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 2004-06-13

00:00:29***Saving seen data "./dancer.seen"
00:05:16 Join NibbIer [0] (
00:39:20 Join xam [0] (
00:40:23xamHi, I have one question: why are *.ucl not included in the bleeding edge builds?
00:48:46scott666bleeding edge or dailys?
00:48:56xambleeding edge
00:49:22scott666not sure, ask bagder
00:50:18scott666whats in the bleeding edge today?
00:52:57 Join AciD [0] (
01:11:08 Quit NibbIer (Read error: 54 (Connection reset by peer))
01:35:54 Join diddystar5 [0] (
01:36:12diddystar5hi pfavr
01:52:13 Quit xam ("Leaving")
01:52:37 Join xam [0] (
02:00:30***Saving seen data "./dancer.seen"
02:19:53diddystar5bye bye
02:20:00 Quit diddystar5 ("Leaving")
02:23:46 Quit dwihno_ (Read error: 54 (Connection reset by peer))
02:24:28 Join dwihno [0] (~dw@
02:47:27 Quit AciD ("")
02:53:05 Join NibbIer [0] (
03:09:51 Quit NibbIer (Read error: 54 (Connection reset by peer))
03:17:02scott666can anyone tell me what the wire on the bottom left corner of this picture ( that goes to the headphone jack is?
03:17:30scott666and if its important that the inner insulation got pulled out while i was trying to strip it
04:00:34***Saving seen data "./dancer.seen"
04:13:07 Join scott666_ [0] (
04:13:07 Quit scott666 (Read error: 104 (Connection reset by peer))
04:30:53 Join BC|coding [0] (
04:32:38BC|codingpfavr, you're up late
04:34:15xamwanna see a tight furry pussy? here: furry pussy!.jpg
04:36:02scott666_you need to convert the spaces
04:36:12 Nick scott666_ is now known as scott666 (
04:37:07xamscott: well, konqueror can handle it, and I just did a copy'n'paste. but you are right ... I should have converted the spaces into %20
04:38:14scott666mozilla can handle it too, but trillian didnt make the rest of it a link
04:38:29BC|codinghey scott, I got my new gfx lib going :)
04:38:41BC|codingand we found one of the bugs in audio-3587
04:38:47scott666im fixing my jack
04:38:54BC|codingkewl :)
04:38:56scott666the thing with the off code?
04:39:51xamBC is there any estimate when the audio-3587 will be incororated/compatible with the mainstream rockbox?
04:40:46BC|codingit's at the top of my htlist
04:41:08BC|codingI needed to fix the gfx lib for it, so I did that while waiting feedback
04:41:17BC|codinggot a few bugs to fix and a couple of mods to make
04:48:46xambtw, does the audio-3587 actually improve the sound quality? or does it "just" allow you to fiddle with more settings of the MAS decoder?
04:49:37BC|codingboth i think is the best answer
04:49:44*xam still hasn't found his usb cable ... sigh
04:52:03BC|codingmake one from carboard and tin foil
04:52:25BC|codingsome old bell wire from the telephone extension cable and some paper glue
04:52:43BC|codingyou may want to get a responsible adult to help you with the cutting - LOL
04:52:43scott666its the jack thats the tricky part
04:52:57BC|codingwrap tinfoil around matched :)
04:53:15BC|codingdamn this azerty keyboard
04:53:19 Join NibbIer [0] (
04:53:41scott666where is that from?
04:53:42BC|codingmy point precisely!
04:53:56BC|codingah dammit you got me now ...I don't know
04:53:57scott666oh god...maybe i should bring a keyboard with me to france
04:54:13scott666dunno if i could live without qwerty
04:54:15BC|codingyeah their keyboard all have wine spilled in them
04:54:19xamazerty? that's french keyboard layout
04:55:02xamwell, arabic and french then ;)
04:55:04scott666yes, azerty = french
04:55:27BC|codingoh top link in google says arabic
04:55:33BC|codingyeah, take a keyboard
04:55:44BC|codingsend me your address and I'll post you a spare
04:56:24xamBC well, it's african layout in general ...
04:56:56BC|codingthta M would be the killer I think
04:58:22xamit's obviously the same as the french layout here:
05:00:59xamyes, I was once on a coding fest, and I didn't bring my laptop (was broken) ... a french guy said I could use his laptop during the event ... that were the worst 5 days of coding in my life
05:01:25BC|codingscare you for life, an expereince like that
05:01:37xambasically every 3rd letter a typo
05:02:19xamyes. no wonder there are not many good french coders out there ;-P
05:03:20BC|codingall the best coder i know nodays are german
05:03:44BC|codinggenerally speaking of course
05:04:06xamluckil they comment in english most of the time ... at least in OSS
05:04:27scott666nice the way that works
05:04:46BC|codinghmm, I've seen plenty of french commented code
05:04:54BC|codingtoo damn close to the place
05:05:18xamBC: something along the lines "merde merde merde ..." ?
