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 2017-01-24

00:00:28furrywolfnah, it works rather well. it scans the filenames for any position, measured either from the start or from the end of the filename, looking for a number that's unique
00:00:56furrywolfso if it has to measure from the end to find the track number, it prints that message. :)
00:01:52furrywolfand if that fails, then it can import a track list and fuzzy match to it
00:01:54chrisjjIf anyone can suggest any way to make more solid the following test of ZEN BSoPO, do say.
00:02:08chrisjj1 Mirror .zip's .rockbox to device 2 Disconnect from PC and connect charger 3 Ensure no SD card or audio jack plug 4 Reset device 5 Power-up device 6 Set Repeat = All 6 In Files, select first of 500 WMAs (to play) 7 Launch plugin battery bench
00:02:12furrywolfbrb, some form of stupidity in the road out front
00:02:49 Quit Senji (Ping timeout: 252 seconds)
00:03:49furrywolfback. someone having a right-of-way contest with the maillady.
00:04:00furrywolfmaillady has my 128gb microsd card. :)
00:05:38 Join Senji [0] (~Senji@
00:05:41chrisjj9 If LED goes out, press non-power keys. If no response, press power. If device powers-up, record the test result as a BSoOD, and record battery_bench.txt last time.
00:07:40chrisjjThis worked fine here on recent pamaury builds (the last being cd8b33327_nowdt, result: BSoPO on both units, Q and G)...
00:08:03chrisjj... but on latest dev build 28bf763-170121, the LED is initially off!
00:09:09chrisjjAnd the backlight is going to sleep.
00:10:03chrisjjpamaury, have you changed the LED to now be Off when charging?
00:10:21jhMikeSMihail: I pushed the revised patch
00:11:00pamaurychrisjj: yes I pushed the led framework and the LED are initially set to off
00:11:12pamauryyou can set the LED to anything using the debug menu
00:11:19chrisjjWhy the change?
00:12:05pamauryif rockbox wants to control the led, it needs to init the hardware to a known state
00:12:24 Join Senji_ [0] (~Senji@
00:13:50[Saint]furrywolf: The most interaction I'll ever have to do with MBP is telling it "No, this/these track(s) are from *this* release of $ALBUM, not /that/ release of $ALBUM"
00:13:50[Saint]furrywolf: I should also note that it will also strip metadata, either completely, or based on RFC defined standards. It can also optionally add replaygain tags (or apply mp3gain), and optionally embed cover|booklet|disc artwork and/or place that artwork alongside the tracks or albums, with user defined bit depth, colourspace, resolution, and container.
00:13:50[Saint]furrywolf: it can also apply timecode synced .lrc lyrics...
00:13:50chrisjjPreviously the LED was On until RB shutdown. The LED acted as a power indicator.
00:14:01[Saint]There's really not a hell of a lot it won't or can't do.
00:15:02[Saint]it's an automated batch tagging one-stop-shop for metadata and file naming.
00:15:04pamaurypreviously rockbox did not touch the lcd at all
00:15:18pamauryand now you can set it in the debug menu if you want
00:15:28 Quit Senji (Ping timeout: 252 seconds)
00:15:55Bilgus_^that is pretty awesome I almost wish I had a LED now
00:16:05chrisjjRB not touching the LED was fine. the LED acted as a power indicator.
00:16:57*[Saint] really thinks chrisjj should learn some embedded C and ARM ASM so he can either put up or shut up
00:17:05*pamaury doesn't like LEDs, they shine a lot and should have at leasta setting to disable them
00:17:21chrisjjThe fact the user can turn it on an off manually (and I've no idea why an end user would want to) isn't much compensation for the loss of the power indicator.
00:17:50Bilgus_Are you like more than one person or something? multiple personalities I'm pretty sure I sat here a week ago and watched as 30pages of text scrolled by of you asking for this feature
00:17:51[Saint]pamaury's just too polite to tell you that he's not your personal caged Rockbox developer and general whipping boy.
00:17:58[Saint]I, however, don't give a shit.
00:18:46pamauryI think you kind of missed the point. It went from "it can whatever color the creative bootloader decided to, which may or may not, probably not, have any kind of relationship to anything, except that it's on" to "if someone writes a patch, rockbox can set the led to whatever if wants in 2 lines of code"
00:19:21 Quit paulk-collins (Quit: Leaving)
00:19:32[Saint]Bilgus_: personalities N through Q
00:20:05chrisjjWhy break the first to get the second? RB could just leave the LED in its current state until you future patch asked for different.
00:20:30pamauryI'm not breaking anything
00:21:15chrisjjThis change has broken the LEDs indication of power state.
00:21:43pamauryI'm not seem anywhere in the manual that rockbox uses the LED as a power indicator
00:21:48chrisjjYou said your change is turning the LED off. That's sufficient explanation for the breakage.
00:22:13Bilgus_the set it to on?
00:22:28chrisjjI too have not seen it in the RB ZEN manual. I've not seen a RB ZEN manual :-)
00:22:29TorCBilgus_: You beat me to saying it.
00:22:34pamaurythe fact that the LED was on was a pure side effect of how I implemented the port, it was never a feature
00:22:35Bilgus_it will never be on when the device is off
00:22:50chrisjjIt was never a feature you knew about.
00:23:19chrisjjWhy not just remove your new init of the LED?
00:23:41[Saint]Why not just...shut your face?
00:24:51chrisjjBilgus, if you think you saw /me/ ask for this feature, you're mistaken.
00:25:05chrisjjWhat I asked for was LED blink on *PANIC*.
00:25:50furrywolf[Saint]: I have no interested in album covers or lyrics. heh.
00:26:13furrywolfalso, I started out with \[ to tab-complete your nick... I've been using bash too much. lol
00:27:15Bilgus_you do realize control of the LED is the first step for that??
00:28:41chrisjjNo, pranging the current effect of the LED is not the first step for LED blink on *PANIC*.
00:28:42 Quit pamaury (Ping timeout: 255 seconds)
00:28:56 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 50.1.0/20161208153507])
00:29:06MihailjhMikeS: you fix typo too fast :) In overall looks good.
00:29:39chrisjjYou see any reason the LED can't stay on, indicating power, and when there's a PANIC, blink?
00:29:57__builtinchrisjj: how about instead of constantly annoying people, you −− I don't know −− try to actually do something yourself?
00:30:18chrisjjI'm doing pamaury's requested tests.
00:30:24jhMikeSMihail: was going through it to check goofs. I finally discovered the utility of 'git diff ... −−color=auto' to get unwanted whitespace changes out
00:31:06furrywolf[Saint]: also, I'm not opposed to metadata... my experience is that between file sources, players, etc, it usually doesn't work.
00:43:06MihailjhMikeS: I reply in gerrit
00:45:34***Saving seen data "./dancer.seen"
00:47:41[Saint]furrywolf: that's...odd. I have no such experience. I mean, ID3v*, vorbistag, etc. ...they're not magic.
00:48:53[Saint]If the software and/or hardware you're using is screwing up a readily defined RFC then maybe you should look at replacing it or look at your implementation?
00:54:41furrywolf[Saint]: two of my three car head units don't seem to support them at all. xmms does, but doesn't do a great job.
00:54:53furrywolfand people making them do a really bad job. :)
00:55:29 Join MrZeus2 [0] (~MrZeus@2a02:c7f:7008:3400:d18d:a239:f7ff:5dcd)
00:57:08jhMikeSMihail: you mean get the bit setting out of ascodec_int_audio_cb and just test it later?
00:57:19 Quit MrZeus1 (Ping timeout: 255 seconds)
00:57:21[Saint]I haven't has a car head unit that didn't support ID3 tagging since the late 90s.
00:57:21 Quit girafe (Read error: Connection reset by peer)
00:57:52furrywolfheh, my oldest one that does mp3s does support them... and shows you the first 8 characters of them just fine.
00:57:54Bilgus_I'm trying to consolidate a bunch of time settings down to a single list of values, expressed in H:M:S with auto formatting of the values so 1-59s shows as seconds 1-59 minutes shows 01:00 - 59:59min and 1-... hours shows 1:59:59 hours currently I have a range of values for the tables expressed in seconds {-1,0,1,2,3,4,5,6,7,8,9,10,15,20,25,30,45,60,75,
00:57:54Bilgus_ 90,105,120,135,150,165,180,195,210,225,240,255,
00:57:54Bilgus_ 270,285,300,315,330,345,360,375,390,405,420,435,
00:57:54DBUGEnqueued KICK Bilgus_
00:57:54Bilgus_ 450,465,480,495,510,525,540,555,570,585,600,900,
00:57:56Bilgus_ 1800,2700,3600,4500,5400,6300,7200};
00:57:57furrywolfand only the one tag you select
00:58:08furrywolfthe other two are no-brand chinese mechless units
00:58:17furrywolfwhich show the entire filename just fine
00:58:22MihailjhMikeS: yes, as we already test it later
00:59:04Bilgus_I don't want to do 0-7200 seconds in a list as that takes way too long to scroll through anyone have an Idea on a good sampling of values for times in backlight, peak hold, clip hold etc
00:59:46jhMikeSMihail: CHG_ENDOFCHG wouldn't stick though
01:00:41Bilgus_keep in mind portions of this list can be used where it makes sense
01:02:09furrywolfI assume rockbox, like all sane software, will be happy with my sdxc card as fat32?
01:02:17jhMikeSMihail: the interrupt only sets the bit and it has to stay until the charging code reads it
01:03:13Mihailyes, you right
01:03:25furrywolfBilgus_: that list sounds fine to me
01:05:34[Saint]furrywolf: Yes.
01:07:18 Quit ender` (Quit: I used to believe in reincarnation, but that was in a previous life.)
01:07:21[Saint]The guts of it is that they're electrically identical and any software htat does care about such things is deliberately crippled, be it be ineptitude, deliberacy, or plain ignorance.
01:07:35furrywolfyes, I know.
01:07:38MihailjhMikeS: how about separate endoch and chg_status as it was in my code?
01:07:56[Saint]If they existed (they never, ever will), Rockbox would happily address the full 4TB the SD spec defines.
01:08:14furrywolfthe sd spec requires sdxc-compliant devices to reject a >32gb card with anything but exfat on it, and sadly many things follow this requirement.
01:09:42furrywolfif you want to put the little sd logo on your product... you have to be an asshole.
01:09:42Bilgus_[Saint], Do you feel that smattering of time values have enough granularity for most of the settings in RB?
01:10:43furrywolf4tb might exist... that's only 8 times larger than what's currently available.
01:11:25furrywolfactually, sandisk says they're putting out a 1TB card.
01:11:53furrywolfso 4tb isn't too far in the future
01:14:10MihailjhMikeS: IMHO all related to IRQ_RTC should be removed as we don't use IRQ_RTC
01:17:58Bilgus_that is sec 1,2,3,4,5,6,7,8,9,10,15,20,25,30,45, 1:00m 1:15, 1:30, 1:45, 2:00, 2:15, 2:30, 2:45...9:00, 9:15 9:30, 9:45, 10m, 15m ,40m , 45m 1:00:00, 1:15:00, 1:30:00 1:45:00 and 02:00:00 hr
01:21:34furrywolfvolume control of the line out dock on the fuze is broken
01:21:58furrywolfthe first 1/4 or so of the volume control is the entire range of volume, with everything over that being the same volume
01:22:50furrywolfeither declare line out shouldn't have volume control, or make it behave the same as the headphones. heh.
01:27:14CtcpIgnored 1 channel CTCP requests in 0 seconds at the last flood
01:27:14*furrywolf decides it's a very minor bug
01:31:24furrywolfah, looks like the code uses the ADC volume control below a certain volume, then the HP volume control over a certain volume. the ADC volume affects line-out as well, while the headphone volume of course does not.
01:31:31jhMikeSMihail: I guess
01:31:54furrywolfwhile line out is fixed at 0db, and only changes when the adc volume is changed
01:35:06furrywolfbbl, doing useful things rather than dicking around with mp3 player. heh.
01:36:55[Saint]furrywolf: that'
01:37:02[Saint]s not broken, per se.
01:37:07[Saint]That's a hardware limitation.
01:37:46furrywolfwell, to make the line out adjust like the headphone volume would be trivial
01:38:02furrywolfbut having it fixed at 0db is probably sane for line out
01:38:11[Saint]To even want to do so represents a massive misunderstanding about lineout.
01:38:21[Saint]Line out should be line level, period.
01:38:36furrywolfjust causes the volume control to behave very non-inuitively, with part of the volume control working, and part of it not.
