<wolfspraul>
I renamed the string of sdram.sch, the string from xue-rnc.sch
<wolfspraul>
I changed it from "DDR Banks" to "sdram"
<wolfspraul>
the history now shows a deletion and creation of a new page
<wpwrak>
yup. it currently tracks one both not both names. that's what i meant when i said that it'll break in one way or another.
<wolfspraul>
hmm. why does it care about the string name at all? since it follows .sch file renames, that should be enough, no?
<wolfspraul>
there's a -f switch to do just that :-)
<wpwrak>
if i recall right, xue had some file renaming in the past, that's why the heuristics are like they are
<wpwrak>
let's see what happens with -f :) if my memory is correct, some other problem should then pop up :)
<wolfspraul>
I think the -f documentation is wrong, it says 'identify sheets by their file name, not the sheet name'
<wolfspraul>
but reading the sources I think it's the other way round
<wolfspraul>
ah wait, no, I'm confused
<wpwrak>
ah yes. it's what goes into the title
<wpwrak>
i added it for one of the student projects, which had some ugly semi-binary stuff in the name
<wpwrak>
(some temporary name kicad assigns by default, i guess)
<wolfspraul>
hmm
<wolfspraul>
I added -f but I suspect we now get some other breakage
<wolfspraul>
because when the project was still called CHIE they probably did not use proper git mv to rename the files
<wolfspraul>
but instead they deleted and readded them under new names.
<wpwrak>
i think this is the same for git
<wpwrak>
its rename detection is based on looking for new files that are similar to deleted files :)
<wolfspraul>
git will record a rename/move
<wolfspraul>
maybe it has that in addition somewhere, but I'm pretty sure it has genuine support for move/rename
<wpwrak>
i think it doesn't record it but tries hard to report a delete/add pair as a rename :)
<wolfspraul>
strange, I am still confused in schhist2web
<wolfspraul>
-f says "identify sheets by their file name"
<wolfspraul>
default is false
<wolfspraul>
when -f is set it becomes true
<wpwrak>
did you run with -f ?
<wolfspraul>
in line 265, it says elif $use_sch_name && ... subschname2file ...
<wolfspraul>
so if use_sch_name is true, it calls subschname2file
<wolfspraul>
subschname2file will look for F0 (the name string), so it's the one that goes from a string name to a file name
<wolfspraul>
I'm misreading something somewhere
<wolfspraul>
before I ran without -f. -f was only used for SIE_CEIMTUN
<wolfspraul>
now I'm rerunning with -f just to see how it looks
<wolfspraul>
meanwhile I'm reading the script trying to understand what it is doing
<wolfspraul>
maybe we only need to introduce an else case in 276
<wolfspraul>
line 276
<wpwrak>
hmm .. lemme think ...
<wpwrak>
let's see what -f yields. i'm not entirely sure anymore about what failure mode to expect
<wpwrak>
it may pass, but use "ugly" names
<wpwrak>
but maybe there's something else i'm not thinking of at the moment
<wolfspraul>
am I misunderstanding -f ?
<wolfspraul>
the documentation and what I read in the script do not match
<wolfspraul>
maybe I misunderstand the syntax somewhere?
<wolfspraul>
&& continues when the left side is true
<wolfspraul>
if it is false, it aborts. right?
<wolfspraul>
so if -f is set, it will go to the subsch2name script, which uses the string name to lookup the file name
<wolfspraul>
but the -f docu says 'identifies by their file name'?
<wolfspraul>
introducing an else case in 276 is probably a good idea, no reason to give up there
<wolfspraul>
but let's see...
<wpwrak>
it takes the "string name" to get the file name. then it continues working with the file name
<wpwrak>
before it used the "string name" to identify the sheet
<wolfspraul>
yes
<wolfspraul>
but the -f documentation says the opposite
<wolfspraul>
"identify sheets by their file name, not the sheet name"
<wolfspraul>
don't you think it's "identify sheets by their sheet name, not the file name"?