05:05:33BC|codingi wouldn't mind, but I only ever have to deal it because it doesn't work!
05:05:48BC|codingcrap code with useless comments - lol
05:05:58BC|codingsounds like the klondike alpha code
05:06:27scott666my shuffling algorithm still there?
05:06:56BC|codingyour random number generators are still there
05:07:14scott666no riffle?
05:07:32BC|codingi optimised a load of code, and sorry, the suffles were optimised
05:08:02BC|codingiirc all three riflles are now done by one routine
05:08:07BC|codingyou still got your name in there
05:08:31BC|codingright above the code you helped with
05:09:09scott666so whats with this code tainting business
05:09:26xamspeaking about optimization ... I have to speed up my graph matching algo by a factor 60 ... at least 60 ... mhhh
05:09:29scott666am i disqualified for submitting code?
05:09:31BC|codingsomething bagder (iirc) said
05:09:59BC|codingI have no idea what's going on for real any more dude, I'm just waiting for it all to go away
05:11:28scott666and then what if the core guys looked at it?
05:11:43BC|codingwho can say?
05:11:48scott666björns head would explode...
05:12:05BC|codingI guess the rule would change or be lied about - only logical choices
05:12:06xamscott the thing is, once you have looked at code/documentation that falls under a NDA or is patented, you cannot/should not partitipate in a clean-room implementation, since you might use/apply knowledge of the things you have seen (and not allowed to do use)
05:12:33BC|codingxam: the thing is there is no NDA code
05:12:40scott666or patent
05:12:50scott666bluechips just plain scum :-)
05:13:08BC|codingnot even finest green pond scum?
05:13:20BC|coding"'tis a lump of pure green m'lord"
05:13:24xamBC is the stuff (API/docs/whatever) publically available?
05:13:35BC|codingfor what?
05:14:28xamwell, it doesn't even have to be a NDA ... it is sufficient if it is proprietary. That is that the code is not licenced under a public license
05:14:52scott666but it IS gpl, isnt it?
05:15:02BC|codingfor what?
05:15:11BC|codingapi for what, my gfx lib?
05:15:21BC|codingnot posted it yet, but it will be
05:15:22scott666BC: in general
05:15:33BC|codingis the api for everything in general available - lol
05:19:09xamBC generally speaking, if s.o. uses info that is not publically available (to everyone) in order to code something, then he/she cannot publish it under a public license. Eg. you worked at a company, left it, and then you write a plugin that uses internal API which is not documented ... that could cause problems ... however if s.o. else reverse engineered the code and found out how to write a plugin, that would be okay
05:20:45scott666xam: what if someone writes GPL code but refuses to disclose their name and someone else uses ideas from it in an open project
05:21:03BC|codingeverything that was done in audio-3587 is clearly extracted from the pdf from the rockbox site - so all is well in this case
05:21:27BC|codinggood on 'em
05:21:34BC|codingthat's the spirit of free
05:21:57BC|codingoh right, think i misread one
05:22:24BC|codingyeah, that is probably the point that was made - i do not contest it - i was just explaining it to scot
05:22:28xamSo as soon as the person has seen code that was for internal use only (and not published), he shouldn't write a open-source implementation similar to what he has seen. Even if it doesn't implement the code he's seen, he might have been 'inspired' (at least the company can claim that ... and he cannot proof the opposite)
05:22:50xamBC yes, in that case everything is fine
05:23:25BC|codingThe things I have security clearance for are not available for discussion - reading pdf's from a gpl site should be safe here
05:23:54xamI just only wanted to point out the general issue to scott (and you) in case it was unclear
05:24:16scott666i understand the general issue
05:24:21scott666its the specific one thats confusing
05:24:25BC|codingoh god no, i've done clean room work myself
05:24:44BC|codinggood fun :)
05:24:59BC|codingpaid for hacking :)
05:27:52scott666well im back to my headphone jack
05:28:03scott666hopefully ill bbs
05:29:06xamBC the problem lies in the fact that you obviously have seen some stuff that falls under NDA (and you most likely signed a NDA I assume). As long as the stuff you are coding is nowhere near (ie. not related to) the stuff you contribute to rockbox, it's all fine. From what I've read one the mailing lists, the other rockbox developers are just not really sure whether you might contribute stuff that could be related to your NDA stuff ...
05:29:17BC|codingi have signed no nda
05:29:41xam... since this would taint a clean-room implemenation and thus endanger the project.
05:29:45BC|codingsorry, you mean, generally in life? or with regard to rockbox?
05:30:09xamBC well, you said you have a "security clearance"
05:30:16xamBC what does that mean?