01:38:52[Saint]You're using lineout because you're delegating volume control to other hardware.
01:39:10[Saint]If you want to use the device for governing the volume, use the headphone out.
01:39:17[Saint]Lineout should be 0dB.
01:39:38furrywolfthe headphone out seems equalized in some fashion... midrange and high-bass heavy. or maybe I'm imaging things.
01:39:49furrywolfI spent $14 on a line out dock to fix this. :)
01:39:57[Saint]it's not and you are.
01:39:59*furrywolf should do an a/b test to make sure it wasn't just the selected track
01:42:24furrywolfI can't swap plugs fast enough. I'd need to rig up an actual switch. heh.
01:42:27[Saint]personally I have the max volume fixed at 0dB, I did submit a change a while back to make that the default and to make people specifically enable overdriving above line level but I'm not sure what happened to it.
01:42:48furrywolfthe volume of the line out is fixed at 0db
01:42:50[Saint]I don't particularly care about people's hearing, if they screw their ears up that's their own fault.
01:42:58furrywolfthe headphones should be able to be turned all the way up to maximum
01:43:12[Saint]But people get all pissy about clipping at high ranges and don;t realize they're causing it.
01:43:19furrywolfdifferent headphones have very different properties, and some need more power than others.
01:43:50furrywolffor example, my koss porta-pros have low enough impedance that my netbook can't drive them above a marginal volume.
01:43:58[Saint]people who care about that should be perfectly capable of tickling a system setting.
01:44:17furrywolf(insufficient output current capability of the netbook's amp)
01:44:30[Saint]Joe Average will push it all the way to +12 dB with no dynamic range or clipping protection, and then blame us for it sounding shit.
01:45:01[Saint]And, if you do need to drive high output or low Ohm monitors...get a damn external amp. lol
01:45:05furrywolflol, googling for koss porta-pro impedance, the first result is "Why are the Koss Porta Pro headphones so quiet".
01:46:04furrywolfthey suggest they have abnormally high impedance, which is odd, as it sure seemed like insufficient drive current when I was playing with them. using a y plug and a second pair caused them to get reallly quiet, strongly suggesting a current issue, not a level issue.
01:46:07[Saint]I'll still stand by the fact that Joe Average has no business driving headphone out above 0dB.
01:46:25[Saint]If it's too quiet, that's exactly what replaygain is for.
01:46:40[Saint]Overdriving the output isn't the sane solution to this.
01:47:26furrywolfeh, I think any output amp should be able to be adjusted to its maximum gain.
01:47:38[Saint]And if you're not matching the impedance of your monitors correctly, you're bound to be fucking up a whole bunch of others aspects as well.
01:47:44jhMikeSMihail: I think I might use RTC later anyway to set a flag that the RTC has changed (no need to read it right away). That way it's not up to a second wrong.
01:48:32jhMikeSI guess a change like that would be a different update though
01:48:40[Saint]furrywolf: if people weren't absolute idiots and had any metric or basis of understanding in this regard I might be tempted to agree.
01:49:02[Saint]To make it clear, I'm not talking about disabling the ability to overdrive completely.
01:49:32[Saint]I am just talking about setting the max volume level feature we already have but don't apply a default value for to 0dB.
01:50:34[Saint]If a user actually understands what they're doing, and the repercussions of theie actions, more power to them...that type of user is perfectly capable of changing a user-facing setting.
01:50:50MihailjhMikeS: ok, as you wish
01:51:41furrywolfheadphones differ too widely to declare something as standard.
01:51:55furrywolfit's not like line out, which has only a couple well-defined levels.
01:52:32furrywolfwhat's unreasonable for a pair of earbuds might barely be audible in a pair of pro over-ear headphones.
01:52:40jhMikeSMihail: I'd like to update that for all targets than can use it. While we don't see it on the clock, it makes a difference for creating file times and dated filenames
01:54:20 Quit Bray90820 (Ping timeout: 264 seconds)
01:54:49furrywolfbbl, need to get things done before it's dark
02:00:22TorC[Saint]: Considering the matter, I think you are probably right that the max should by default be 0db. Especially when the standard volume display is a tiny little icon, not an actual number. I think on some targets there isn't even any obvious warning when crossing 0db either.
02:00:58TorCI keep my volume display at the actual dB level, just so I know exactly where I am.
02:02:57[Saint]There's two things I do with my targets surrounding volume and both revolve around a function that I'm pretty confident 90% of users have no idea exists.
02:03:30[Saint]I am willing to bet that a few people know about the max volume setting, but I sincerely doubt many people know about fixed.cfg
02:03:32 Join Bray90820 [0] (
02:04:16[Saint]I set a fixed.cfg to ensure that the max volume level is always 0dB, and then the volume level is always returned to a safe value on boot.
02:04:45[Saint]I generally set it to about halfway, so 32dB or thereabouts.
02:05:19TorCNo, I hadn't come across fixed.cfg. Sounds worth looking into.
02:05:32 Quit ZincAlloy (Quit: Leaving.)
02:05:34[Saint]fixed.cfg is great, it follows the exact same format and uses the exact same strings as config.cfg, but the values in fixed.cfg will always override config.cfg
02:05:58[Saint]So that even if you do a 'factory reset' of the config, you can still have a subset of settings applied.
02:06:03TorCThough I often keep the same volume across days, so I'm not sure I'd want it for that.
02:06:18[Saint]I use it primarily to make sure that I don't blast the shit out of myself with high volumes.
02:06:50[Saint]And I also use it to reset the quickscreen values because it is absurdly easy to accidentally change the quickscreen values and/or actions accidentally.
02:07:44[Saint]It's one of those really neat features of Rockbox that doesn't really get a hell of a lot of attention.
02:08:06TorCI'll have to think about setting up a fixed.cfg. Since I don't use the quickscreen, that doesn't matter to me, though.
02:10:12 Quit zoid (Quit:
02:10:40 Quit MrZeus2 (Ping timeout: 255 seconds)
02:11:39 Join zoid [0] (~zoid@
02:12:10 Quit zoid (Changing host)
02:12:10 Join zoid [0] (~zoid@unaffiliated/taxationistheft)
02:23:07 Quit Mihail (Quit: - A hand crafted IRC client)
02:33:43 Quit skapazzo (Quit: leaving)
02:36:20 Quit Bray90820 (Ping timeout: 264 seconds)
02:43:47 Join Bray90820 [0] (
02:45:38***Saving seen data "./dancer.seen"
03:10:22*__builtin needs some ideas for what SDL apps to port
03:32:35 Quit Rondom (Remote host closed the connection)
03:41:30 Join Strife89 [0] (
03:42:04furrywolfblah. I knew I should have spent the extra money for the Extreme microsd card. this Ultra one is sitting here writing at 10MB/sec. It's too large for that.
03:45:10Bilgus_10mbps isn't too bad for sd card but yeah gonna take foooorrrever
03:45:37furrywolfmy 32gb Extreme one manages 90MB/sec. heh.
03:46:09 Quit Strife1989 (Ping timeout: 276 seconds)
03:51:10Bilgus_90 is pretty damn fast
03:51:31furrywolfsandisk has 250MB/sec ones now, but they're ungodly expensive.
03:51:47furrywolfthey use a second row of contacts transfer data faster. I have the reader for them, but haven't bought any of the cards yet.
03:56:28Bilgus_huh wonder what the lifetime is on those :o
03:59:03furrywolf that look like that
04:01:37Senji_why don't we just start putting ssds in things
04:02:34Bilgus_read write cycles would be my guess
04:04:40Bilgus_not to mention I wouldn't keep anything important on a SD card
04:27:20furrywolfI formatted my 128GB card fat32, and put a bunch of files on it. It's not recognized by rockbox, at all.
04:27:55furrywolfdoesn't even show <microSD1> in the files like it does with my 32GB card.
04:29:51furrywolfit does show up in the disc info debug screen.
04:30:05furrywolfwith sane-looking information.
04:30:59 Quit Bray90820 (Ping timeout: 258 seconds)
04:34:15 Join Bray90820 [0] (
04:34:18furrywolfwhy is it not reading the fs? heh
04:36:41furrywolfbah! even the stock firmware reads the card just fine, despite it having a supposed 32gb limit.
04:37:13furrywolfof course, it's going to be "refreshing your media" for about an hour at the rate the progress bar is moving... seems it wasn't optimized for 110GB of mp3s.
04:44:01chrisjjfurrywolf: A few days back you mentioned an RB unshuffle operation. Did you mean just setting Player Settings Shuffle to No?
04:45:39***Saving seen data "./dancer.seen"
04:57:56 Quit vflyson (Quit: ZNC 1.6.3 -
04:58:13 Join vflyson [0] (
05:01:55chrisjjPresumably Shuffle -> Yes must keep a copy of the current playlist from which Shuffle -> No restores.
05:08:17furrywolfbah, the stock firmware died halfway through "refreshing your media".
05:08:49furrywolfhow do I troubleshoot why rockbox won't recognize this card?
05:10:15furrywolfit shows up correctly in the disk info debug screen, but not in the files menu.
05:15:01Bilgus_how many entries do you have per directory
05:16:00furrywolfI doubt any directory has more than 100.
05:16:54Bilgus_ok how many directories then
05:17:59Bilgus_do you see the <SD Card> in your file browser at all?
05:19:46Bilgus_IIRC fat32 allows something like 35535 entries in a single directory
05:19:54Bilgus_ro9ot incld
05:19:54furrywolfactually, I was wrong. a directory I copied off an ipod has 1029 files in it. the most my other directories has a 103 files.
05:20:29furrywolfa total of 2058 directories on the card
05:20:37Bilgus_hmm do you have directory cache on?
05:20:49furrywolfI do not know. I have not changed any options related to that.
05:20:57 Quit Strife89 (Read error: Connection reset by peer)
05:23:10Bilgus_try turning it off the restart
05:23:31furrywolfstill doesn't show in files
05:23:52 Join Strife89 [0] (
05:23:53furrywolfstill does show in disk info
05:25:21furrywolfI formatted it with mkdosfs, and it's recognized by the stock firmware, despite the stock firmware claiming a 32gb limit.
05:26:09Bilgus_hmm if it works in OF it should surely work in RB
05:26:19furrywolfand yet it doesn't. :)
05:26:58Bilgus_are you using 3.13 still?
05:27:04furrywolfno, 1/20 devel
05:27:41Bilgus_huh well ya got me I'm sure someone knows but I haven't used anything above 64gb
05:28:10furrywolfI could try deleting files from it, in case it's hitting some limit there, but I waited half the bloody day to copy files to it at 10MB/sec. heh.
05:28:43Bilgus_nah I wouldn't yet there shouldn't be a limit on RB OF yes RB no
05:32:22furrywolfI probably should wipe the card and try with just a few files
05:32:43jhMikeSBilgus_: 65536 directory entries (2MB). How many files depends on how many 32-byte entries each filename takes.
05:33:03Bilgus_ahhh good point
05:35:02furrywolfI would imagine linux would have complained rather than created an invalid filesystem.
05:36:01furrywolfI just wiped a big chunk of the card. now it contains a total of 1028 directories, the largest of which has 103 entries.
05:36:39furrywolfstill does not show up in files at all.
05:37:02jhMikeSif you max-out the length of every filename, 3120 files will fit
05:37:21furrywolfI do not believe any fat32 limits are the issue.
05:37:46furrywolfboth because linux would complain when hitting them, and now the most I have in any directory is 103 entries.
05:39:47Bilgus_well FFS try just a file or two then
05:40:03furrywolfthe partition is:
05:40:12furrywolf /dev/sdc1: sticky x86 boot sector, mkdosfs boot message display, code offset 0x58, OEM-ID " mkdosfs", sectors/cluster 32, Media descriptor 0xf8, heads 255, sectors 249704448 (volumes > 32 MB) , FAT (32 bit), sectors/FAT 60960, serial number 0x2eb26ca7, label: " "
05:40:43furrywolfI'll reformat it for a clean start.
05:42:11furrywolfok, fresh format, one directory with ~15 files.
05:42:22*jhMikeS wonders if there could be an overflow in calculations in fat32?
05:42:39furrywolfdoesn't show up in files.
05:42:43Bilgus_If that doesn't work try doing it the hard way
05:43:03furrywolfhard way?
05:43:20Bilgus_also whats name the label something 8 chars
05:43:31furrywolfplays perfectly with original firmware
05:44:24furrywolfgood to know, the original firmware has no problems with a 128GB card... that should be posted on a forum somewhere.
05:44:43furrywolf... why would I check the newly created filesystem for errors?
