#milkymist IRC log for Friday, 2011-05-13

--- Fri May 13 201100:00
azonenbergSo I just decided to upgrade to a spartan-6 from the spartan-3a i was designing my platform before... huge improvement01:33
azonenbergBut now i want a Virtex so i can do 1080P01:33
kristianpaulwhy?01:34
kristianpaulxiangfu: hi there03:03
kristianpaulmorning btw :-)03:03
xiangfukristianpaul: Hi03:05
kristianpaulscript is update to last flickernoise revision isnt?03:06
kristianpaulman, too many libs now to hack/compile by hand :-)03:07
wolfspraulkristianpaul: hey I am trying to make a collection of m1 screenshots03:26
wolfspraulif you have anything please point me to the url...03:26
wolfspraulhttp://en.qi-hardware.com/wiki/Milkymist_One_screenshots03:26
wolfspraulhere's what I have so far03:26
kristianpaulk, not that i have right now, but i'll keep in mind :-)03:27
wolfspraulyes, keep in mind03:27
wolfspraulpress F2 once in a while :-) (was it f2?)03:27
kristianpaulwell, for now i'm using the rtems shell more often :-)03:28
wolfspraulI need to go through all patches, take a screenshot of each one, etc.03:28
wolfspraulthen label them all, and so on. argh. lots of work :-)03:28
kristianpauloh yes..03:28
wolfspraulof course the best would be video including sound, but that's going to take a while to implement03:28
wolfspraulI mean the m1 could compress and stream it right away...03:29
wolfspraulso for the time being, I will focus on screenshots03:29
xiangfuCtrl + F2 for screenshot.03:48
xiangfukristianpaul: Hi. there is a command 'make milkymist-git-update' for update your local milkyist gits.03:49
xiangfuhttp://en.qi-hardware.com/wiki/Milkymist_One_screenshots just added two more screenshots. :)03:50
kristianpaulxiangfu: http://paste.debian.net/116818/05:29
kristianpauland btw mupdf version is now 0.8.165 or get the 0.8 from archive..05:31
kristianpaulLIBPNG_VERSION -> LIBPNG05:33
xiangfukristianpaul: (mupdf) everytime they release a new version. they just move it to the 'archive/' folder.05:36
kristianpaulahh mupdf patch broken.. better get vesion from archive..05:36
xiangfuI will update my make file.05:36
xiangfuyes05:36
kristianpaulyup05:36
xiangfukristianpaul: Please update your local git. just committed.05:37
kristianpaulhmm05:38
kristianpaulthe bot dint  said nothing about that :-)05:38
kristianpaulah,, this is now qi ;-)05:38
kristianpaulok xiangfu05:38
Action: kristianpaul wanted to be optional not compile all this gui dependent stuff when i just wanted a shell, network and ftp/telnet working..05:39
kristianpaulnow expat...05:41
kristianpauloh well..05:41
kristianpaulah .tar.gz again..05:47
kristianpaulxiangfu: i think you have similar issue with expat as when mupdf05:50
kristianpaulalso a missing ufpr.dl. may be?..05:51
kristianpaulhmm no05:53
kristianpaulargg mirros..05:58
kristianpaulhttp://paste.debian.net/116820/05:58
xiangfukristianpaul: have you change the 'makefile" it should not  "expat-expat-2.0.1.tar.gz"06:00
xiangfukristianpaul: sorry .there is a typo :(06:01
xiangfuline 29606:01
xiangfukristianpaul: please update local git again. sorry for those errors  :(06:02
kristianpaulnp06:03
kristianpaulphew... finally compile06:04
kristianpaulhttp://paste.debian.net/116821/06:05
kristianpaulhttp://paste.debian.net/116822/06:06
kristianpaulhttp://paste.debian.net/116823/06:06
xiangfupaste automatic :)06:07
kristianpaulhttp://paste.debian.net/116824/06:07
kristianpaullol06:07
kristianpaulno06:07
kristianpaul:(06:07
xiangfukristianpaul: you want paste all file ? :)06:07
kristianpauli just dont want commit a hack06:07
kristianpaulthose files are hacked to NOT load gui/mtk whatever..06:08
kristianpauljust console06:08
kristianpaulbut06:08
kristianpauli got this error06:08
kristianpaulhttp://paste.debian.net/116825/06:08
kristianpaulahh, may be i just miss do git diff > file06:09
kristianpaulanyay06:09
kristianpaulerror is related to flashvalid.c but i dint touched that one..06:10
xiangfukristianpaul: you may still needs those '-lz -lm"06:10
kristianpaulahh06:10
kristianpaulwhat is that for?06:10
Action: kristianpaul dint read the manual about gcc06:10
xiangfuzlib and math.06:10
kristianpaul;-)06:11
kristianpaulsame error..06:11
kristianpaulthere a especial order? i think so..06:12
kristianpaulahh wait06:13
kristianpauli forgot make clean :-|06:13
kristianpaulhttp://paste.debian.net/116826/06:20
kristianpaulthanks for help xiangfu06:20
