libera/#neo900/ Thursday, 2018-08-16

* chomwitt checking mastadon... found fosstodon instance01:00
chomwittis the virtual machine with the kicad still operational?01:02
houkimeit ivirtual machine01:02
houkimewhat vm?01:02
houkimeomg, my sausage fingers01:03
houkimewhat vm you're talking about?01:04
chomwitti think one year ago Joerg-Neo900 had started a debian-iso with the kicad version + fixes?01:07
houkimedon't bother. Any kicad from year ago is obsolete.01:08
houkimeyou have troubles installing kicad?01:08
houkimeright now i'm running git version01:10
houkimebesides you probably don't wanna run kicad in a vm. kind of resource-heavy01:13
houkimestarting from kicad5 it finally started to use graphics card I believe. non-accelerated legacy toolset was real sluggish01:19
houkimerecently joerg started using suse build servers to build development versions. i personally use Arch and its PKGBUILD scripts01:23
houkimekicad5 is already released though so you shouldn't really need dev stuff01:26
houkimeneo900 does use kicad5 specific stuff but none of kicad6 yet01:26
houkimeFor now I don't really see any 6->5 backward incompatibility ahead.01:29
houkimeit has a bit more ergonomic interface but that's all that i notice so far01:30
chomwitthttp://neo900.files.dev-1.org01:31
chomwittthats were the devuan ! iso i think is. but as u say since kicad5 is out maybe is obselete.01:31
chomwittagain.. the issue arise .should there is a 'start here developer' guide to lower the barried and be updated?01:32
houkimeof course it should be such a guide. i have already mentioned it earlier here01:33
chomwitta new issue then!01:34
houkimedo you want to open it yourself?01:35
houkimeyou're welcome if so01:36
houkimeit is an issue for a planning repo rather then a technical one01:37
chomwittok!01:37
houkimeit is done as in github, you just get simple acc for notabug and open your stuff01:38
chomwittacc ?01:41
houkimeaccount01:42
chomwittok.  issue 101:46
houkimegreat.01:54
houkimewould you like to create (probably dummy at first time) CONTRIBUTING.md yourself?01:55
chomwittin the wiki page of Neo900-planning ?02:01
houkimewiki is usually used for very long sofisticated instructions in addition to CONTRIBUTING.md02:04
houkimeCONTRIBUTING.md considered a required minimum02:05
houkimebecause it is a file in the repo you can propose changes to it like to any other file02:05
houkimeright now ther's no such file so one needs to create it02:08
Joerg-Neo900chomwitt: yes, the kicad VM s still the reference kicad release for the prokect02:10
houkimebtw once we have a wiki server running (which we can have any time) I can redirect a repo wiki link to neo900.org wiki02:10
chomwittso wikipage is not part of the repo ?02:10
chomwittJoerg-Neo900: thanks02:10
houkimewikipage can be a part of the repo and may be a separate thing which is linked via redirection02:10
houkimechomwitt: check the kicad version there. You might have problems if it is not 502:11
chomwitthoukime: ok. i'm not a hardware person , put curious and intrigued like you from the educational aspects of inspecting a so complex project.02:13
chomwitthoukime: well Contribute.md should have valid info. Readme.md says get the sources from Joerg-Neo900 .ok . and then ? according to Joerg-Neo900 he must install the kicad VM02:16
chomwittbut where is the kicadVM ?02:16
houkimechomwitt: ignore joerg, he is not working with files right now02:16
houkimechomwitt: generally, right now a stable release of kicad5 should be ok02:18
houkimechomwitt: and i don't see a single reason to make contributor to downgrade02:19
chomwitthoukime: but ..we say .. go to Joerg-Neo900 to get the updated schematics.. , but dont listen to how we'll work with them? couldnt a diff in what version of kicad is used create issues?02:20
* chomwitt scratch his head..02:20
houkimechomwitt: there is no contradiction The schematics was bugfixed and upgraded to kicad5 by me02:22
houkimechomwitt: you need joerg to access a gitserver02:22
houkimechomwitt: then you checkout my crispy branch and rock-n-roll02:23
chomwitthoukime: didnt know that02:24
chomwitthoukime: if u have access to the gitserver why you cant give access to it ?02:26