05:44:52furrywolfon, fdisk, not fsck
05:44:54*furrywolf mis-read
05:45:22furrywolfwhy would I change the partitions when it already has one full-size partition that works fine?
05:46:00Bilgus_delete the partition
05:46:10Bilgus_create a new one
05:46:18furrywolf... I can, but why? heh
05:46:44furrywolfyou think it has a bad partition table?
05:47:08chrisjjfurrywolf, did you find why Shuffle -> Off sometimes fails to restore the original order?
05:48:03Bilgus_well i'm not sure what all this mkdosfs is doing so do it the manual way and see if it works
05:48:34Bilgus_don't forget to set the partition active
05:49:21furrywolfmkdosfs is just a fat format utility
05:49:32furrywolfand I will be using it to format the new partition
05:50:30Bilgus_use mkfs -t vfat /dev/sd.
05:50:58furrywolfno partition? format the block device?
05:51:20Bilgus_you haven't finished with the first part yet
05:51:35Bilgus_hold on a sec and ill run you through it
05:51:40furrywolfheh, mkfs -t vfat, calls mkfs.vfat, which calls mkdosfs. :P
05:53:03furrywolfhah, it does show up. so rockbox has some bug that the partition table was triggering.
05:54:35furrywolfunfortunately, I did not image the card before repartitioning it, so short of buying a new card, I can no longer help debug this bug.
05:55:00*jhMikeS wonders what bpb_is_sane would have returned
05:55:04Bilgus_what does the partition show up as now?
05:55:19furrywolfsame thing it always has... one partition the size of the drive, vfat.
05:56:08Bilgus_./dev/sdc1: sticky x86 boot sector, mkdosfs boot message display, code offset 0x58, OEM-ID " mkdosfs", sectors/cluster 32, Media descriptor 0xf8, heads 255, sectors 249704448 (volumes > 32 MB) , FAT (32 bit), sectors/FAT 60960, serial number 0x2eb26ca7, label: " "
05:56:19furrywolfoh, you mean the filesystem. sec.
05:56:36furrywolf /dev/sdc1: sticky x86 boot sector, mkdosfs boot message display, code offset 0x58, OEM-ID " mkdosfs", sectors/cluster 32, Media descriptor 0xf8, heads 255, sectors 249737153 (volumes > 32 MB) , FAT (32 bit), sectors/FAT 60960, serial number 0xde34bd8e, label: " "
05:56:58Bilgus_well the sn changed lol
05:57:08furrywolfand the size, by a few sectors.
05:57:47furrywolfthere must have been something with the original partition table. I formatted it exactly the same way.
05:58:13furrywolfhowever, the original partition table was acceptable to both linux and the original firmware, so I would say it is by far most likely to have been a rockbox bug.
05:58:37furrywolfbut, since I didn't save a copy of the original partition table, I can't do anything to troubleshoot it now.
05:59:12jhMikeSreformat the first way and dump it?
05:59:21Bilgus_really not that big of a deal unless it happens next time
05:59:35furrywolfI only ever formatted it one way.
05:59:51furrywolfthe difference between not working and working was using cfdisk to create a new partition table
06:00:05Bilgus_well how did you do the partition table the first time>
06:00:06jhMikeSdoing it twice with the same program and settings gave different results?
06:00:19furrywolfBilgus_: it's a deal, because new users who have not repartition their sd card will find it simply doesn't work.
06:00:53Bilgus_surely the card came as exfat or other some shit
06:00:57furrywolfI didn't partition it the first time. the first partition table was the one the card came with from sandisk. I just reformatted it to fat32. (yes, with the proper partition type)
06:01:25furrywolfit was originally exfat, which is evil.
06:01:39jhMikeSRB won't accept any extended partitions (it could have been that)
06:01:51furrywolfno, it had a single primary partition, type vfat.
06:02:15Bilgus_then why differing sizes?
06:02:20furrywolfagain, the stock firmware handled it fine. it's a bug. but until I buy a new card, it's useless to try troubleshooting the bug.
06:02:43Bilgus_no the stock FW quit part way through
06:02:45furrywolfdunno. the original partition left a few mb of free space. maybe they want to make sure the controller has scratch room or something...
06:03:07Bilgus_nah you wouldnever see the extra blocks
06:03:25furrywolfthe stock fw quite halfway through when I had 110GB of music on it. it might have finished, I was just sick of how slow the progress bar was moving. when I had only a few files on it, the stock fw loaded and played it fine.
06:03:42furrywolfthe stock fw takes ungodly long to scan the disk... even my 32gb card takes a very long time.
06:03:56Bilgus_yeah it builds a damn index
06:04:18Bilgus_and I always forget to pull the damned SD card before I boot OF
06:04:26jhMikeSyou sure it was a single partition? that extra space could have been something rockbox didn't like
06:04:51furrywolfjhMikeS: cfdisk reported it as free space, and linux only showed a single partition.
06:06:04furrywolfI doubt there was anything wrong with the original partition table. I doubt sandisk is shipping cards with broken partition tables, linux had no errors using it, the stock firmware had no problems using it, and cfdisk read it fine. it's far more likely there's a bug in rockbox. :)
06:06:05Bilgus_well sorry you wasted 200 minutes to have to do it again
06:06:34furrywolfhowever, until I buy a new card, and it has the same problem, nothing I can do to figure out why.
06:07:03Bilgus_I think I'd rather put a note to reformat after deleting all partitions in the manual then try to track down a bug like that
06:07:59Bilgus_you could try reformatting to exfat then do your same steps and see if it reproduces the problem
06:08:38Bilgus_although I've always re partitioned sd cards before I use them alon with a bad block test
06:08:52furrywolfI just repartitioned it again, with a bit of free space, and it works fine, so it's not free space breaking it.
06:08:56Bilgus_same with SSD
06:09:35jhMikeSfurrywolf: yeah, that shouldn't bother it. it should skip over partition types it doesn't understand too.
06:09:57Bilgus_did you format to EXFAT then try your same steps?
06:10:53furrywolfI don't have any way to format exfat, or do anything else to exfat.
06:11:10furrywolfand anything that only affects partition contents is eliminated when you reformat fat32.
06:11:22furrywolfI'm copying music back onto it now.
06:12:07furrywolfas I said before, I don't think it's possible to troubleshoot it anymore.
06:12:13furrywolfthe original partition table is gone for good.
06:12:27furrywolfI could buy a spare card, if someone wants to send me cash. :P
06:12:50furrywolfgah, it's after 9pm. not copying music to it tonight.
06:13:29Bilgus_I'm sure if it is a RB bug the issue will come here again
06:14:00furrywolfcard was $47... not sure I want to order a spare.
06:14:49 Quit TheSeven (Ping timeout: 245 seconds)
06:15:21 Join TheSeven [0] (~quassel@rockbox/developer/TheSeven)
06:16:14Bilgus_good thing you didn't buy the extreme
06:16:31furrywolfwhy's that?
06:16:50Bilgus_the material that goes into those chips it's like printing money lol
06:17:02Bilgus_extreme would be more
06:17:03chrisjjAh, found one situation of Shuffle -> Off messing up the order every time: If you did Insert Next near the start of the list, Shuffle -> Off puts that track at the end.
06:17:07furrywolfit would have made all these steps a lot faster, and I wouldn't have to wait another half a day to fill it again. :P
06:17:35Bilgus_yeah but how often do you actually fill a card
06:17:57Bilgus_I do it once then buy another
06:18:14furrywolfchrisjj: I noticed two shuffle issues... the easily reproducible one is that if you use insert with shuffle on, they're not shuffled, and you have to turn shuffle off and back on. it does this every time. sometimes turning it off and back on still wouldn't cause shuffling, but I can't reproduce it.
06:21:37 Quit naleo (Ping timeout: 256 seconds)
06:22:21furrywolfI really wouldn't mind having another card, so I can use it in place of the (undersized) 64gb I use for music in the shop... but I don't know if I want to spend so much money.
06:24:00furrywolffor only $208 I can buy the 275MB/sec read, 100MB/sec write version. :)
06:28:09furrywolfbbl, wolfy bedtime
06:34:43 Join naleo [0] (~naleo@unaffiliated/naleo)
06:45:40***Saving seen data "./dancer.seen"
06:54:10 Quit Senji_ (Ping timeout: 255 seconds)
06:55:55 Quit furrywolf (Read error: Connection reset by peer)
07:06:57 Quit naleo (Read error: Connection reset by peer)
07:08:05 Join naleo [0] (~naleo@unaffiliated/naleo)
07:10:51 Join MrZeus_ [0] (~MrZeus@
07:11:56 Quit MrZeus (Read error: Connection reset by peer)
07:22:00 Quit jhMikeS (Ping timeout: 240 seconds)
07:33:03 Quit Bray90820 (Ping timeout: 258 seconds)
07:43:42 Join Bray90820 [0] (
07:47:09 Quit sparetire_ (Ping timeout: 240 seconds)
08:02:22 Quit naleo (Read error: Connection reset by peer)
08:05:59 Quit alexweissman (Remote host closed the connection)
08:07:07 Join alexweissman [0] (
08:10:14 Quit Bray90820 (Ping timeout: 258 seconds)
08:14:15 Join parchd [0] (~parchd@unaffiliated/parchd)
08:14:30 Join Bray90820 [0] (
08:29:09 Join ender` [0] (
08:37:53 Quit alexweissman (Remote host closed the connection)
08:45:44***Saving seen data "./dancer.seen"
08:49:40 Join petur [0] (~petur@rockbox/developer/petur)
08:54:26 Join wodz [0] (
09:03:00 Part robertd1
09:16:05 Join pixelma_ [0] (~pixelma@rockbox/staff/pixelma)
09:16:05 Quit pixelma (Disconnected by services)
09:16:10 Join amiconn_ [0] (~amiconn@rockbox/developer/amiconn)
09:16:10 Nick amiconn is now known as Guest18090 (~amiconn@rockbox/developer/amiconn)
09:16:10 Quit Guest18090 (Killed ( (Nickname regained by services)))
09:16:10 Nick amiconn_ is now known as amiconn (~amiconn@rockbox/developer/amiconn)
09:17:59 Quit Bilgus_ (Remote host closed the connection)
09:18:39 Join Bilgus_ [0] (~Bilgus@gateway/tor-sasl/bilgus)
09:26:22 Join robertd1 [0] (
09:47:52 Quit [Saint] (Quit: Quit)
09:57:12 Join dys` [0] (~dys@2003:5b:203b:100:6af7:28ff:fe06:801)
09:57:17 Join idonob_ [0] (~Owner@
10:02:13 Join sparetire [0] (~sparetire@unaffiliated/sparetire)
10:07:29 Join paulk-collins [0] (
10:14:25 Quit rela_ (Quit: Leaving)
10:14:40 Join rela [0] (~x@pdpc/supporter/active/rela)
10:18:41 Quit Bray90820 (Ping timeout: 260 seconds)
10:27:29 Join xorly [0] (
10:30:14 Quit mc2739 (Ping timeout: 248 seconds)
10:38:45 Quit rela (Read error: Connection reset by peer)
10:41:37 Join amazoniantoad [0] (~golden@2601:481:8400:a139:6040:c934:3c59:162)
10:41:48amazoniantoadIs down for anyone else?
10:42:33gevaertsIt'll come back. It's been on and off for a while now.
10:42:48gevaertsWork is being done on replacing the entire server, but that'll take a while still
10:42:59amazoniantoadI just received my ivue pack and the center panel button is missing >:C
10:43:56amazoniantoadWell I guess I can get my question answered here. New to rockbox, is there a theme that looks like iZune?
10:44:39 Join Bray90820 [0] (
10:45:03 Join elensil [0] (~edhelas@2001:1c02:1903:d800:b54b:cb14:8fad:a9e)
10:45:48***Saving seen data "./dancer.seen"
10:52:09 Quit Bray90820 (Ping timeout: 240 seconds)
10:55:06 Join Bray90820 [0] (
10:57:01 Quit paulk-collins (Read error: Connection reset by peer)
10:57:35 Join paulk-collins [0] (
11:00:39 Quit Bray90820 (Ping timeout: 240 seconds)
11:04:20 Join Bray90820 [0] (
11:06:46 Join [Saint] [0] (~sinner@rockbox/staff/saint)
11:07:30 Join pamaury [0] (~pamaury@rockbox/developer/pamaury)
11:09:18 Quit Bray90820 (Ping timeout: 240 seconds)
11:20:38amazoniantoadI'm looking for a place for tutorials on adding internal bluetooth to a 7th gen ipod class. Any ideas?