<wolfspraul>
actually it looks like "identify sheets by their sheet name, if file name not found"
<wolfspraul>
because this only kicks in when the file name is not found, if I see correctly
<wolfspraul>
if this is true, there should be no reason to not make -f the default
<wpwrak>
the first check is for the root sheet
<wpwrak>
the root sheet is special
<wpwrak>
for the others, we come from the .ps files. kicad uses the "string name" for the .ps files, not the .sch file's name
<wpwrak>
(not my idea ;-)
<wolfspraul>
ah
<wolfspraul>
I will make the two the same, once we can avoid schhist breaking down...
<wolfspraul>
so the file name is the sheet name, at this place
<wpwrak>
sheet name = foo.sch or the thingy inside foo.sch ?
<wpwrak>
let's say we have foo.sch which calls itself (inside) "bar"
<wpwrak>
how we get some project-bar.ps ("project" is the basename of the top-level sheet, project.sch)
<wpwrak>
then we check if foo.sch happens to be project.sch
<wpwrak>
err, sorry
<wpwrak>
we check if "bar" happens to be "project"
<wpwrak>
if it is, then we're looking at the top-level sheet. we track top-level sheets elsewhere, so we already know we're good (else we wouldn't have made it this far)
<wpwrak>
now, if -f isn't set, we just use "bar"
<wpwrak>
otherwise, we dig a little deeper - and try to obtain the real file name, in this case foo.sch
<wpwrak>
now, through all this, we're at some point in the project's history
<wpwrak>
with the name foo.sch, we have something git can relate to. so we ask it for the latest name of this file, considering all the renames that may have happened since. this is what gitwhoareyounow does.
<wpwrak>
finally, we use "foo" or any "foo-today" git may have found as our name of the sheet
<wolfspraul>
phew
<wolfspraul>
good thing this is logged for eternity
<wolfspraul>
if I need to go back in there I will paste it as comments, still waiting for -f results
<wolfspraul>
wpwrak: wow nice!
<wolfspraul>
congratulations!
<wolfspraul>
-f is magic :-)
<wpwrak>
whee ! :)
<wpwrak>
now that the names are "nice", it looks quite decent
<wolfspraul>
ok I will continue to fix the strings, hopefully that won't break schhist again
<wolfspraul>
maybe -f should be the default? well, for now we are good...
<wpwrak>
maybe do it all in one commit, to avoid this long string of changes in the history
<wpwrak>
if the file names are useful (which they are now), -f is good
<wpwrak>
that's semi-generated. you get it when you associate the components with cvpcb
<wolfspraul>
ok understood
<wolfspraul>
but .net can go?
<wpwrak>
.net can go
<wpwrak>
.cmp may or may not be redundant. but in any case, if it's not there, something (pcbnew, i think) will complain all the time
<wolfspraul>
and the 4 cache files too, unless they accidentally moved original data in there (which I saw before, hope it's not the case here)
<wolfspraul>
sure we can cleanup slowly
<wolfspraul>
hmm
<wpwrak>
(".cmp may or may not") what i mean is that it can be 100% redundant (if you set all the footprints in the schematics), but it can also contain original date (if the schematics don't have all the details or if you manually override them). so it's not generally safe to delete .cmp files.
<wolfspraul>
the .sch files all list LIBS:xue-rnc-cache
<wolfspraul>
should I remove that line from the .sch files? (together with the -cache file)
<wpwrak>
i think this LIBS line is just some output that isn't of much use anyway
<wpwrak>
i wouldn't bother :)
<wolfspraul>
yeah, a LIBS:-cache entry is in mmone-jtag-serial cable too
<wolfspraul>
ok so I just delete the cach files from the repository first
<wpwrak>
it's just spam :)
<wolfspraul>
that should be safe, and if not we need to rescue the original data from them
<kyak>
might be interesting for you and easy to port, too
<xiangfu>
kyak: yes. thanks. I already start a little. not commit yet.
<kyak>
how does chinese input work with latin keyboard?
<kyak>
is there some corrsepondence between chinese symbols and latin letters?
