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 2004-09-21

00:03:20bagawkamiconn, ???
00:03:26bagawkrecording woes?
00:03:49amiconnNo, just found a possible cause for the totally non-working MMC access on the Ondio
00:04:26 Quit bagawk ("Leaving")
00:05:04 Join bagawk [0] (lee@bagawk.user)
00:06:09amiconn"Dear Rockbox, please don't mess with my serial port!"
00:07:26amiconnThat means, I have to disable the serial (!) remote code for the Ondio, since the serial port is used for the MMC. Howgh!
00:11:49amiconnThere is nothing strange with that. The CPU has 2 serial ports. One of them is used for feeding the mas (playback), the other is used for the remote control on jukeboxes. However, this is needed for MMC access on the Ondio, so I have to disable all remotre control code for it.
00:17:20amiconnAhaaa! Still not working correctly, but *much* better this way :-)
00:17:44 Quit AciD ("Read error: 666 (Connection reset by Satan)")
00:18:27 Join LinusN [0] (
00:18:34amiconnhi LinusN
00:19:03amiconnJust found one cause for non-working mmc access :)
00:19:05bagawkhey linus :)
00:19:15LinusNamiconn: saw that
00:19:19LinusNbagawk: yo
00:19:24amiconnLog peeker ;)
00:20:07amiconnLinusN: I wonder if debug builds make sense at all on the Ondio, since there is no spare serial port?
00:20:35LinusNprobably not
00:21:59amiconnHmm. I don't know how to handle that. I think this has to be accounted for in the Makefiles?
00:24:48LinusNyes, the configure script creates a makefile that sets the DEBUG macro
00:51:02***Saving seen data "./dancer.seen"
00:52:55LinusNphew, finally back to normal after the disk crash
00:53:11bagawkdisk crash?
00:53:35LinusNa failed partition copy operation made my pc unbootable
00:53:56LinusNso i have had some fun hours restoring the partition table and grub config
00:54:04bagawkyou could have installed a boot loader onto a floppy
00:54:18LinusNi booted a linux rescue disk
00:54:23bagawkboot-us has saved my partitiions a few times
00:54:30bagawkahh that works :)
00:54:55LinusNtime to sleep
00:55:00LinusNcu around guys
00:55:00bagawksee you LinusN :)
00:55:05 Part LinusN
01:01:11 Join dwihno_ [0] (~dw@
01:19:12 Quit dwihno (Read error: 110 (Connection timed out))
01:22:01 Quit bagawk ("Leaving")
01:22:27 Part amiconn
01:57:04plokWe got an iRiver from ebay.. how exciting! :)
02:12:36 Join bagawk [0] (Lee@bagawk.user)
02:37:27bagawkbye bye
02:37:50 Quit bagawk ("umount /dev/brain")
02:51:04***Saving seen data "./dancer.seen"
03:07:58 Quit scott666_ ("i'll be back...eventually...")
04:37:16 Join scott666_ [0] (
04:51:07***Saving seen data "./dancer.seen"
04:57:44 Join bagawk [0] (Lee@bagawk.user)
04:58:27bagawkhey scott666_
04:58:40bagawkan Sebulba02
04:59:45 Join ashridah [0] (
05:00:38bagawkhi ashridah
05:20:57bagawkhave to reboot for my scanner to be detected :(
05:21:05 Quit bagawk ("umount /dev/brain")
05:27:12 Join bagawk [0] (Lee@bagawk.user)
05:36:08 Quit bagawk ("umount /dev/brain")
05:43:36 Join bagawk [0] (Lee@bagawk.user)
06:03:03 Quit scott666_ ("i'll be back...eventually...")
06:16:35 Quit ashridah (Connection timed out)
06:39:14 Quit bagawk ("umount /dev/brain")
06:44:41 Join LinusN [0] (
06:46:26 Join bagawk [0] (Lee@bagawk.user)
06:49:48 Quit bagawk (Client Quit)
06:51:08***Saving seen data "./dancer.seen"
07:18:50 Join AciD [0] (
07:56:50 Join ashridah [0] (
07:58:23 Join amiconn [0] (
08:08:07 Join Zagor [242] (
08:23:57 Join [IDC]Dragon [0] (
08:25:25[IDC]DragonLinusN, Zagor: I implemented the FAT16 up to some level, but now face architectural problems
08:26:18[IDC]Dragonwit FAT16, the root dir position and size is fixed, after the FATs, but before the first data sector
08:27:05[IDC]Dragonour FAT driver works with clusters internally, not sectors. That translation is done later.