kristianpaultomorrow i hope bug you a bit about rtem drivers :-)06:20
kristianpaulbut may be for now i'll end using pointers..06:20
kristianpaulhttp://paste.debian.net/116826/ <- some size comparaciones06:21
wolfspraulkristianpaul: what do you think? http://en.qi-hardware.com/wiki/Milkymist_One_screenshots06:27
wolfspraulI ordered them a bit by what I think makes it an interesting gallery06:27
wolfspraulit's a start...06:27
kristianpauli liked the two first ones06:39
kristianpaullooks good i think06:39
wolfspraulkristianpaul: you only like the first two?06:48
wolfspraulI rearranged a bit more, final for now.06:48
wolfspraulI need to go through all patches we have and take shots :-)06:48
wolfspraulany final feedback on our lgm news release?07:28
wolfspraulhttp://en.qi-hardware.com/wiki/News_Release:_Milkymist_One_video_synthesizer_shown_at_6th_Libre_Graphics_Meeting_in_Montreal07:28
wolfspraulunfortunately we still have no picture of the m1 in action, but whatever. this release is mostly to go through the exercise of a news release, whether it's picked up or not.07:28
wolfspraulI trust Jon that the actual m1 use and party happened as we all think it did...07:29
wpwrakwolfspraul: i think mm1 day is just today. so it hasn't happened yet08:03
wolfspraulI think it was yesterday (Thursday)08:04
wolfspraulno?08:04
wolfspraulanyway that thing is going out now08:04
wolfspraul:-)08:04
wpwrakyou should read the channel backlog ;-)08:07
wpwrakrejon said he had all the other projects to worry about so far and only today will he focus on mm108:07
wpwrakif you want to release this now, they you should at least reword is as a belated announcement of an upcoming presentation08:09
wpwrakof course, this would have made a lot more sense had it been posted several days if not weeks before, as i suggested ;-)08:09
wolfspraulcall it the chaos of the newsroom, I've finished my edits and it's out now08:10
wolfspraulthis is it08:10
wolfspraul:-)08:10
wolfspraulI don't expect a huge uptake anyway out of this, it's more to run through the exercise once, and remind some journalists that Milkymist exists/is still around.08:10
wpwraki think i'll spare you my opinion on the logic of all this ;-)08:11
wolfspraulyay, this is pretty close to real news now!08:11
wolfspraul:-)08:11
wolfspraulwhy do you think it's all so completely wrong :-)08:11
wolfspraulI've added some nice screenshots, we have a nice template, we have Jon's quote08:12
wolfspraulonly thing missing is picture of m1@lgm08:12
wolfspraulwhich according to wpwrak is an impossible thing to have because it's a picture from the future :-)08:13
wpwrakyeah. either you report about the past, you try to predict the future, or you write a work of fiction :)08:15
wpwrakby the way, the screenshots, while nice, all suffer the problem of not showing input (or if they do, it's not clear that they do)08:16
wolfspraulcorrect, because I have no better screenshots right now08:18
wpwrakspecifically, there's no indication that effects are modulated by incoming audio, and there's no indication that effects can manipulate and respond to incoming video08:18
wolfspraulsure, the audio thing could only become more clear in a whole video08:18
wolfspraulwhich would be great to have, but it requires vp8/theora support in the fpga...08:18
wolfspraulor maybe we find some vga grabber in the meantime08:18
wolfsprauland for video, well we simply don't have them yet08:19
wolfspraulthere is one screenshot where it's more or less obvious, I just titled it 'me' :-)08:19
wpwrakthis is something that also came up in the /. comments. people are having a hard time understanding what the product does, not only because it's an oddball thing with vj and fpga rolled into one, but also because its feature set is unclear08:19
wolfspraulhttp://en.qi-hardware.com/wiki/File:M1_screenshot_cam3.png08:19
wpwrake.g., those images could very well come from some C64 lovers convention08:19
wolfspraullooks like a picture from a Bosnian mass grave to me, but I like it08:19
wolfspraulwell the really weird thing of the news release is that a picture of m1 in action at lgm is missing08:20
wolfspraulmaybe I can still get one a little later, then I will fit it in08:20
wpwrak"me" doesn't make this very obvious. you can interpret that this is a modification of incoming video but it may very well also be something synthetic08:21
