#milkymist IRC log for Friday, 2012-05-11

wolfspraulcan a spartan-6 drive TTL logic?00:16
wpwrakin case anyone has their printer choke on the AD-generated M1r4 schematics: the magic conversion is   psnup -Pa1 -pa4 -f -1 <input.ps >output.ps00:26
wpwrakthe stuff is in A1 (!!!) and has X/Y swapped (but has "portrait" orientation) for extra fun00:26
GitHub38[board-m1] adamwang pushed 1 new commit to master: https://github.com/milkymist/board-m1/commit/97b7bd5182825c6a9d9dd7cc082468c7537e602303:42
GitHub38[board-m1/master] Misc.sch: 1. mirror few leds to not overlap '2' pin number. 2. added lables for leds - Adam Wang03:42
GitHub77[board-m1] adamwang pushed 1 new commit to master: https://github.com/milkymist/board-m1/commit/2de62f1fd6ec5e1f09878bb1a3f558cd1456006f03:44
GitHub77[board-m1/master] Misc.sch: removed duplicated lable. - Adam Wang03:44
mischief6sebastian just gave a presentation at noisebridge :-)05:16
larscwpwrak: i'm highly skeptical until i've seen it for real06:09
wpwrakwell, for compilers, this was pretty much the standard for the last five decades or so. wouldn't be too bad if also synthesis tools finally make at least a token effort at catching up.06:11
larsci meant wolfspraul06:22
wolfspraullarsc: just from what I read on the web, I have the feeling xilinx is very serious about vivado06:38
wolfspraullooks like the plan is to slowly let ISE fade away (which will take years, with all the inertia among their customers, even longer)06:38
wolfsprauland throw it all behind the super-uber great vivado, integrating *everything* and turning xilinx from an 'fpga company' to a 'can-do-everything chip company'06:39
wolfspraulsomething like that06:39
wolfspraulgood thing they have their faithful fpga customers finance all this - it will take years and maybe go nowhere. we see. for sure all this will be slow, vivado won't even be released until later this year, etc.06:39
wolfspraulthere are only 2 high-end fpga makers left now, altera & xilinx06:40
wolfspraulthe endgame is near06:40
wolfsprauland after that, they will have more latitude investing outside of their old (fpga) core business06:40
wolfspraulthat's how I see vivado, guessing from the outside...06:40
wolfspraulI hope one day we see some action in llhdl and a free toolchain again06:41
larscme too06:57
larscanyway, off to work06:58
larsconly 2 minutes left ;)06:58
larscnah, didn't make it in time :/07:15
guyzmoI got a board that has a quite old firmware to flash19:10
guyzmowhich is runnnig flickernoise 0.1 (20101119 built)19:10
guyzmoand this one does not support internet connection19:12
guyzmoshall I use a sdcard ?19:12
guyzmoor shall I go using flterm ?19:14
guyzmoand can I upgrade directly from a very old version to the newest using its flash procedure ?19:19
mwalleguyzmo: do you have the jtag board?20:38
mwallein that case the old flash conent is irrelevant, you can just program it via jtag20:39
guyzmomwalle - I found the reflash_m1.sh21:10
guyzmoand I'm reflashing all the M1 I have :)21:10
guyzmothanks to xiangfu liu :)21:11
sb0is that reflashing the release, or some git snapshot with bugs that could be avoided?21:12
guyzmolastest qi release21:12
guyzmoI don't know which one is better21:12
guyzmothough rtfm told me to use --qi21:12
sb0best is to use stuff from milkymist.org/releases21:17
sb0oops milkymist.org/updates21:17
guyzmois it that different ?21:18
guyzmoI just reflashed 4 MM1 from qi's repository :)21:18
sb0what version were they running before? if they're not too old, you should be able to just run the web update ...21:19
sb0and don't open the case etc.21:19
guyzmothey were running 0.1021:19
sb0there's no 0.1021:20
guyzmobut now I upgraded all to firmwares from qi that are dated april, 4th21:20
guyzmoflickernoise 0.1 (20101119 built)21:20
sb0ha, yeah, that's very old21:20
guyzmoit's now running flickernoise 1.2 dated21:21
sb0not sure if it has the "no flash filesystem" bug... maybe21:22