08:27:55[IDC]Dragonforthe root dir I would get negative clusters, since it is before the data start
08:28:16[IDC]Dragon(not allowed currently, and the values are unsigned)
08:29:01[IDC]Dragonand worse, the root dir size doesn't necessarily match a cluster integer, needs only to be sector aligned
08:29:30[IDC]Dragonalthough we could refuse to mount such partitions
08:29:54[IDC]Dragon(so much for my problem report)
08:31:08[IDC]Dragonhow was this done in the original FAT16 code from LinusN?
08:31:29Zagorwe need to add a fat_open_sector() which takes a sector number instead of cluster. and then we need exceptions in a couple of places to handle cluster<0
08:32:26[IDC]Dragonso you'd say the clusters should be able to get negative?
08:32:51[IDC]Dragonclusters don't make too much sense for the root dir
08:33:01[IDC]Dragonwhen it's not aligned
08:33:11Zagorno but for all other files and dirs, and the whole driver uses the file concept (since fat32 does)
08:33:55Zagorjust open a file at cluster -1, sector 4 (or whatever) instead of the normal cluster 42, sector 0.
08:34:40[IDC]Dragonanother way would be to switch to sectors
08:34:42[IDC]Dragonnot that I'm in favour of it, though
08:35:33plokIs there a good starting point anyone can recommend for getting into the hardware side of things eg, tracing, soldering, measuring voltages etc.. without doing an Electrical Engineering degree?
08:36:32[IDC]DragonZagor: I mean, make the internal housekeeping work with sectors instead of clusters
08:36:57Zagorwe need to work with cluster. everything other than fat16 root dir works with clusters.
08:37:10Zagorso make the change for fat16 root dir, not everything else
08:37:52Zagorin fact fat16 root dir works with clusters too, it just doesn't start on sector 0 in the cluster
08:38:42[IDC]Dragonfor the ata driver, it'll get translated to sectors anyway, it would "just" mean to thaat earlier (but as stated, I don't like it either, just looking for ways out of the mess)
08:39:10ZagorI just explained the way... :-)
08:41:09[IDC]Dragonok, ok, I'll look into it
08:41:55[IDC]Dragonso a potential fat_open_sector() would be able to open a file at a non-cluster boundary?
08:46:47LinusNplok: i'm sure there are lots of pages on the web with beginner guides to electrical stuff
08:47:02 Join [IDC]Dragon2 [0] (
08:48:23[IDC]Dragon2Zagor: sorry for bugging, I didn't realize the existing structures could be used to open a dir (file) at such a position.
08:48:39[IDC]Dragon2Will try and tell. :-)
08:48:59Zagorit may need some adjustments, but I think that is the solution which requires the least amount of code or structure changed
08:49:21[IDC]Dragon2that's what I'm looking for, yes
08:50:33[IDC]Dragon2(the root dir is always unfragmented, doesn't use FAT chain, has a fixed size limit)
08:50:50 Quit ashridah ("ircII EPIC4-2.0 -- Are we there yet?")
08:51:11***Saving seen data "./dancer.seen"
08:51:58[IDC]Dragon2c u later, thanks
08:52:03 Quit [IDC]Dragon2 ()
08:52:21 Join webguest33 [0] (
08:55:16 Quit webguest33 (Client Quit)
09:05:42 Quit [IDC]Dragon (Read error: 110 (Connection timed out))
09:06:41 Join webguest85 [0] (
09:10:05 Quit webguest85 (Client Quit)
09:13:36 Join Bagder [0] (
09:19:16Zagoririver ships one firmware for 120/140 and a separate one for 110. Indicates hardware differences.
09:25:14Zagorso we need to buy a broken ihp100 too ;)
09:28:03plokthere was someone in here earlier implying s/he had a broken iRiver to donate
09:28:10 Join Headie [0] (
09:28:25*Zagor checks the log
09:31:23plokDoctorObvious at the end of the 20/09 log
09:31:31ploklooks like it's a 140 though
09:31:45Zagorwell we're not picky :)
09:32:27LinusNhe said he'll mail us, but i haven't received any mail :-(
09:32:50Zagorme neither. yet...
09:39:05 Join amiconn_ [0] (
09:39:43LinusNany protests if i remove the shutdown option from the recorder main menu?
09:49:22BagderI've just realized why the deps stuff is re-run when you do make clean
09:49:38Bagderits an internal make rule that attempts to create missing include files
09:53:08Bagderso its really hard to get rid of the rebuild of the deps when you do make clean twice
09:56:13LinusNi have no rewritten the A/D driver once again
09:56:25LinusNit now converts all channels every tick
09:56:43 Quit amiconn (Read error: 110 (Connection timed out))
09:56:43 Nick amiconn_ is now known as amiconn (
09:59:08 Join PaulS [0] (
09:59:09LinusNi have tested it on my recorder and it works fine, but it *might* reveal some issues with the buttons
09:59:37PaulSHey guys.
