commit. i totally messed up my local repo then pushed it. and after pushing it realized it was jacked.
19:37:13 Quit yxo (Quit: Page closed)
19:37:28 Nick yxoo is now known as yxo (59a984ae@gateway/web/freenode/ip.
19:39:00 Quit LambdaCalculus37 (Quit: LambdaCalculus37)
19:55:47amayercan i push a branch besides master to gerrit?
19:55:50amayerand would the mess up merging after the patch is accepted?
19:57:48 Join Gallomimia [0] (
19:58:14 Join prof_wolfff [0] (
20:03:51 Join liar [0] (
20:04:16 Quit tjb0607 (Ping timeout: 244 seconds)
20:07:18 Join tjb0607 [0] (~tjb0607@
20:07:52amayerim looking for a more efficent way to work on the theme site. ive never worked on more then one patch at a time before
20:09:08 Quit Topy (Quit: Leaving)
20:13:27bluebrotheramayer: you can setup a repository somewhere else (like github) and push there.
20:13:59amayeri dont think im explaining this right
20:14:27bluebrotherok, then you might want to rephrase your question ;-)
20:15:10amayerlets say i want to work on file1 and make changes
20:15:13amayerat the same time i want to work on file 2 and make changes
20:15:14amayerthey are for seperate commits.
20:15:16amayerif i commit file1 and then commit file2 i can no longer make ammends to file1
20:15:19amayerhow do you handle work like this?
20:15:36bluebrotherdo you really need to amend the commit to file1?
20:16:06Bagderyou ammend commits, not files...
20:16:48Bagderwith rebase -i you can ammend earlier commits too, or you can cherry-pick etc into a new branch
20:18:12bluebrotherwhat I don't understand is why you (seem to) need to amend commits often. Amending is something you should need rarely.
20:18:28bluebrotherthe only exception are gerrit reviews.
20:18:43amayerim talking about gerrit reviews
20:18:57bluebrotheryou can also use different branches for this
20:19:09amayer...thats what i asked in the first question
20:19:27bluebrothernot exactly :)
20:19:43bluebrotheryou asked about pushing to different branches. You want branches. I.e. local ones
20:19:53 Quit Gallomimia (Quit: Gallomimia)
20:20:02bluebrotherall of those branches push to refs/for/master, so they end up as gerrit review
20:22:59amayerok. so i can have multiple local branches and gerrit will move them to the correct spot to be merged with the master of what ever project i am working on?
20:23:58bluebrothergerrit recognizes existing reviews using the Commit-Id line. Otherwise pushing an amended commit wouldn't work.
20:24:26bluebrotherit doesn't know about your local branches −− you're just pushing some commit(s).
20:25:04bluebrotherand since a commit on a local branch b has a different Commit-Id than a commit on branch a (unless you cherry-picked from b) it will create a new review.
20:25:23bluebrotheror update a different review if it's already existing
20:25:58bluebrotherhowever, if you're working on multiple things it might be easier to simply create a repository on github or similar and push to that.
20:27:07bluebrotherand then nag people to look at your patch series. Gerrit makes a review out of each commit which might not be the best thing in such a situation.
20:27:37bluebrother(and if you say made 7 commits you can still push the patch series to gerrit, which will then create 7 reviews)
20:30:30amiconnwodz: pong (probably timed out by now ;) )
20:32:47 Quit y4n (Quit: PANTS OFF!)
20:33:04Tornebluebrother: please don't encourage people who are doing a perfectly reasonable thing to push stuff to github instead
20:33:13 Join Strife89 [0] (~Strife89@
20:33:18soapare the forums (in particular the admin panel) very slow for anyone else or just me?
20:33:33amayerbluebrother: thanks that helps. i still dont fully get it but if i can push different branches and get them reviewed thats all i need
20:33:39 Join Topy44 [0] (
20:34:19amayerTorne: did you see my message ^
20:34:28AlexP_soap: Yes, the admin bits are very slow
20:34:32AlexP_e.g. banning people
20:34:34Torneit's much better not to make new changes, though
20:34:44Torneyou could've just uploaded a new patch to the old one
20:34:57Tornemaking new ones and abandoning the old ones means nobody who was watching the old one knows about the new one
20:35:09amayeri overwrote it with different files
20:35:19Tornethat doesn't matter
20:35:22amayerthats why i was asking about branches
20:35:23Torneupload a third patch that makes it right again
20:35:53bluebrotherTorne: hmm? Pushing stuff to github doesn't make it more problematic to create gerrit reviews. And given the slow reaction on gerrit reviews I had the impression it slowed amayer down.