wpwrak(and even monochrome)08:21
wolfspraulI'm not so worried about the slashdot comments08:21
wolfspraulthose guys cannot be converted into buyers, no matter what, imho08:21
wolfspraulor even into contributors08:21
wpwraki think the /. comments are useful for finding holes in the story08:22
wolfspraulwe need to get the product in front of real music lovers08:22
wolfspraulor people interested in playing with video08:22
wolfsprauland in the slashdot comments they do show through here and there08:22
wolfspraulsome airmchair general saying "bla bla bla", and then the response "you have no idea what you are talking about. I'm a musician and I ..."08:22
wpwrakyes, and then they comment that this is something that was state of the art a decade or longer ago08:23
wolfspraulespecially the theorists that are comparing specs, yes08:23
wpwrakof course, these too :)08:23
wolfspraulif you go to really creative people, they work with all sorts of stuff, decades old. who cares. the result matters, and being quick, and expressing yourself, and getting things done.08:23
wolfsprauljust go to any studio08:23
wolfspraulmay look like a museum to some08:24
wolfspraulso what...08:24
wpwrakno, but there are several comments that sound as if the intention was friendly, yet the people still don't see a particularly desirable product08:24
wolfspraulsure. we need to get that across.08:24
wpwrakand it's usually because the feature set is under-advertized08:24
wolfspraulcorrect08:24
wolfspraulbig marketing challenge08:24
wolfspraulaccessories are not clear either08:24
wolfspraulit's all way too abstract right now08:24
wpwrakmaybe make a (synthetic) picture as follows:08:24
wpwrak- MM1 in the middle08:25
wolfspraulwhich is partially also because the whole thing is still develping so quickly08:25
wolfspraulupdate over ethernet - wip08:25
wolfspraulaccessories list - not finalized, sourcing, evaluating, thinking08:25
wpwrak- visible "live" people (crowd on the dance floor, etc. can be symbolized)08:25
wolfspraulbox - not started yet08:25
wolfspraulmanual/tutorial - impossible to start because too many features still changing08:25
wolfspraultwitterwall - just added last week or so, I never saw it in action08:25
wpwrak- an image that applies some good-looking effect on the "live" image08:26
wolfspraulI see the glass half full, I think there were some good comments on slashdot08:26
wpwrak- arrows from the "live" image and from audio (put speaker, a dj's mixer table, microphone, etc., as visual indicator)08:26
wolfspraulmy friend pem is already calling me an asshole now though :-)08:26
wpwrak- arrow to the generated/merged image08:26
wolfspraulhe continues to be the master at jumping to conclusions, skipping over steps in his thought process, etc. fun guy.08:27
wpwrak(pam) have you established a communication channel with him outside the /. comment section ?08:27
wpwraks/pam/pem/08:27
wolfspraulhe said "assholes like you have given open source such a bad name"08:27
wolfspraul:-)08:27
wpwrakpem = psychotic egomaniac misfit ?08:28
wolfspraulhe doesn't even see the irony, oh well08:28
wolfspraulnah, no communication channel08:28
wolfspraulhe did something useful, pointing to the proprietary/confidential headers08:28
wolfsprauleverything else from him is rubbish08:29
wpwrakhave you established communication with lattice legal then ?08:29
wolfspraulno, I wait for pem08:29
wolfspraulthis is all a no-issue imho08:29
wolfspraultotal nonsense08:29
wolfspraullattice legal will not get into this08:29
wolfspraulif pem doesn't want to use Mico32, I think nobody will care less than lattice :-)08:29
wpwrakhis objective is mud wrestling. if you don't establish a non-public channel, you will not get anything from him08:29
wolfspraulI don't need it.08:29
wolfspraulI think we should leave the header as-is.08:30
wpwrakyou still have the issue of the problem headers08:30
wolfspraulnobody is proud of the mico32 core anyway, it's just something open source we took and built an soc around it.08:30
wolfspraulthere is no problem, aside from the license being a maze08:30
wolfspraulit is very clear in my mind that those sources are considered 'open source', including use, modification, manufacturing, by Lattice08:30
wolfspraulwhat we are doing now is climbing around those legal constructs, for no good08:31
wolfspraulI'm not happy about them, they should be clearly BSD/MIT/APache or GPL. but that won't happen.08:31