mwallewhaaa its harder to get a patch into uboot than into linux21:22
guyzmosb0 - when shall it happen ?21:23
sb0either way that reflash_m1.sh script should use milkymist.org/updates by default21:23
guyzmoI can test it21:23
sb0if it does not.... xiangfu should fix it :)21:23
guyzmoit's the wiki that shall be fixed : http://www.milkymist.org/wiki/index.php?title=Flashing_the_Milkymist_One#Automated_JTAG_method_.28for_developers.2921:24
sb0guyzmo, I don't know exactly, and I don't care either - it's fixed in head now, and the releases don't have it.21:24
guyzmo $ sudo ./reflash_m1.sh --qi21:24
guyzmosb0 - well, tomorrow I'm doing a Milkymist workshop with many veejays21:24
guyzmoso I hope it won't fail21:25
guyzmobut latest release from updates are from march 1st, 201221:26
guyzmolet's hope it's ok21:26
Action: guyzmo crosses fingers :)21:26
guyzmosb0 - and the handbook and the http://milkymist.org/wiki/index.php?title=Flickernoise_Patching_Language wiki page appart21:26
guyzmodo you know where I can find useful informations about flickernoise patching21:27
guyzmoI'm gonna need as much documentation I can have tomorrow21:27
sb0there isn't any that I have not sent to you, ie check the milkdrop guides and ml archives. the two people who should have written it did not.21:28
sb0or three if you count me21:28
wpwraksb0: (no flash filesystem) you men the permission issue ?21:29
wpwrakokay. that's very easy to detect :)21:29
wpwrakguyzmo: once you have them updated, if they can render _anything_, they don't have that file system problem21:30
guyzmowpwrak - ok, then it's already tested :)21:30
wpwrakperfect :)21:31
guyzmoso sad I only got one camera for tomorrow :/21:31
wpwrakah, clarification: render = run any patch. only the gui doesn't count.21:32
guyzmowpwrak - of course, but it rendered a patch at boot just after the flash21:32
guyzmothe one with the white dots21:32
guyzmoI considered it tested once I saw that one21:33
wpwrakyeah. then it's fine.21:34
sb0(my opinion about this doc not being written is that it's more a symptom than a cause, and the project needs more radical change than just good documentation)21:38
guyzmosb0 - you're talking about the flickernoise's documentation ?21:39
guyzmowell, I'm not a big fan of graphic processing programmation21:40
guyzmoand I'm neither a fan of veejaying :)21:41
guyzmoso I can't really tell, about that :)21:41
guyzmobut all I know, from veejays I talked with, is that they'd love to have a MM1-like box, that can works with digital output, and with what they can make the same kind of effects as with after effects, or illumin8 and many others I don't remember the names21:43
guyzmoto be able to do vectors, 3d and mapping in real time is really what makes them tick21:43
sb0and did these people have any precise idea of what FN does before saying that?21:47
sb0probably not :) and that's part of the problem.21:47
guyzmowell, usually they have as much a precise idea as what I can tell them21:48
sb0I don't think it makes much sense to copy after effects, either21:48
sb0then Apple will do it better21:48
guyzmobut well, tomorrow I'll see what the people say about it21:49
guyzmoand this time those people will take the time to sit and try it out21:49
sb0what it does should be made obvious by the design of the next-gen M1/FN21:49
guyzmosb0 - and what would be the improvements that are to come, so I can tease a bi tomorrow ? :)21:51
guyzmo(I don't have time to read thoroughly the mailing list)21:51
sb0try to find the notacon video (in nickfarr style...) - got bags to pack21:54
sb0going home tomorrow21:54
guyzmowell, I've packed everything up too21:55
guyzmoready for tomorrow :)21:55
guyzmogood night and good trip ;)21:55
sb0it's 3pm where I am now :p21:57
Action: guyzmo &22:02
wpwrakhmm, symptom ...22:21
--- Sat May 12 201200:00

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