10:00:30LinusNZagor: since the button A/D channels are updated more often, there is a possibility that the button driver behaves differently
10:00:42LinusNbut i think it will be for the better :-)
10:00:46LinusNPaulS: hi
10:01:13PaulSAre we talking about the IHP button input?
10:01:24LinusNno, the archos
10:03:10PaulSOh, interesting that it uses an A/D. Is it wired up similarly to the IHP?
10:04:01PaulSI'm from the old school where folks used scan codes. This resistor ladder / A/D thing is new on me.
10:04:46LinusNthe bad thing about it is that we can't combine keys
10:05:05Zagorhow many channels does the iriver use?
10:05:37PaulSIt's interesting as well that the Archos and iPod both used a serial interface for the remote, whereas the IHP uses another resistor ladder. I spent quite a while with a logic analyzer being confused about how it worked.
10:06:08Zagorso the spi is only for the display?
10:06:14LinusNhehe, a logic analyzer wasn't the best of tools in that case :-)
10:06:32 Join [IDC]Dragon [0] (
10:06:40PaulSZagor: So far I've identified 3 channels of the ADC: 1 for external remote, one for internal buttons, and one for the battery status.
10:07:06Zagordamn. if all main unit buttons use a single channel, it will seriously hamper the interface
10:07:18LinusNthat sucks
10:08:02PaulSZagor: That's right. I started getting suspicious (and scared!) when hanging a 50kOhm scope probe on the remote line caused the player to skip tracks.
10:08:32PaulSThe "Play" button on the remote is wired up to a separate (GPIO?) input.
10:09:02[IDC]Dragonsorry for OT: there's rumours that a Windows Media 9 Audio Decoder is somewhere available in source code?
10:09:16Zagor[IDC]Dragon: yes
10:09:24Zagoruh, no sorry. source code?
10:09:32Zagori read "in iriver firmware"...
10:09:42[IDC]Dragonyes, source
10:10:05LinusNZagor needs glasses
10:10:52Zagorhaven't heard that. can't say i'm very interested either, though :)
10:10:58[IDC]Dragonany WMA in Linux?
10:12:12Zagori don't know
10:12:37Zagori know there are players who use the microsoft dlls, but I don't know if there are any native players. i doubt it.
10:13:01PaulSYou should definitely scope out the main unit buttons though. I'm much more confident about the remote's function since I've ohmed it out. It does look to me, however that at least in some cases the same function gets used for both sets of buttons, with only the ADC mux index as an argument. This may leave the possibility that the Play button on the main unit is also separate.
10:13:21PaulSIn fact, I'd guess that it is, considering that it is used to wake the device from "off" mode.
10:21:13LinusNPaulS: in your schematic, where is the Stop button?
10:21:20PaulSAny interesting progress or questions I can answer before I head to sleep?
10:21:47PaulSThe Stop button is S0, and R0 is 0 ohm.
10:22:17LinusNand RS?
10:22:23PaulS(I put a resistor in at R0 just to retain consistency in the discussion.)
10:23:56PaulSI actually never checked the value of RS. At the time I hadn't opened my IHP yet, and any voltage measurement I made at the remote end (which I did have open) would include some assumptions of VP which I also do not know.
10:28:37PaulSHave a good night, folks.
10:28:50PaulSOkay, I'm calling it a night..
10:29:00 Part PaulS ("Zzzzz")
10:34:25 Part Lynx_
10:36:34 Join ashridah [0] (
10:41:47Zagoramiconn: did you ever do any work to get grayscales in the simulator?
10:46:06Zagormethinks the simplest way to add color support to the lcd interface is to add lcd_setcolor() and lcd_drawimage()
10:48:12 Quit Bagder ("Leaving")
10:51:14***Saving seen data "./dancer.seen"
11:14:17ZagorLinusN: are you adding a filter to button.c as well?
11:14:58LinusNi can do that
11:15:25LinusNi wonder how it should be done...
11:15:56Zagorjust compare the two latest, if they differ more than X throw them away
11:16:00 Join pike [0] (
11:16:16 Join midk_ [0] (
11:16:37LinusNso i should keep the last two readings of the three a/d channels
11:16:51[IDC]DragonLinusN: I have already implemented that
11:16:59LinusN[IDC]Dragon: good
11:17:07LinusNthen i leave it to you
11:17:28 Quit midk (Read error: 104 (Connection reset by peer))
11:17:56[IDC]DragonI once did it, and made a little survey
11:18:58[IDC]Dragonbut perhaps it was void because of the reduce ADC poll frequency, I didn't know that
11:28:00Zagorwhat kind of display does the h300 have? 8-bit, 16-bit, 24-bit?