wpwrakvery well, but it's lattice's explicit statement, right on top of them, that they aren't08:31
wolfspraulthe statement says you need a license08:31
wolfspraulLICENSE.LATTICE08:31
wolfspraultheoretically we have a long wish list for lattice legal :-)08:32
wolfspraul1) release those files directly, downloadable from lattice.com without registration, without being the output generated from proprietary software08:32
wolfspraul2) use an established license, not a new one. Apache/MIT/BSD.08:32
wolfspraul3) cleanup the headers to reflect all this08:32
wolfspraulwon't happen though, waste of time08:33
wpwrakyou see a nice piece of green. you know it's a public space and that in your city, walking over public spaces is permitted for anyone. there's a sign "no trespassing. trespassers will be shot without warning.". will you feel confident marching over it ?08:33
wolfspraulit won't change08:33
wolfspraultake it or leave it08:33
wolfspraulthere are tons of people picknicking on that space, btw08:33
wolfspraulbut of course you are free to not join them because of the signs08:33
wpwrakwill you feel confident sending others over this area ?08:34
wolfsprauland the owner of the space made clear in a big ceremony with the mayor that anybody could use the space08:34
wolfspraulhe just forgot to take down the stupid signs08:34
wolfspraulabsolutely, of course08:34
wpwrakwell, if it's so clear-cut, you should be able to work out a solution with lattice08:34
wolfsprauldon't underestimate bureaucracy08:34
wolfspraulthis is Lattice _LEGAL_ :-)08:34
wolfspraulit has that name for a reason08:35
wolfspraulthe mico32 core, based on the open source license, is used in many products08:35
wolfspraulit's so obvious that this core was released as 'open source' imo08:35
wolfsprauljust read the mico32 homepage on lattice.com ... we discussed it before.08:35
wolfspraulthe intention is clear08:35
wolfspraulthe wording and precise implementation is unfortunate08:36
wolfspraulalso with the mico32 system builder download, generated files, etc. etc.08:36
wolfspraulall bad08:36
wolfspraulwon't change though08:36
wolfspraulmaybe if google would come and say "we don't use the mico32 core in an upcoming chip unless you clean this all up", they would do it08:36
wolfspraulfor us - never08:36
wolfspraulwe can take it or leave it, they won't care08:37
wpwrakyou're just trying to weasle around addressing the issue. people like pem put their criticism in words that make it difficult to agree with them, but the problem is a fairly clear one08:37
wolfspraulthere is no problem08:37
wolfspraulthose that see one cannot join08:37
wolfspraulthe party on the public space continues...08:37
wolfspraulthe only solution would be to switch to another core08:37
wolfspraulit's 8000 lines of code08:38
wolfspraulif milkymist is successful, it will be replaced one day08:38
wpwrakwell. lattice themselves tell me that something is proprietary and confidential. wolfgang who doesn't even talk to lattice tells me it isn't. wolfgang wants to sell me a product that he can only sell if what he says about lattice's work is true. he can't sell the product if what lattice themselves say about their work is true. now, seriously, why should i believe this guy ?08:39
wpwrakyou have a big credibility problem there08:40
wolfspraulyou know that you take 1 word out of context08:41
wolfspraulif you want to do that it's your problem08:41
wolfspraul2 words that is08:41
wpwrakwhy should i invest my time and money into a technology where a major IP issue that came up right after a few hours of people looking into it is deal with nothing but denial ?08:41
wolfspraulif you start reading from teh homepage, and you walk through the exercise, it is clear that those files are covered under a bsd-like open-source license, including manufacturing rights08:41
wolfspraulit is not a major IP issue08:41
wolfspraulonly for the paranoid08:41
wolfspraulyou take 2 words out of context08:42
wolfspraulthat's a major problem for those that do it08:42
wolfspraulthe real question is this:08:42
wolfspraulhave we misunderstood lattice?08:42
wolfspraulis there a risk in lattice using legal action against open source users of the mico32 core?08:42
wolfspraulthose are the real questions08:42
wolfspraul(to me)08:42
