#qi-hardware IRC log for Friday, 2010-12-17

qwebirc58347were can i get the wifi chip at for the nanonote00:09
kristianpaulrxvt: hi00:22
kristianpaulcheck this http://en.qi-hardware.com/wiki/Ben_NanoNote_Wi-Fi00:22
rxvtkristianpaul: i couldnt find anything on there about buying one of the chips00:22
adamwangwolfspraul, wpwrak : you there?00:24
adamwang4.65mm - 3.6 mm = 1.05 mm, 7.6 mm + 1.05 mm = 8.65 mm00:26
adamwangnow P1 (Digi -Key) S7002-ND the height is 8.5mm00:27
adamwangJ5 (2.6mm + 6mm = 8.6mm), but actually only 6mm, so i need to find a P1 with 6.05 mm (8.65mm - 2.6mm = 6.05mm)00:30
adamwanghope I don't calculate in wrong00:30
wolfspraulhope we can find this00:34
wolfspraulnow we have already shipped out m1, so if anything we need to make the adjustment on the jtag-serial side00:34
adamwangyes, i am finding..:)00:34
wolfspraulmaybe clip off the 4-pin serial header on jtag-serial? that's what yanjun luo ended up doing00:34
wolfspraulbut it's a bit hard to clip it off, because it's a female version so it has metal, plastic, all together00:35
adamwangyeah...it's a female type00:35
adamwangwill ugly with poor quality if crop it on jtag/serial brd.00:36
adamwangwell..let find digi-key parts first :)00:36
adamwangwolfspraul, http://www.mouser.com/Interconnects/Rectangular-Connectors/Headers-Wire-Housings/_/N-7aunt?P=1z0wxp7Z1z0z63xZ1z0wxp6Z1z0wxp8Z1yzv7x101:28
adamwanggo to the lowest two links(datasheet)01:28
adamwangsorry, the last second and third one01:29
wolfspraulok go on01:31
adamwangor this (low profile from 3M)http://www.mouser.com/ProductDetail/3M-Electronic-Solutions-Division/929984-01-04/?qs=sGAEpiMZZMtsLRyDR9nM19IUZhCvwtX27JEp18A0Hio%3d01:33
wolfspraulactually I don't know how boom can generate a shopping list for multiple distributors01:34
wolfspraulright now it only displays one list, and the p/n are digikey numbers01:34
adamwangthe P1 I calculated is 6.05mm AT MOST. So we should can use 3M low profle01:34
wolfspraulso you will get this from mouser?01:35
adamwangi know, this P2 temporarily we may pick from other distributer01:35
wolfspraulhow about the jtag (7*2) connector?01:35
wolfspraulalso mouser, or that one digikey?01:35
adamwangcan use that one in Digi-Key (79109-1006)01:35
adamwangof course if later we found omst of shipping list can buy from mouser, then we buy from mouser.01:36
adamwangright now , i don't know. i just tried to find the parts I am worrying first.01:37
wolfspraulyes sure01:38
adamwangother electronic parts should be ok.01:38
adamwanghow do you think?01:38
wolfspraulI can discuss with wpwrak how we can extend boom to multiple distributors (it's already possible, just needs a bit of work at the surface)01:38
wolfspraulyes sure, all good01:38
wolfspraulonly keep us posted for any differences from the shopping list boom generates right now01:38
adamwanglet me first review all parts done in shopping list and memo the new links I found first.01:38
wolfspraulso for everything on that list that you actually don't buy, or buy in another quantity, or from another source, I would like to update boom01:39
wolfspraulbut you should definitely move ahead and just get these things produced01:39
wolfspraulboom has to be able to handle reality :-)01:39
wolfspraulso no waiting01:39
adamwangyes..i just recorded on my BOOKSHELF first then we determine later...01:40
wolfspraulthose connectors will always need to be matched manually anyway, I'm sure. that's not worth trying to automate in boom.01:40
wolfspraulgood, yes01:40
adamwangwpwrak, wolfspraul when i use 'dsv' to review bookshelf, the bookshelf itself only shows few components(or just classify simply already).02:53
adamwangfrom bookshelf it seems that I can not sure for example resistors or capacitors will be missed? is it possible?02:54
adamwangbecause i can see the final 'make show' parts list with every part without doubting.02:56
wolfspraulthat was too fast for me02:58
wolfsprauldsv and boom are two totally separate tools02:58
wolfspraulwhich one do you want to talk about - dsv or boom?02:58
wolfsprauladamwang: got disconnected...03:01
wolfsprauldsv and boom are 2 totally separate tools03:02
wolfsprauldsv works with BOOKSHELF, that's all03:02
wolfsprauladamwang: is your question about dsv or boom?03:03
adamwangi'm trying to understanding the system more. when I review footprint of 1206/0603/0402 package03:04
adamwangi have many ways can review: one from layout/or gerber file, one from shopping list (digi-key)backward to check if the part link is correct.03:05
adamwangso i am now from the shopping list backward to review actually footprints used in PCBnew.03:07
wolfspraulwe should pull all links to PDF files into BOOKSHELF03:07
adamwangwow..damn it. current bookshelf itself doesn't have resistors and capacitors' pdf link.03:08
adamwangso you're saying i need to do this now?03:08
wolfspraullet's do one together03:09
wolfspraulwhich one are you looking at?03:09
wolfspraulit's not that bad :-) BOOKSHELF has 12 entries right now03:10
wolfspraullet's add one...03:11
wolfspraulfor example the 33pF capacitor03:11
wolfsprauldigikey p/n is 311-1020-1-ND, as shown by boom03:11
adamwangyes, go on03:12
wolfsprauldigikey says the datasheet is http://www.yageo.com/documents/recent/UPY-GP_NP0_16V-to-50V_4.pdf03:13
adamwangok...now you don't need to do this03:13
adamwangi do this on my site03:13
wolfspraulwait a sec03:13
qi-bot[commit] Wolfgang Spraul: added 33pF 0402 capacitor to BOOKSHELF http://qi-hw.com/p/mmone-jtag-serial-cable/34eaee803:17
wolfsprauladamwang: hmm03:17
wolfspraulit worked03:17
wolfspraulso I added a new entry to BOOKSHELF, also with the aliases C1 and C303:17
wolfspraulthen I ran "dsv setup BOOKSHELF" to download the new file03:17
wolfspraulnow I can do "dsv C1"03:18
adamwangyes, i know your way.03:18
wolfspraulthe I dit git commit -a and git push.03:18
wolfspraulso right now, dsv and boom are totally separate.03:18
wolfspraulthat means boom can choose another capacitor, and dsv will not know about it.03:18
wolfspraulfor example when we add mouser to boom.03:18
wolfspraulor a new capacitor manufacturer03:18
wolfsprauldsv is simply a small tool to manage pdf files03:19
wolfspraulI think this is good right now, we can start like this.03:19
wolfspraulwe cannot design all tools perfect at the beginning, we have to start using them, and then slowly improve.03:19
wolfspraulso right now, you just add those PDF links manually into BOOKSHELF, even though boom does not know about BOOKSHELF.03:19
wolfspraulthen we can discuss with Werner how to improve the tools. Slowly because we have a long todo list of things we want to improve :-)03:20
wolfspraulwhat do you think?03:20
adamwangyes, agreed.03:20
adamwangbut my gaol today is to determine order all them from Digikey or Mouser.03:21
adamwangif to do all of them...man.03:21
wolfspraulnot that many missing03:22
wolfspraulI can add them if you like, including the aliases for schematics references03:22
wolfspraulno problem03:22
wolfspraulshould I do that?03:22
adamwang311-1020-1-ND from shopping list it says 10 pcs for one set?03:23
wolfspraulwe need 2 for one jtag-serial, and probably the price for 10 is lower than for 203:23
adamwangbut from schematic I only see TWO.03:23
adamwangam i wrong?03:23
wolfsprauldigikey only sells 1003:23
wolfspraulremember that boom uses a real digikey database03:24
wolfspraulincluding volume discounts03:24
wolfspraulso if we need 6, then maybe buying 10 is cheaper than 6*103:24
wolfspraulthen boom will buy 1003:24
wolfspraul(for example)03:24
wolfspraulthe things boom does not consider right now are:03:25
wolfspraul1) shipping costs03:25
wolfspraul2) lead time (it looks at stock level though)03:25
wolfspraul3) minimum USD amount per order03:25
adamwangwow..ok..that's the boom's 'thought' now03:25
wolfspraulbut it does consider volume discounts03:26
wolfspraulif you want me to update BOOKSHELF just let me know, I am glad to be able to help you03:26
adamwangi hope there's one column which can show exactly how many 33pF we used in schematic03:26
wolfspraulyou mean in the shopping list?03:27
adamwangno need? like I made mistakes in rc203:27
adamwangi forgot to add some resistors actually.03:27