20:35:55Tornealso, your db class change patch was fine and you could've just committed it
20:36:12amayerlet me explain the context
20:36:14Tornebluebrother: nobody was interested in looking at ameyer's changes at all
20:36:21Torneit's nothting to do with it being on gerrit
20:36:31Torneit's weird php hacking in the ugly themesite code
20:36:35Tornewhich most people are not interested in :)
20:36:45amayeri made changes to themesite.class.php
20:36:47amayercommited them to branch master
20:36:59Tornebluebrother: github encourages you to learn a workflow that's completely different from the one we use
20:37:01amayerthen i made the changes to db.class.php and commited it to branch master
20:37:21Torneso, telling someone who isn't very confident in what they are doing in git to go use that instead is unhelpful
20:37:24bluebrotherit does? I've never seen github as doing so
20:37:37wodzamiconn: pong
20:37:46Torneamayer: what i mean is, i approved your other change
20:37:51Torneso you could've just pressed "submit" on gerrit for it
20:37:53Torneit was approved
20:38:02bluebrotherrather as "a place where I can push things (and allow people to look at it)"
20:38:25soapAlexP_, for what it's worth those last two were from the same IP address, and the same IP address as earlier ugg boot spam this morning.
20:38:30amayeri didnt know i had to do anything. i dont have commit access.
20:38:45amayer(at least not to my knowledge)
20:38:48AlexP_amayer: No, but a committer had approved it
20:38:56AlexP_soap: Did you ban the ip?
20:39:27amayeroh. i didnt have to do anything to my last patch. i think someone did that for me and didnt tell me i had to do it
20:39:38Torneoh, actually it's not set up to let people commit their own changes
20:39:40Tornenevermind :)
20:39:44Tornesorry. maybe it should be.
20:39:50AlexP_yes, the committer can also submit changes
20:40:01AlexP_Only a committer can submit them then :)
20:40:11soapAlexP_, 6 days
20:40:17AlexP_soap: cool
20:40:58amayerwell Torne can you reaccept the new one and i will look for a submit button
20:41:00Torneamayer: anyway. no matter what mistake you have made, you can always just undo it
20:41:08scorche`hrm - forums do seem to be a bit slow...
20:41:08Tornethere is never any need to start from scratch
20:41:11wodzOk, what to do with elfloader stuff? ML thread didn't attract much attention.
20:41:12Torneit's just less convenient for everyone :)
20:41:44Torneamayer: also, in general if you want someone to review your change, add them as a reviewer :)
20:41:44AlexP_scorche`: They are OK in general for me, it is just banning anyone (or similar admin) that is really slow
20:41:50amayerTorne: im watching a tutorial on Git from
20:42:19amayerso i should learn alot more. i know the theory now. but i need to know how to execute what i want to do
20:42:33amayerall i know how to do now is commit, ammend, and push
20:42:36 Join Gallomimia [0] (
20:42:52Tornethat's sufficient :)
20:43:29amayerwell i would like to learn branches
20:43:34amiconnwodz: Regarding the paste links you mailed - yes it looks like gcc used word sized jumptable entries there, hence the bug doesn't hit. Did you use -Os, and what gcc version was that?
20:44:01amayerthat way i could have worked on the db class and themesite class at the same time and had sepperate reviews that i could make changes too
20:44:15Torneamayer: i wouldn't've bothered making those separate changes in the first place
20:44:28Tornethe db cleanup is so tiny it's not really necessary to make it a separate patch :)
20:44:40wodzamiconn: both 4.0.3 and 4.6.3 with -Os. I did get byte jumbtable with -O2 but the code is a bit weird as it reshuffles the order of cases also.
20:45:41amayeri thought for the sake of the log i would put it in there so people could see that is where the database class was last edited
20:45:42amayer(hears echos in the themesite cave he is standing in)
20:45:44amayer...or just for my own sake
20:46:28amayerTorne: you merged the patch?
20:46:52amiconnIt may reshuffle the order of cases to shorten the maximum offset
20:47:34amiconnIt probably won't do that if the (numerically) last case does have the biggest block of code
20:47:57amayershould i keep putting "themesite:" at the begining of commit messages?
20:48:07*amiconn wonders what's different between this test case and the sound setup function which has been crashing due to that bug back then
20:49:04amayerTorne: so now i do a rebase and commit my next change?
20:49:32Tornethe themesite is a completley separete repo, so i don't see the need to repeat it in the commit message
20:49:40AlexP_ah, forgot that
20:49:51Torneincluding it just makes the oneline summary longer :)
20:50:07amayerTorne: thats what i was thinking
20:54:30amayerTorne: when i add you as a reviewer does it send you an email or something?
20:56:06wodzamiconn: gcc-4.6.3 with -O2
20:59:09amiconnIs this patched or unpatched?