wpwrakwhat i'm saying is that your story is weak. so weak that you're even afraid of contacting lattice legal to see if this can be resolved.08:42
wolfspraulI agree with the problem to newcomers.08:42
wolfspraulnot afraid, waste of time08:42
wolfspraulafraid people will have another 100 things to be afraid of08:42
wolfspraul"reverse engineering fpgas is illegal"08:43
wolfspraul"this is all patented"08:43
wolfsprauletc. etc.08:43
wolfspraulI agree that it's unfortunate, and bad for newcomers.08:43
wpwrakyou spend months making that product, spend days arguing about why this isn't an issue, but you're completely overwhelmed by the task of writing a short mail to lattice legal to inquire ?08:43
wpwrakthere are plenty of other small issues, agreed08:44
wpwrakbut this is a quite in-the-face contradiction08:44
wolfspraulif the probability of progress in one activity is 1%, and in the other 0%, then you can still do the one with a 1% progress probability at infinite and do the right thing08:45
wolfspraulso even if we discuss this another 1000 hours, the 10 minutes to email lattice legal are still a waste of time08:45
wolfspraulyou understand that the issue is entirely absurd?08:45
wolfspraulif the mico32 core is indeed 'confidential and proprietary', then all lattice webpages talking about it would be badly misleading08:46
wolfsprauldon't you think?08:46
wolfspraulthis header is a result of the stupid mico32 system builder code generation08:46
wolfspraulwhich raises a whole rats nest of legal issues anyway08:46
wolfspraulregistration, running proprietary software, license for that proprietary software, and so on08:46
wolfspraulso the issue is not that 1 line in the header08:46
wolfspraulthe issue is the whole genesis of those files08:46
wpwrakno. mailing lattice legal can strengthen your argument. if they simply ignore you, you can say that a) there the license granting you these right, and b) lattice have for all practical purposed abandoned the project08:47
wolfspraulwhat do you think we should ask lattice?08:48
wolfspraulif anything, the best would be if they could offer just the sources for download directly08:48
wolfspraulbut that won't happen, ever08:48
wolfspraulthe lawyers cannot even do it, because those files are output files generated by the system builder, and so on08:49
wolfsprauland all downloads require a registration too, it seems08:49
wolfspraulask them to allow us to change this header?08:49
wolfspraulask them to write a new license for us?08:49
wolfspraulask them to create a new tarball with new license, and email that to us as an attachment?08:49
wolfspraulseriously. nothing will ever happen.08:49
wolfspraulask for a clarification? which clarification exactly?08:50
wolfspraulthe mico32 system builder makes it very difficult to address the real issue08:51
wpwraktell them that you're using their core in a open source product. tell them that you're using it because of announcement (link) and license. tell them that you noticed the inconsistency in the files that are subject to the license. tell them this is a problem, because you cannot make an authoritative representation of lattice's legal position on this before your customers on lattice's behalf. ask them how this can be solved.08:51
wolfspraulwhich is an issue of discomfort on our side, not so much a legal issue08:51
wolfspraulwell. except for crazy people like pem, everybody would take the lattice mico32 homepage to be an 'authoritative representation', no?08:52
wolfspraulwhereas pem of course dimissed it right away, trying to find something he can start his rant on08:52
wpwraka statement that these files are indeed to be considered as open source as of appendix c irrespective of what their header says would already be sufficient. particularly if they allow you to post this publicly.08:52
wolfspraulhave you read the mico32 homepage?08:52
wpwrakdoes the home page mention specifically the files with the conflicting terms ?08:53
wpwrakor does it just talk about the bundle08:53
wolfspraulyou have to be pretty paranoid to dismiss that as wrong (!), and then get stuck on some lines that are a result of the automatic code generation process, and clearly superseded by the license file08:53
wpwrakdoes it say that every last bit in the bundle is open source ?08:53
wolfspraulof course not08:53
wpwraksee08:53
wolfspraulbut you talk about 'authoritative representation'08:54
wolfspraulfor whom now?08:54