adamwangso will 'boom' happen?03:28
wolfspraulwhen you are in the bom/ directory, try this command: cat usb_jtag.ord03:28
wolfspraulit will list the schematics references behind each item, just for a quick look now03:29
wolfspraulwe can ask Werner what he thinks about adding a count, or list of references, to the 'pretty printed' 'make show' shopping list03:29
adamwangok, so this .ord file will show exact qty brd used?03:29
wolfspraulI know you want that for cross-referencing/double-checking.03:29
wolfsprauldid you do 'cat usb_jtag.org'?03:30
adamwangi saw it03:30
adamwangit's good03:30
wolfspraulthere is no count, but the schematics references are listed. you can count yourself :-)03:30
wolfspraulthis can easily be formatted differently, when werner is back we see what he thinks03:30
adamwangwell...that's why I want a quick 'count' data of column then I can see it...03:31
adamwangthe i don't need to see other .org file...agreed?03:31
adamwangwaste time03:31
wolfspraulthe .ord shows you the schematics references03:31
wolfspraulup to you, the .ord is always there, it's part of the boom processing03:32
wolfspraulit's generate automatically when you run 'make clean show'03:32
adamwangpls add me to jtag member, tks.03:36
wolfsprauladamwang: done03:39
adamwangcan i just use "git add BOOKSHELF" only?03:50
adamwanghm..i saw help, should be no problem.03:51
wolfspraulmost important is git commit, then git push03:52
adamwangfatal: The remote end hung up unexpectedly<<< when i 'git push'04:07
adamwangi have long time no use 'git', i should successfully 'git commit -a' with [master 44d1c56] trying to add R1504:08
adamwang 2 files changed, 21 insertions(+), 1 deletions(-)04:08
wolfsprauldid you do a new git clone after I added you as member?04:10
adamwangok...clone again04:11
wolfspraulwait one second04:11
wolfspraulah yes it should be OK04:11
wolfspraulremember I copied you my boom parts database, but it's in another folder04:11
wolfspraulso I think you can safely delete the mmone-jtag-serial folder, and clone it again, with04:11
wolfspraulgit clone git@projects.qi-hardware.com:mmone-jtag-serial-cable.git04:11
wolfspraulthen you need to create the usb_jtag.lst file by going into eeschema, and clicking on the create bom button at the top04:12
adamwangum...hope those parts database will still there.04:12
wolfspraulalso if you edited the BOOKSHELF, copy it somewhere so you don't need to reenter04:12
adamwangyeah..thanks reminder04:12
wolfsprauljust rename the whole mmone-jtag-serial-cable folder to mmone-jtag-serial-cable.old or so04:13
adamwangi still got fatal after 'git push'. Does my current 'password' not same as before you create?04:20
wolfspraultry: git pull04:21
wolfspraulthat works?04:21
adamwanggit pull works04:22
adamwang# On branch master04:22
adamwang# Your branch is ahead of 'origin/master' by 1 commit.04:22
adamwangi think the BOOKSHELF is being committed just no 'push' successfully.04:23
wolfspraulok wait04:23
adamwangadam@adam-laptop:~/mmone-jtag-serial-cable$ git commit -a04:24
adamwang[master 1e6288c] trying to add R15 and other P1 parts found04:24
adamwang 1 files changed, 14 insertions(+), 0 deletions(-)04:24
wolfspraulyou did the git clone git@... ?04:25
wolfspraultry: cat .git/config04:26
wolfspraulthere should be a line url = ...04:26
wolfsprauldoes it say url = git@...04:27
adamwangyes, it does04:27
adamwang[remote "origin"]04:28
adamwangfetch = +refs/heads/*:refs/remotes/origin/*04:28
adamwangurl = git://projects.qi-hardware.com/mmone-jtag-serial-cable.git04:28
wolfspraulhe :-)04:28
wolfspraulit does not04:28
wolfspraulit says git://04:28
wolfspraulbut we need git@04:28
adamwangoah...what's difference? sorry I don't know.04:29
wolfspraulno problem, it's a bit confusing04:29
wolfspraulthere are 2 ways for each project04:29
wolfsprauldepending on whether you are read-only (public, anybody in the world), or read-write (only members and admins of the project)04:29
wolfspraulthe read-only looks like this: git clone git://projects.qi-hardware.com/mmone-jtag-serial-cable.git04:29
wolfspraulthe read-write looks like this: git glone git@projects.qi-hardware.com:mmone-jtag-serial-cable.git04:30
adamwangso i did a read-only clone?04:30
adamwangoah..i see04:30
wolfspraulthere is a way to change it, but I don't know how, and xiangfu is just out04:30
wolfspraulso just do the git clone again04:30
wolfspraulnew tree04:30
adamwangsorry, delete and try again04:30
qi-bot[commit] Adam Wang: add R15 and P1 for other source http://qi-hw.com/p/mmone-jtag-serial-cable/238057904:37
wolfsprauladamwang: congratulations!04:39
wolfspraullekernel: how do you like this: http://en.qi-hardware.com/wiki/File:AC%28EU%29_to_DC_adapter_Fairway_WN10B-050_with_Milkymist_4_pixels.png04:53
lekernelnot as bad as on the side :)04:54
wolfspraulok good04:55
wolfspraulso we go with that for now04:55
wolfspraulthe dc-10 is slightly better too now, the one for bearstech was still glued :-)04:55
wolfspraulwe have ordered adapters with US connector as well04:55
wolfspraullekernel: another thing, I want to remove all usage of microSD, *SD* everywhere. we discussed a bit yesterday here, and our proposal is to call it 8:10 card04:56
wolfspraul8 pins, 10 mm (rounded)04:56
wolfspraulon the side we can say something like (*) microSD-like, but primarily we say 8:10 card first04:57
wolfspraulwhat do you think?04:57
lekernelI say "memory card"04:58
lekernelI've already removed usage of the "microSD" brand04:58
lekernelpeople will immediately understand "memory card", but "8:10 card" means nothing04:59
lekerneleven in the source code it's called "memory card"04:59
wolfspraulboth works for me04:59
wolfspraulok good that is great, main point is to stop using the *SD* trademarks05:00
wolfsprauladamwang: did you see lekernel's feedback on the milkymist sticker?05:01
wolfspraulthis is OK for now05:01
lekernelif we say "memory card reader that works with microSD cards", is that trademark infrigement?05:02
wolfspraulin such a sentence - probably not05:02
adamwangwolfspraul, ok with 4 pixels.05:03
wolfspraulbut the risk is that our own people simplify this and write "microSD" everywhere05:03
wolfspraulthat's why a new 'name', even if it is bulky like 8:10 card, helps align the home forces05:04
wolfspraulwhen speaking about 'our' side, the Milkymist One, the most we can say is that it is microSD-like05:04
wolfspraulin your sentence, you are describing that our 8:10 card holder happens to also work with microSD cards05:04
wolfspraulthat shouldn't be a trademark infringement05:05
wolfspraulthere is a line somewhere05:05
wolfspraulwe don't want to dance around with this line too much05:05
wolfspraulthe best is if microSD, or *SD*, just doesn't appear anywhere in our literature, web sites, etc.05:05
wolfspraulthat's also the easiest to enforce and maintain05:06
wolfspraulhe he05:07
wolfspraulmaybe it's even better to say "memory card reader that might work with microSD cards"05:08
wolfspraulwe probably must not claim that we can guarantee compatibility05:08
lekernelsounds a bit too unprofessional05:08
wolfspraulbecause without license we have no access to the standard05:08
wolfsprauland without access to the standard we cannot claim compatibility05:08
lekernelthere's the "simplified" standard available for free05:09
wolfspraulyes but you ask about my opinion. I'm not a lawyer but from past cases this is the kind of stuff you will hear from a lawyer.05:09
wolfspraulif you say "that works with microSD cards" - that may be too much already05:09
wolfspraulbecause you don't have access to the standard05:09
wolfspraulso how can you say that?05:09
lekernel"we tested it"?05:09
lekernelI don't really want to say "might work" either05:10
wolfspraulno I was kidding05:10
wolfspraulbut I realized the standard and compatibility problem05:10
wolfspraulfor example 'compatible' is a strong word05:10
wolfspraulwe cannot use it05:10
wolfspraulwe will be shut down legally05:10
wolfspraulyou cannot say microSD compatible05:10
wolfspraulthe most is microSD-like05:10
wolfspraulso I took that a little further to make it 'maybe' :-)05:11
lekernel"designed for microSD cards"?05:11