* Oksana shrugs and hopes that "bugfixed and upgraded to kicad5": 1) didn't take too much time; 2) will not make it difficult to edit the schematics, or manufacture the boards02:26
OksanaThere is difference between being an admin and being a contributor, I guess.02:26
chomwittOksana: ok02:26
* Oksana 's experience with git is limited to occasional glances at github, though02:27
chomwittso Joerg-Neo900 is the admin and will create a git account with 'push' rights02:27
houkimechomwitt: i just have a write access to my branch. joerg makes decisions and atk manages access rights02:28
houkimeatk is admin02:28
chomwitthoukime: so we assume that your branch is the one where work will be done from now on ?02:30
chomwitta rhetorical question.. i just hoped that there would be blessing from all.. like in ceremonies.02:32
houkimeOksana: 1) It atually took a fair bit of time. 2) I believe what WAS there created much more difficulties. And downgrading to kicad4 now when kicad5 is out is an extra difficulty02:32
Oksana2) the difficulties were from "this was made in kicad4, kicad5 doesn't understand it well" or from "this isn't drawn right, the wires look nice but aren't actually connected properly"?02:34
OksanaAs in, did upgrade to kicad5 create additional overhead (not that necessary, but preferred because kicad6 is out), or highlight existing problems which would have needed to be fixed anyway?02:35
houkimeOksana: 2) both. however  mostly not like illusionary connections but grounds not sorted out between boards, sch pins and even pinnumbers not corresponding to actual fps and actual datasheets etc02:38
houkimetons of nasty stuff that made a sync between board and sch inpossible02:39
houkime*pincounts02:39
houkimeimportant correction, sorry about that02:40
houkimei meant the total headcounts of pins02:40
houkimeyou see, for kicad to work as intended and bugs to be rare the proper sync between board and sch is very important and speeds up work02:41
houkimealso for example for routing diffpairs there're naming rules for pins02:43
houkimeotherwise kicad won't recognize them as diffpairs02:43
OksanaYou lost me at diffpairs, but yes, real-time sync between schematics and board is important, as far as I can guess02:45
houkimeoksana: if you have seen kicad working sessions you actually have schematics in the other window while working on the board and kicad finds and shows you in sch stuff you selected on board automatically02:47
houkimeetc etc02:47
houkimeso actually even the situation when the board was kicad5 (since metacollin) and ach in kicad4 is very inconvenient02:48
houkime*sch02:49
OksanaOkay... I installed kicad on my laptop, but didn't open it yet - it could be kicad6, for all I know02:50
houkimeit is not kicad6 if it is not a devbuild02:53
houkimelatest stable is 502:53
houkimewhich is ok02:53
OksanaOkay, will likely be kicad5 then, good02:54
houkime6 is actually also ok. 4 is not02:54
houkimenow, my branch is not master yet. also joerg doesn't want to open layout files yet.02:55
houkimeso files from neo900.org are from master branch and they are for kicad4. which is autoupdatable and viewable but not relly representative02:56
houkimeand they are old02:57
houkimeto proceed surther to sth constructive you will need to ask Joerg and atk to give you an access02:58
houkime*further02:58
* Oksana nods, but also notes lack of time in schedule, for at least the next month or so02:59
houkimeoksana: doesn't matter. I also lack it.03:02
houkimewell, actually maybe changes were not so big. They were more of subtle-but-important. you can still get the idea from open docs while you wait but for actual work you need access.03:06
houkimeI actually should compose some list of useful electronics resources to kickstart guys03:08
houkimeand do more viseos about subsystems and stuff03:11
houkimeso everything is as clear as it can be03:11
houkimeit actually helps debugging also03:12
houkimechomwitt: Are you still writing the early version of contributing guide? Don't go into much details just yet, we will add them later.03:15
houkimejoerg-neo900: will oksana with 7 preorders be allowed to have a read access to the gitserver?03:17
houkimechomwitt: my branch is called houkime/layout if you need a name03:20