wolfspraulfor a lawyer?08:54
wolfspraulfor a paranoid engineer?08:54
wolfspraulfor a layman, the issue is clear08:54
wpwraktowards your customers, distributors, etc.08:54
wolfspraulunless your paranoia level is somewhere at 9/10 or 10/10, this would never come up08:55
wolfspraulmico32 is used in many places08:55
wpwrakinstead of giving them clear evidence, you're asking them to make a judgment call about the intentions of parties they don't know08:55
wolfspraul(under the open source license)08:55
wolfspraulthe best evidence I have is the mico32 homepage, and license.lattice appendix c08:55
wolfspraulif that's not enough, sorry I cannot help any further. they need to go elsewhere.08:55
wpwrakmaybe they're all in violation of the license and lattice just hasn't bothered to attack them yet08:55
wolfsprauloh, other evidence is that if you google, you will find many others agreeing with that position. (which of course pem immediately dismissed as well)08:56
wolfspraulof course, possible08:56
wolfspraulno risk no fun08:56
wolfspraulit's a paranoid argument08:56
wpwrakno, really. what you need is a statement from lattice. if the situation is as clear as you think it is, that should be no problem08:56
wolfspraulmaybe google has hidden some patents in vp808:56
wolfspraulwe won't get that08:56
wolfspraulso we have only 2 options:08:57
wpwrakif legal stonewalls, you can also ask them for a contact that can make a decision08:57
wolfspraul1) use mico3208:57
wolfspraul2) replace it with something else08:57
wpwrakis 2) a realistic option with the present release schedule ?08:57
wpwraki don't think you have an alternative to 1). else you wouldn't afraid of approaching lattice legal.08:59
wolfsprauloh sure I think #1 is totally fine and great09:00
wolfspraulthis is a completely idiotic issue09:00
wolfspraulsimilar to "there are hidden patents in xxx"09:01
wpwrakand i think (leaving rethorics aside now) you are afraid. namely that they tell you in more or less clear terms these files aren't as Free as they seem to be09:01
wolfspraulargh :-) that's about the last thing.09:01
wpwrakbut see the positive side. if lattice choose to maintain this threat, then this also strengthens the motivation for 2)09:01
wolfspraulyou don't seem to understand how many people have made this potentially wrong assessment then09:01
wolfspraulit's not just about qi/milkymist and mico3209:02
wolfspraulit's about lattice vs. world09:02
wolfsprauldo you think we are the first ones to use the mico32 core, open source or proprietary (but under the os license)?09:02
wolfspraulyou have to admit that your theory is quite paranoid09:02
wolfspraulplease keep the mico32 homepage in mind09:03
wpwrakwell, there are numerous companies that use gpl-licensed software yet blatantly violate the gpl. most of then probably haven't even heard that you can get successfully sued for that ;-)09:03
wolfspraulshould I quote from it?09:03
wolfspraulif Lattice turns around and sends Sebastien or qi-hardware.com a cease-and-desist letter, that would be awesome09:03
wolfspraulwe could turn that into news for Milkymist for months09:04
wolfspraulbut it's all completely nonsense, I'm sure. I won't do more in this direction, it's just too stupid for me.09:04
wpwrakas an example of one entrepreneur choosing to ignore an obvious warning sign and getting burned for it ?09:05
wolfspraulplease let's focus on reality. reality is that they opened the core, while keeping a lot of peripherals closed.09:05
wpwrakwhere were the news for months about openmoko vs. sisvel ? ;-)09:05
wolfspraulso milkymist doesn't use any of those peripherals09:05
wolfspraulthey were hoping for wider adoption of the open core09:06
wpwrakif it's all so clear, why don't you ask lattice to confirm this ?09:06
wolfspraulthis is all no news, this is what they've been doing since 200609:06
wpwrakyou're contradicting yourself09:06
wolfspraulno I've always said the same thing, which is that it's clear imo09:06
wolfspraulonly unfortunate and legally irrelevant snippets from the code generation process09:07
wolfspraulwhich are superseded by the license agreement09:07
wpwrakeither it's perfectly clear, without room for interpretation. then you'll have your answer quickly enough.09:07
wolfspraulperfectly clear to me09:07
wolfsprauland won't get an answer either09:07
wolfspraulthe header is unfortunate though, for newcomers09:07