wolfspraulsounds very dangerous05:11
wolfspraulthat's exactly what you cannot claim05:11
wolfspraulwe can say that we bought a lot of cards, some of which claim to be microSD compatible05:11
wolfspraulour tests show that all of those cards work05:11
wolfspraulbut you cannot say that you know that you are microSD compatible05:12
wolfspraulor that you 'support' microSD05:12
wolfspraulor that you 'have' microSD05:12
wpwrakgood morning :) let's see what's in the backlog today ...05:12
wolfspraulafaik the maximum is microSD-like05:12
wolfspraulwpwrak: sebastien is not very fond of our beautiful 8:10 card05:13
wolfspraulhe wants to have a microSD card05:13
lekernelyeah, or maybe not mention "microSD" at all05:13
wolfspraul(joking, joking...)05:13
wolfspraulof course that's the best05:13
lekerneland keep "memory card"05:13
wolfspraulfine by me, it's not a very central feature of Milkymist One anyway05:13
lekernelexternal people will eventually say "microSD" anyway05:13
wolfspraulwe just need to keep ourselves clear from trademark infringement05:13
wolfspraullekernel: the defense line goes like this:05:17
wolfspraulsomeone says "you need a SD HALA license"05:17
wolfspraulwe say "no, we don't support any SD standard or protocol"05:17
wolfspraulthey say "yes, you support microSD cards"05:17
wolfspraul"no, we don't"05:17
wolfspraul"but what is that connector there"05:17
wolfspraul"that's a 8:10 card holder"05:17
wolfspraulit's difficult they have patents, copyrighted stuff, trademarks. but we need to start somewhere and staying clear of the trademarks is relatively easy.05:19
lekernelpatents? besides the connector, there's nothing that's patentable (at least in EU)05:22
wpwrak(boom & multi-distro) yes, you can have multiple distributors. see gta02-core. you have to modify the output generator to see them, though. (in gta02-core, you have a different makefile target for each, but that's probably not the most elegang way)05:24
wpwrak(boom: count/ref) good idea. let me add it to my to do list :)05:30
wolfspraullekernel: we seem to have a solution for now. We are purging *SD* everywhere. You say 'memory card', I say '8:10 card' when pressed or in tech specs, otherwise also memory card or flash memory card when explaining.05:36
wpwrak(usd/8:10 - claim compatibility) i don't see why a compatibility claim would be a problem. it's an assurance you make about your product. as long as you don't say it's certified ...05:38
wolfspraulno, 'compatible' is definitely no good05:39
wolfspraulthat word has been 'taken' by the other side :-)05:39
wolfspraulwe cannot claim compatibility unless sanctioned to do so by the other side05:39
wolfspraulwe absolutely cannot claim microSD compatibility05:40
wpwrak("memory card" vs. "8:10") i think "memory card" works fine in MM1 "product documentation", for descriptive purposes, because there's only one card anyway. the 8:10 is more something we need to identify cards if there could be other possibilities.05:41
wpwrak("compatibility") ah, language is tricky ;-) i meant the "works with microSD" or maybe "use with", so you give the customer instructions, without making any explicit claims about your product05:43
wpwrakwolfspraul: if you really want to take this to the end,  then have some "8:10 card" in your shop ;-)05:43
wpwraklekernel: (card name) one problem is that any names that "sound familiar" are dangerous. it just sucks, no matter what you do.05:44
wolfspraulyes all sounds good, I think we are all on the same page05:47
wolfspraullet's see how it goes05:47
wpwraknow, the dreaded connectors ... what are the possible actions available ? replace any part with one with a fitting (let's avoid "compatible" :) footprint ? replace only some parts ? re-layout JTAG ? ...05:47
wpwrak(JTAG) the JTAG board i mean05:47
wolfspraul"use with microSD" might just be OK still, especially if it is in a larger explanatory text05:47
wolfspraulwe have to stay away from saying that _OUR_ stuff is somehow *SD*, and we have to stay away from 'compatible', because that is understood to be a mutual thing05:48
wolfspraulfor the rest, we do as discussed here, should be OK05:48
wolfspraulwpwrak: no re-layout now, because we sent the gerbers out already05:49
wolfspraulthe connectors need to be positioned without through-hole, or soldered manually05:49
wpwrakso which parts can be replaced with different ones ?05:50
wpwrakdifferent but same-footprinted05:50
wpwrak(avoiding "compatible" is fun ;-)05:50
wpwrak"commensurate" is a good word :)05:51
wpwrakcommensurate parts :)05:51
wolfspraulI think adamwang found solutions for the connectors already.05:53
wolfspraulI just want to make sure we keep updating the bom and bookshelf so our KiCad files are in great shape :-)05:53
wpwrakwolfspraul: (8:10 card) you should post reason and proposed name on the list, perhaps also mentioning that in a reduced context like MM1 one can also use things like "memory card". then let's wait a few days. once the storm of indignation passes, we can start renaming the few references we already have05:54
qi-bot[commit] Werner Almesberger: modules/usd-card.fpd: added measurement of total length http://qi-hw.com/p/kicad-libs/c4f059205:55
lekernelthere are no references to SD in milkymist afaik05:55
lekernelI used "memory card" everywhere05:55
wolfspraulwhat does the leaflet say?05:56
lekerneland no one said a thing btw :)05:56
lekernelmemory card05:56
wpwraklekernel: "memory card" works well in the context of a specific product, because you already know that it has to be congruent with uSD05:57
wpwrak(hmm, all this avoiding of "compatibility" makes me wonder if some of the strange translations we get in the manuals of asian products may actually be intentional ;-)05:58
wpwraklekernel: "memory card" wouldn't work when specifying/discussing a new design, where is may also be full-size SD, CF, memorystick, or maybe a DIMM.06:00
wpwraklekernel: that is, unless you want to explain each time that it has to agree with uSD06:02
lekernelno big deal...06:02
lekernelsee, it's about the first time I hear about this "problem" when there are already 46 milkymist boards out06:03
lekernelall using the "memory card" terminology06:04
wpwraklekernel: the "problem" right now are more those 1000+ bens out there that proudly proclaim "microSD" on the case ;-)06:04
lekerneljust rename it to "memory card", I'd say06:05
lekernel8:10 card really does mean nothing06:05
wpwrakit's a name that is descriptive yet hasn't been taken yet. so yes, it doesn't mean anything _yet_06:06
wolfspraullekernel: the point of 8:10 card is that it means nothing06:06
wolfspraulthis is not meant to force a re-education upon people06:07
wolfspraulit's a placeholder for some tech specs that will give us a clean legal shield against claims of trademark infringement06:07
wolfsprauland it's a good name. 8 pins, 10 mm. we can use it in other places later.06:07
wpwrakthink of it like "the glass". if you're asking someone you're eating with to pass their glass, you won't bother to specify what the glass is like. if you go to a shop and are looking for one that matches your existing collection, you will use something more descriptive.06:07
wolfspraulof course we are not using 8:10 to explain anything06:07
wolfspraulso we are all on the same page I think06:08
wolfspraulmemory card is fine, it's great06:08
wpwrakwolfspraul: eventually, 8:10 will become a meaningful name by itself ;-)06:08
lekernelyeah, like GNU/Linux and Iceweasel...06:08
wolfspraul8:10 card is the fallback for places where we don't want to say memory card. how about that? :-)06:10
wolfsprauland for Milkymist One, we can probably say memory card almost everywhere, maybe everywhere.06:10
wpwrakjust wait until places like amazon get so many customes asking for "8:10 cards" that they add them as a product category ;-))06:10
wolfspraulI think we are all on the same page actually.06:10
wpwrakhah, better example. "meat". at a barbecue, everybody will understand things like "is the meat ready ?". but you'd look rather foolish ordering "meat" at the butcher shop.06:12
wolfspraullekernel: wpwrak works on a 8:10 card that has an Atmel RF IC on it. not quite a 'memory card'. but it wouldn't fit into the m1 board anyway... :-)06:12