Oksanahoukime : 6 pre-orders, and that's not my last name ;-)03:22
houkimejoerg-neo900: ok, 6 preorders. I think from any possible angle she has the right to have look even if she doesn't end up helping.03:24
houkime*won't03:24
* Oksana points houkime to http://maemo.org/profile/view/wikiwide/ 03:26
houkimechomwitt: sorry i missed your question about my branch being the development focus. Yes it is a development focus, because it is the only branch progressing and it is gone veeery far03:33
chomwitthoukime: i cant ignore Joerg-Neo900 opinion 'the kicad VM s still the reference kicad release for the project'. i dont even know the guy or kicad.03:34
houkimechomwitt: you can consider this opinion but only as formal since following it for real will introduce too much problems.03:36
Oksanachomwitt : try both ways (kicad VM, and kicad5), in parallel, if you can?03:38
chomwittbut wouldnt be strange if a newcomer ,following a Contribution guide' , asks Joerg-Neo900 for git access but then the same guide stating 'dont pull his branch and dont listen to him' ?03:39
* Oksana shakes head and thinks that whyever metacollin switched board to kicad5, there better be a good reason for that03:42
houkimechomwitt: sorry. i didn't really mean ignoring joerg completely. joerg is a nice ceo. he knows a lot and he keeps things running. He just doesnt' work with files currently and thereby may suggest impractical things.03:42
chomwittsorry.. 'a nice ceo' ... 'suggest impractical things' ??03:43
* Oksana hopes that PCB proceeds to manufacturing soon - because this kicad4-or-kicad5 stage may take up long enough time that kicad6 gets released into stable, which will not help any03:44
chomwittyou wont me to write a guide inviting people to a project where in their very first step will have a strange feeling like i do now.. (being in a battle or sth..)03:46
chomwitt..hey come to eat in the restaurant , dont bother that chef says that you shouldnt try the new recipies?03:47
chomwittJoerg-Neo900: should step forward and be more clear i think. is he ok with a guide 1) get access from you 2) choose what repo to work on ?03:50
* Oksana takes a look at http://neo900.org/stuff/kicad/proto_v2/2016-11-20/review-howto.html 03:52
OksanaGoing by http://neo900.files.dev-1.org/vagrantboxes/kicad_1920x1080__root-devuan_compile-kicad/kicad-xfce-devuan.box , I should install VirtualBox on my laptop03:52
houkimechomwitt: my branch is not rogue or anything. It does disobey some of Joerg recommendations for practicality reasons and contribution ease.03:53
chomwitthoukime: i didnt even imply sth like that.03:53
chomwitti just say that i feel ..uneasy.03:54
OksanaWell, as long as it's not KiCad nightly [it isn't] and everybody else can easily view/edit/manufacture from your files03:54
Oksanachomwitt : Understandable. I miss wpwrak, haven't seen him talking in months?03:56
chomwittanyway i'have to sleep. i'd prefer Joerg-Neo900 to state his opinion for the Contribution guide .03:56
houkimeoksana: I miss him too03:56
houkimeoksana: it was always interesting to talk to him.03:57
houkimechomwitt: good night03:58
houkimeoksana: it was kicad nightly at first, but now there's no reason for that.04:00
OksanaWhy was there need for kicad nightly in the first place?04:01
houkimeoksana: metacollin. Also to be fair the difference between 4 and 5 is drastical. 4is a slow buggy mess04:01
houkimeoksana: and metacollin spoke about oval shapes that were not in kicad404:02
OksanaSpeed difference because of hardware graphics acceleration, or lack thereof?04:02
houkimeoksana: graphical acceleration and ergonomics also04:05
houkimeoksana: it is just way more convenient for various reasons04:06
houkimeoksana: this whole situation reminds me of recent Linus Torvalds hurricane about users who shouldn't need to downgrade under no circumstances04:13
houkimehttps://www.youtube.com/watch?v=FUsJ-Lr4kww04:13
houkimeAnd also reminds me of my familiar Arch system where "package which forces users to downgrade=dead package"04:16
Oksanahoukime : Please, don't link youtube. Neither K-meleon [in my current configuration] nor MicroB [on my Nokia N900] can play its videos. And if it is about Linus Torvalds, then surely, it's also available on a better site?04:17