<xiangfu>
kyak: sorry. what's the meaning of "corrsepondence ... "
<kyak>
sorry, i mean that latin characters are somehow mapped into chinese
<kyak>
how do you remember that?
<xiangfu>
kyak: I know to kind of chinese input. one is  base on "pronounce". another is base on the "the write of chinese character"
<xiangfu>
I know two.  [s/to/two]
<kyak>
i can understand the "pronounce". How does the "write of chinese character" work?
<xiangfu>
kyak: like `}(ni hao) split to [»]  [] and [s] [P], chinese is  assemble by base small module. so easy latin mapped to several "base modules"
<kyak>
it looks like you are not choosing an easy way :)
<xiangfu>
kyak: when I start to learn type, the easy way is not easy in fact, but now. those input method can remember what you type. so it's a little intelligent.
<xiangfu>
kyak: I think for now no one will try to study the WUBI.
<kyak>
there are online web pages to try wubi :) pretty funny
<xiangfu>
kyak: you know "Chinese" ??
<xiangfu>
kyak: I mean you understand the "Chinese" ? then maybe we can talk in Chinese :)
<wpwrak>
oddly, there's no chloride smell. i just reused some etchant from yesterday, added the peroxide to make it reactive again (HCl+H2O2 gets stale very quickly. there's supposedly a more stable state further down that process, but it doesn't sound as much fun as the initial mix)
<lekernel>
doesn't it get green?
<wpwrak>
(ionocraft) nice ! didn't know that ion propulsion was practical at the bottom of earth's gravity well
<wpwrak>
yup, it gets green. that
<wpwrak>
's the copper
<lekernel>
mh. mine was green right away because of the dissolved chlorine gas
<lekernel>
(or, at least, I guess so)
<lekernel>
wpwrak: not sure how practical that is; the device cannot lift its own power supply
<wpwrak>
that would be quite a lot of gas
<lekernel>
it's pretty inefficient too, it was using like 20W power to fly a few grams
<wpwrak>
i wouldn't have expected it to be able to even lift itself :)
<wpwrak>
well, add a H2O2-based first stage plus a solar panel and you're ready to conquer the stars ;-)
<wpwrak>
hmm, about 3 minutes total etching time. still not bad at all.
<kristianpaul>
el nanonote actualmente cuenta con sofware de openart y jlime(open emdebbed) alli puedes tener mas info de que esofware esta disponible con detalle
<seplveda>
claro, por programas matematicos wolfram
<kristianpaul>
zrafa: thos can be re-used with the last oe (muffiman) ?
<zrafa>
kristianpaul: nope. I think that I already said which is the difference of jlime at qi.. and jlime :)
<kristianpaul>
yes yes
<kristianpaul>
(bad memory)
<zrafa>
patented
<kristianpaul>
oh yes thats bad
<zrafa>
jlime uses OE and :)
<zrafa>
kristianpaul: hey.. how is the life ?
<kristianpaul>
zrafa: good :)
<kristianpaul>
you?
<zrafa>
great
<zrafa>
very well as well
<kristianpaul>
zrafa: this jlime you pointed, are you plannin to document how was built?
<kristianpaul>
I saw same docs for free of patent owrt used in nanonote as well :)
<wpwrak>
(jlime docs) great !!
<zrafa>
kristianpaul: how was built: nah... that is old I would say.. and I think that users do not want to build the same stuff which is already built :P.. But let me give you a chance.. I know you like how qi works with development.. and you have asked me several times about how to built the same muffinman version already built. So If you really needs that I can write.. WARNING: I will write the current OE docs explaining which exact commit I used from OE git a
<zrafa>
how I added the extra applications
<kristianpaul>
zrafa: well i know how build muffiman, you help on that, i just was talking about documentation in qi wiki :)
<zrafa>
kristianpaul: ah.. did I? :P.. I am old and forget
<kristianpaul>
i hope OE have good layers ;)
<kristianpaul>
zrafa: yes you did
<zrafa>
kristianpaul: well, I do not have a lot of free time to write that :(
<zrafa>
sorry
<kristianpaul>
np
<kristianpaul>
but
<zrafa>
I would prefer to write the stuff users still needs
<kristianpaul>
i'm afraik as that is missing the image will get old as time pass..