11:21:10 Join ZincAlloy [0] (~Adium@2a02:8108:8b80:1700:a127:13a6:2efd:6ac6)
11:26:46pamauryamazoniantoad: you mean physically adding bluetooth to the device ?
11:26:58amazoniantoadpamaury, that's the goal
11:27:05amazoniantoadI want to attach a transmitter
11:27:08amazoniantoadBut internally
11:27:26amazoniantoadI found a thread for the 5g but I can't follow it since it isn't my model
11:27:43pamauryI don't think it's really possible, in the sense that even if you attach a transmitter, you need to tell the transmitter which speaker to pair to and that requires software
11:27:58pamauryhow do you plan to control the transmitter ?
11:28:17amazoniantoadIt's possible, it's been done before
11:28:36amazoniantoadWell I have to 3D print a custom back for the ipod so that I can turn the bluetooth on/off
11:28:43amazoniantoadI have a 3D printer so I have that covered
11:31:33 Join asymsucon [0] (bcaf24b1@gateway/web/freenode/ip.
11:31:34shrizza_So it's pretty much just a lineout to bluetooth transmitter solution, but making it look neater.
11:33:14TorCamazoniantoad: Do you need the space inside the custom back for the bt module? If not, why not just mill an access hole to control the bt module?
11:33:30asymsuconHi, unfortunately the forums are down again, Mihail, you're here?
11:37:12 Quit amazoniantoad (Ping timeout: 255 seconds)
11:37:29 Join Bray90820 [0] (
11:41:34 Quit Bray90820 (Ping timeout: 240 seconds)
11:49:05 Join Bray90820 [0] (
11:50:22[Saint]asymsucon: check the nicklist
11:50:24 Join amazoniantoad [0] (~golden@2603:3018:600:a000:980a:457d:519a:8b90)
11:50:30[Saint](hint: he's not)
11:51:36 Quit Rower (Ping timeout: 255 seconds)
11:52:35 Join Rower [0] (
11:52:57amazoniantoadshrizza_, yes
11:53:02amazoniantoadSorry. I think I lost connection
11:54:31 Quit asymsucon (Quit: Page closed)
11:54:37 Join skapazzo [0] (~skapazzo@
11:57:20 Quit Bray90820 (Ping timeout: 264 seconds)
11:57:25[Saint]amazoniantoad: there's any number of about a dozen places you can get access to line level PWM and -3.3V/-5V traces from.
11:57:36wodzpamaury: Are you going to push your hwstub/mips work?
11:57:56[Saint]amazoniantoad: or you could be lazy as shit and just steal input from the 3.5mm jack
11:58:32[Saint]I mean, ideally you'd want to be pushing line level, but for all I know you might /want/ to have the iPod govern the volume.
11:58:39[Saint](IMO, you don't...but, hey)
11:59:25wodz[Saint]: Hey, as you are ipod expert. I have n2g with f*** up software on it in a state that it fires up emcore on boot and thinks it has 8gb despite being 4gb device. I tried to connect it to itunes in disk mode but aparently itunes does not detect it. Any trick needed to force recovery?
11:59:27pamaurywodz: yes, I have a small change to make
11:59:51pamauryI was chasing an undefined behavior in socdesc lib that was making the register file randomly shuffled
11:59:56pamauryfound it yesterday ;)
12:00:01[Saint]amazoniantoad: Depending on your BT device, there's a very good chance you might have to shift the logic from 3.3V to 5V and back again as well.
12:00:11pamaurywodz: I reworked a few things in hwstub
12:00:29pamauryhe you want to review them, they are on gerrit, otherwise I'll just go for it
12:00:40wodzpamaury: I saw. I crawled through gerrit tasks this morning
12:00:57pamauryI found a bug in data abort catching on arm, we were using an unpredictable nstructio
12:01:06pamauryit was more or less pure luck
12:01:15[Saint]wodz: what iTunes version?
12:01:24wodzpamaury: I saw your conversation with Mike.
12:01:31wodz[Saint]: Latest I think
12:01:42[Saint]wodz: iTunes is _very_ picky, and pretty much everything past 9.2.1 will fail.
12:02:23[Saint]ideally you want 9.2.1, but that's old enough to fail the lookup of the firmware so you can't rely on getting the firmware file from iTunes servers.
12:02:39pamaurywodz: did you have a look at all of them and have comments or can I freely push them this afternoon ?
12:02:59[Saint]You need to download it and when you're on the restore page, press [Shift], which allows you to provide an .ipsw file locally.
12:03:06wodz[Saint]: Oh shit. I always thought itunes is the p'n'p solution
12:03:40 Join Bray90820 [0] (
12:03:40[Saint]wodz: there's also
12:03:43wodzpamaury: I skimmed all of them. I don't have comments.
12:03:52pamauryok great
12:03:53[Saint]but ipod_scsi is a pain in the tits on Linux.
12:04:09pamaury[Saint]: why ?
12:04:46wodz[Saint]: Not working basically. The current state of this ipod is side effect of try to restore without itunes
12:05:26[Saint]pamaury: various distributions either have broken or nonexistent implementations.
12:05:48[Saint]I'm guessing wodz probably bounced off this very thing.
12:06:10pamaury[Saint]: how is that possible ? you only need linux kernel to send scsi commands
12:07:09wodz[Saint]: Basically I bounced off proper combination of emcore and emcore host tools + unable to time sync with TheSeven to guide me.
12:11:09[Saint]pamaury: well, it's a pain in the tits compared to 'just work'-ing, the only distributions I can think of that even offer an ipodscsi package are Fedora and Arch, and last time I tried Arch's it was entirely broken.
12:11:10[Saint]If you compile your own on the debian-esque distros you'll have luck, but by the time most users spend with that rigmarole you may as well have just fired up iTunes 9.2.* in a container.
12:11:52 Join mc2739 [0] (~mc2739@rockbox/developer/mc2739)
12:11:53pamauryah there is a package for it
12:12:51[Saint]Obviously it's subjective based on skill level, but for the majority of people dicking around with finding or building ipodscsi is an order of magnitude more difficult than just delegating to iTunes.
12:14:11 Quit amiconn (Quit: - Chat comfortably. Anywhere.)
12:14:11 Quit pixelma_ (Quit: .)
12:14:22 Join pixelma [0] (~pixelma@rockbox/staff/pixelma)
12:14:23 Join amiconn [0] (~amiconn@rockbox/developer/amiconn)
12:14:41wodz[Saint]: Is itunes ok?
12:15:04[Saint]The only problem with it is that to be able to reliably restore a iLoader/emCORE/ipod6gboot'ed iPod Nano 2G or iPod Classic 6G in iTunes you need a very specific iTunes version and to supply your .ipsw file manually.
12:15:10[Saint]wodz: it should be, yes.
12:15:21wodz[Saint]: Anyway proposed recovery action without itunes didn't work for this particular ipod definitely.
12:16:39[Saint]wodz: you'll need to pull the ipsw file from
12:17:11[Saint]iTunes 9.* has been neutered and cut off from the iTunes servers.
12:17:27[Saint]So you've got to supply the firmware manually from a local path.
12:18:00*wodz downloading stuff
12:20:20wodziTunes Library.itl cannot be read because it was created by a newer version of iTunes :P
12:20:40wodzWhere the hell does it sit?
12:20:40 Join Mihail [0] (25d4ab9c@gateway/web/cgi-irc/
12:21:22 Join Piece_Maker [0] (
12:22:13wodzsquashed !
12:24:17 Quit Acou_Bass (Ping timeout: 256 seconds)
12:24:17 Nick Piece_Maker is now known as Acou_Bass (
12:24:28wodz[Saint]: Now I have ipod in disk mode, how do I force it to recover in itunes?
12:26:09[Saint]there should be 'check for updates' and 'restore $device' options under the device header in the settings sidebar.
12:27:05[Saint]if you select the restore option, and hold shift, it'll prompt you for a local path to supply the ipsw file from instead of doing a lookup for it from the Apple servers.
12:28:02wodz[Saint]: You mean left hand side (where are 'Library', 'Store' and so on)?
12:28:49[Saint]yes. right up the top of that it should show the device, like 'wodz's iPod' or so.
12:29:08wodz[Saint]: Nop, nothing there
12:29:24[Saint]Ah...shit. Interesting.
12:29:31wodzMaybe I should reboot VM after installing itunes
12:30:02[Saint]does the VM have guest additions installed to be able to do USB passthrough?
12:30:09wodz[Saint]: yes
12:30:44[Saint]you /might/ have to hand off control of the iPod to the VM manually.
12:30:53wodz[Saint]: Apple iPod [0002] is passed to VM manually
12:31:48[Saint]Hmmm, right. Sorry, it's hard to switch out of #freemyipod-support mode where I primarily deal with people who don't know their ass from their elbow.
12:32:11[Saint]I should give you more credit than that. Apologies.
12:33:19*user890104 is not sure if the "sans-iTunes" method of restoring works on n2g
12:33:50user890104it works on n3g/n4g afaik
12:33:56[Saint]user890104: I've definitely got it to work before, but I can't quite remember the esoteric recipe for it.
12:34:31user890104ipodscsi won't be needed on n2g, as you already have complete access to the whole block device
12:34:36user890104using disk mode
12:34:40wodzuser890104: itunes doesn't work as well (at least for this combination of device/state)
12:34:55user890104wodz: do you have emcore loader installed?
12:34:59[Saint]Well, I should say "I've definitely got it to work before with very heavy prodding from TheSeven"
12:35:49wodzuser890104: on boot it says emCORE v0.2.3 r859
12:36:06wodzuser890104: unless forced to boot in disk mode of course
12:36:22user890104wodz: what about the split second just before that? can you reset it, and immediately lock the hold switch while connected to usb?
12:36:44[Saint]Regarding the virtual machine I think I ended up setting up a rule to automatically pass through all devices with the O5AC VID because the USB PID changes at least once during the restore and the virtual machine didn't like that one bit.
12:37:23wodzuser890104: I don't understand
12:38:28user890104wodz: there's recovery mode of emcoreldr, which is triggered by booting with the hold switch locked (similar to rockbox's restore defaults)
12:38:39[Saint]wodz: he's saying force a hard reset while connected over USB and immediately toggle hold on
12:39:03user890104using this, you can upload a new emcore installer, then reboot it, then select Uninstall from emcore's menu
12:39:13wodzuser890104: If I engage hold during apple logo I still get emCORE line but with faded backlight
12:39:38user890104wodz: in this state, can you check if the message says emcore loader or emcore?
12:40:06user890104hm, strange
12:40:37user890104ok, can you communicate with emcore using
12:40:59wodzuser890104: I could before, need to dig up the tool on this machine
12:44:01user890104you should be able to upload a new installer, which should overwrite the current emcore installation
12:44:15user890104and then you can use Tools -> Uninstall emCORE
12:44:33wodzuser890104: So I'd say yes, I can communicate
12:44:40user890104ok, looks good
12:44:45[Saint]user890104: that's only a part of the problem though
12:45:13[Saint]user890104: the bulk of the issue if that somehow an 8GB firmware ended up on a 4GB device.
12:45:32[Saint]that'll still be true with emCORE installed or no.
12:45:42user890104uhm, the firmware doesn't have a option for selecting the capacity
12:45:52***Saving seen data "./dancer.seen"
12:45:52user890104it auto-detects it from the flash chip
12:46:04user890104unless someone manually overwritten the nand flash
12:46:20[Saint]yes, that's what I meant, sorry - shitty wording.
12:46:43[Saint]"firmware magic from an 8GB device", not an 8GB firmware.
12:47:02wodzuser890104, [Saint]: The device was deliberately overwritten with payload from 8GB device, yes.
12:47:07user890104i'm not sure how these devices behave if someone /dev/zero'es the nand :)
12:47:39user890104in case of emcore, it should be able to boot anyway, and even show the default menu
12:48:01[Saint]I seem to recall this fucks up a bunch of magic if the firmware magic isn't what iTunes expects.
12:48:10wodzFTL was completly broken and TheSeven guided me.
12:48:11[Saint]Which is probably why iTunes isn't willing to detect it.
12:48:16user890104then you need to initiate a firmware update, which should upload an AUPD image on the nand
12:48:29user890104and after rebooting, it should do its thing (tm)
12:48:51[Saint]user890104: yeah, right - but what we're bouncing off at the moment is in the screwy state iTunes isn't detecting it.