05:30:19BC|codingnot at archos - LMAO
05:31:01BC|codingor mas, or any (afaik) even vaguely related companies
05:31:17xamBC no, it just need to be something related to the hardware ... e.g. internal knowledge of the MAS chip that was never published
05:31:33BC|codingnope, none of that
05:31:48BC|codingwhy do you think I should have access to top secret rockbox-y type stuff?
05:32:47xamBC well, after all you once said you could get the code (or you had the code) to reprogramme the MAS to be able to play something else than MP3 ...
05:34:21BC|codingwe have the codec for the wav file for a parallel wired chip (which ours is not) which I was allowed and have the email that says it was released openly and knowingly by the mirconas support team knowing it was for an open project
05:34:58xamBC I see, thanks for the clarification
05:35:08BC|codingthe closest to getting something useful was found in hungarian as a university course
05:35:24xammhh .. hungarian
05:35:41BC|codingif you squint you can start to rad it after a hile
05:36:04*BC|coding bangs side of keybaord to wake up little man
05:36:18kaboofalisten to this to wake up :
05:36:19kaboofa286. 01_Ludwig_Van_Beethoven_-_romanticheskaya_klassika_-_simfoniya_N_5_do_minor,_soch._67._Allegro_Con_Brio.mp3 [7:55 / 192kbps]
05:36:35BC|codingyes, tis nice :)
05:36:50kaboofaMy dream is to drive down the autobahn at 999999999KM/h listening to that
05:37:15BC|codingNot the spice girls then?
05:37:48xamkaboofa: I once had a accident (total crash) while listening to "Loss of Control" from Van Halen
05:38:02xamkaboofa: not kidding
05:39:23kaboofaMy friend blared some song on an airplane
05:39:32BC|codingHaving studied hypnotism during the last couple of years, that really doesn't surprise me - most poeple are more suggestible than they would like to believe
05:39:53kaboofait was like "flying high/ in the sky/ about to meet my maker/ white clouds, black ground / about to make a craiter"
05:39:55kaboofaand he shouted it out
05:40:11scott666god damn that tiny PCB
05:40:18scott666and its 4 tiny little wires
05:41:09BC|codingwire for spares in place and then join them
05:41:25xamkaboofa: was this before or after 911?
05:43:25xamkaboofa: lucky him
05:43:32kaboofahe did it after that sgt. skippy thing
05:43:33kaboofahold on
05:43:57BC|codinganyone here coding?
05:44:02kaboofaNot allowed to sing "High Speed Dirt" by Megadeth during airborne operations. ("See the earth below/ Soon to make a crater/ Blue sky, black death/ I'm off to meet my maker")
05:44:41xamkaboofa: there goes free speach
05:45:54BC|codingMust never call an SAS a 'Wanker'.
05:47:23BC|codingthat WOULD potentially be fatal
05:47:29kaboofa Putting red 'Mike and Ike's' ® into a prescription medicine bottle, and then eating them all in a formation is not funny.
05:47:53kaboofaThe following items do not exist: Keys to the Drop Zone, A box of grid squares, blinker fluid, winter air for tires, canopy lights, or Chem-Light ® batteries.
05:48:15kaboofa No military functions are to be performed .Skyclad'.
05:48:18kaboofathat's a witch one
05:48:26kaboofaskyclad = naked for ritual
05:48:39xamkaboofa: hehe ... blinker fluid ... classic one
05:48:56xamaswell as winter air
05:49:23kaboofaMust not make T-shirts up depciting a pig with the writing "Eat Pork or Die" in Arabic to bring as civilian attire when preparing to deploy to a primarily Muslim country.
05:49:43*xam knows a blonde that actually fell for winter air
05:50:14scott666there was a blicker fluid joke in red vs blue
05:51:26scott666 'Napalm sticks to kids' is *not* a motivational phrase.
05:51:36xamI don't understand all this religions stuff (I didn't want to say crap, since it could offend some people here in the chan)
05:52:11kaboofawiccan, but i don't care if you say anything to 'piss me off'
05:52:35xamBC kind of
05:52:44scott666unitarians leave burning question marks in peoples yards
05:53:35xamBC what I wanted to say is I don't know why so many people care about religion, and sometimes in such a fanatic way
05:54:02BC|codingserious question?
05:55:12BC|codingthere is a great article by Dick Sutphen called something like "their fight for you mind"'s quite short, good reading and all over the net - take a read of it - it is VERY interesting
05:56:50scott666its organized brain-washing
05:57:18BC|codingin a way, yes, but is that so wrong?
05:57:45xamI understand the need for religion (for many people), especially in poor countries or where people suffer. It's something they can hold on, something that gives them hope. But it doesn't explain those fanatics (eg. those who want to convert everybody. Not neccessarily the dangerous ones)
05:58:27xamscott yes, brain-washed is correct.
