#milkymist IRC log for Sunday, 2011-12-11

GitHub74[migen] sbourdeauducq pushed 3 new commits to master: https://github.com/milkymist/migen/compare/d3127fd...af74a8900:15
GitHub74[migen/master] convtools: insert reset on variables - Sebastien Bourdeauducq00:15
GitHub74[migen/master] fhdl: remove broken fragment iadd - Sebastien Bourdeauducq00:15
GitHub74[migen/master] corelogic: timeline module - Sebastien Bourdeauducq00:15
wpwrakhmmm, i think we shouldn't allow "." as synonym for 000:28
wpwrakRovastar & Idiot24-7 - Balk Acid.fnp has a typo at line 5300:30
wpwraki guss thus should be  wave_g+0.20 ... ?00:30
wpwrakthe funny thing is that the current parser will treat it as  wave_g*0.20...   so, after fixing it, the patch will look different00:31
wpwraknow, what on earth should the last line of  Idiot & Rovastar - Altars Of Madness 2 (X42 Mix).fnp00:37
wpwrakdo ?00:37
wpwrakperhaps    per_vertex=rot = if(above(box,1),sin(0.885*time),0)*(x+y)-rad*sin(q5)*0.5   ?00:37
kristianpaulhours of not ending experimentation to make something looks cool? :-)00:41
wpwraklook there - it finally becomes clear why the altars of madness are called that ;-)00:42
wpwrakwell, all the patches look a bit like that :)00:53
wpwrakbut what i meant was that the line had a syntax error00:53
wpwrak(in the original)00:53
wpwrak /nick lurkersbane01:10
wpwrakhmm, i'm only 35 commits ahead of origin. need to make more changes.12:12
wpwrakbarbu-uucp: is it intentional that you feed TOK_ERROR into Parse in fpvm_parse ? (parser_helper.c)12:22
xiangfuwpwrak, Hi I can help on flicekrnoise 's commits.12:24
wpwraksebastien: okay if xiangfu commits ?12:25
wpwrakxiangfu: also milkymist.git ? the beginning of the patch series is there. then things move over to flickernoise.git12:26
xiangfuwpwrak, no commit right in milkymist.git, only flickernoise.git.12:27
wpwrakah, then we have to wait12:27
xiangfuwpwrak, better Sebastien review a little on soc and milkymist.git code. I have no idea about soc.12:29
xiangfuwpwrak, one little thing. the code that fix my rapoo 1800 also fix the full-speed mouse. just FYI. in case you have some clues :)12:30
wpwrakyeah, i know what to do with your devices. need to write a report parser. that way, the USB stack can understand the message format better.12:31
wpwraki'll do that after i'm done with the first batch of compiler changes. it's already looking pretty decent, so hope that won't be long now. maybe 1-2 days.12:32
xiangfuwpwrak, (report parser) what can I do on this.12:32
xiangfuif it not that complicate like 'compiler changes' maybe I can just finish it. :D12:33
xiangfuwpwrak, it maybe also help on full-speed right?12:33
wpwrakif you understand how the report descriptors defined in the HID specification work, you could write that parser, yes. they're a little tricky, though.12:35
wpwrak(full-speed) doesn't matter. both full- and low-speed use the same protocol12:35
xiangfuwpwrak, ok. I will try first.12:35
wpwraki think we have a low-speed speed regression that causes some devices that used to work with the old firmware to have timing issues with the current version. not sure yet where the problem is. alas, none of my devices exhibit that issue. so i think i'll go mouse shopping next week. see if i can find anything that doesn't work.12:37
xiangfuwpwrak, my stuff works fine , no errors so far.12:40
wpwrakyeah, the bugs only happen to those who can't fix them :)12:41
barbu-uucpwpwrak: i'll have a look at your patches tomorrow. thanks for them! I'm focusing a bit on Migen atm12:44
wpwrakok, thanks !12:45
barbu-uucpTOK_ERROR... I don't remember. this is february 2009 code :p12:45
barbu-uucpxiangfu: maybe you can fix the FTP bugs? shouldn't be a very difficult task...12:46
wpwrakokay. i'll move it and see what happens :)12:46
barbu-uucpbut I don't think it should make a difference... we're destroying the parser immediately after seeing this token12:47
xiangfubarbu-uucp, will try to fix the FTP bugs next few days.12:48
wpwrakyeha. it just tries to force a parser error in addition to the scanner error12:50
wpwrakmy regression tests like the change :)12:55
xiangfuthe ftp bug relate to recently change on stack size. trying now.13:13
wolfspra1lwpwrak: any news regarding the rc4 package to you?13:31
wolfspra1lI just emailed you a more detailed invoice in case that helps you13:31