12:48:53user890104so ipodscsi might actually work
12:48:59[Saint]So, chicken<->egg
12:49:22user890104if you can enter nor dfu mode, uploading a dfu image is easy
12:49:30wodzok guys. I'd prefer at most 3 steps recipe how to recover this ipod :-)
12:49:49user890104i just need to remember how (without opening and shorting a pin on the flash)
12:49:52[Saint]yeah, that involves actually opening the device though does it not?
12:50:02user890104there's a software method iirc
12:50:07[Saint]IIUC there's not a way to trigger NOR DFU from emCORE?
12:50:23[Saint]Oh. Hmmm. Interesting. Well, I'm not aware of it if there is one.
12:50:27wodzI could open it if needed but I prefer to avoid this.
12:50:30[Saint]I only know the brutal method.
12:50:45user890104wodz: no, it should be possible by software
12:51:09wodzcould it be just jump to some particular address or something?
12:51:16[Saint]wodz: yeah, these devices aren't particularly forgiving when opened. So fingers crossed user890104 digs up a software method to achieve this.
12:51:24user890104that's the info i have, looking for more in my archives
12:51:48[Saint]I didn't even know there was a method to do this from emCORE, so he's one up on me at this stage certainly.
12:52:07[Saint]I only know of the brutal but effective pin-shorting method.
12:52:13user890104well, since i was experimenting with this method, i guess there is one
12:52:37user890104[20:12:26] <user890104> emcoreldr was launched from nor dfu, if it matters <<< this
12:52:42[Saint]Oh, yeah. I don't doubt it. I just wasn't aware it could be done on N2G.
12:52:44user890104haven't opened my n2g ever
12:53:36[Saint]Lucky boy. I have one with a 2-pole dome switch broken out to the back of the case, lol.
12:54:38[Saint]I think I did that during the period where Rockbox would reliably trash the FTL on hard reset.
12:54:38user890104time for log grepping!
12:54:49wodzI replaced broken screen in this unit so it is possible to open, close and have still functional device. Then I played with it hard and this trashed FTL
12:55:29wodzuser890104: So do you suggest to run this 4 commands at the end of pastebin?
12:55:44user890104wodz: no, they are *after* you enter NOR DFU mode
12:56:15user890104and yes, they are the commands you need basicly
12:56:59wodzare this binaries prebuild or I need to compile from source?
12:58:28wodzuser890104: And doesn't exist standalone nowadays
12:59:09user890104(root)/emcore/trunk/tools/ - Rev 854
12:59:21user890104it's in the SVN trunk
12:59:49user890104you can use emcore-installer from here:
13:01:41user890104no idea about the version of the last two, but they should be good for the job
13:01:57user890104you can of course compile the whole firmware
13:02:38wodzuser890104: ^
13:04:35*wodz apologizes
13:05:05wodzuser890104: So if I understand correctly the last missing bit is entering nor dfu mode
13:05:21user890104yes, just found some instructions
13:07:34user890104ok, this isn't very helpful
13:09:08user890104neither is this:
13:10:31user890104wodz: found it:
13:10:42*wodz reads
13:11:36user890104enter NOR DFU mode (hold left+play after resetting)
13:11:41user890104that was the missing part
13:12:34wodzuser890104: How does nor dfu mode manifest?
13:14:31wodz\o/ nice logo with text :-)
13:15:17user890104and a dock connector?
13:16:01user890104you can follow TheSeven's instructions from this log, to completely wipe the nand
13:16:27user890104then you can enter this mode again, and use + apple's restore file for your ipod model
13:17:25wodzuser890104: Isn't it enough to connect to itunes in nor dfu mode?
13:17:56user890104wodz: i would wipe it first, in order to not confuse itunes
13:18:15wodzok, lets make the magic happen
13:18:40user890104it shouldn't trust the partition layout for obtaining flash memory size, but who knows...
13:19:25user890104after discovering that itunes sends a random memory buffer over usb after restoring a device, i don't trust them :)
13:19:40 Quit parchd (Ping timeout: 240 seconds)
13:23:50wodzthat gonna be bumpy road
13:24:56user890104ah, the info was on the wiki...
13:25:01user890104The Nano 2G also has a DFU mode, but this mode can only be entered by shorting testpoints on the circuit board or flashing the NOR with an image with a wrong signature/hash. There's a NOR DFU mode though, that can be entered by holding down BACK+PLAY right after rebooting the device.
13:25:57user890104wodz: what's your python version?
13:26:02wodzuser890104: 2.7
13:26:08user890104looks like 2.x, the scripts have been ported to 3.x
13:26:14wodzuser890104: ok
13:27:10wodzuser890104: Not that much better
13:27:49user890104you need to upload the .dfu file
13:28:00user890104not the .bin
13:28:27wodzstupid me
13:29:11wodzuser890104: That worked, one step further now
13:32:50wodzuser890104: Where to get emcore-ipodnano2g-r725.bin (or any other working r)?
13:34:06user890104this should be a trunk build with minimal (if any) modifications
13:35:46 Quit Mihail (Quit: - A hand crafted IRC client)
13:36:14wodzuser890104: Thats point 6 in
13:37:59user890104wodz: point 6 is execute
13:38:08 Quit Rower (Ping timeout: 264 seconds)
13:38:11user890104you mean 7?
13:38:20wodzyes, 7, my bad
13:38:34 Join Rower [0] (
13:38:36user890104can you retry uploading?
13:38:48user890104if not, please reset the ipod and start over
13:39:05user890104this part is fragile, as it's not well tested
13:39:37 Quit paulk-collins (Remote host closed the connection)
13:39:51pamaurywodz: re your comment on gerrit: I removed the verbose argument because it was a duplicate :)
13:40:04pamaury(there already was an entry for 'verbose' -> 'v')
13:40:10wodzpamaury: ok
13:40:34pamauryand nice catch for the header include guard
13:43:03wodzuser890104: On point 5 it says Uploading emcoreldr-ipodnano2g.bin to 0x22000000.... done but doesn't drop back to shell. Is it normal?
13:45:03user890104emcoreldr is slow, it could take a couple of seconds
13:46:34wodzyou men on upload or exec?
13:46:52user890104shouldn't take more than 2 minutes though
13:47:19wodzI'll wait then a bit longer. This 'done' message is misleading
13:48:15user890104ah, it says done
13:48:27user890104then it should be completed
13:48:40user890104just ctrl+c it
13:50:30wodzuser890104: should ipod reenumerate after execute 0x22000000 0 ?
13:51:58user890104yes, after a while, but the hold switch needs to be locked so it can enter emcore loader mode again
13:53:33wodzuser890104: seems like it doesn't reenumerate after execute. hold switch is engaged.
13:59:00wodzuser890104: btw. If I emcore is present on device (remember it defaults to emCORE r859) can't I simply run point 9 ?
14:00:11user890104wodz: in theory yes, i just don't know how messed your current instllation is
14:00:41user890104if you only messed up the nand flash, emcore should be intact
14:04:55user890104uhm... no idea about this one, maybe someone who knows python can help
14:05:23user890104can you try getting it back to DFU mode, and restoring with itunes?
14:05:57wodzheh, running it as python2.7 works
14:10:55 Join jhMikeS [0] (
14:12:43wodzuser890104: Now the ipod displays briefly 'Use itunes to restore' and enters disk mode. It is still not seen by itunes. If I force nor dfu mode and connect it, itunes says 'iTunes cannot recognize this iPod at this time'
14:16:13user890104i'm looking for the dfu image which restores nano 2g
14:17:15 Quit xorly (Ping timeout: 260 seconds)
14:18:15wodzFor me point 5 or 6 doesn't work correctly. After execute nothing happens but device stops responding and does not reenumerate.
14:19:05user890104wodz: check your private message
14:19:20user890104try to upload this file using
14:19:40user890104it is the original apple NOR loader
14:19:52user890104and it should show a different error message
14:20:05user890104so itunes can hopefully recognise it
14:20:34user890104if you setup a filter to redirect usb devices to your itunes VM, keep in mind that the usb ids for different modes are not the same
14:22:20user890104iirc the update process on nano2g is initiated by creating an AUPD image on the firmware partition
14:22:32user890104and ipodpatcher from rockbox should be able to do this
14:22:37wodzuser890104: The effect is the same as I described. It briefly shows 'Use itunes to restore...' and then it enters disk mode
14:22:40user890104just need to find the correct sommand line
14:24:02user890104wodz: what are the usb ids in lsusb?
14:24:41wodzuser890104: in which mode? In disk mode it is Bus 002 Device 052: ID 05ac:1260 Apple, Inc. iPod Nano 2.Gen
14:26:01wodzuser890104: in dmesg I also have New USB device found, idVendor=05ac, idProduct=1240 but don't know if this is nor dfu or this brief period of 'Use itunes...' message
14:31:10user8901041260 is normal mode
14:31:18user8901041240 is WTF mode
14:31:29user890104(yes, it's called WTF)
14:32:02*pamaury discovers that 0 is often the smallest unsigned integer
14:32:36wodzWTF :-P
14:32:41gevaertsThat's been known to happen, yes!
14:32:54wodzuser890104: so WTF it is?
14:33:30user890104in this mode you should be able to upload the file i sent you using
14:34:01user890104check if you can get the right timing (e.g. while [ 1 ]; do python ....; done)
14:35:01wodzuser890104: the file uploads just fine, apple logo appears and then 'Use itunes to restore...' and disk mode again
14:36:23user890104let me look for an ipodpatcher solution
14:38:41user890104−−write-partition (-w) filename.img
14:38:41user890104Restore a backup of the firmware partition. This can also be used to install a Firmware-X.Y.X firmware upgrade from Apple.
14:38:55user890104wodz: can you get ipodpatcher working?
14:39:04user890104then use this option
14:39:57user890104then download the most recent firmware from
14:40:10user890104unzip the ipsw and use the Firmware-xxxx file inside
14:41:55wodzuser890104: ipod is unrecognized by ipodpatcher
14:42:03user890104wodz: what is the usb id?
14:42:12user890104should be 1260
14:42:17wodzBus 002 Device 054: ID 05ac:1260 Apple, Inc. iPod Nano 2.Gen
14:42:59user890104can you post the output of ipodpatcher?
14:45:56***Saving seen data "./dancer.seen"
14:46:24user890104wodz: alternatively, you can restore a 4GB MBR and then dd the whole Firmware-xxxx file to the firmware partition
14:46:39user890104and format the data partition with fat32, just to make it happy
14:50:18wodzdang, considered me spammer :P
14:51:42wodzuser890104: for alternative I'd need 4GB MBR
14:51:44user890104uhm, does fdisk -l see it?
14:52:37 Quit bluebrother (Read error: Connection reset by peer)
14:52:37 Quit fs-bluebot (Read error: Connection reset by peer)
14:53:41 Quit Rower (Ping timeout: 248 seconds)
14:54:02 Join Rower [0] (husvagn@
14:54:33user890104wodz: try the 2gb one:
14:54:42user890104you should be able to resize later
14:55:18user890104then, after the partitions appear, ipodpatcher should see it
14:55:30wodzuser890104: lets try then
14:56:08 Quit idonob_ (Ping timeout: 264 seconds)
14:57:52wodzuser890104: Now ipodpatcher refuses with error '[ERR] Firmware partition doesn't contain Apple copyright, aborting.'
14:58:27 Join bluebrother [0] (~dom@rockbox/developer/bluebrother)
14:58:27user890104just dd the Firmware-xxx file to /dev/sdb1
14:58:38user890104or whichever one is the smaller partition
14:58:51user890104then "eject /dev/sdb"
14:59:00user890104and it should reboot & reinstall apple's firmware
14:59:14user890104if not, give it a push by rebooting manually
15:00:04wodzuser890104: Now I spotted there are two versions of firmware called 19/1.13 and 29/1.1.3. Which should I use?
15:00:15user890104i think it doesn't matter
15:00:21user890104haven't found a difference so far
15:04:50wodzNo dice. On boot it enters 'Use iTune to restore' mode but on connecting usb it immediately goes to disk mode
15:05:03wodzuser890104: ^
15:06:11wodzuser890104: ipodpatcher still complains about the lack of apple copyright in apple's partition
15:08:54 Join Rondom [0] (
15:08:55 Join fs-bluebot [0] (
15:09:17wodzuser890104: success - I forgot to unpack Firmware file from .ipsw. Now after dd if=Firmware-xxx device updated itself and booted to regular apple firmware
15:10:14wodzthe only thing left is that it thinks it is 2GB version
15:13:35user890104wodz: it should be safe to delete the data partition
15:13:53user890104then recreate it using the maximum allowed size, but keeping the start sector the same
15:14:02user890104then you can format it with mkfs.vfat
15:14:59 Join paulk-blaze [0] (
15:17:27wodzuser890104: seems to be working :-)
15:19:06wodzuser890104: Thanks a bunch
15:19:08fs-bluebotBuild Server message: New build round started. Revision 6ef3f7c, 255 builds, 13 clients.