06:00:37***Saving seen data "./dancer.seen"
06:01:24BC|codingyou only believe that because the environment you come from brain-washed you in to believing it were not born with that opinion
06:01:53scott666i think any large group of people that will just do whatever a religious leader tells them to without thinking for themselves is a bad idea
06:02:11BC|codingoffice work
06:02:40BC|codingit is an inherent part of mans need to be superior
06:03:01BC|codinghow can you be "cool" if there are no "dweebs"?s
06:04:17scott666and the converse being that its not possible to be cool without also being a sheep?
06:05:00BC|codingthe dichotomy
06:05:13BC|codingaka "damned if you do, damned if you dont"
06:06:30BC|codingby the very nature of religeon it is defined by the individual to fulfill any needs they cannot find in manifest - it's where you go when you want MORE
06:07:37BC|codingmight i venture so far as to suggest that it is one of the less self-indulgent forms of greed?
06:07:58xamBC I go to the local porn video store if I want more ... or to the super market ... depending on what I want more
06:08:34BC|codingthose things can all be found in manifest (you can touch it)
06:08:42xamBC fair enough
06:08:50scott666extermemly literally on the former
06:09:14BC|codingwhat about absolvement from sins - if you have never done something you were ashamed of you haven;'t lived
06:09:30scott666what if you just dont feel ashamed?
06:09:50BC|codingthen "you" don't need that part of religeon
06:09:54xamBC but what about knowledge? you could also satisfy your greed by seeking knowledge instead of simply believing in god(s)
06:10:17BC|codingxam: yeah, that's probably my vice!
06:12:51kaboofaI like my religion. Mostly because it's fun
06:13:04kaboofaBecause you get to run around a fire and stuff
06:13:10kaboofayeah, i'm weird, i'll go back to code.
06:13:45BC|codingI know a few witches, they seem happy with their choice
06:14:12BC|codingthe bc_gfx and bc_rnd libs are pretty much ready if you wanna play
06:14:43xamkaboofa: fire stuff?
06:15:01kaboofaxam: may 1st there is usually a huge bonfire and music
06:15:32xamreminds me of burning man
06:15:37BC|codingcooking marsh mallows, singing kum by jah, bruning children, the usual stuff
06:15:41kaboofai want to go to burning man
06:15:55scott666kaboofa: my moms a witch
06:15:59kaboofabut we do cook children :)
06:16:04kaboofaok, i'm not serious there
06:16:05kaboofabut it's fun
06:18:01xam*homer simpson voice* hmmm .... nice tits! (
06:18:36BC|codingornithology and pornithology all on one page - how nice
06:27:02 Join midk [0] (
06:34:13kaboofasleep better than human interaction
06:34:16 Nick kaboofa is now known as tom-sleep (
06:38:49midkyo bcbc
06:41:13BC|codinghmm int x[0][3] any bets on whether the compiler will gpf?
06:42:12xamBC what do you mean by compiler will gpf? a compiler usually doesn't gpf
06:42:25BC|codingheheheheheh you're not trying hard enough ;)
06:42:44BC|codingfound some awesome bugs in gcc in my time
06:43:05xamunless the compiler has some real stupid bugs, it woun't cause a segmentation fault
06:43:14xamBC you mean the gcc ICEs
06:43:30BC|codingICE? integrated circuit emulator?
06:43:40xamBC internal compiler error
06:44:06BC|codingnah - I got the rockbox devkit to kill it's dos box the other night
06:44:31xamBC go to the gcc mailing list (I'm subscribed), and you will see quite regulary ICE reports, that true ... but never a gpf aka. segfault)
06:44:33BC|codingint i [x][y] = { looooooooooooooooooooads of data }
06:44:46BC|codingthen make one of the number wrong
06:44:51BC|codingx or y
06:45:18BC|codingif you're low enough on resources it will eventually kill windows
06:45:25xamBC do you mean gcc segfaults, or the compiled programme?
06:45:45xamBC I see, you mean the resulting programme causes a windows gpf ...
06:45:51BC|codingno sigsegv ...death
06:46:15BC|codingwell in that example there was no reported memory violation
06:46:20xamBC well, it's not gcc fault then, is it? :-P
06:46:47BC|codingbut they're there if you go to the their little hidey holes
06:47:16BC|codingoff codig again...
06:54:38xamBC yeah those nasty 'memory violations' are only caught if you access data that's in a different PTE ... and the page size under windows usually 4kb ... that's why you want to use 'malloc debuggers'
06:56:25BC|codingwe use ICEs to debug BIOSes
06:56:56BC|codingan entire pc between the cpu and the board :)
06:57:29xamBC mhh, nice. I suppose expensive
06:57:50BC|codingyou need to be the right person to buy them too
07:28:41scott666my headphone jack seems to be totally dead
07:28:41 Quit NibbIer (Read error: 54 (Connection reset by peer))
07:28:52scott666i got everything connected, it just doesnt work
07:30:23midkgood WORK scott
07:30:34scott666stfu midk
07:30:45BC|codingtotal dead unit?