21:00:46amiconnwodz: In this example, the largest offset is 0xc4, hence the bug won't hit
21:01:46*amiconn is building unpatched gcc 4.0.3 in order to put together an example that triggers the bug
21:02:13wodzamiconn: I know, but it is hard to predict the offset if gcc changes the order
21:02:25wodz4.6.3 is vanilla
21:04:00 Quit Gallomimia (Quit: Gallomimia)
21:04:09scorche`saratoga: ping
21:07:45***Saving seen data "./dancer.seen"
21:13:38 Quit liar (Remote host closed the connection)
21:14:37wodzamiconn: I am unable to find combination which increases the offset near problematic boundary - if I increase size of one case it is reshuffled, if I increase sizes of more cases it switches to word offsets. This is all with -O2, -Os always use word jump entries with gcc-4.6.3.
21:15:09wodzat least thats what I get
21:16:02 Join pretty_function [0] (~sigBART@
21:17:50 Join Horschti [0] (~Horscht@xbmc/user/horscht)
21:20:34 Quit pretty_function (Ping timeout: 244 seconds)
21:21:13 Quit Horscht (Ping timeout: 240 seconds)
21:22:08 Quit mortalis (Quit: KVIrc 4.1.3 Equilibrium
21:29:58 Quit kevku (Quit: KVIrc 4.2.0 Equilibrium
21:31:19 Quit wodz (Quit: Leaving)
21:38:19 Quit ender1 (Quit: Punch a psychic today: ask them if they saw it coming.)
21:41:39 Quit pamaury (Read error: Operation timed out)
21:43:33 Join pamaury [0] (~quassel@rockbox/developer/pamaury)
21:44:59amiconnhrrm, he's gone
21:48:32saratogascorche`: pong
21:50:02 Quit XavierGr ()
21:52:13 Quit tjb0607 (Ping timeout: 255 seconds)
21:53:31 Join XavierGr [0] (~xavier@rockbox/staff/XavierGr)
22:13:17 Join tjb0607 [0] (~tjb0607@
22:20:37 Quit amayer (Remote host closed the connection)
22:34:09 Quit n17ikh (Ping timeout: 260 seconds)
22:34:56 Join n17ikh [0] (~n17ikh@
22:34:57 Quit n17ikh (Changing host)
22:34:57 Join n17ikh [0] (~n17ikh@unaffiliated/n17ikh)
22:37:52 Quit Strife89 (Quit: Heading out.)
22:52:19 Quit tjb0607 (Ping timeout: 248 seconds)
23:03:16 Join JdGord [0] (~AndChat80@
23:05:53 Quit JdGord (Read error: Connection reset by peer)
23:07:35 Join Gallomimia [0] (
23:07:48***Saving seen data "./dancer.seen"
23:09:44 Join JdGord [0] (~AndChat80@
23:10:37 Quit Rower85 (Quit: Hmmm...)
23:10:43 Join JdG [0] (
23:13:55 Quit JdGord (Read error: Connection reset by peer)
23:18:55 Join tjb0607 [0] (~tjb0607@
23:19:59 Quit yxo (Quit: Page closed)
23:20:44 Join amayer [0] (~alex@
23:26:48 Part PathosesNorms
23:28:18 Join ender1 [0] (
23:29:50 Quit XavierGr (Ping timeout: 245 seconds)
23:31:13 Quit derk0pf ()
23:41:37 Nick ParkerR_ is now known as jmad2011 (ParkerR@unaffiliated/parkerr)
23:41:53 Nick jmad2011 is now known as ParkerR_ (ParkerR@unaffiliated/parkerr)
23:43:15 Nick ParkerR_ is now known as ParkerR (ParkerR@unaffiliated/parkerr)
23:43:33 Nick ParkerR is now known as ParkerR_ (ParkerR@unaffiliated/parkerr)
23:49:40 Join lebellium_ [0] (
23:50:22amayerwhen you select order by "Rating" "Descending" should themes be ordered by rating alphebeticaly or themes with 10 ratings of 5 stars be in front of themes with 6 ratings of 5 starts?
23:51:13 Quit lebellium (Ping timeout: 240 seconds)
23:51:20 Nick lebellium_ is now known as lebellium (
23:55:01AlexP_by rating alphabetically?
23:55:17AlexP_rating is I believe average rating
23:55:18 Join JdGord [0] (~AndChat80@
23:55:36amayerim talking about the order on a page when sorted
23:55:42AlexP_and ascending/descending is low to high and vice versa
23:56:02AlexP_yes, but I don't know what rating alphabetically means
23:56:08gevaertsI'd say when equal, most ratings wins
23:56:24amayerTheme "A" has 5 ratings
23:56:26amayerTheme "
23:56:28AlexP_yes, if the average is identical
23:56:34amayerB" has 6 start
23:56:40gevaertsAnd when still equal then, go alphabetical
23:56:44amayerhold on let me retype
23:57:47 Quit bertrik (Ping timeout: 244 seconds)
23:58:03copper(A: 10 * 5) > (A: 6 * 5) > (B: 6 * 5)
23:58:10 Quit JdG (Ping timeout: 252 seconds)
23:58:51AlexP_imo of course :)

Previous day | Next day