<kristianpaul>
sure
<kristianpaul>
i understand and end user is very important :)
<zrafa>
kristianpaul: what is old? :)
<kristianpaul>
just ignore me :p
<zrafa>
kristianpaul: I do not think that it will get old
<kristianpaul>
zrafa: why?
<zrafa>
kristianpaul: two things:
<wpwrak>
zrafa: if there's a specific timestamp, revision, or such, that identifies the upstream revision, that may be useful to mention. (but i doubt there a single identifier, or is there ?)
<kristianpaul>
wpwrak: is some webpage i cant mention because i dont have a layer ;-)
<kristianpaul>
s/is/it is
<wpwrak>
layer ?
<kristianpaul>
some text files, very usefull :)
<kristianpaul>
s/layer/lawyer
<wpwrak>
ah :_
<kristianpaul>
my fauly english sorry :)
<kristianpaul>
faulty*
<zrafa>
kristianpaul: 1- what does old mean?.. if you mean "packages" will get old and newer versions will come then it will get old completely. You will need to build a new repository completely. But I think that a sane and good repositoy like jlime has would be useful for a couple of years.. ANd 2-
<zrafa>
kristianpaul: 2- This jlime version for qi also brings a bootstrap version,
<zrafa>
which,
<kristianpaul>
oh bootstrap sounds good
<zrafa>
you can use for building your own version from scratch (if you mean about old like you would like to see another GUI or way)
<zrafa>
you just take bootstrap and install on that the packages you want to have the GUI you want, etc
<wpwrak>
zrafa: (couple of years) hmm, probably not. we have to make sure things are easy to keep up to date. basically automate what you did manually. qpkg tries to go in that direction - i just have to finish it ;-)
<kristianpaul>
wpwrak: (qpkg) great !
<kristianpaul>
i really want to move OE based free of patent distro fr my nanonote, but i'm afraid automate building for long term
<zrafa>
wpwrak: I think that we are talking about many things.. What is exactly "getting old"?.. all gets old..
<kristianpaul>
zrafa: okay lets replace get old to follow upstream :)
<kristianpaul>
s/to/with
<zrafa>
kristianpaul: well, I am not going to work with OE and free of patents stuff.. no time for that. In fact, I already said that I am not the best fan of OE.. so if I do something.. I will try to have a repository of packages.. just that. I will not use OE to build rootfs, or similar kind of weird things. I like OE for the repo it builds
<kristianpaul>
zrafa: (best fan of OE) ohh wow i you're, i missed that backlog :)
<zrafa>
kristianpaul: follow upstream is really hard. Jlime is not working with "free of patents" versions
<kristianpaul>
s/i/i tought*
<zrafa>
kristianpaul: nah.. I worked with OE because I like how easy you can build a nice repository for embedded devices. No more
<zrafa>
kristianpaul: and I used it because JLime uses it.. if jlime devs likes Conectiva Linux I will use (still if it is dead) :D
<zrafa>
wpwrak: I would like to have the few tools around jlime really useful for all (like qpkg, or the theora player, or the wikireader, etc).. But that is different with maintain documentation for development/OE built repositories/rootfs OE building documentation up to date
<wpwrak>
zrafa: (get old) i mean that things need updating from time to time. so it's best to have a process that makes this painless.
<kristianpaul>
zrafa: wikireader :) sure go ahead !
<wpwrak>
zrafa: my idea is that you could just flag problem packages as undesirable and then qpkg could tell you which packages can no longer be installed. or it could generate a list of packages that can, depending on what you're after.