07:31:02scott666it works perfectly at doing everything but playing mp3s
07:31:08scott666actually, it does play them
07:31:14scott666i just dont have a way to listen to them
07:31:31BC|codingcan you hard wire in a set of headphones?
07:32:08scott666i dont have the time knowledge or resources to do that
07:32:13scott666mostly time
07:33:01BC|codingwish i were there
07:33:45scott666ok, so i can get an independant study credit for making a journal of the trip, right?
07:33:58scott666i can just make a bunch of recordings!
07:36:53scott666midk: ever get anywhere with breakout?
07:37:07scott666i just deleted the shitty demo you sent me from my archos
07:49:00BC|codingshame the new audio controls will not be ready in time
07:49:16BC|codingyou'll have to find a "cafe internet"
07:49:26BC|codingand a "cable usb"
07:51:04BC|codingcan you even touch the wires in the right places and get a sound?
07:51:16scott666not that i can tell
07:51:27scott666i soldered all the wires back where they belong
07:51:33scott666no sound at all
07:51:42scott666jack must be fried
07:51:54BC|codingconnect the headphones direct
07:52:47BC|codingbypass the jack
07:53:29scott666thats where the knowledge part comes in
07:53:39scott666i dont know where the 2 wires need to go
07:53:47BC|codingthree wires
07:53:51BC|codingeft right ground
07:53:56scott666and i also dont have a pair of headphones i feel like destroying
07:54:01scott666stupid ground...
07:54:06BC|codingthen no can do
08:00:38***Saving seen data "./dancer.seen"
08:05:58 Quit scott666 ("i'll be back...eventually...")
08:21:22 Join amiconn [0] (
08:57:59 Join NibbIer [0] (
09:13:03 Join LinusN [200] (
09:13:48LinusNhey, BC|coding, I had a friend of mine translate the 35xx presentations
09:15:09LinusNdidn't give any useful extra information, was just a summary of what's already in the manual
09:15:44BC|codingbut nice going dude
09:16:12BC|codingI would do most anything for wav playback
09:16:24midkLinusN: i was just looking at the docs page... think it needs to be a bit reorganized
09:16:43midkfor example, feature comparison chart fits better under "users" than "advanced users"
09:16:54BC|codinglol, yes never understood how downloads made it to advanced documentation
09:16:58midkand the GPL would probably be better for hackers instead of advanced users...
09:17:11LinusNif you can come up with a better arrangement, i'll be happy to change it
09:17:20midkthe help texts too would also fit better under end users :0
09:17:27BC|codingput the links to all the util on the downloads page :)
09:18:44BC|codinglinus, are you aware that libplugin does not work in sim mode
09:19:10LinusNyes, saw that you discovered that
09:21:22BC|codingdid i already mention it somewhere?
09:21:38BC|codingnp, there's a simple bodge around it :)
09:21:56LinusNi think you posted a q on the mailing list about it
09:22:14BC|codingwont post another then - lol
09:22:17LinusNanyway, the kids are starving, gotta make them breakfast
09:22:21LinusNcu around
09:22:26BC|codingenjoy - l8rz dude
09:22:30 Part LinusN
09:23:38 Quit xam (Remote closed the connection)
09:25:25midkyo bcbc
09:27:13amiconnDamn, missed Linus
09:33:30amiconnBC|coding: Two questions:
09:33:48BC|codinghi, yes
09:34:00amiconn(1) Is the added "synchronous" option to gray_deferred_update() of any use?
09:34:40BC|codingsorry, I've not tried it, I ran out of RAM and had to remove the greyscale stuff, and I've been, errrrr, quite tied up with "other" stuff recently
09:34:48amiconn(2) If yes, could you please test it with an additional change to make the wait more multi-thread friendly?
09:35:28BC|codingahh, you've hit yield issues too :) That got me a right cropper ;)
09:36:40amiconnYes, after the (quick) implementation of the synchronous option I also found that busy-waiting isn't good style in a multi-thread environment
09:36:52BC|codingIt's probably going to take me a couple of days to recover my recent lost time, but once Audio_3587 is patched you can count on me to see if it fixes my old complaint :)
09:37:50BC|codingwhen the digital distortion warning kicked in, the program fell back to getkey(false) ...which I had expected to contain a yield - whoops
09:38:38amiconnBtw, I've got an ideas yesterday (caused by a comment of Jörg) how to make the (core) bitswap routine 70% faster and saving at least 100 bytes of precious IRAM at the same time
09:39:17*BC|coding bows respectfully
09:39:23BC|codingnice find dude!
09:39:56amiconnThe bitswap will no longer use a fliptable, but use a completely different algorithm. Expect this next from me
09:40:35BC|codingI bet the DSP has exactly what we need :(
09:41:01BC|codingcare to explain your algorithm?