11:33:39LinusNi have no idea
11:34:02amiconnZagor: I once wanted to add grayscale to the sim, but then found that the x11 sim (unlike the win32 sim) doesn't use a bitmap, but sets/resets individual pixels. I see no way to do it without the x11 sim being changed to use a bitmap too.
11:37:15amiconnSince I'm not familiar to X11 programming, I have to leave that for someone else
11:48:26 Quit ashridah (Read error: 60 (Operation timed out))
12:47:07 Join webguest08 [0] (
12:51:17***Saving seen data "./dancer.seen"
12:51:25 Quit webguest08 (Client Quit)
13:47:06[IDC]Dragonhooray, first FAT16 file access!
13:51:35dwihno_Go Jörg!
13:51:37 Nick dwihno_ is now known as dwihno (~dw@
13:56:50[IDC]Dragonbut don't try to create root files, etc.
14:04:40LinusN[IDC]Dragon: way to go!
14:05:04[IDC]Dragonindeed, some way to go.
14:07:18 Quit pike (Read error: 54 (Connection reset by peer))
14:07:20 Join pike|| [0] (
14:07:24 Nick pike|| is now known as pike (
14:15:25 Join Bagder [0] (
14:16:57 Quit pike (Nick collision from services.)
14:17:13 Join pike [0] (
14:33:45 Join ashridah [0] (
14:41:54midk_wooo! zagor101 got it! YEAH!
14:45:13 Join grogro [0] (
14:45:57 Join R3nTiL [0] (
14:47:10 Join MrMoo [0] (~me@
14:47:26Zagormidk_: hehe, good morning :)
14:51:20***Saving seen data "./dancer.seen"
14:55:38amiconnZagor: Got my X11 "inquiry"?
14:56:16Zagorabout the sim and colors, yes
14:56:32amiconnI need a bitmap to work with...
14:56:55amiconn[IDC]Dragon: Your are progressing faster than me :-|
15:03:25[IDC]Dragonnot really, this was the easy part
15:04:10[IDC]DragonZagor: negative clusters for root dir have the side effect of not always being negative
15:04:21amiconnNow that I disabled the serial remote code, I get consistent responses from the card (although not those that I want yet)
15:04:25Zagorouch. how come?
15:04:27[IDC]Dragonbecause the true clusters start a 2
15:04:32Zagorah, right
15:04:46midk_Zagor, morning :)
15:04:47[IDC]Dragonso the dir runs from minus something to 1
15:05:13[IDC]Dragoncrossing the zero, which has a special meaning in varoud parts
15:05:18 Quit maikeul (Read error: 110 (Connection timed out))
15:06:39[IDC]Dragonamiconn: sorry for the serial, I should have disabled that, I knew it's in the way
15:07:55[IDC]Dragonmy FAT16 file system currently only works if you stay within the first cluster of the root dir
15:08:03 Nick midk_ is now known as midk (
15:08:08[IDC]Dragonsubdirectories should work as usual
15:08:33Zagor[IDC]Dragon: what's the problem with crossing 0?
15:10:47[IDC]Dragoncluster 0 somewhere means not initialized
15:12:12amiconn[IDC]Dragon: ...and you swapped the chip selects, although this was no real problem, since I intend to select/deselect on access
15:12:45[IDC]Dragonsorry ;-)
15:13:02[IDC]Dragonwas the twiki doc correct?
15:14:34 Part LinusN
15:14:50amiconnI'll get an MMC tomorrow, so I can test with external as well
15:15:25[IDC]DragonZagor: I found such a case at line 1683 of fat.c
15:15:50[IDC]Dragon || !cluster
15:17:30[IDC]Dragonfor clusters <2 (the root dir), I have to just count up, ignoring the fat
15:18:19[IDC]Dragonin fact, not reaching 2, because the the root dir would be full
15:18:30[IDC]Dragonthen the
15:20:33[IDC]Dragonhowever I'll handle the case that nothing can get appended to the root dir
15:29:19[IDC]Dragonwhat else could FAT-wise happen to the root dir? Can we seek it? Will we truncate it?
15:30:10[IDC]Dragon(we won't delete it, hopefully ;-)
15:31:09Zagorwe can't, since there is no pointer to it anywhere
15:31:52Zagorwe don't seek dirs, we load whem whole at once
15:32:30[IDC]Dragonso fat_readwrite() is all I should take care of then?