<zrafa>
wpwrak: yes, you are right.. BUt we need to convice more people to help :).. Currently JLime has around 7, 8 devs.. for 4 or 5 platforms/devices.. Just me was trying to have something to use on qi/resellers.. SO when I work on that, jlime lose onde dev, because other devs are not trying to work with this kind of stuff
<wpwrak>
zrafa: or maybe add some dependencies into upstream (OE or "full jlime", whichever works best) and tell qpkg to knock out some dependency and generate the resulting list of available packages.
<zrafa>
wpwrak: wait...
<wpwrak>
zrafa: you need to do a bit more propaganda :-) the problem is not unique to qi-hw.
<zrafa>
wpwrak: how would qpkg relate with OE full jlime?.. It was clear that versions for qi/resellers should not include links to complete repositories, still if tha package manager will not install problematic packages.
<zrafa>
tha=the
<wpwrak>
zrafa: my idea is that you basically take some upstream package set, add your extra wishes, and let qpkg produce a subset of the upstream
<zrafa>
ah.. nice.. I see now
<wpwrak>
zrafa: the subset would be self-contained, like the one you've produced manually so far
<wpwrak>
zrafa: the form in which those extra wishes would appear is open to debate. once the infrastructure for resolving dependencies is there, it's relatively easy to do other things with it.
<wpwrak>
zrafa: qpkg can already solve many installation dependencies. what it doesn't do yet is handle cases where you need some complicated backtracking and it doesn't sort its output by dependency
<wpwrak>
(both need the same underlying algorithm - alas not an easy one. will take me a few days to figure out)
<wpwrak>
zrafa: it basically marks packages a being about to be installed, but then finds out that it can't do this, and when trying some alternative, it doesn't take into account that the packages it considered for installation weren't
<wpwrak>
zrafa: not sure if this is constellation is very likely in real life, but it's something that can be expressed, so i better handle it
<zrafa>
wpwrak: I need a while to understann all you wrote :)
<wpwrak>
it's a little bit complicated :)
<wpwrak>
lunch time :)
<zrafa>
haha :D
<kristianpaul>
zrafa: what do you use your nanonote for?
<kristianpaul>
Btw i wonder if decemenber in argentina is as noissy as in here :S
<zrafa>
kristianpaul: no much use these days.. I like it to play music files, wikireader when traveling.. and games
<wpwrak>
kristianpaul: you mean december in general ? :)
<kristianpaul>
wpwrak: yes
<wpwrak>
naw, pretty normal. except for the 24th and a few small test runs before and after
<kristianpaul>
damn my neighboors noise is goint to kill me
<kristianpaul>
he "small test runs"
<zrafa>
kristianpaul: your neighboors do many parites?
<zrafa>
parties
<wpwrak>
(we have fireworks on the 24th. military grade :)
<kristianpaul>
noise
<kristianpaul>
parties are 24 and 31 i guess
<zrafa>
kristianpaul: which kind of noise?
<wpwrak>
in fact, i'll get some armament tomorrow :)
<kristianpaul>
hihg volume music (bad electronic, and musica de cantina, repetida ;))
<kristianpaul>
reggueton :S also
<kristianpaul>
argg
<wpwrak>
so strike back. which movie was that .. apocalypse now ? where they played wagner from the helicopters
<kristianpaul>
no idea, i'm not good with holliwood-like movies :-)
<zrafa>
kristianpaul: wpwrak : when I was working few days in Caracas.. my company rented an apartment.. well all the nights the neighbors houses were doing parties listening reggeton music and laugh all the time the whole night
<zrafa>
well all=well, all
<zrafa>
kristianpaul: so that sounds similar like you are suffering there :)
<kristianpaul>
caracas .ve ?
<zrafa>
yes
<kristianpaul>
oh yes they are party makers before born :)
<zrafa>
wpwrak: do you remember how we installed toolchain and set the proper links for reporitories? :) .. I need the doc of that for the last link here : http://en.qi-hardware.com/wiki/Jlime
<zrafa>
"Jlime Toolchain Documentation"
<wpwrak>
i really ought to visit .ve some day ;-)
<kristianpaul>
ah travel is so relaxing :)
<kristianpaul>
brb i'll try Fedora and milkymist devel env on it
<wpwrak>
zrafa: (repo link) hmm, i think you had some file for me to download
<wpwrak>
zrafa: i think my notes are on the dead disk. but the discussion ought to be in the irc log
<wolfspraul>
ah, before that I will remove the AVR, FAN and ft2232
<wolfspraul>
andres-calderon: is it ok that I delete docs/xue-bom.ods ?