09:41:18amiconnYes, Linus make a comment once that probably the best place to do it would be inside the mas itself
09:41:42BC|codingoh for the programmer docs
09:41:54amiconnThe basic idea of the algorithm is simple
09:42:21amiconnThe flipping is done for 4 bytes at once (longword in a register) in 3 stages:
09:42:39amiconn(1) flip the nibbles within each byte
09:42:53amiconn(2) flip the dibit groups of each nibble
09:43:05amiconn(3) flip the bits of each dibit group
09:43:28amiconnAll this is done with logical and shift operations only
09:43:47BC|codingby flip, do you mean swap?
09:44:52BC|codingI'm looking forward to seeing that - I spent ages working on that problem once - never came up with a good solution
09:45:23BC|codingas luck would have it some1 stumbled over a register that did it in the cpu
09:45:26amiconnI got the idea from my research on how to speed up grayscale
09:45:49BC|codingi'm sitting here with ALL my fingers crossed - LOL
09:46:08amiconnOnce you got the basic idea, it is relatively simple
09:46:56BC|codingthat will have a direct impact on the video will it not?
09:46:59BC|coding(a positive one)
09:47:18amiconnWhile the cpu does not have a single instruction to do this, the 1-bit and 2-bit swaps can be done in 6 cycles (nibble swap will take 8)
09:47:43amiconnNo, the bitswap has nothing to do with video stuff
09:47:56BC|codingrighty, got my memories crossed
09:49:37amiconnIf my preliminary calculations are correct, one loop iteration will take 29 cycles, flipping 4 bytes at once
09:49:59amiconnThe "old" loop takes 25 cycles for flipping 2 bytes
09:50:37BC|coding50% ...nice one :)
09:51:22BC|codingCode optimisations was my passion for many years, It is nice to meet someone else who enjoys the same challenges :)
09:51:41BC|codinghere's an interesting one...
09:51:46amiconnThere is one thing left to solve (but I already have an idea of how to do it) - swapping the leading/trailing bytes that are not a complete longword
09:52:26BC|coding#define x "long string" is IDENTICAL to static const char x[] = "long string" matter how many times the string is referenced
09:53:15amiconnImho this is caused by compiler optimization (a nice one in this case) - "constant folding"
09:53:45BC|codingvery clever, but most annoying, as I did the entire string table before I tried it - D'oh!
09:54:23amiconnIn what manner could this be annoying?
09:54:46BC|coding100+ lines of code I didn't need to write
09:54:52BC|codingplus endless code mods
09:55:15BC|codingconsider print ("−−"x"−−");
09:55:31BC|codingwond work with const char[]
09:55:57amiconnSome other optimizations the compiler comes up with are really annoying - even caused bad crashes for me
09:56:03BC|codingyou need to printf("−−"); printf(x); print("−−");
09:56:35amiconnNo, you don't, just write printf("−−%s−−", x);
09:56:51BC|codingthen I have to write a printf routine
09:57:19BC|codingoh whoops - ityped the f above
09:59:32amiconnThere is snprintf() readily available from the core...
10:00:40***Saving seen data "./dancer.seen"
10:01:10BC|codingstill have to go though and mod all the code, and after defining a buffer and calling snprintf(), I'd be in memory deficit most liekly
10:01:28BC|codingI was LITERALLY scrabbling for BYTES at the end
10:01:52BC|codingThe code was actually 32,762 bytes long at one point
10:02:27amiconnBtw, if I look at your example with the string concatenation - in this case may be the #define will no longer be identical to the const char[]
10:02:30BC|codingAnd I had an uber-cool greyscale screen saver all lined up
10:03:19BC|codingyes, from a coding point of view there is a subtle difference of use, but from a code point of view, the output changes not
10:04:25amiconn..since if you write "−−"x"−−" the string will get concatenated by the compiler, yielding "−−long string−−", which is obviously different from "long string" and will be stored separately, increasing code size
10:04:52BC|codingbut it doesn't!
10:05:23BC|codingI thought the same too, hence speding ages recoding all my #defines to const char[]'s
10:05:33amiconnDid you have a look at the binary with a hex editor confirming that it really doesn't?
10:06:22BC|codingno, i didn;t go that far in honesty, but changing some 100 or so strings of varying lengths with a varied useage frequency made 0bytes difference to the resultant code size
10:07:51amiconnHmm. Maybe I should have a look at the source. Btw, I am wondering why your audio_3587 does need that much code (did not try it out yet)
10:08:10BC|codingI think both blocks are still in the source
10:08:57 Quit midk|gone (Read error: 104 (Connection reset by peer))
10:14:10amiconnBC|coding: Yesterday you wrote something about needing a tiny font for use in audio_3587
10:14:17amiconnTwo ideas for that:
10:14:59amiconn(1) Did you try out my rockfont-8? This one is tiny, proportional, and complete ISO-8859-1
10:15:42BC|codingmy font it 6 high
10:16:11amiconn(2) Did you know that you can run pfaedit under cygwin if you have x11 support installed?