15:24:00 Quit wodz (Ping timeout: 240 seconds)
15:25:17 Quit paulk-blaze (Ping timeout: 258 seconds)
15:25:26amazoniantoad[Saint], could you tell me more? I'm desperate to get this done.
15:27:50[Saint]amazoniantoad: well, if you're content with using the device to govern the output volume, stealing L+R+G +/- 3.3V from the 3.5mm TRS should be pretty self explanatory.
15:27:53amazoniantoad[Saint], I'd really appreciate it if you could show more more of what you're talking about.
15:28:15amazoniantoad[Saint], I can program. I don't know anything about electrical engineering
15:29:32[Saint]It really depends on factors that are unknown to me, and quite possibly unknown to you. Is the bluetooth radio you have active or passively powered?
15:29:49[Saint]Is it 3.3, 5, or 6 volts?
15:30:31amazoniantoad[Saint], Bluetooth from this tutorial
15:31:07amazoniantoadI got the exact model. I have the ipod opened up as I'm going to 3D print the back and was going to apply the iVue to the front of the ipod. But the person that sent me the parts left out something. So I have to give them a call
15:31:18amazoniantoadIf there is a way to determine its voltage I can do that on my end.
15:34:21amazoniantoad[Saint], I figured I could just do exactly what this guy did as he posted nice photos but it's a 5th gen ipod. Internals are different.
15:35:38[Saint]Yeah, that was a really dangerous assumption. The two are absolutely nothing alike internally.
15:35:57amazoniantoadWell thankfully I didn't go through with anything :p
15:36:07amazoniantoadI was pretty upset last night once I realized my mistake
15:37:15amazoniantoadDoes [Saint] here is the bluetooth device:
15:39:59fs-bluebotBuild Server message: Build round completed after 1252 seconds.
15:40:02fs-bluebotBuild Server message: Revision 6ef3f7c result: All green
15:40:03fs-bluebotBuild Server message: New build round started. Revision fdb98c2, 255 builds, 14 clients.
15:40:25[Saint]Yeah, I got that far, but it doesn't really tell me anything at all other than the fact that it is actively powered, but annoyingly has its own internal battery. It doesn't however give any indication as to the makeup of that cell.
15:40:37pamauryphew, that was a lot of commit
15:41:03amazoniantoadWould showing you the battery help at all?
15:41:09amazoniantoadI have it opened up.
15:41:47[Saint]The battery should have the voltage printed on it unless they're complete dicks.
15:42:06[Saint]It's almost certainly going to be 3.7V, but I wouldn't like to assume that.
15:42:46amazoniantoad[Saint], 3.7V 100mAh N30AMW
15:54:41fs-bluebotBuild Server message: Build round completed after 877 seconds.
15:54:42fs-bluebotBuild Server message: Revision fdb98c2 result: All green
15:54:43fs-bluebotBuild Server message: New build round started. Revision a1d1832, 255 builds, 13 clients.
15:57:59[Saint]amazoniantoad: sorry - I was buggering around for a bit on another project. ok, that makes things easy then.
15:57:59[Saint]You don't really need to touch the mainboard at all, honestly, I'm not entirely sure why the "genius" that came up with this did it that way, considering the choice is between potentially screwing up a combined total of $40 of hardware, or $150 of hardware.
15:58:33 Join xorly [0] (
15:58:37amazoniantoadYou're fine. So you're suggesting I do it differently than he did?
15:58:56[Saint]You can steal 3.7V +/- right off the iPod battery, and 3.3V +/- from the 3.5mm TRS jack.
15:59:20[Saint]Screwing up either one of those is considerably cheaper.
15:59:20amazoniantoad[Saint], How do I find the TRS jack? Do you mean the audiojack?
15:59:40[Saint]I do. TRS == Tip Ring Sleeve
16:00:50amazoniantoadSo my next question is, this guy appeared to solder two speaker wires, one to one component and the other to another component on the motherboard. How do I determine what is left speaker v. right speaker?
16:01:38user890104i think you can swap them in rockbox if you mix them up
16:01:50amazoniantoadI have rockbox already loaded on there.
16:02:21 Join cc___ [0] (~ac@2001:910:113f:1:6a05:caff:fe1c:1627)
16:03:46amazoniantoadSo how do I find *where* to steal power from the battery and *where* to solder onto the TRS jack.
16:04:42user890104do you have a multimeter?
16:04:51[Saint]Battery has very obvious wires coming out of it that are equally obviously colour coded for pos/neg and thermistor.
16:05:00[Saint](thermistor is of no value to you)
16:05:14[Saint]TRS stereo is entirely standardized.
16:05:38[Saint]Tip == left, ring == right, sleeve == ground
16:05:53amazoniantoaduser890104, yes
16:06:17amazoniantoad[Saint], thanks very much
16:06:26amazoniantoadI really appreciate your advice. Both of you guys
16:06:48amazoniantoadDo you stay here regularly? In case I need advice in the future with regards to this project?
16:06:57fs-bluebotBuild Server message: Build round completed after 737 seconds.
16:06:58fs-bluebotBuild Server message: Revision a1d1832 result: 5 errors 0 warnings
16:07:25user890104we do usually
16:07:36amazoniantoadThanks again guys
16:07:41amazoniantoadThis sounds very possible now :p
16:07:48[Saint]The battery isn't particularly ambiguous at all, but if you have any doubts whatsoever about the TRS you can use your multimeter and the balance feature of Rockbox to swing between left and right to derive left and right.
16:07:49user890104so, set your multimeter to 20 V DC, and measure the battery cables
16:07:50amazoniantoadPretty excited. T
16:08:09user890104or to DC Voltage, if the scale is automatic
16:08:17[Saint]beat me to it.
16:08:32user890104red should be positive, black should be negative
16:08:51user890104you'll get 3.5-4.2 V depending on the charge
16:09:13amazoniantoadPretty excited about this
16:09:14[Saint]tip /should/ be left, and ring /should/ be right.
16:09:20[Saint]sleeve is alwats ground.
16:09:55[Saint]Apple has been known to flip TRS L|R channels before, so you might want to use that multimeter to make sure.
16:10:09amazoniantoad[Saint], do I need to remove these parts from the back in order to see the tip?
16:10:47user890104what's your ipod model?
16:11:56amazoniantoaduser890104, it's a 7th gen ipod classic
16:12:06amazoniantoadSorry. 6g
16:12:30[Saint]technically speaking 7g doesn't exist, people kinda invented it.
16:12:52amazoniantoadYou're the expert. I'll go with whatever you say at this point :p
16:13:55user890104soldering wired to the filtering capacitors should work
16:14:04user890104as in the original post
16:14:17user890104you just need to find which are these
16:15:00[Saint]user890104: my suggestion was to target the 3.5mm audio/hold assembly, as there's (albeit very fine pitch) testpoints on it.
16:15:07[Saint]And it's much cheaper to fuck up.
16:15:22[Saint]one sec.
16:16:00user890104this guy removed the filtering caps and used line out
16:16:27user890104so here are the locations of them
16:16:42[Saint]you'd need to use the multimeter to figure out which testpoint was what exactly, but this is the way I did it for my 5.5G
16:16:56amazoniantoaduser890104, what does this purple wire do exactly?
16:17:22 Join furrywolf [0] (
16:17:33user890104he bypasses the headphone amplifier, in order to connect an external (and probably better) amp outside the ipod
16:17:41user890104and still use the 3.5 mm jack
16:18:03user890104you don't need to do this, just use it as a reference of where to soler your wires
16:18:22user890104or take [Saint]'s approach, and solder them to the headphone assembly
16:18:39amazoniantoadI think I will stick with that just because I'm so inexperienced
16:18:48user890104i would be careful with the latter one, as it looks (and probably is) very fragile
16:18:50amazoniantoadI really don't want to screw anything up
16:19:08[Saint]user890104: I'm not entirely positive that's a 6g mainboard in that image.
16:19:23user890104ah.. sorry
16:19:27user890104that's an ipod video
16:19:57[Saint]yeah, I thought so, the layout is entirely different.
16:20:57user890104i can try to find these on my ipod when i get home
16:21:01user890104and take a picture
16:21:11user890104no promises though
16:21:17amazoniantoadAwesome. Thanks
16:21:17[Saint]That would be useful, if only for information's sake.
16:21:34user890104an easier way would be using the headphone jack, or the dock connector
16:21:47user890104sacrificing an old chinese apple dock cable
16:22:14[Saint]that would make it external though.
16:22:25amazoniantoadI'm doing all of this to avoid that
16:22:27[Saint]...which actually, isn't a bad idea.
16:22:38[Saint]but if you /want/ it internal...
16:22:53amazoniantoadWell remember I'm already going to print out my own back casing. So the signal should be fine
16:23:04user890104or, you can cheat and solder to the dock connector :)
16:23:25amazoniantoadI'm going to try using the audio jack first.
16:23:27user890104you need pin 1 or 2 for ground and 3/4 or 5/6 for audio signal
16:23:38[Saint]the way I did it with my 5.5G was to target the audio and hold assembly using 26 AWG wire and fine solder paste.
16:23:46user890104if you choose line-out, you'll need to control the volume using bluetooth
16:24:16 Join Senji [0] (~Senji@
16:24:17user890104if you choose headphones (5/6 pin), you can control it from rockbox, and the "side effect" described in the OP will work
16:24:19 Join Senji_ [0] (~Senji@
16:24:27user890104ah, sorry
16:24:32user8901045/6 is audio IN, not OUT
16:25:03[Saint]soldering from the dock connector on the logic board does indeed give a better platform for stability, and the option of using either line level or audio out.
16:25:40user890104[Saint]: looks like there's no amplified audio output on the dock connector
16:25:47 Join girafe [0] (
16:25:53user890104only line-in/mic
16:26:45[Saint]that can't be true...else they wouldn't work in iPod docks, and they most certainly do.
16:27:00[Saint]my clock radio knows that well.
16:28:17user890104you can control the volume of the radio from the ipod?
16:31:06[Saint]though it's possible that they're doing it with some IAC magic.
16:32:11[Saint]I honestly have no idea. It's actually highly probable they're using the data lines and IAP magic, come to think of it.
16:32:49[Saint]Because the iPod screen becomes a nice fullscreen digital clock when docked using the OF.
16:33:21 Quit girafe (Quit: Leaving)
16:37:41 Quit xorly (Ping timeout: 248 seconds)
16:39:13user890104[Saint]: i guess they're streaming the audio over usb, involving IAP
16:39:30user890104that's how it works on Kia Ceed's default radio
16:39:34[Saint]yeah, you're probably right.
16:39:50[Saint]it didn;t immediately occur to me.
16:39:54user890104it uses a regular usb-to-dock cable
16:40:17 Quit Rower (Ping timeout: 240 seconds)
16:40:21user890104too bad i didn't have a usb 2.0 sniffer at hand
16:40:44 Join Rower [0] (
16:40:49user890104tested my ipod colletion on a friend's car, managed to lock it up when connecting nano 2g
16:41:18user890104restarting the ignition fixed it
16:42:28user890104the car's iap implementation just shows a small b/w logo in the center of the screen
16:44:45 Quit cc___ (Ping timeout: 255 seconds)
16:46:00***Saving seen data "./dancer.seen"
16:48:46 Quit Bray90820 (Ping timeout: 276 seconds)
16:49:16 Quit Senji_ (Quit: KVIrc 4.9.2 Aria
16:53:01 Join Bray90820 [0] (
17:09:45 Join cc___ [0] (~ac@2001:910:113f:1:6a05:caff:fe1c:1627)
17:20:25 Join alexweissman [0] (
17:21:20 Join asymsucon [0] (bcaf24b1@gateway/web/freenode/ip.
17:26:08 Quit Rower (Ping timeout: 264 seconds)
17:26:16 Join Rower [0] (
17:28:57 Join paulk-collins [0] (
17:30:52 Quit petur (Quit: Connection reset by beer)
17:34:20 Quit Bilgus_ (Remote host closed the connection)
17:34:45 Join Bilgus_ [0] (~Bilgus@gateway/tor-sasl/bilgus)
17:38:05 Join mutnai [0] (6db91733@gateway/web/freenode/ip.