wolfspraulthat is sad09:08
wpwrakthe only certain way for not getting an answer is not asking :)09:08
wpwrakwell, actually it isn't09:08
wolfspraulpem already helped us09:08
wolfspraulthey have to contact us now, if there is an issue09:08
wolfspraulif they don't - all clear09:08
wpwrakwhat will happen instead is that the likes of pem will communicate with lattice on your behalf, bringing all their social skills and balanced views of the issues at hand into the conversation09:08
wolfspraulno problem, he talks to legal09:09
wolfspraulthey are used to people who are a little hot tempered :-)09:09
wolfspraullet's do this: I wait another 1-2 weeks, and then I send a friendly email to that address as well.09:10
wolfspraulhow about that?09:10
wpwrakthey don't have to contact "us". they only have to answer to pem. you haven't even established a communication channel with that guy.09:10
wolfspraulI don't want to be in one boat with this social outcast.09:10
wpwrakheh ;-)09:10
wpwrakbut you should ask him to pass on whatever he gets from them09:10
wolfspraulno but he pointed them to their 'confidential' source file on a qi-hardware.com server09:10
wolfspraulso if that is a problem to lattice legal, they HAVE TO go after that server09:10
wolfspraulthey cannot just tell pem "it's bad", and not go after the server. that would be a big mistake that lattice legal will never make.09:11
wolfspraulI wait 1-2 weeks09:11
wolfspraulI ahve nothing to do with this guy.09:11
wolfspraulthen I will write them a friendly mail asking for a cleanup (need to think what I actually want though)09:12
wolfspraulI will cc you, how about that? :-)09:12
wolfspraulthere is no way in this world that this email will lead to _ANYTHING_09:12
wolfspraulif I loose, the next Asado in Buenos Aires is on me, all meat, everything09:13
wpwraki'll start looking for some fine winery then ;-)09:17
wpwrakthe problem with leaving this hanging in the open is that anyone can make up their own interpretations, however it suits their motives.09:19
wpwrakgetting this cleared up, at least a bit more, is just about the same as making sure your device meets FCC/CE emission standards09:20
wolfspraulit's not hanging in the open09:20
wpwrakit's inconvenient and may not really add any value for you, but it's necessary nevertheless, to protect those who use/sell your product09:20
wolfspraulit's about as much 'hanging in the open' as people saying the GPL is unconstitutional, or saying whatever09:21
wolfspraulit's really really far fetched this theory09:21
wolfspraulthere is a whole big license09:21
wolfspraulthere is an even bigger homepage09:21
wolfspraulI am sure there are press releases09:21
wolfspraulbut then you find two words and you think they are overriding everything else?09:21
wolfspraulINSANITY09:21
wolfspraul:-)09:21
wolfspraulwhy do those 2 words override everything else?09:22
wolfspraultotally idiotic, really09:22
wpwrakwell, let's say i put a big text "all the code i write is open source" on my web page. then you and i agree, just among ourselves, that i write a bit of closed source for you. would my broad public announcement supersede the specific private contract ?09:22
wolfspraulwhat I am really interested in whether there is a massive misunderstanding here09:22
wolfspraulbut in the tech industry, that must be one of the biggest misunderstands ever09:22
wolfspraulI don't care about crazy arguments like "oh my god, there is a word 'confidential' here, did you see that? this word clearly has super-power and overrides everything else"09:23
wpwrakno, but having that word on the file it apples to overrides some general statement elsewhere09:24
wolfspraulwe are not discussing a release that was made yesterday, it was made in 200609:24
wpwrakparticularly since it's on public record that you're aware of this09:24
wolfsprauloh no. the files go on to mention that you need a license to use this file.09:24
wolfsprauleven better (public record). the whole world thinks the mico32 core is 'open source'.09:24
wolfspraul2 weeks, I send a short and nice mail09:25
wolfspraulgotta go now09:25
wpwraka pretty big part of the world believes in creationism and similar concepts. so what does that prove ? ;-)09:25
wolfspraulwpwrak: let's focus on reality - core vs. peripherals09:25
wolfspraulsome peripherals are closed09:25
wolfspraulbefore emailing them, I would probably need to study the mico32 system builder closer, argh09:25