10:16:26BC|codingi have a minimum cygwin install :(
10:16:59amiconnThis is how I did two fonts. Although being a bit unstable, it is usable for that
10:17:08BC|codingi gave up with the cygwin installer, i have 20GB of hdd space, and that's kinda not much nowdays :(
10:17:38BC|codingcant believe there's nothing else out there!
10:17:48amiconnThe cygwin installer isn't that bad after you first got the idea how it works.
10:18:37BC|codingthanks for the ideas, maybe I will install cygwin proper one day and bin this devkit idea
10:18:56amiconnIt is able to auto-update an installation, detecting the components which need updating, and while being very flexible in selecting packets, it handles dependencies well
10:20:26amiconnMy cygwin installation takes 250 MB currently (not counting the home dir with all those rockbox sources)
10:21:32BC|codingmine takes about 30
10:21:46BC|codinganther .2GB to find ;)
10:22:11BC|codingWindows is currently lying that I have 800MB free drive space - lol
10:22:23BC|codingit will give up around 300-400 MB
10:22:42BC| it often does - LOL
10:23:06midkBC|coding, try a scandisk
10:23:09midkthat usually fixes it
10:24:08BC|codingconnecting to an XP box solves it, but it reverts back for some reason
10:34:29 Join AciD [0] (
10:42:19 Quit AciD (Read error: 54 (Connection reset by peer))
11:01:15 Quit midk ("yo yo yo cya later YO YO YO wasa wasa!")
11:27:32 Part amiconn
11:27:32 Quit NibbIer (Read error: 54 (Connection reset by peer))
12:00:42***Saving seen data "./dancer.seen"
12:42:35 Join AciD [0] (~acid@
13:11:28 Join NibbIer [0] (
13:42:32pfavrBC|coding, no, just left the IRC logged in - now I'm awake and having brunch
13:42:33 Quit NibbIer (Read error: 54 (Connection reset by peer))
14:00:46***Saving seen data "./dancer.seen"
14:46:19 Quit pfavr ("ChatZilla 0.9.52B [Mozilla rv:1.6/1]")
15:16:33 Join NibbIer [0] (
15:34:07 Quit elinenbe (Read error: 104 (Connection reset by peer))
16:00:47***Saving seen data "./dancer.seen"
16:05:07 Join scott666 [0] (
16:05:07 Quit NibbIer (Read error: 54 (Connection reset by peer))
16:06:31 Quit adi|home (Client Quit)
16:42:16 Quit AciD (Read error: 54 (Connection reset by peer))
17:49:31 Join NibbIer [0] (
18:00:51***Saving seen data "./dancer.seen"
18:26:49 Quit scott666 ("i'll be back...eventually...")
18:42:28 Join AciD [0] (
19:39:24 Join top_bloke [0] (
20:00:54***Saving seen data "./dancer.seen"
20:25:30 Join Weyoun [0] (
20:25:41WeyounHello all
20:27:12 Quit Weyoun (Client Quit)
20:29:39 Quit AciD (Read error: 54 (Connection reset by peer))
20:32:29 Join Smooz [0] (Smooz@
20:32:36Smoozsup guys
20:32:43Smoozi nedd some help again
20:33:04Smoozhow do u know when your 20g is used up
20:33:23top_blokecheck the free space
20:33:37top_blokerockbox info
20:33:47top_blokei think
20:33:51Smoozk i did that and it said i got tlike 10 gig left
20:34:02Smoozbut it wont let me add one more file
20:34:24top_blokesee what your PC says when u USB
20:34:36top_blokecheck the drive space
20:34:46Smoozit says i dont have the high speed usb
20:34:59Smoozbut it has let add most my music
20:35:09top_blokeyeah mine does that too
20:35:28Smoozk well why cant i add any more to it
20:35:49top_blokewhat does it say when u try
20:36:26Smoozit says....
20:37:01Smoozcan not copy track13 there is not enough free disk space
20:37:33 Join lImbus [0] (
20:37:37top_blokecheck the space on the drive using your pc
20:37:47top_blokeor run scandisk
20:38:19Smoozsays its full when i do that
20:38:56top_blokewell then i guess its full
20:40:55top_blokeselect ALL the folders and files in root and hit alt+enter
20:42:36lImbusselecting all is done with Ctrl-A
20:54:12 Join TwoD [0] (
21:07:05TwoDWould anyone be interested in creating something like Rockbox for MamboX P353? I know I've asked this before but I got no reply...