zearwell, aren't SD cards more or less a dominent card standard on the market?06:12
wolfspraulnow we can test it!06:12
wolfspraulzear: we don't support SD06:13
zearyou mean you support uSD instead?06:13
wolfspraulwhat is uSD? do you mean microSD?06:13
wolfspraulwe don't support microSD either06:13
zearyes, i mean micro06:14
wolfspraul(sorry need to warn you, you are in a guinea pig test right now...)06:14
wolfspraulzear: no, we don't support microSD06:14
wolfspraulthat's a proprietary standard we don't have access to06:14
wpwrakwolfspraul: at least we're not "compatible". maybe we're "friendly" ? :)06:14
zearwolfspraul, i just want to say that here all the non-tech people call SD cards simply by "memory card"06:14
wolfspraulyes we are friendly to microSD, definitely06:14
wolfspraulwe support many memory cards06:15
zearsince nobody used any other format06:15
wpwrakzear: your digital camera uses uSD ?06:15
wpwrakzear: like the Ben uses full-sized SD ? :)06:15
lekernelyeah, exactly, that's why the "memory card" term is good (for M1 at least, where I don't plan any other use of the card connector)06:15
zearwpwrak, yes, i use uSD for my digital camera06:16
wolfspraulWerner is raining on my parade06:16
wolfspraulI didn't see any of this.06:16
wpwrakwolfspraul: we're very supportive of memory cards ;-)06:16
wolfspraulI have never seen or heard about anything SD06:16
wolfspraulmemory cards06:16
wolfspraulthat I know06:16
lekernelthough I'm not shocked of "wireless adapters" that plug into memory card connectors either06:16
lekernelSD means "secure digital", so it's not a good term anyway06:17
wpwraklekernel: "wireless" is a good example for a term that's easily used in an overly generic way. e.g., it would be borderline fraud if you sold the ben-wpan card simply as a "wireless card"06:17
zearwhen you say "a floppy disk", nobody thinks about 8 inch ones, am i right?06:18
wpwrakSD = Synchronous Data ;-)06:18
zearthe same way when you say "memory card" nobody thinks of anything else than SD family06:18
andres-calderonnewcomer...  What is the problem with the uSD connector? legal issue?06:18
wpwrakzear: i remember the days where you had the choice of 5.25" and 3.5", HD and DD, nicely on the rack next to each other06:19
wpwrakandres-calderon: yes. the name "microSD" is protected (trademark)06:19
zearwpwrak, lucky you, i never had a chance to mount a floppy on linux :P06:20
wpwrakzear: "sd family". perhaps. but you still don't know which member of the family.06:20
andres-calderonok, SDIO is protected  too?06:20
zearwpwrak, yes, but you have adapters06:20
wpwrakzear: i wrote some code in floppy.c ;-)06:20
zearwpwrak, nobody buys full SDs anymore, they usually sell microSD + fullSD adapters06:20
wpwrakzear: (adapters) like for putting that full-sized SD into your phone ? :)06:21
zearwpwrak, no, the other way out one ;)06:21
zearalthough i have such one06:21
wpwrakzear: you can still buy full-sizes SD pretty much everywhere06:21
zearwpwrak, they're mostly leftovers06:21
zearat least here06:21
wpwraksee what "memory card" you get at a sony shop ;-)06:22
zeari think they still sell that high speed fullSDHC for cameras, for any other task they sell microSDs06:22
zearwpwrak, why would you want to shop there? :D06:22
zearthe last time i went there, i returned with the only sony bravia tv that wasn't linux based06:23
zearquite a disappointment06:23
wpwrakone criterion for the name is also that it has to be non-interactive. "a memory card" followed by "no, the other kind" is not an option.06:23
zearwpwrak, then call them "es-dee cards"06:24
wpwrak(tv without linux) see, that's what happens if you're not specific enough ;-)06:24
wpwraki suggested 5D ;-)06:24
zear5D sounds great06:25
wpwrakthe enemy lawyers would have feeding frenzy :)06:26
wpwraks/have/have a/06:26
wolfspraullekernel: exactly, 'secure digital' is a trademark, not a name06:27
zearwhy not call them "insecure digital" then06:27
wolfspraulandres-calderon: we call them 8:10 cards now06:27
wolfspraul8 pins, 10 mm (rounded)06:27
wolfspraulah no, not 'them'06:28
wolfspraulour cards _are_ 8:10 cards06:28
wolfspraulwe don't know whether they are compatible with microSD or not06:28
andres-calderonwhere i can found  any  document to read about the uSD issue?06:28
wolfspraulandres-calderon: the issue is that 'Secure Digital' and anything *SD* is trademarked06:29
wolfspraulin addition they claim to hold patents, and copyrighted whatever06:29
zearwolfspraul, what about microMMC cards, are they different?06:29
zearin size/number of pins06:29
wolfspraulyes I think that's owned by someone else06:29
wolfspraulare you asking a legal or technical question?06:30
wolfspraultechnically it's probably 'the same' :-)06:30
zearthough i guess the driver can support them as well if they only can fit the slot06:30
wolfspraullegally it's not the same06:30
wolfspraulwe are mostly talking about legal stuff here06:30
zearbut 8:10 cards would point to two different things then ;)06:31
wpwrakmicroMMC has a different size. not sure about number of contacts/protocol06:31
wolfspraulandres-calderon: this is the issue: http://www.sdcard.org/developers/join/licensing/hala/06:31
wolfspraulas copyleft hardware, we try to make hardware that is as little dependant on such proprietary technology as possible06:31
wolfspraulotherwise the licensing requirements, even if Sharism Ltd. would sign the license, would perpetuate upon anybody making use of the copyleftness of our work06:32
wolfspraulso it would taint the copyleftness of our work06:32
wolfspraulhence we work on removing such taintedness06:32
zearbtw wolfspraul, i remember you mentioning somewhere once that you guys had 2 or 3 cases of nn warranty replacement due to weird stuff happening on screen, am i right?06:32
zearbecause my nn seems to change the LCD colors when i move with the hinge06:33
wolfspraulnot good :-)06:33
zearand i'd like to know what can be the problem06:33
zearmine is the pre-retail unit, so i have no warranty, and the make quality is probably worse as well ;)06:33
wolfspraulideally it's just a problem with the fpc going into the mainboard06:33
wolfspraulthen you could open that, reseat the fpc, and it would be stable again06:34
zearah, sounds like something an unskilled person could do06:34
wolfspraulworse, it could be a problem with the soldering of the connector the fpc goes into, or on the other side (the side of the lcm), where the cable is soldered onto the pcb right away06:34
wolfspraulyou have to be careful a little06:34
andres-calderonSDA sowed terror in the linux kernel community a  time ago.  But after reading the legal terms they found a simple solution: "we have not signed anything"06:34
wolfspraulhave you ever opened your nanonote? taken off the plastic around and above the keyboard?06:34
zearwolfspraul, nope, never06:35
wolfspraulhow bad is the color problem?06:35
wolfspraulif it's not very bad, maybe fix it when it gets worse06:35
wolfspraulwe'll fix it, no worries06:35
zearwolfspraul, the colors are sort of reversed06:35
zearthey're not mirror reversed, so black -> white06:35
zearit's more like what happens when you try to run on a different color depth06:35
zearand only in certain hinge positions06:36
zearand not always, but there's about 50% chance of this occuring when you open the case and move with the hinge06:36
wolfspraulfirst thing would be to try to reseat the fpc cable on the mainboard side06:36
zearmy nanonote is still pristine, i wish i had skills to fiddle with it a bit06:39
zearlike, to install that serial console chip i got from wpwrak via rafa06:39
zearunfortunatelly certain people forgot to take soldering tools with them for the jlime meeting, so they weren't able to install it into my nn ;P06:40
wpwrakzear: yeah. real pity that one.06:52
zeari'd love to play with the serial console, but as long as i don't find someone with hardcore soldering skills, that won't happen :(06:52
wpwrakzear: here's your chance to become a hardcore soldering man yourself ;-)06:53
zearnah, i already toasted my pc by resoldering a broken gamepad's cable and plugging it to the usb port06:54