wpwrakwolfspra1l: (package) thu and fri were holidays. so tomorrow is the first day i can go to customs and try to get it13:36
wolfspra1lok maybe the invoice I sent you helps13:36
wpwrakpolling ...13:37
wolfspra1lsorry about this, it's definitely a mistake on my end to not put a real invoice in in the first place. I got complacent with so many packages without problems...13:37
wolfspra1lgood reminder to never let down13:37
wolfspra1lreally sorry to waste your time13:37
wolfspra1lI hate those trips to customs offices as well, must have done a lot in my life in Germany, and that cannot possibly be as bad as in Argentina, though the time-waste aspect is similar I guess.13:38
wolfspra1lfeels like in a time machine13:38
wolfspra1llike going to an economic museum or so, the way they are equipped and think13:38
wpwrak(time machine) yeah. the one for regular mail also gives you a nice taste of how it must be / have been in communist countries13:39
wpwrakthe good thing there is that they're generally set up to let things pass. so it's inconvenient and a waste of time, but low risk13:41
wolfspra1lwith the invoice I sent you now it would probably have passed13:42
wolfspra1ltoo late though13:42
wolfspra1l100% my mistake13:42
wolfspra1lin Munich it's quite funny as the customs museum and customs office are in the same building complex13:43
wpwrakthe one at the airport is inconvenient, a waste of time, and high risk. also because it seems that regular people are only supposed to approach them for a small set of transactions (like lugagge). for anything else, you should go through a custom broker. so if they insist on this, it can get arbitrarily messy13:44
wpwrakhmm .. still no mail .. checking the spam folder ...13:44
wolfspra1lso the customs museum explains you how things were in the past, with trains arriving and all sorts of job titles who have long disappeared processing the goods through this and that stage13:44
wolfspra1land then you can just proceed left down the floor, up a stay, to the right, to the left, and ...13:45
wolfspra1lthe remants are still alive :-)13:45
wolfspra1lup the stairs13:45
wpwrakah. now it arrived.13:45
wolfspra1lbeautiful old building though, very nice architecture etc. late 19th century13:45
wolfspra1lonly the few guys still wandering the halls and nagging people today are a bit fallen out of the times13:45
wolfspra1lthe extension of the museum, so to speak13:45
wolfspra1lthey must be looking forward to retirement13:46
wpwrak(invoice) great, thanks !13:46
wpwrak(customs in munich) ;-)))13:47
wolfspra1lit sounds like fedex did hand it over to a broker, not sure which one though13:47
wolfspra1lmaybe you can just email/fax that invoice to the broker and all else goes remotely? probably not...13:47
wpwraki think they just don't have a category for "left in limbo"13:48
wolfspra1loh fedex happily declares it as "delivered"13:48
wolfspra1lgood for statistics I'm sure :-)13:48
wpwrakwhat normally happens is that the shipment sits at the customs depot and you then have to try to get it out yourself or find a broker. there's a building where they roost just a short walk away13:49
wpwrak"delivered" = documents delivered :)13:49
wpwrakwell, you should see how this works with DHL. there, they only bring you a letter. then you have to go to the DHL office and pay them some USD 100 to get the real paperwork.13:50
wpwrakfedex at least have the decency of not adding to your misery13:50
Action: kristianpaul had bad experience with DHL in past13:52
kristianpaulThey want squeezy your last penny,like adding to taxes value the shipment it self !! (knowing they we're the carrier!!)13:53
wpwrakah, but that's how such things are often taxed13:53
kristianpaulbut not in fedex i remenber13:53
wpwrakthat's not DHL but your government who puts the tax value on value of goods plus cost of shipping13:53
wpwrakyeah, also here, those calculations vary :)13:54
kristianpaulThats why i like airmail,13:56
wpwrakairmail and EMS here have the problem that they often don't process customs for you. so you have to go there yourself.13:59
wpwrakfedex usually do that. (unless there's a problem)13:59
kristianpaulthe max i have to do about aimail/EMS was pay taxes when a product weight was more than 2KG14:02