wolfspraulunless I want to write stupid mails like pem does09:26
wpwrakcan't hurt to do this cleanly. maybe you'll even find more spots in need of clarification.09:26
wolfsprauland all this has no legal or economic value, only to satisfy the paranoid, even more argh :-)09:26
wpwrakcompanies opening a previously closed product with homebrewn licenses are fun, aren;t they ? ;-)09:27
wolfspraulexactly09:27
wolfspraulthat's what happened here09:27
wolfspraulyou know this, I know this09:27
wolfspraulthere is no way Lattice would 'retract' the 'open core'09:27
wolfspraulbut unfortunately they will also not clean it up, I'm sure09:27
wolfspraulit's too difficult09:27
wolfspraulonce you got this wrong at the beginning, it stays wrong09:27
wolfspraulwe should have emailed them in 2005 or 200609:28
wolfspraulthere are way too many customers relying on this open core now09:28
wpwraki also know that companies aren't necessarily as stubborn as they may seem. but you may have to be persistent, because things can easily get dropped on the floor, simply because someone couldn't figure out what to do and then forgets09:28
wolfsprauleverybody would be worried if they even changed 1 word09:28
wolfspraulbbl09:28
wpwraki've seen this happen at philips. there was that big webcam issue with a driver in the linux kernel09:28
topi`hi. is sebastien here?09:33
topi`i'm interested about the milkymist board design. how big a footprint does the latticeMico32 core eat from the LX45?09:38
wpwrakmay be a bit hard to catch sebastien today. he's traveling, had an event yesterday in amsterdam and one tomorrow in paris.09:50
kristianpaulwolfspraul: (like only first two) not only, but seems for me the most interesting, not winamp-like too much :-)12:36
kristianpaulterpstra: ping12:57
kristianpaulI wonder if you read the entire backlog about lattice "open source license here..12:58
kristianpaulIn any case, do have have similar concern about this topic?12:58
wolfspraulkristianpaul: yes agreed, they are definitely the most interesting.13:01
kristianpaulterpstra: I'm hoping you may be a legal team also there ;-)13:01
wolfspraulI think once we focus to document all patches we already have with screenshots, and we focus on generating interesting shots with the camera, and then we focus on new interesting patches, then it should be really cool :-)13:02
kristianpaul(camera) yeah, i need take real screenshots, also i have the black cloth now13:02
kristianpaulterpstra: http://www.ohwr.org/projects/lm32/repository/entry/cores/lm32/rtl/lm32_top.v just in case poeple at cern dont bother or get paranoid with source code havint the words "proprietary" in the header13:04
wolfspraultrust me, if cern asks their lawyer about this, they will start reading documentation and legal matters from the beginning, not from somewhere inside after doing a 'grep' :-)13:12
wolfsprauland if they do that, they will come to the inevitable conclusion. but... the more eyes...13:12
kristianpaulThats true (grep paranoid)13:15
juliusb_where's lekernel?15:06
juliusb_i hvae a question about revision control of cores in milkymist15:06
juliusb_i want to do something like that for orpsoc15:06
juliusb_or at least, detecting the version is what at runtime15:08
wpwrakjuliusb_: he must be somewhere between amsterdam and paris :)15:31
juliusb_ah, hard life16:02
juliusb_:)16:02
lekernelhi16:09
lekernelazonenberg: what platform?16:09
lekernel(with spartan 3a/6/virtex)16:09
lekernel"It seems there are a lot of people (here at least) who seem to have some very fundamental disagreements with other people making a nice product." lol16:40
wpwraklekernel: it's called "penis envy" ;-)16:41
kristianpaulno, la donacion es opcinal ud elije cuanto pagar16:44
kristianpaulargg16:44
kristianpaulsorry16:44
azonenberglekernel: the softcore MIPS i'm practicing verilog on18:39
lekernelwell I think you can definitely do 1080p on spartan619:05
lekernelif you use the serializers you can reach 1 ghz/pin19:05
kristianpaulokay, even the system config in flickernoise is "linked" to graphics...22:13
kristianpaulargh22:13
azonenberglekernel: Nice23:12
azonenbergWhy is mm1 limited to 1024x768 then?23:12
azonenbergand why VGA rather than, say, DVI?23:13
azonenbergspartan6 has TMDS, right?23:13
--- Sat May 14 201100:00

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