zeari don't want to experiment with soldering anymore :D06:54
wpwrakzear: naw, don't worry. it's easy - and wolfgang likes to sell more bens ;-)06:54
zearonly if he accepts 100 ubuntu 5.04 cds as currency :)06:55
zeari once ordered 212 of them, but managed to giveaway only 11206:56
wpwrakah, the tragedy of the commons :)07:00
zearthis was back in 2006 or something, people weren't used to linux yet07:01
wpwrakwe live in very different timescales, it seems ;-)07:02
wpwrakif you had written 1996, maybe ...07:02
zearwell, back then there were no netbooks and linux wasn't popular on mobile phones07:03
zearnow most of the people i talk to heard of linux and are interested in getting free printed cds07:04
lekernelhow does canonical get money, btw?07:09
viricin the firefox, in ubuntu, they put "client=ubuntu" in the google requests.07:10
zearshuttleworth is also a multimilionare07:10
lekernelviric: that's all?07:11
lekernelzear: yeah, but this won't last undefinitely (except maybe through the magic of "financial instruments")07:11
viricno idea :)07:11
lekerneland why did he create canonical as a commercial company, if the objective is to operate at a loss?07:13
zearlekernel, well, he was the second space tourist, if he's still a millionaire after that, he must have a nice money flow07:13
lekernela non-profit organization would have been more appropriate07:13
lekernelwell, afaik he got all of it by selling thawte to verisign07:15
lekernelthat was a huge shot of cash, not really a flow07:15
andres-calderonPaid developers power the Linux kernel: http://news.cnet.com/8301-13846_3-20024219-62.html  (old news)07:39
viricandres-calderon: like openbsd's paid developers07:45
viricis it? :)07:45
andres-calderonviric: I like to get paid for doing what I like.07:47
viricI refer to the FBI related funds :)07:49
wpwrakviric: maybe andres would like to work for the FBI. well, CIA in his case. i'm sure they're recruiting ;)07:53
andres-calderonviric: wow07:53
rxvtok ive been looking for hours and i can not find a mico sd wifi card anywere, if anyone knows were i could buy one of these and could point me in the write direction it would be nice07:54
andres-calderonwpwrak: jeje07:56
andres-calderonThat risk will always exist in open source SW model. But, we just can  imagine the nice collection of  backdoors, troyan, etc in the most popular closed OS07:57
andres-calderonthe big brother also write code.07:59
qi-bot[commit] Adam Wang: Add alias names for some parts http://qi-hw.com/p/mmone-jtag-serial-cable/d222b3f08:01
wolfspraulexcellent, Adam is committing :-)08:03
wolfspraulandres-calderon: we need to jump into the kicad files too! :-)08:03
zrafalekernel: if IIRC canonical offers IT linux services like red hat and others.. so they admin servers, do virtualization solutions, etc.. Nothing special I would say, the same like red hat, suse, ibm, hp.. etc etc.. If you ask them to put a pet shop for you they will do08:04
wolfspraulI was slowed down by some work on snapshots and server rescue stuff, but it's almost done now, then back to more the stuff at the surface...08:04
lekernelah, yeah08:04
kristianpaulif i want remove or move files, later sync with git, should i do care of soemthing?12:51
kristianpaulor just git add as usual and he will realize new folder and deleted files?12:52
wpwrakgit rm  and  git mv   ?13:12
kristianpaulahhh ;-)13:13
kristianpauli'm a dumb with git :(13:15
wpwrakADHD ;-)14:28
Action: wpwrak wonders if jlime has as much trouble with their menu system as openwrt has with gmenu2x ...14:32
kristianpauli dont think so wpwrak14:32
kristianpaulwell it run on X :)14:32
Action: kristianpaul is at home today :)14:33
kristianpaulwell comparative is not fair14:34
kristianpaulmatchbox is not a menu14:34
kristianpauland gmenu2x want to be more that a menu ;)14:35
wpwrakwell, "launcher"14:35
kristianpaulah yes14:36
wpwrakthere seem to be a heck of a lot of very basic issues with gmenu2x. well, some may be just the result of an attempt to optimize for the last byte ...14:36
kristianpaul6Mb in ram is too much, for just a wallpeper and some icons14:37
kristianpaulmay be i wallpaper fault ;)14:38
kristianpaulOh look what i found http://wiki.openmoko.org/wiki/Launcher14:39
wpwrak"It features events notification too" ... "It is aimed at FSO framework" ... probably has some nasty dependencies14:41
kristianpaulthings get simpler when you elminate Xilinx dependent cores :)14:49
kristianpaultalking about sram > buffer used on SIE shared bus14:49
wpwrakportable is good :)14:50
kristianpaulI learn that from milkymist code :)14:50
kristianpaulwhoa it sinthetized at first try :), let see still working14:52
wpwrakwhat's that burnt smell ? (-:C15:04
kristianpaulplease join #qi-hardware15:08
kristianpaulwait wait not yet ;)15:08
kristianpaulwpwrak met Jay715:08
Jay7morning ppl :)15:08
kristianpaulfrom kexecboot.org15:08
kristianpauldevel* right?15:08
Jay7lead devel ;)15:09
Jay7and single seems atm :(15:09
kristianpauli just think you should met as i read lots of times abot kexec here from wpwrak  :)15:09
Jay7I've seen something like from qi people :)15:09
wpwrakhi Jay7 ! ah yes, i love kexec :)15:09
Jay7but there are a LOT of kexec-based shell loaders :)15:09
kristianpaulBtw Jay7 are you aware of Ben Nanonote, thats what we need the bootloader15:09
Jay7kristianpaul: even more15:10
Jay7I got nanonote from CELF to port kexecboot there :)15:10
kristianpaulJay7 also point me this http://labs.o-hand.com/psplash/15:10
Jay7as part of proposed work15:10
kristianpaulJay7: ahh !!15:10
kristianpaulYou already suscribed in the list?15:10
Jay7still in my todo queue.. I hope this will happens around christmas holidays15:10
Jay7kristianpaul: qi's one?15:11
kristianpauldid you tell this before, will be neat if you do15:11
kristianpaulJay7: yup15:11
Jay7no, point me :)15:11
kristianpaulJay7: http://en.qi-hardware.com/wiki/Main_Page <--  Discussion List - developers and users Please fill Signup15:12
Jay7ok, mom15:12
kristianpaulgood :)15:16
kristianpaulBtw do you have screenshots of psplash ?15:16
kristianpaulJay7: when i asked for app lanucher how kexecboot fits that question?15:17
Jay7kristianpaul: hm.. any of angstrom images have psplash running :)15:17
kristianpauloh let me see15:18
Jay7kristianpaul: kexecboot is C program that scans available partitions on available medias, builds FB menu and allows to select one to boot from15:18
kristianpaulwpwrak: not burned out, but i'm getting some not expected readins, i may missed timing with my implementation of Read-only ram :)15:18
Jay7then kexec is launched15:18
Jay7so.. there is all code you need :)15:18
Jay7fb menu + executing code15:19
Jay7but fb code is changing.. I'm refactoring it slowly15:19
wpwraka screenshot would indeed be nice15:20
Jay7here is our git: http://git.linuxtogo.org/?p=groups/kexecboot/kexecboot.git;a=summary15:20
Jay7here are some15:20
Jay7latest is old mockup, now we have other outlook15:21
Jay7sorry, have no screenshots of new UI15:21
Jay7I should do some for site15:21
wpwrakthanks ! looks nice !15:22
kristianpaullooks good15:22
Jay7new UI is even better ;)15:22
kristianpaulAnd is avalible for nanonote as well? :D15:22
Jay7it should be15:22
Jay7I've just not tried15:22
kristianpaulthanks !!15:22
wpwrakkewl. i guess it tool some inspirations from the PS3 version ? as far as i remember, they were the first (several years ago)15:23
kristianpaulYou can write to qi list15:23
Jay7it even was compiled last time iirc15:23
kristianpauli'm sure you'll get beta testers (including me)15:23
Jay7wpwrak: thesing (Thomas Kunze iirc) is original developer15:23
Jay7may be it was inspired by PS3 something :)15:24
Jay7but kexecboot was written almost from scratch (except FB code was taken from psplash)15:24
Jay7now I've changed most part of that code to do UI faster15:25
Jay7but I have at least one bug there :)15:25
Jay7kristianpaul: I should try to run it on nanonote at least before :)15:26
kristianpaulPlease :)15:26
Jay7well.. I'll try as soon as possible15:26
Jay7kristianpaul: what was your original problem about UI?15:27
Jay7some launcher for what?15:27