17:38:42 Quit asymsucon (Quit: Page closed)
17:43:55 Join Senji_ [0] (~Senji@
17:46:32 Quit Senji (Ping timeout: 264 seconds)
17:56:00 Quit cc___ (Ping timeout: 240 seconds)
18:02:11 Quit [Saint] (Remote host closed the connection)
18:03:32 Join [Saint] [0] (~sinner@rockbox/staff/saint)
18:06:00 Quit elensil (Quit: Leaving.)
18:06:31 Join petur [0] (~petur@rockbox/developer/petur)
18:08:06 Join naleo [0] (~naleo@unaffiliated/naleo)
18:19:00 Quit dys` (Ping timeout: 240 seconds)
18:24:48 Quit mutnai (Ping timeout: 260 seconds)
18:32:19 Quit Rondom (Remote host closed the connection)
18:33:35amazoniantoad[Saint], I have the thick 6th gen ipod. I remove the battery but all I see coming out of the battery is a single ribbon.
18:37:15 Join Strife1989 [0] (~Strife89@2602:306:bce1:8c20:5a94:6bff:fe41:c78)
18:37:39 Nick Strife1989 is now known as Strife89|work (~Strife89@2602:306:bce1:8c20:5a94:6bff:fe41:c78)
18:39:15dyspamaury: they appear to have a model that ist just an ADC+AMP without the DAP functionality. Is has the same Blackfin DSP. Maybe the interesting things happen there, and they put the A9 in there to handle display+sdcard and feed the blackfin with it.
18:39:26dysstill doesn't explain the lack of strings in the firmware :-/
18:40:42dysalso, I got a cheap used one in the meantime. This thing ist HUGE. About half the size of a portable CD player.
18:41:33 Quit Rower (Ping timeout: 240 seconds)
18:41:42 Join Rower [0] (
18:46:04***Saving seen data "./dancer.seen"
18:46:39 Quit Senji_ (Ping timeout: 240 seconds)
18:49:13 Join Senji_ [0] (~Senji@
18:49:56dys(used ones are rather cheap. they routinely get abysmal reviews describing the many ways the stock firmware sucks.)
18:54:09 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
19:08:34 Join xorly [0] (
19:10:01 Quit xorly (Client Quit)
19:27:04 Join xorly [0] (
19:27:39 Quit Rower (Ping timeout: 240 seconds)
19:27:40 Join parchd [0] (~parchd@unaffiliated/parchd)
19:27:47 Join Rower [0] (
19:33:10 Join edhelas [0] (~edhelas@
19:36:28 Join lebellium [0] (
19:44:18 Quit Bray90820 (Ping timeout: 255 seconds)
19:48:36 Quit edhelas (Quit: Leaving.)
19:55:50 Join Bray90820 [0] (
20:00:49 Join MrZeus2 [0] (~MrZeus@2a02:c7f:7008:3400:7053:4c15:1b68:c818)
20:03:24 Join edhelas [0] (~edhelas@
20:04:25 Quit edhelas (Client Quit)
20:11:01 Quit furrywolf (Read error: Connection reset by peer)
20:16:09 Quit parchd (Ping timeout: 240 seconds)
20:18:11 Join cc___ [0] (~ac@2001:910:113f:1:6a05:caff:fe1c:1627)
20:46:05***Saving seen data "./dancer.seen"
20:48:28 Join furrywolf [0] (
20:52:20 Join girafe [0] (
20:59:37 Join shdwprince [0] (~textual@
21:01:28amazoniantoaduser890104, are you there?
21:01:38 Quit thomasjfox (Quit: Konversation terminated!)
21:03:22 Join thomasjfox [0] (~thomasjfo@rockbox/developer/thomasjfox)
21:06:55 Join johnb3 [0] (
21:07:47 Quit Bray90820 (Ping timeout: 255 seconds)
21:08:34amazoniantoaduser890104, so I am looking at the battery and audiojack. They don't appear to be what you both described
21:08:50amazoniantoadThe battery just has a ribbon and the audio jack is just this metal box sitting on top of a ribbon
21:10:04user890104can you take a picture of your board, preferably with a flash using a decent smartphone or a camera?
21:10:21amazoniantoaduser890104, yes. plz hold
21:12:20 Quit johnb3 (Ping timeout: 264 seconds)
21:15:37amazoniantoaduser890104, sent the pics to my email. it will take a minute or so
21:20:30 Join Bray90820 [0] (
21:24:33 Join johnb3 [0] (
21:24:39 Join asymsucon [0] (bcaf24b1@gateway/web/freenode/ip.
21:25:38 Quit Bray90820 (Remote host closed the connection)
21:26:00 Join Bray90820 [0] (
21:28:51user890104amazoniantoad: can you take a picture of the ipod's main circuit board?
21:29:09user890104soldering on these ribbon cables isn't a good idea
21:31:15dyspamaury: ugh, you were right with your doubting of the A9
21:31:52dysI thought that pcb shot from the russian site was part of the pcb-sandwich that the others didn't bother to unsolder
21:32:10dysit turns out it from some other player they compared to this one
21:32:25dysso there's only the blackfin in there
21:33:26 Join mutnai [0] (6db91733@gateway/web/freenode/ip.
21:33:29 Quit asymsucon (Quit: Page closed)
21:33:39dysguess i'll build a blackfin-binutils tomorrow to take another look at the firmware images
21:36:35pamaurydys: but what about instruction set ? it looked like thumb code, and blackfin has its own instrution set I think
21:37:11dysyea, it looked suspicously like thump2. all that push instructions with the LR register last looked like a calling convention
21:37:33dysI started doubting again when I couldn't find any of the rockchips MMIO magic values
21:37:47 Quit johnb3 (Quit: Nettalk6 -
21:37:55dysthey would have to be in the assembly somewhere
21:39:10dyswith that pcb off the table, there's only a 128Mb RAM and an soic-8 flash in there
21:39:17dysnow the firmware size makes size, too
21:39:38amazoniantoaduser890104, any ideas?
21:41:01user890104yes, looking at other board pictures at the moment
21:41:06amazoniantoadah. k
21:41:39pamaurydys: you have the player now right ? so what the processor on the pcb ?
21:42:24dysI tried to take it apart, but the LCD is glued to the acryl which is glued to the case :-/
21:42:53dysaccording to the PCB shots from the net (which I'm pretty sure now are in there), theres a ADSP-BF606 (2-Core 500MHz Blackfin DSP)
21:45:45dysthere's a non-official gcc/binutils port on sourceforge, so I haven't lost all hope
21:46:15Bilgus_amazoniantoad, user890104 there is a PTC 'fuse' to the right in that picture why not solder to that?
21:46:36amazoniantoadBilgus_, I know nothing about electronics. Can you circle it in the pic for me?
21:46:47amazoniantoadBilgus_, also what does that fuse do?
21:47:13Bilgus_you do have a multimeter correct?
21:47:32Bilgus_the fuse will either protect the board or the battery
21:47:54*pamaury looks at ADSP-BF606
21:48:41user890104amazoniantoad: can you take a picture of the clickwheel side?
21:48:48amazoniantoaduser890104, yes
21:48:50user890104the dock connector pins should be visible there
21:48:55pamauryat least it has 32-bit registers
21:49:29user890104you can use 1 or 2 for audio ground (sleeve), 3 and 4 for left and right audio signal (tip and ring)
21:50:22amazoniantoadBilgus_, what does the blue v red do?
21:50:28Bilgus_amazoniantoad, the red circled is the PTC and the blue circle should have ground
21:51:03amazoniantoadBilgus_, so PTC provides power?
21:51:08user89010415 and 16 should also be usable as ground for the power output
21:52:18amazoniantoaduser890104, can you show me what you mean? I'm sorry I don't quite understand what you mean by 1/2/3/4
21:52:44user890104look at the picture and diagram at top left
21:53:00amazoniantoadIs this compatible with 6g?
21:53:04amazoniantoadThis says just 2/3
21:53:54user890104all ipods with dock connector are compatible, there are slight differences in which ones outputs video, and in one of the other pins, about indicating charging current
21:54:11user890104also iphones, ipod touch, basicly every apple product with a dock connector
21:54:19amazoniantoadso how do I access these pins? At the docking location at the bottom oft he ipod?
21:54:47user890104the pins should be visible on the front side of the board
21:54:52Bilgus_the ptc will have power you just need to figure out which side is the supply and which side is the protected I imagine you want the side with the '+' judging by the pic
21:54:55user890104but i'm not sure which ones are they
21:55:11pamaurydys: the ADSP-BF606 doesn't seem to support usb boot though, that makes experimenting more tricky, but it does support rsi(emmc/sd) boot, so potentially you could boot from micro-sd card (assuming it has such thing). I assume there is spi flash on the pcb ?
21:55:41amazoniantoadBilgus_, so if I use that for power how do I get audio? user890104: you mean the things underneath the click wheel?
21:55:45Bilgus_I picked the Ptc and gnd side of those diodes because it was nice large pads
21:55:50dyspamaury: it does have a card and what I think is an soic-8 flash
21:56:07Bilgus_audio don't know yet
21:56:37 Join TheLemonMan [0] (~root@irssi/staff/TheLemonMan)
21:56:51 Quit naleo (Ping timeout: 276 seconds)
21:57:40pamaurydys: the lack of string is puzzling though
21:58:58 Quit Rower (Ping timeout: 245 seconds)
21:59:10user890104amazoniantoad: can you check the voltage between the pins which Bilgus_ marked?
21:59:47 Join Rower [0] (
21:59:49amazoniantoadthis will be my first time so bear with me
21:59:55 Join Massa [0] (4fe8ddcd@gateway/web/freenode/ip.
22:00:17dyspamaury: ja. I'll try some bit-shuffling now. only 40320 ways to arrange 8 bits into a byte :-)
22:01:03Bilgus_as for the audio I'd probably go on the connector that the cable plugs into but thats some pretty fine pitch for a novice
22:01:25pamaurydys: ah good idea. I tried XORing, but after an hour playing with it, I am not convinced it uses XORing
22:01:33 Join naleo [0] (~naleo@unaffiliated/naleo)
22:02:09dyssince this a device from japan, maybe they use utf-16 or something. that would mess up strings() pretty good, too
22:03:05amazoniantoadBilgus_, I'm very good when provided instructions :p
22:03:12amazoniantoadIf you guys wouldn't mind a bit of hand holding
22:03:16amazoniantoadhold on setting stuff up
22:03:36pamaurydys: but ascii characters would still be encoded as xx 00
22:03:41Bilgus_flux would be the key to not bridging the whole damn thing
22:04:03 Quit xorly (Ping timeout: 240 seconds)
22:04:32pamauryI tried "strings" utility with various encodings to no avail
22:05:32dyspamaury: ah, I see
22:06:10pamauryI'm looking at the datasheet and programming manual of the dsp
22:07:29amazoniantoadSo uhm....don't laugh
22:07:48amazoniantoadred/black wires and I have options 10A, mA, and COM
22:07:53amazoniantoadWhat goes where?
22:09:00amazoniantoaduser890104, Bilgus_
22:09:24user890104red into ma, black into com
22:09:29Bilgus_the mA one should have V on it too
22:09:34user890104multimeter set to 20 V DC
22:09:47user890104or DC voltage, if the scale is automatic
22:10:10Bilgus_whatever you do don't set it to A or MA that will be a direct short
22:10:19 Join xorly [0] (
22:10:41amazoniantoadabout to show a pic for confirmation
22:11:22amazoniantoaduser890104, Bilgus_
22:12:11amazoniantoadso I'm going to focus on the red circle first:
22:12:31amazoniantoadRed wire goes to +? black wire goes to the other side?
22:12:43Bilgus_you will put the red lead on there and the black on the blue circle
22:13:22pamaurydys: I'm reading blackfin hardware reference for ADSP-BF06x and it describes the boot format. It's not encrypted or anything. Thus I conclude that the flash possible contains a bootloader that further loads the firmware. And this firmware might be encrypted
22:13:24Bilgus_only way doing it with it on both side of the red circle would be if the fuze was blown
22:13:40amazoniantoadBilgus_, there are several pieces of metal in the blue circle. Which metal piece?
22:14:11Bilgus_directly on the side of the two larger black rectangles
22:14:57amazoniantoadso touch both of the pieces of metal?
22:15:02amazoniantoadfrom both black squares?
22:15:47Bilgus_one or the other hold on a sec
22:16:10dyspamaury: ja, I guess it's some primitive scrambling, since the entropy of the firmware is very low
22:16:29pamauryI'm still reading, I may have missed something
22:17:17amazoniantoadBilgus_, user890104 did that and got no reading. Did red on red circle, on the + side. Did left black square, then right, then both. No reading.