21:09:53TwoDAnyway, my email is the_dragonwizard at hotmail dot com. If you feel up to it then please mail, or atleast mail bomb me
21:10:13TwoDfor suggesting something like the above lol
21:11:38 Quit TwoD ("- nbs-irc 1.8 - -")
21:12:06lImbuswhy do everybody quit before I am finished their stuff
21:12:18lImbus...I am finished reading their stuff
21:13:50 Join TwoD [0] (
21:14:13lImbushey TwoD
21:14:16TwoDlol I quit because I clicked the wrong freakin' button... (was reading the log)
21:15:04lImbusI've never heard about that Mambox, but usually "Open Source" means the one who wants it implements it
21:16:51TwoD, I would try it myself if I knew how...
21:18:00lImbusfurthermore, the photobank / sd-cardreader is something completely new, so "porting" rockbox would not be enough
21:20:11TwoDThought so... but would it be posiible to port rockbox without the cardreader functions and maybe fix that later?
21:20:54lImbusdunno. depends on the hardware beeing used.
21:22:42lImbusporting would be easier (would be in fact rather be porting than rewriting it to have the same look and feel) if it had at least the same processor (sh1) and a similar mp3-decoder hardware, MAS
21:23:58TwoDDon't know much about it, I've only had it for a while. Where would I go to learn more about the hardware, I suppose asking MamboX is out of the question..
21:24:54lImbusI am afraid you either have to know someone from MamboX or either open the device
21:25:09lImbusthe first would be information leakage
21:25:24TwoDand the second would be risky..
21:27:14lImbusgotta go eat now
21:27:27TwoDbtw, does the Archos boxes read the playlists generated by Winamp (pls and m3u)? The MamboX doesn't and it is really annoying...
21:33:40TwoDnm, I got to go too... will see if I can dig up more info about the P353 since I really want a new firmware =)
21:34:00 Part TwoD
21:35:39 Join xam [0] (
21:39:35 Quit Ka_ (Read error: 104 (Connection reset by peer))
21:39:49 Join midk [0] (
21:46:54Smoozif u go to rockbox info is it alway accurate about how much space u got
21:47:34midkrarely accurate
21:47:37 Join Ka_ [0] (
21:48:13Smoozman i cant even get 250 cd on it
21:48:34midkum, how high bitrates are you ripping at
21:48:39 Join hardeep [0] (1098@
21:49:23Smoozrecorder v2
21:49:30midksounds about right
21:49:55Smoozshould i rip at a lower rate
21:50:47midkif you want
21:51:03Smoozwhat that do to the quality of music
21:51:19midklower bitrates, lower quality
21:51:24midk160 would be OK
21:51:27midk128 is a bit too low
21:51:45Smoozhow much more space u think that would give me
21:52:11midk160 might give you350 CDs
21:52:23midk250 at 192k is sort of little
21:52:30hardeepwhy not replace the hard drive with a larger one instead of re-ripping your CDs?
21:52:30midkare you sure you can only fit 250?
21:52:48midki think 250 @ 192 is a bit off
21:53:03Smoozwell mabey i got a few more but it filled up
21:53:15Smoozi got like over 3000 files
21:54:38Smoozill check how many files
21:56:01Smoozall mp3
21:56:24Smoozu there
21:56:38Smoozhow do i replace the harddrive
21:58:54hardeepthat's for the old v1 recorders... not sure if it's different for the v2
22:00:58***Saving seen data "./dancer.seen"
22:01:02 Join mat___ [0] (
22:01:39mat___just posted a question about the sleep timer and thought i'd ask here too.
22:02:42mat___Does the sleep timer wait for the drive to finish spinning before powere is shut off?if not can it?
22:05:25midkhardeep, would this work in the jb
22:06:04Smoozwho should i get to swap it
22:06:46midki don't know of anyone who would..
22:07:10hardeepmidk: but the 5400rpm is overkill.... you can find 4200rpm drives for cheaper
22:07:32midkit seems cheaper than some of the other ones i saw
22:07:46hardeepSmooz: do it yourself, it's easy
22:08:20Smoozso get a 4200rpm rather than 5400
22:08:35hardeepboth will work just fine
22:08:38midkgot to go
22:08:40midksee you alter
22:08:44Smoozthanks man
22:08:48 Quit midk ("yo yo yo cya later YO YO YO wasa wasa!")
22:09:40Smoozwill the one on the link that he just put up work
22:10:09Smoozi got the recorder v2
22:10:16Smoozbut its easy to swap
22:12:01hardeepi don't know about the v2, but i'm guessing it wouldn't be much harder then the v1
22:13:45 Quit Smooz ()
22:23:24 Quit mat___ ("ChatZilla 0.9.61 [Mozilla rv:1.7/20040514]")
22:29:03 Join AciD` [0] (
23:10:04 Quit NibbIer (Read error: 54 (Connection reset by peer))
23:13:22 Quit AciD` (Read error: 104 (Connection reset by peer))
23:14:04 Join amiconn [0] (
23:22:53 Join midk [0] (
23:37:16 Quit top_bloke (Read error: 54 (Connection reset by peer))
23:48:44 Join adi|home [0] (~adi|

Previous day | Next day