kristianpaulCurrently owrt is used on default nanonote image15:27
kristianpaulinittab lanuch an app called gmenu2x, originally taken from dingox15:28
kristianpaulother xbusrt based device15:28
kristianpaulbut gmenu2x well have some issues, one i'm afraid as i realized is that uses 6Mb of memory15:28
kristianpaulwich is a  LOT i think15:29
kristianpaulAs far i know one of the nanonote sofware goals is fast boot15:30
kristianpaulthat cold be achieve15:30
kristianpaul1: removing uboot and replace it by kexec15:30
kristianpaul2: replace current app launcher for one lighter and faster to run,15:31
kristianpaulthe last one will be nice if support wallpaper and icons for each apps15:31
wpwrakkristianpaul: before 1, you need some simple first-stage loader15:31
kristianpaulthats the original problem Jay715:31
kristianpaulwpwrak: true15:32
zrafakristianpaul: why do you think that 6MB is a lot?.. which applications you want to run which needs those 6MB of ram+free ram?15:32
wpwrakzrafa: 6 MB is almost 20% of your memory :)15:32
zrafawpwrak: yes, I know15:32
zrafawpwrak: I want to know which applications people want to run using the ram15:33
wpwrakOpenOffice, of course ;-)15:33
zrafaI read all the time : it is slow and it eats a lot of ram.. but nobody seems to use the ram15:33
zrafawhich something more than ls, top, moc, vim15:33
zrafawith something*15:34
wpwraki saw someone mention allocation failures. was that with emacs ?15:34
kristianpaulzrafa: because is ~21% from our 28M ram memory :/15:34
zrafakristianpaul: ??15:34
Jay7I can create some form of laucher from kexecboot15:34
Jay7it should be easy enough15:34
kristianpaulzrafa: wpwrak already told that, i dint notice it :)15:34
Jay7but I can start only from 1-5 Jan15:34
zrafakristianpaul: and what wpwrak said does not answer the question15:34
Jay7so if you need something before - do it yourself ;)15:35
zrafathe main question is which application users want to run to use the whole ram15:35
kristianpauli was afraid for that answer Jay7 ;-)15:35
Jay7we can use kexecboot as second-stage loader to run kernel from any available media15:36
kristianpaulJay7: We thanks your help when you can :)15:36
Jay7then run some tool based on it to run apps15:36
Jay7interface will be almost the same :)15:36
Jay7kristianpaul: if Qi can pay me for that work, I'll do it faster ;)15:37
kristianpaulwpwrak: (first stage) i was thiking a reduced/cuted/hacked uboot code..15:37
zrafawpwrak: one answer could be : users want to run 1000 ls at the same time, 50 tops and 100 vim15:37
Jay7first stage loader should just load kernel from NAND15:37
Jay7all other thing will be done by kexecboot15:38
kristianpaulJay7: (pay) dont know, but as if you can help us later will be nice, all we here are unpaid :)15:38
kristianpaulfrom qi15:38
Jay7kristianpaul: well, np :)15:38
kristianpauli meant15:38
Jay7it was (semi-)joke :)15:38
zrafa"I will not but that house, the extra room is small".. okey, now tell me: "who will use that extra room?".. mmehhh.. "nobody.. but I will not buy that house, the extra room is small"15:38
kristianpaulzrafa: well i dont care too much for ram, but if i can get my menu faster that now, i will be happy guy  (and others as well i think, endusers??? :))15:39
Jay7so, I'll add this channel to znc and will hang here15:39
kristianpaulWellcome then :)15:40
Jay7my working timezone is like CET :)15:40
kristianpaulJay7: i'm adding #oe too to my irssi15:40
Jay7so, feel free to contact me :)15:40
wpwrakzrafa: naw, uselessly wasting memory is always bad. the sooner or later, there will be something that fails because of it. remember "640 kB is enough" ? :)15:41
Jay7kristianpaul: you can look at boot.cfg sample and try to imagine some config for apps launcher15:42
wpwrakkristianpaul: (reduced u-boot) naw, u-boot is a mess. don't go there. i think lars already has some small loader.15:42
kristianpaulwpwrak: ok15:42
Jay7just to have it before15:42
zrafawpwrak: yes, I know, but I would not be so worry about those 6mb gmenu uses.. if nobody is going to use those 6MB yet I would try to complain with other kind of stuff :)15:43
Jay7multiple sections is needed here.. but I'll do the same for multiple kernels15:43
kristianpaulJay7: hey looks  simple setup (from what i can read)15:49
Jay7yes, it is15:49
Jay7bad thing that current implementation is only for single kernel (single boot.cfg) per partition15:50
Jay7but multiple kernels per partitions (i.e. multiple sections inside boot.cfg) is also proposed work for CELF15:50
Jay7so I'll do this anyway..15:50
kristianpaulhmm the forth oriented CPU (J1), interesting aprouch, when talking about CPU oriented to certain programing language17:02
kristianpaulnow wich program language to choose :) ?17:02
kristianpaulargg i still getting randon reading, i think i missed a verilog wire17:03
wpwraki'd pick a cpu optimized for C ;-)17:07
kristianpaulFPGA can beat current CPU in that way i think, plus the reconfiguration stuff17:09
wpwrakyou, switch from C to verilog :)17:09
kristianpaulwell is not the same aprocuh (procedural vs cocurrent)17:10
wpwrakactually, once synthesis is opened. i wonder how long it would take for someone to come up with a C compiler that programs the FPGA17:10
kristianpaulI was pointing, ie, from you pick, C is the assemlby language17:11
kristianpaulbut is hard, you dint program a FPGA at all17:11
kristianpauljust load a hardware model17:11
kristianpauli still been hardware some how, you still needing the sofware logic glue17:12
kristianpaulmostly because the *soft* part17:12
wpwrakthink of compilers for vector processors. they also need to split into vectorized and scalar part.17:13
kristianpaulNice times are coming  :)17:13
wpwraksome cpus also have some very low-level run-time core. e.g., look at the propeller.17:13
wpwrak(or just any microcode)17:13
kristianpaul(vector processors) yummy :)17:15
kristianpaulparallax propeller, looks interesting, i dint knew it of it17:18
wpwrakit's an interesting concept. probably not really worth the trouble for real-life applications, though.17:43
Jay7kristianpaul: I remember some java-bytecode-compatible processors17:55
Jay7and I'm sure we can see .net-compatible in the future17:55
wpwrakare the java cpus any commercial success ?18:00
kristianpaulis java a sucess?18:02
kristianpaulno, are not  (cpu)18:02
kristianpaulcommercially talking is hard to define sucess from techincal point of view18:03
Jay7wpwrak: seems no :)18:04
wpwrakJay7: then i wouldn't hold my breath when it comes to .net cpus ;-)18:08
Jay7well.. there will be microsoft behind ;)18:08
wpwrakkristianpaul: i suspect that, by the time the things are mature enough, regular cpus will have caught up with them :)18:08
wpwrakJay7: well, M$ have a long history of "strategic" projects that are epic commercial disasters ;-)18:09
wpwrake.g., Windows HPC anyone ? ;-)18:09
Jay7I hope this will one of that :)18:09
wpwraki don't see how it could fail to be ;-)18:10
Madridmanhi folks18:32
Madridmana question, what's the meaning of "qi" on the room's name?18:33
wpwrakMadridman: it's about open source/copyleft hardware: http://en.qi-hardware.com/wiki/Main_Page18:49
wpwrakwolfgang: small detail - if you go to the wiki, there's no direct way to find out where the "qi" comes from :-)18:50
wpwrakwolfgang: fun fact: www.sharism.cc -> the latest "community news" are from april 29. i must have hallucinated all the more recent ones ;-)18:52
kristianpaulMadridman: Qi is Chinese  related, look: https://secure.wikimedia.org/wikipedia/en/wiki/Qi19:13
kristianpaulhe now i read zeros, well better than randon :)19:41
wpwraksounds a bit like my misadventures with the latest atusb board :-( the beast was happy enough to begin its first programming, but then failed the checksum and stubbornly refused any further updates19:43
wpwrakvoltages look good. soldering looks good. no shorts i can see. very mysterious.19:47
wpwrakcurrent consumption is perfect19:50
kristianpaulmcu issues?..19:50
wpwrakcould be broken, yes. strange, though.19:51
wpwrakprogrammer works perfectly with a different kind of board. so it's really this critter that's having issues.19:52