<wolfspraul>
then I will slowly start with boomification, first with an empty bom/ directory and a little Makefile structure, filling in more later
<wolfspraul>
in sim/verilog/micron, we have 2 .zip files committed
<wolfspraul>
maybe as a first step I will unpack them and commit what's inside instead
<andres-calderon>
Tomorrow I will fix the ERC errors and warnings.
<wpwrak>
yes ! 2nd try works much better :)
<wolfspraul>
andres-calderon: do you see anything wrong with what I've posted?
<andres-calderon>
wpwrak: I will be more conservative.
<wolfspraul>
I am currently working in a mode of "when I don't hear anything, that means silent approval"
<andres-calderon>
wolfspraul: not yet.
<wpwrak>
wolfspraul: (library -> components) it would be consistent with my use, but i don't have a strong opinion on it. kicad terminology may change a little anyway, as these things get rearranged.
<wolfspraul>
yes I know, but let's go for consistency at least right now
<wolfspraul>
(good action on KiCad btw, nice to see the progress there...)
<wolfspraul>
andres-calderon: what about those 2 zip files?
<wolfspraul>
I will first unpack them, peel the onion as werner says
<wpwrak>
andres-calderon: (2nd try) ah, that was about my little ben-wpan board ;-) yesterday, i had a very frustrating day with a board that doesn't work for some obscure reason. today i made a new one and it seems to work.
<wolfspraul>
andres-calderon: also, in the root xue/ directory there is now a Makefile that will make some things easier
<wolfspraul>
you can do 'make sch' or 'make brd'
<wolfspraul>
I will add more
<andres-calderon>
Are micron files (IBIS files), used in to automatic kicad component generation
<wolfspraul>
I would strongly argue for removing them, not on technical facts but looking at our manpower, risk that it won't work or backfire, manufacturing complexity, etc.
<wpwrak>
(this uses the tools i developed in the last days, to let me automate more of the pcb making)
<andres-calderon>
wpwrak, wolfspraul:Â Â I've been converted. I think that simplicity is the name of the game.
<wpwrak>
hehe ;-)
<wolfspraul>
good
<wpwrak>
andres-calderon: maybe keep the series/sense resistor but make them 0R. that way, you can remove them and insert a current meter/probe in the lab.
<qi-bot>
[commit] Wolfgang Spraul: removed xue-bom.ods which only had a few web links for memory chips http://qi-hw.com/p/xue/c0089b7
<andres-calderon>
I will discard the  onboard measurement of energy. Maybe add  a set of test points (plus 0 ohms resistors)
<qi-bot>
[commit] Wolfgang Spraul: renamed kicad/library to kicad/components for consistency http://qi-hw.com/p/xue/99067ec
<wolfspraul>
let's see whether schhist survives that
<wolfspraul>
we have a few empty .dcm files in kicad/components, unless there is a reason against it I would propose to delete them
<wolfspraul>
andres-calderon: from 12 .fpd files in kicad/modules, 11 are machine generated
<wolfspraul>
what is the plan with that now? where will future edits be made?
<andres-calderon>
wolfspraul: I'll make a makefile
<wolfspraul>
to create the .fpd ?
<wolfspraul>
or to go from .fpd to .mod?
<wolfspraul>
we have 22 empty .mdc in kicad/modules, if it's safe to delete them I will probably delete them
<wolfspraul>
(I'll wait until Werner is back he knows KiCad best...)
<andres-calderon>
I will review the makefiles of the GTA02 core project
<wolfspraul>
that's quite old by now, I suggest ben-wpan
<andres-calderon>
wolfspraul, ok
<wolfspraul>
andres-calderon: how were the .fpd generated?