houkimeoksana: btw does peertube work for you?04:19
Oksanahoukime : Checking... Doesn't seem to work in K-Meleon, should try to fool it with user agent sometimes...04:22
OksanaAnd I wish ircs:// links were common on the Internet, instead of leaving it to users to copy-paste names of server and channel manually04:23
houkimethe video is yt-only, but here's the source https://lkml.org/lkml/2018/8/3/62104:29
OksanaFunny. If my ancient laptop is strong enough to run a VirtualBox, perhaps, I will have opportunity to try out kicad5 files in kicad4. Because bug-fixes should not break user...04:33
houkimewhy do you need a vm?04:35
OksanaTo try out kicad4 VM?04:35
OksanaI understand that installing kicad5 and kicad4 on the same computer without VM is possibly possible, but I would try VM first.04:36
houkimeoksana: kicad5 files are not really readable by kicad4. not downgrade but upgrade is needed to allow reading just new kicad5 files so it sorts of ok. kicad5 itself can devour and autoupgrade kicad4 files.04:45
houkimetechnically it isn't breaking the user i wouldn't say.04:46
houkimeoksana:need to sleep04:48
OksanaAnd if user, for whatever reason, doesn't wish to upgrade to kicad4? And yes, I am familiar with issue of "old version cannot read files generated by new version" - only in context of different software.04:49
OksanaGood night :-)04:49
* brolin_empey wanted to ask houkime about using Vimeo.05:28
brolin_empeyOksana: You can use youtube-dl, which is a cross-platform Command-Line Interface Python application, to download a local copy of videos from YouTube and many other video sharing Web sites.  You should be able to run youtube-dl on an N900 if you have a sufficiently new version of Python and any other dependencies.05:30
brolin_empeyAlso, I thought K-Meleon is a Windows application.  Are you using Windows?05:31
brolin_empeyYes, apparently K-Meleon is only released for Windows.05:36
* enyc meows05:59
Joerg-Neo900a particular problem with switching toolset furing project runtime is: anythingchecked and validated in e.g. kicad4 has to be considered unknown-state and needs re-verification after you opened/coberted the project files to kicad506:33
Joerg-Neo900simple made-up example: kicad4.1.6 has a bug in computational rounding, so the devel works around this by adding a 0.001mm constant to every footprint/layout distance ending in digit "5". Now some devels switch to kicad4.1.7 which has this bug fixed, while others don't. Good luck ever sorting this again after a dozen open+modify+save cyles done by 4 devels using different versions of kicad06:39
enycJoerg-Neo900: makes sense, what does this mean in termso f  versinos-of-toolage  available at pcb-manufacturers/assemblers, though ... ?06:44
Joerg-Neo900so the project wide rule been: "we have a golden sample (see VM) kicad version, and everything neds to load and get verified on this golden version. Devels are somewhat free to use newer kicad versions (though they are discouraged to do so) but then need to check their work for correctness in golden version before distributing / pusjing"06:45
Joerg-Neo900enyc: PCB houses will accept kicad4 generated Gerber just happily06:46
Joerg-Neo900afrer all Gerber isn't a kicad-'proprietary' format but rather a industry standard06:47
Joerg-Neo900since our leyouters all have been keen to use kicad5 aka 'nightlies' so far, the plan been to switch the project to kicad5 officially as soon as (or shortly after) there's an official stable LTS kicad5 release. Then we'd have to do re-validation of schematics and footprints and whatnot only *once*06:49
enycJoerg-Neo900: kicad 5.0.0 is officially released, but i have no idea how buggy it proves to be =)07:02
Oksanabrolin_empey : Yes, at Uni workplace I am using Windows. Thanks to IT, the only other choice is Mac - they don't offer Linux on their workstations, yet.07:07
Oksana>> "but then need to check their work for correctness in golden version before distributing / pushing" << So it's a good idea to see if kicad4 VM can open files from kicad5...07:09
OksanaDo not see anything about KiCAD LTS, though.07:12