15:32:54 Quit R3nTiL (Read error: 54 (Connection reset by peer))
15:33:27Zagoractually we do seek dirs, in the fat driver itself
15:34:24Zagori guess an exception in add_dir_entry will be needed too
15:36:24[IDC]Dragonand the name handling also moves around
15:41:52 Join R3nTiL [0] (
15:49:32 Join R3nTiL1 [0] (
15:49:43[IDC]DragonI think I better offset the root dir cluster, such that they don't cross zero
15:49:55[IDC]Dragonelse too many problems
15:53:19 Quit R3nTiL (Read error: 104 (Connection reset by peer))
16:00:41 Quit R3nTiL1 ()
16:00:52Zagoryou mean offset it negatively, like -10?
16:06:00[IDC]Dragonalready done, I now let it reach to -1 at max
16:06:18[IDC]Dragonbefore, it went up to 1
16:06:32Zagorsounds good
16:06:38[IDC]Dragonbecause 2 is the start of the data area
16:06:48[IDC]Dragonyes, and works :)
16:06:54Zagorthat's good too :)
16:07:56[IDC]Dragonneeds an if in first_sector_of_cluster(), to offset the positive clusters with 2, not no offset for the negative
16:09:12[IDC]Dragonso my root can reach beyound its first pseudocluster now
16:09:26[IDC]Dragonstill need to handle the seek
16:24:33 Quit [IDC]Dragon ("CGI:IRC")
16:29:17 Join methangas [0] (
16:41:34 Quit ashridah ("sleep")
16:44:04 Part Zagor
16:51:21***Saving seen data "./dancer.seen"
16:56:45Bagdermy take at another cleanup of the makefile/configure stuff
17:03:02 Join mecraw__ [0] (~lmarlow@
17:20:02 Nick grogro is now known as gromit` (
17:37:12 Quit AciD (Read error: 104 (Connection reset by peer))
17:41:08 Join webguest91 [0] (
17:41:19webguest91hi all
17:42:46webguest91I have a request for the website: a link for "download all irc logs to date as a single .txt .gz .zip"
17:43:03webguest91Is that reasonable?
17:44:10*webguest91 realises he's just typing into the logfile...
17:45:19 Quit webguest91 (Client Quit)
17:55:26 Join AciD [0] (
18:09:02 Nick AciD is now known as AciD` (
18:24:30 Nick AciD` is now known as AciD (
18:51:24***Saving seen data "./dancer.seen"
19:28:55 Join R3nTiL1 [0] (
19:28:55 Quit elinenbe (Read error: 104 (Connection reset by peer))
19:37:27 Quit Headie (Read error: 60 (Operation timed out))
19:40:29 Quit pike (Read error: 104 (Connection reset by peer))
19:41:56 Join pike [0] (
20:20:15 Quit AciD (Read error: 104 (Connection reset by peer))
20:24:35 Join Headie [0] (
20:38:11 Join AciD [0] (
20:43:09 Quit R3nTiL1 (Read error: 110 (Connection timed out))
20:51:25***Saving seen data "./dancer.seen"
20:59:55 Quit mecraw__ (Read error: 104 (Connection reset by peer))
21:00:35 Join mecraw [0] (~lmarlow@
21:52:37 Join scott666_ [0] (
22:23:16 Join [IDC]Dragon [0] (
22:36:23 Quit methangas (" HydraIRC -> <- irc client ownage!")
22:49:10[IDC]Dragonamiconn: u there?
22:49:41[IDC]Dragonhow're things?
22:50:14amiconnTried some modification of the transfer routines. Now it hangs :(
22:50:31[IDC]DragonI shoudn't have asked
22:50:40amiconnAnd: I'm facing 34 printed pages of asm source
22:51:20[IDC]Dragonand still think reverse engineering is fruitful?
22:51:29***Saving seen data "./dancer.seen"
22:52:32amiconnHmm, really don't know yet. My transfer routines just look right, but they refuse to work. Obviously I must have overlooked something (like an & or such)
22:55:03[IDC]Dragonwhat does it look like on a scope?
22:55:11amiconnNo clock :(
22:55:23amiconnI already was there, I know
23:03:57[IDC]Dragonsorry to leave
23:04:06 Quit [IDC]Dragon ()
23:13:10 Quit scott666_ (Read error: 110 (Connection timed out))
23:13:11 Join scott666_ [0] (
23:22:54 Quit mecraw (Read error: 104 (Connection reset by peer))
23:23:28 Join mecraw [0] (~lmarlow@
23:24:10 Quit mecraw (Client Quit)

Previous day | Next day