<wolfspraul>
was it a one-time concversion, and future edits will be made in .fpd, or will the source stay the source for edits
<wolfspraul>
right now they say "don't edit"
<wolfspraul>
ah no, they don't say that :-)
<wolfspraul>
but they say "MACHINE GENERATED"
<andres-calderon>
with the FPED editor
<wolfspraul>
they were created with the fped editor?
<wolfspraul>
I'm wondering about the 'MACHINE GENERATED' comment
<kristianpaul>
Hello i'm back now from Fedora (MIlkymist dev mainlly reason) :)
<kristianpaul>
Lets see how it goes
<wolfspraul>
Fedora has SELinux enabled by default, right?
<kristianpaul>
dont know
<wolfspraul>
never worked with that before, had to get used to a few bumps
<kristianpaul>
ah yes i got some auto-bugs pop ups from SELinux..
<wolfspraul>
yeah. first time for me, I was always happy that it passed me, but now I guess it's time :-)
<wolfspraul>
wpwrak: there you are :-)
<wpwrak>
troublesome connection
<kristianpaul>
hmm this is because xilinx installer, tricky
<wolfspraul>
how do you feel about deleting empty .dcm and .mdc files?
<wolfspraul>
when fped creates a .mod file out of a .fpd file, does it also create the .mdc file?
<wolfspraul>
the alternative would probably be micro-usb right away, and those platforms are normally so closed that it will be very hard to get atusb to work
<wpwrak>
(the general shape would be that of the first one)
<wpwrak>
no micro-usb :)
<wolfspraul>
there are very few devices that have a mini-usb host receptor
<wolfspraul>
like the Neo FreeRunner :-)
<wpwrak>
micro-usb is for idbg. let's not confuse things ;-)
<wolfspraul>
I am thinking from the other side - what to plug it into...
<wolfspraul>
for the full-size, it will be notebooks, netbooks, maybe open tablets
<wpwrak>
it's basicallly either pc directly or via cable
<wolfspraul>
for mini-usb, I doubt many devices exist at all
<wpwrak>
with cable
<wolfspraul>
for micro-usb, a lot of devices exist with usb-on-the-go, but they will be too closed to have fun with atusb
<wolfspraul>
sure cable is another option
<wpwrak>
pro: can find better placement. contra: if you can't get rid of the cable
<wpwrak>
there are also full-sized USB cables, but they're not so easy to find
<andres-calderon>
wolfspraul:Â Â I already deleted the mod. files (and has been a makefile based in the ben-wpan modules makefile)
<wpwrak>
(digi-key has them, though ;-)
<wolfspraul>
andres-calderon: I saw it, great
<wolfspraul>
when you committed, did you see the merge commit?
<wolfspraul>
there is a way to avoid that, I think before you push your commits, you do a 'git fetch origin' and then 'git rebase origin/master'
<andres-calderon>
wolfspraul: I dont know anything about the  mdc files. They are used when the module is added to the design...
<wolfspraul>
sure don't worry those are tiny details
<wolfspraul>
wpwrak: I'm ok with 'cam' right now for mechanical stuff
<wolfspraul>
I don't need it urgently anyway, but it's a good first step
<wpwrak>
so .. full-size usb A plug or mini-usb B receptacle ? :)
<kristianpaul>
wpwrak: full-size i think
<kristianpaul>
still more universal for usb adapters
<wpwrak>
any other votes ?
<wolfspraul>
full-size
<wpwrak>
seems that we have an unanimous vote :)
<wpwrak>
now i have to figure out why my full-sized usb resets all the time ...
<kristianpaul>
not reset in mini-usb?
<wpwrak>
the design i made in august, with mini-usb has no usb problems. (the rf side is a disaster, though)
<wpwrak>
the new one, with full-sized USB, has big troubles staying up. not sure why - the usb circuit is basically the same as in "cntr" and very similar to the old atusb, idbg, etc.