wpwraknot a good day so far. the first board i made had an irreparable soldering problem (a solder bridge that didn't go away, no matter what i tried)19:54
wpwrakthen i found that the ben had a ubifs corruption that prevented it from booting. so, a little detour with a full reflash.  (reflash_ben.sh works nicely, particularly when used in combination with idbg :)19:56
wpwrakand now the damn mcu won't flash :-(19:56
kristianpauli was reading an int, so make sense now !21:03
kristianpaulgood !21:07
kristianpaulnow, time to dump to a file, lets see what option is better21:07
wpwrakwhich are the options ?21:12
kristianpaulas this need to be fast i tought:21:15
kristianpaul1: allocate some Mb of memory and dump to it, when finish , write to disk21:15
wpwrakthat's what i'd do21:16
kristianpaul2: write to disk, wich is kind slow i think, i tough use memory flash 10:8, but tahts no recommended in SIE21:16
wpwrak2: no way :) it also adds stray delays and such, which will probably make you overrun your buffer21:17
kristianpauloops :)21:17
wpwrakdo you have an indication for buffer overruns ?21:17
wpwrakmaybe add one. otherwise, you'll have no end of trouble with strangely incorrect data21:18
kristianpaulok lets see how to do it :),21:18
kristianpaulbtw are you using some indication for buffer overruns in atusd code? :D21:19
wpwrakworse yet, even if your data is perfectly good but something else is wrong, you'll always suspect the data :)21:19
wpwrakright now, a new frame would just overwrite the old one21:20
kristianpaulah no dump21:20
kristianpaulyou may consider one ! ;-)21:21
wpwrakthere's the possibility of disabling the receiver until the frame has been received, though21:21
wpwrakdump ?21:21
kristianpauldata dump?21:21
kristianpaul( disabling the receiver) is that a question for me?21:22
wpwrakhmm ? when the chip has received the frame, it generates an interrupt. then i transfer the frame. if i'm too slow and a new frame is being received, it may overwrite the first one21:22
kristianpaulah ok, nv last question :)21:22
wpwrakno, talking about atusd/atusb21:22
kristianpaulyes yes21:22
wpwraki'll probably let the microcontroller do the first response to interrupts on atusb. otherwise, it'll just sit around lazily and get fat ;-)21:23
wpwrakon atusd, there's no mcu, do it's the ben who has to react21:24
kristianpaulafter all i can have a decent 8bit IO in SIE plus 13bits of addresing (multiplexing), just deleting the sram and adding some case staments21:30
kristianpaulsimple :)21:33
wpwrakwhatever you say ;-)21:38
kristianpaulbetween SIE and FPGA i mean—*21:40
wpwrakthat's the part i understood. the rest sounds mysterious ;-)21:40
wpwrakkristianpaul: how do you like my latest soldering accessory ? http://downloads.qi-hardware.com/people/werner/tmp/hotplate.jpg21:57
wolfspraulone day I report all of this to homeland security21:59
wpwrakfor completeness, here's today's opponent: http://downloads.qi-hardware.com/people/werner/tmp/nemesis.jpg22:03
kristianpaulGreat !22:05
kristianpauladam will want one :-)22:05
kristianpaulYou have nice workspace bte, (white) all here is brown :)22:06
kristianpaulwow there are really tiny resistors in nemesis22:07
kristianpauls/resistors/pasive components22:07
kristianpauldo you glue it some way before reflow?22:08
kristianpauloh i can see a Ben naked board on the right botton corner :)22:09
wpwrak(workspace) all DIY :)22:09
wpwrak(reflow) soldered them manually, no reflow22:10
kristianpaulso how helped the hotplate?22:10
wpwrakif i did reflow, i wouldn't need to glue them. they don't move around too easily. but i think i need to make a stencil for proper reflow. just depositing solder paste with a syringe doesn't produce nice results.22:11
kristianpaulah ok22:11
kristianpaulis nemesis the bad-mcu guy?22:12
wpwrakthe hot plate heats the board to about 100 C. so when for melting the solder, i only need to overcome a temperature difference of about 70 K, not 150 K. for simple things, this doesn't make a big difference. but in this case, i have a lot of ground and a gazillion vias connecting both sides. the ground planes suck away the heat. so there are several trouble spots where a 270 C iron is difficult or insufficient.22:13
kristianpaulwhy nemesis have that green-like wire, is not enought the pcb atenna?22:14
wpwrakwith the hot plate, these problems disppear :)22:14
wpwrak(wire) this is for the scope, to monitor the voltage :)22:14
wpwrak(nemesis) yup, that's the problem board22:15
wpwraki'm now making two new boards. i hope they'll end up being better.22:15
wpwrakwolfspraul: i should actually integrate the iron into the table :) and while i'm at it, add also an area with under the table lighting, for aligning transfer sheets and such. ah well, so many ideas, so little time :)22:17
kristianpaul(270) wow which component needs that ?22:17
kristianpaul(integrate the iron into the table) good idea22:18
wpwrak270 C is my regular setting with lead-free solder. not sure how hot the thing really gets. the tip is quite worn, so it's probably colder.22:18
wolfspraulwhy do you use lead-free solder?22:21
kristianpaulless hot to apply no?22:22
wpwrakwolfspraul: to give customs no excuse ;-)22:23
wpwrakkristianpaul: no, it's actually very inconvenient to work with. needs more heat and generally doesn't handle as well as leaded solder.22:23
wolfspraulmanual rework is always exempt from RoHS22:25
wolfsprauljust fyi22:25
wpwrakkristianpaul: mechanical properties after soldering probably aren't so great either. e.g., i think it's more brittle. so things may break off under stress.22:25
wolfspraulthere are many exceptions, but a simple claim of 'manual rework' is already enough22:25
wpwrakwolfspraul: also without the "re" in "rework" ?22:25
wolfspraulI'm not aware of any country that would really go after every gram of lead. it's quite political, and the important idea is 'self-enforcement' and to start with the big amounts somewhere, the stuff that is handled by the machines...22:26
wpwrakinteresting exemption. they're nicer than i would have thought.22:26
wolfspraulwell I'm also not aware of anybody who would use that excuse and keep their machines running with leaded solder while claiming the rework thing.22:27
wolfspraulmaybe people are more honest than thought, sometimes22:27
wolfspraulbut for that reason, alsmost all rework people use leaded solder still, because apart from you I have never heard anyone saying lead-free is easier to work with22:27
wpwrakwell, customs have to process the package either way, so if they feel bored, they may cause trouble. also, they don't know if this is really a unique piece or whether there's a million more to come, each shipped individually (like bens :)22:28
wolfspraul(ah sorry, I misread you)22:28
wolfspraulyou also say it "doesn't handle as well as leaded solder"22:28
wpwrakyeah. it's a pain.22:28
wolfspraulno I doubt there is any such check ever22:28
wolfspraulyou think too complicated again :-)22:28
wolfspraulRoHS is a great success, it was very effective. but that was _because_ the system was setup as a 'self enforcement' system, and with generous exceptions.22:29
wpwrakwell, based on my experience with argentine customs, i expect the very worst22:29
wpwrak... and sometimes still get nasty surprises22:29
kristianpauloh really?22:29
wolfspraulI highly doubt they will go after lead, but OK. every rework station in all of Asia uses leaded solder.22:30
wpwrakkristianpaul: well, i didn't have issues with RoHS so far. but they had me twice with "seguridad electrica"22:30
wolfspraulthat means there is a good chance, < 10 %, that somewhere on every piece of electronic that are a few spots of leaded solder22:30
wpwrakkristianpaul: and they also hinted once that "PCB" could be a banned chemical22:30
kristianpaulahh¡¡ (PCB)22:31
kristianpaulthats the top22:31
wolfspraulwpwrak: so bottom line, in my experience you can safely use leaded solder for your work, and still be rohs compliant.22:32
wolfspraulthere are also exceptions for machine cleaning and such22:32
wpwrakkristianpaul: so what would happen is that they'd hold it at customs. then i'd have to go to the airport, spend about half a day before i can hear what the issue is. then i could of course explain that pcb is the board and not the banned chemical. but they can of course simply reject this explanation.22:32