kristianpaulbut yeah, still that risk tought..14:03
wpwrakoh, anything < 2 kg is free ? that's convenient ;-)14:07
GitHub157[migen] sbourdeauducq pushed 1 new commit to master: https://github.com/milkymist/migen/commit/05d91c7104721c0dc475514cf10aaa7066d5d53b14:08
GitHub157[migen/master] bus: Wishbone to CSR bridge - Sebastien Bourdeauducq14:08
kristianpaulhmm, letss see if there is quick way to disable USB suppport in mm bios..15:55
kristianpaulokay fake IRQ_USB and not calling usb_init hope works..16:02
kristianpaulyeah, m1nor very handy no more hacked batchfiles :) Thanks a lot wpwrak !16:04
lars_barbu-uucp: this is sort of dangerous:16:16
lars_class StatementList:16:16
lars_def __init__(self, l=[]):16:16
lars_self.l = l16:16
lars_x = StatementList()16:16
lars_y = StatementList()16:16
lars_print y.l => [1]16:16
lars_i'm a bit surprised that it hasn't caused any problems for you yet16:18
lars_unless it's what you want, but i somehow doubt it16:19
barbu-uucplars_: it did create some problems, but they're not in the critical path yet, so I have not looked at them closely. what's the solution? :)17:08
lars_def __init__(self, l = None):17:13
lars_if l is None: l = []17:13
lars_quite ugly, but that's how python works17:14
barbu-uucpalright. thanks !17:33
GitHub25[migen] sbourdeauducq pushed 5 new commits to master: https://github.com/milkymist/migen/compare/05d91c7...eee698019:22
GitHub25[migen/master] bank: fix csrgen address decoder - Sebastien Bourdeauducq19:22
GitHub25[migen/master] bus: 14-bit CSR addresses - Sebastien Bourdeauducq19:22
GitHub25[migen/master] bus: fix CSR interconnect data readback - Sebastien Bourdeauducq19:22
ThihiI'm making my first flickernoise patch :P20:28
wpwrakwhee ! :)20:29
ThihiSO far I'm just rotating a wave with a per_frame rot=rot+0.1*sin(bass) and zooming it with a zoom=zoom+0.1*sin(mid/treb) :p20:29
ThihiHaving fun so far :)20:29
kristianpaulSeems you are under control :-)20:30
wpwrakif you have a MIDI controller, you may want to throw this into the mix as well. it's a lot more fun if you can interactively tweak your patch20:31
ThihiI don't have one handy atm.20:32
ThihiBut I will certainly test a kaossilator or something with this :P20:32
kristianpaulyeah, patches with varibles looks more fun20:33
kristianpaulalso when you cant screen enought  to make some sound variables to boost ;)20:34
wolfspra1lThihi: oh wow, good!20:34
wpwrakre2c has amazingly awkward way of adding conditions. unlike lex, where you only instrument the rules you care about, it requires you to instrument them all20:49
kristianpaulno suckless at all then?20:51
wpwrakit's also not very documented. well, it shows that these things don't have the age and maturity of lex and yacc yet20:54
kristianpauli also wonder how suckless-like will they be if reach that maturity :)20:56
wpwraki haven't had them flat out fail, but when you push them, they show weaknesses20:56
barbu-uucpGNU flex and GNU bison have some weaknesses when you push them too ...21:00
wpwraktrue. in different areas, though21:01
wpwraklemon looks kinda promising. it probably just needs better documentation21:02
wpwrake.g., it's not clear how to get out when you have an error. what i've done is use lemon's own parse error exit mechanism21:03
wpwrakand, according to google, i'm not the only one with that problem21:03
ThihiOk, so re there any other variables that one can set for the input video except the video_a ?22:18
ThihiThe guides I've found on making fnps are somewhat lacking in detail ;P22:19
kristianpaulThihi: i could not agre more22:26
wpwrakknown bug :)22:40
wpwrakhmm .. where could a message like   Error: expected tag but found 'sy' instead  originate from ?22:46
wpwrak(displayed in the patch compiler)22:46
wpwrakoh ..maybe pixbuf22:47
wpwraklet's just doa clean build before jumping to conclusions ...22:48
wpwrakThihi: the good news is that the patch language should get a little friendlier soon. all the variables still need better documentation, though22:49
wpwrakthat's from libmtk. interesting.22:53
wpwrakfound it :) error messages are "magic"23:12
wpwraknice. critter runs ;-)23:17
wpwrakbarbu-uucp: what is per_frameinit_1 ?23:28
wpwrakFN seems to simply ignore the init_123:29
kristianpauloh, milkymist2 repo23:42
--- Mon Dec 12 201100:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!