22:17:35Bilgus_you will need the batter hooked up
22:18:50amazoniantoadBilgus_, should the ipod be powered on?
22:18:54amazoniantoadOr should I turn it off?
22:19:02Bilgus_if its on so be it
22:19:10Bilgus_just don't short it out
22:19:49amazoniantoadNo reading
22:21:05amazoniantoadSo what did this tell you?
22:21:12amazoniantoadThat I can get power?
22:22:17Bilgus_yep it has battery voltage now try turning it off and see if it goes to 0
22:23:28Bilgus_I'd want it to turn off automatically personally but I doubt you will get that lucky
22:24:02amazoniantoadWell I plan on allowing myself to turn the bluetooth module on/off
22:24:08amazoniantoadbrb need to do something for a sec
22:26:09 Quit thomasjfox (Quit: Konversation terminated!)
22:26:29amazoniantoadBilgus_, I only have the option to reboot and boot into disk mode
22:26:32amazoniantoadNot to turn it off
22:26:39amazoniantoadShould I just remove battery and test it?
22:26:57Bilgus_oh then no matter nah when you pull the battery it would be off
22:27:24amazoniantoadSo what's next?
22:27:30Bilgus_I'm not an Apple guy just electronics in general
22:28:04Bilgus_next is finding your audio points
22:28:20dyspamaury: another argument against XOR is that there are too many 0000 0000 sequences in the file. so I guess the scrambling would be something that leaves 0 idempotent (like bit shuffling)
22:28:43dysnow i need to get my subset enumeration algorithm foo up to speed…
22:29:36Bilgus_the way I see it you have 3 choices, 1. the ribbon cable where the jack is soldered on, 2 the connectors pins that the cable plugs into, 3 some points on the board that have audio before the amp
22:31:25amazoniantoadWhich do you think would be easiest for me?
22:32:33Bilgus_cheapest if you screw up will be that flex cable, thing about flex cables is that they don't like being over heated
22:32:56Bilgus_do you have flux?
22:33:07amazoniantoadWhat is flux?
22:33:52Bilgus_flux is essentially tree sap or some petroleum derivative that helps carry the heat and make the solder flow
22:34:15amazoniantoadI can go buy some. Would walmart have it?
22:34:27Bilgus_probably not
22:34:45__builtinBilgus_: a.k.a. "rosin?"
22:34:50Bilgus_well they might have plumbing flux but it is a nono for electronics as it has acid
22:35:02amazoniantoadWhat if I use something plasticy?
22:35:11amazoniantoadI forget what it's called...silicon?
22:35:14Bilgus_yes rosin is the typ electronics flux
22:35:26__builtinyou can probably get some from a music store
22:35:39amazoniantoad__builtin, I can get gooy rosin?
22:35:43amazoniantoadI have rosin, just hard rosin
22:35:50Bilgus_well thats for string instruments not realy the form we need
22:36:04amazoniantoadBilgus_, what happens over a long period of time? Won't it dry out?
22:36:10Bilgus_I guess you could heat it up
22:36:43Bilgus_nah you typically was it off with high proof rubbing alcohol when you are done
22:36:59Bilgus_the solder displaces it
22:37:51__builtinI think it shouldn't matter what it's "intended" for
22:38:23amazoniantoadBilgus_, we will get back to that later
22:38:28Bilgus_here ya go a nice primer on flux
22:38:30amazoniantoadSo what's your idea with the flex cable
22:39:12Bilgus_take a picture of the metal box flipped over
22:39:30amazoniantoadBilgus_, you mean the back of the ipod?
22:40:03amazoniantoadk hold on
22:40:08Bilgus_looks like it is affixed to a foam pad
22:46:06***Saving seen data "./dancer.seen"
22:46:15Bilgus_give me a few to find the points
22:47:23amazoniantoadall right
22:47:26Bilgus_will that white plastic come off without destroying it?
22:47:41amazoniantoadI don't think so
22:50:39Bilgus_ok do you have a pair of shitty head phones you can destroy?
22:50:54Bilgus_we just need the headphone plug
22:51:07amazoniantoadhold on\
22:53:07amazoniantoadBilgus_, some old broken bose headphones
22:53:18pamaurydys: but the thing definitely looks periodic, there are lot of very similar patterns
22:53:47Bilgus_sure whatever if you can get to the speakers you can just probe the soldered wires
22:54:00 Quit Strife89|work (Quit: Leaving)
22:54:32amazoniantoadonly one speaker works
22:54:40amazoniantoadshould I get one with two speakers?
22:54:54amazoniantoadI have more headphones
22:55:37Bilgus_probably be a good idea to have a good cable
22:55:44Bilgus_do you have a patch cable?
22:55:55amazoniantoadidk what that is
22:56:07Bilgus_double ended headphone cable
22:56:23Bilgus_or rca on one end and headphone jack on the other
22:56:25amazoniantoadI can go buy one
22:57:03amazoniantoadSo the list of materials you want me to get is a patch cable and some shitty headphones, yes?
22:57:17Bilgus_one or the other
22:57:29amazoniantoadLet me go get some headphones
22:57:58Bilgus_as long as you can get them apart w/o destroying them they'll be fine after
22:58:17pamaurydys: another possibility is that the firmware upgrade is simply the raw boot firmware, scrambled + a header
22:58:31pamaurythat would be helpful
22:59:06amazoniantoadBilgus_, should I get wiring for soldering while I'm at it?
22:59:22amazoniantoadturns out I will need to buy some cheap heaphones. My only other pair is bluetooth
22:59:22dyspamaury: hm, right, the periodicity seems everywhere. Maybe if we guess where a long region of NULLs is in the plaintext, we get a clear picture of the state that's computed in.
22:59:28Bilgus_where are you going?
23:00:23Bilgus_If i were to buy I'd get a patch cable and when done w/ it cut the wires from it
23:00:23amazoniantoadWell I have a lowes and a gas station in the next half mile. There is a best buy and a music store about 10 miles out
23:00:45amazoniantoadSo patch cable and what else?
23:00:46amazoniantoadThat's it?
23:00:54Bilgus_does your BT adapter use a headphone jack?
23:01:18Bilgus_ok s patch cable and we'll cut it in half
23:01:30Bilgus_you really need flux
23:01:39amazoniantoadAny ideas on where to buy it?
23:01:41Bilgus_got a Radio shack still in biz?
23:01:50Bilgus_lowes might have it brb
23:01:51amazoniantoadThe radio shack around here just closed down
23:02:01pamaurydys: yes that's what I tried for XOR and it didn't go anywhere
23:02:21Bilgus_You abs don't want plumbing flux
23:03:02pamaurydys: but assuming it's something more involved, it could be an arbitrary permutation + xoring, probably of small size given the patterns
23:03:31Bilgus_they have it! :)
23:03:41pamaurydys: I don't expect to find many NULL regions though, because the firmware format supports filling regions. You can fill a whole region with one boot "instruction"
23:03:47Bilgus_they will have a patch cable too but price might be high
23:03:56pamaurywe'll only find data regions that are almost-NULL
23:05:21dyspamaury: there are lots of suspicuous binary countes visible if I render it as a pixmap:
23:06:12pamaurydys: I'm not familar with this representation, what does that tell us ?
23:06:49dysevery bit in the image is a pixel in the 128x8192 png
23:06:58dyss/image/firmware image/
23:07:15dys(i chose a part of the image where the binary counters are most apparent)
23:07:33dys(you have to scroll around a bit for the patterns)
23:07:56dysthere's some perfect binary counting going on
23:08:17dysmaybe 1/4 into the vertical range of the image
23:09:34dysbut if it was just a counter added or xored to the plaintext, then the perfect black region doesn't make sense
23:11:51 Join wodz [0] (
23:12:50pamaurydys: which firmware are you looking at ?
23:13:13pamauryI'm looking at DAC_HA300_b19.120 and I don't have this very long NULL region I think
23:14:08 Quit mutnai (Ping timeout: 260 seconds)
23:14:38wodzpamaury: I dug further in icx_si470x_radio_tuner.ko and I have bad news. Sony doesn't follow v4l2 specification. VIDIOC_G_TUNER fills v4l2_type struct (which seems to be standard in terms of mem layout) but the content is non standard.
23:15:33pamaurywodz: ah, that's unfortunate
23:15:39wodzpamaury: rangelow is set to 7600 and rangehigh to 108000 (which means in units of 1kHz) despite advertising 62.5Hz step in capabilities field
23:16:01 Quit TheLemonMan (Quit: "It's now safe to turn off your computer.")
23:17:03wodzpamaury: signal field which should hold rssi in range 0-65535 in fact is set to 1 if tuned and 0 otherwise. And tuned is figured out by comparing reported by chip rssi in register 10 with programmed threshold in register 5
23:18:42 Quit lebellium (Quit: ChatZilla 0.9.93 [Firefox 50.1.0/20161208153507])
23:20:52dyspamaury: I think the long black block visible correspondes to byte offset 2382928 in the file (super lots of FFs not 00s)
23:21:08 Join mutnai [0] (6db91733@gateway/web/freenode/ip.
23:21:45 Join Mihail [0] (25d4bf9c@gateway/web/cgi-irc/
23:22:03dysja, should be right, the whiteish regions that come a bit later are recognizable in the hex view as well
23:22:10amazoniantoadHey Bilgus_ sorry. This phone repair guy is here
23:22:21amazoniantoadI have to put this on hold until he leaves
23:23:41wodzpamaury: audmode is also non standard 0 - mono, 1 - stereo
23:23:54amazoniantoadBilgus_, thanks for helping me btw. I really appreciate it
23:24:00dys$ sha1sum ../firmware/DAC_HA300_b19.120
23:24:00dys9cecfb35886fb84b54156ca6b90a4aa0399ac376 ../firmware/DAC_HA300_b19.120
23:25:14amazoniantoadI'm not finding any patch cable
23:26:13amazoniantoador as I call it an auxiliary cable
23:26:28pamaurydys: I'm confused about the image, I can't quite seem to relate the image with the data ^^ But that's all right, I'm just staring the hex data
23:27:16dyspamaury: I take a bit from the firmware file, put a pixel in the image. left-to-right (wrapping at 128 pixels)
23:27:32pamauryah, it's not the entire firmware!!
23:27:46dys2307 <dys> (i chose a part of the image where the binary counters are most apparent)
23:27:47pamauryok that explains my confision
23:27:54dyssorry about that
23:27:59pamauryok sorry I missed that part ;)
23:29:35dysI tried to render the full binary first, but most picture viewers disagree with images sized 128x21875 :-)
23:29:58Bilgus_amazoniantoad, it looks like that white cover will come off.. thats going to be the easiest place to solder to
23:30:09pamaurydys: do you have the tools to dump what looks like the flash on the pcb ?
23:31:15amazoniantoadBilgus_, tips on removal?
23:31:47Bilgus_use a credit card to pry under it, no metal..
23:32:00dyspamaury: I do. It's not trivial to disassemble the case though, due to the glued display. Don't want to break this thing on the first attempt.
23:32:23Bilgus_walmart will have an aux / patch cable for sure
23:32:24dysIt was rather cheap compared to the USRP, but not cheap enough to sacrifice it
23:33:16amazoniantoadBilgus_, it's off
23:33:23 Part shdwprince ("Textual IRC Client:")
23:33:28amazoniantoadBilgus_, need a picture?
23:33:32Bilgus_awesome can you get me another pic
23:34:21amazoniantoadcoming right up
23:35:59Bilgus_ah looks like they use a 4pin headphone port.. makes sense
23:36:30pamaurydys: I see, maybe we can figure out the encryption then
23:39:40pamaurythe ADSP boot rom documentation is extensive...
23:39:58Bilgus_amazoniantoad, I'm guessing it is these 3 once you have a patch cable we can know for sure
23:44:06wodzpamaury: VIDIOC_QUERYCAP reports driver version 1.0.10 and capabilities = V4L2_CAP_RADIO | V4L2_CAP_TUNER
23:44:29wodzpamaury: nothing special (despite RDS not advertised)
23:46:10pamaurywodz: it looks like it may be better to directly access si470x registers
23:46:18amazoniantoadBilgus_, going to pick up the items now
23:46:38Bilgus_k ill throw you together an instructional pic
23:46:45wodzpamaury: unless radio chip is not different on other families
23:49:39pamaurywodz: you mean different on other families ? I had a quick look and I believe si470x is the only tuner used by Sony
23:50:09wodzpamaury: ah, ok, then it may make sense to access registers directly

Previous day | Next day