wolfspraulso when a factory switches between leaded and lead-free runs, they don't have to go crazy over the cleaning thing22:32
kristianpaulwpwrak: (reject this explanation.) reasons? technically talking22:33
wpwrakkristianpaul: then i could either take a certainly complex and expensive bureaucratic appeal process, or find out how much of a financial incentive they would require for changing their mind22:33
wpwrakkristianpaul: no technical reasons. it's a plausible explanation. of course they know there's nothing wrong. but this way, they can improve their personal income.22:33
kristianpauli hope i dont have to pass for that in colombia22:34
wolfspraulin my experience colombian customs is good22:34
kristianpaul(they can improve their personal income) Wellcome to South America :) !!22:34
wpwrakwolfspraul: (cleaning) nice. all this sounds amazingly reasonable. what happened ? did they exclude the politicians when they made those regulations ? :)22:35
wpwrakkristianpaul: yeah ...22:35
wpwrakkristianpaul: argentine customs are probably among the worst22:36
wpwrakwell, i've heard good things said about russians/ex-USSR customs ...22:36
wolfspraulwpwrak: I think they understood they are dealing with mass production.22:37
wolfspraulso even if you pull one out, of 10,000, and you find something, then what?22:37
wolfspraulso a top-down, fines, punishments, regulations to the last detail approach simply would not have worked22:38
wpwrakstill, amazingly reasonable :)22:38
wolfspraulit needs to start on the other side, where the big machines are being operated22:38
wolfspraulfocus on the x tons of solder first, not on reworks or all sorts of other small stuff happening in a real factory22:38
wolfspraulof course that removes the ability to deduct too much from an individual piece in front of you22:39
wpwrakoh, a system that has lots of traps is good for political purposes. e.g., you can enforce it mainly against unwanted foreign competitors22:39
wolfspraulbut that was not the point, the point was to move mass-production away from leaded solder22:39
wpwrakcollateral advantages :)22:39
wolfspraulso even if 1 piece in front of you is full of lead, that doesn't mean much in terms of RoHS, at least at the beginning22:39
wolfspraulso anyway22:40
wolfspraulyou can relax about that in your home station, although the heating plate of course is very cool22:40
wolfspraulit also helps with large chips, or when you have to work with hot air22:40
wolfspraulit helps to avoid cracks in the vias inside pcbs22:41
wpwraki probably need it anyway even with lead. the problem there are the vast ground planes.22:41
wolfspraulsure, it's good22:41
wolfspraulI was just surprised when you said you use lead-free solder for reworks.22:41
wolfspraulthat sounds like additional trouble/risk for no good22:41
wolfspraulnobody is doing that, afaik22:41
wpwrakall my vias are hand-made. so they're a little special :) one advantage of the process is that they won't crack. (it's a solid wire)22:42
Action: wpwrak welcomes back the 60/40 eutectic leaded solder :)22:43
wolfspraulI'll ask around a bit more...22:43
wolfspraulmaybe some factories are proud to move to a 100% lead-free process? never heard of it though22:43
kristianpaulah wikireader is awsome my mother really liked it :) (no trouble with classic keyboard)22:43
wpwrakwolfspraul: of course, the dark side of my environmental footprint would then be all those heavily leaded counterweights ;-)22:44
Action: wpwrak wonders if tuxbrain did anything with them22:44
kristianpaulI need one22:45
kristianpaulmy ben fall all the time..22:45
wpwraki'm actually surprised that nobody seems to complain about this22:48
kristianpaulhmm may be they are taking it like a faeture? :)22:49
kristianpaulthat could be an indicador of dailly use of it, on the desk22:51
kristianpaulwich  as pocket computer is hard to realize22:51
wpwrakhmm. if you can make people believe this is a feature, then i think we need a nobel price for marketing, and you're the first to get it :)22:57
wolfspraulwpwrak: just googled a bit and what I said about manual reworks and RoHS may be wrong.23:15
wolfspraula good reminder that I need to refresh my RoHS knowledge again, it seems to be constantly evolving (in general tightening)23:16
Action: wpwrak puts the 60/40 leaded solder back into the drawer :-(23:17
wolfspraulha :-) relax...23:17
wolfspraulthere are still many big fat exceptions23:17
wolfspraulit's a moving target23:18
wolfspraulof course the point is not to ever get more creative to define oneself into the exceptions23:18
wpwrakgood for customs. so they can spring surprises23:18
wolfspraulmost people I have seen that deal with rohs take it very seriously and professionally, it's quite nice actually23:18
wolfspraulon the enforcement side, there is little if any aggression, on the production side, there is little if any cutting corners23:19
wolfspraulthe telecommunications industry seems to have been quite active, see exemption 7b for lead23:19
wolfspraul"Lead in solders for servers, storage and storage array systems, network infrastructure equipment for switching, signalling, transmission, and network management for telecommunications"23:21
wpwrakyeah, pretty broad ;-)23:21
wolfspraulso who knows. I still think your manual reworks are totally safe, even with leaded solder.23:21
wpwrakamazingly, no exceptions for automotive23:21
wolfspraulnot reworks23:21
wolfspraulmanual 'works'23:21
wolfspraulbut I need to take this more serious again too, will dig next time I visit a manufacturer.23:22
wolfspraulmaybe they are already further than I realize23:22
wolfspraulon that website (seems UK related), it says about enforcement "We will help those that are aiming to comply and pursue vigorously those that intend to flout compliance."23:23
wpwrakrepair of "new" (> july 2006) devices is also under the directive23:28
wpwrakthere's an exception for demonstration purposes, but the interpretation may be quite narrow, perhaps only for trade shows and such, not even general in-house evaluation23:30
wpwrakwell, it's probably okay: "Prototypes for experimentation may be used providing they are not made available on the EU market."23:35
wpwrakthe next paragraph is confusing, though23:36
wpwrak"Ultimately only the courts can provide a definitive answer on the meaning of the Regulations," yeah :)23:38
wpwrakwe need wikilaws ! then anything with "weasel words" gets deleted ;-)23:38
wolfspraulfrom what I hear, rohs is not about customs trying to hunt down individual products23:39
wolfspraulthat would be totally impossible given the complexity of it23:39
wolfspraulit's about authorities working with the industry23:40
wolfspraulalso I think rohs will be tightened more, it's such a big success23:40
wolfspraulso more substances, processes, what not can be attached to it :-)23:40
wolfspraulwpwrak: what do you think about introducing datasheet pdf urls into boom?23:42
wpwraki think the feature per se is desirable. semanstics aren't trivial, though.23:45
wpwrakfor best effect, the data sheets would come from multiple sources, including distributors.23:45
wpwraksi the digi-key database download would also try to access data sheets, and make them available for the respective digi-key part numbers23:46
qi-bot[commit] Wolfgang Spraul: removed two generated png files, leaving svg originals http://qi-hw.com/p/xue/fb852f723:46
wolfspraulgot it, and sounds good. on the todo list then...23:46
wpwrakthere are several access paths. e.g., by component reference, a) for the "current" data, b) for historical data.23:48
wpwrak(b) could be outside boom, e.g., just archive a snapshot when doing something with the data)23:48
wpwrakthen by item in an order23:49
wpwrakand finally in a yet-to-be-written component browser that would interact with kicad when adding a component and its characteristics in the schematics (or editing, etc.)23:50
wpwrakthat's the one that would make carlos happy :)23:54
--- Sat Dec 18 201000:00

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