brolin_empeyOksana: OK.  I was not criticising your choice but I was surprised that you use Windows.  I use Windows β€” actually, Windows NT β€” too because I prefer it as a desktop client platform, not because I am forced to use it.07:14
Oksanabrolin_empey : I would consider Windows as an option if it didn't force me to frequently discard old computers and buy new ones (because old computer cannot run new Windows, and old Windows is too vulnerable to be connected to Internet). Maybe, ReactOS becomes useful some time in the future...07:17
* Oksana used to have Windows XP on laptop, and then got Linux to dual-boot on it, since XP lacks security bugfixes needed to connect to Internet07:18
brolin_empeyOld computer cannot run new Windows? 0_o  Windows 7 still runs on a Pentium II.  Windows 8 still runs on a Pentium 4.  I use Windows 10 for x86-64 on Core 2 computers from around 2008.07:22
Joerg-Neo900>>So it's a good idea to see if kicad4 VM can open files from kicad5...<< won't07:23
Oksanabrolin_empey : Old laptop, from 2007 or so. To be fair, it maybe could run Windows 10... If 4GB RAM are installed... But it would be about the limit: any RAM-hungry application (such as, web-browser) would make the whole system slow?07:31
* Oksana manages to bring Windows 7 workstation at Uni to terrible slowness sometimes, despite 16GB RAM installed07:32
brolin_empeyOksana: Does your notebook computer from 2007 have an x86-64 CPU?  8 GiB of main memory seems to be the minimum for my use case.  With only 4 GiB, Windows keeps wanting me to exit applications because it is low on memory.  My Dell Latitude X1 model of notebook computer from 2005 with Pentium M CPU runs Windows 8.1 for x86-32 from 2013.  I ran Windows 10 for x86-64 with only 2 GiB of main memory at work today on a Latte Panda x86 SBC.07:37
brolin_empeyOksana: Does said workstation run from an HDD?  If yes then that is normal.  HDDs are a major bottleneck.  If you have never used a desktop/notebook computer running from an SSD then you may not realise how much the responsiveness is decreased by running a computer from an HDD.07:40
brolin_empeys/an SSD/an SSD or a RAM drive/07:43
Oksanabrolin_empey : x64, if I remember correctly from the time when I installed Linux on it. | >>8GB...minimum<< And that's a good example of how Windows gladly places users between vulnerabilities or purchase of a new computer.07:44
Oksanabrolin_empey : Yes, HDD on workstation (and laptop as well). I have heard terrible things about how SSD loses data quickly when powered off.07:44
brolin_empeyβ€œAnd that's a good example of how Windows gladly places users between vulnerabilities or purchase of a new computer.”07:49
brolin_empeyWhat do you mean?  My x86-64 desktop motherboard from 2007 officially supports up to 8 GiB of main memory and unofficially is said to support up to 16 GiB.  I will test it with 16 GiB when the 16 GiB of RAM I ordered a month ago finally arrives.07:49
brolin_empeyMuch criticism of Microsoft comes from persons who are mistaken inclusive-or who do not understand business.07:50
OksanaI just mean that if your use-case happened to be with my old laptop (instead of your old desktop), then 4GB-officially-supported would not have been enough at all07:51
brolin_empeyThe amount of main memory officially supported by a model of computer is determined by the storage capacity of the RAM modules available at the time the computer design is tested/validated, not by Microsoft nor any other operating system developer.  The Apple Macintosh SE/30 from 1989 unofficially supports up to 128 MiB of main memory.  The actual maximum amount of main memory often increases after the computer is released when larger storage capacities of07:57
brolin_empeycompatible RAM modules become available, such as 16-MiB 30-pin SIMMs for the Macintosh SE/30 or 4-GiB DDR2 SDRAM DIMMs for my x86-64 motherboard.07:57
* Oksana thinks that the user manual should state real limitations instead of contemporary "maximum size of RAM module", and looks into the Internet's words of wisdom on Lifebook RAM08:05
brolin_empeyAlso, an x86-64 notebook computer is not old on some scales, such as when considering that I have an 80286 notebook computer from around 1989. :-P08:06
* Oksana nods, but then, I am younger than your 80286 computer... and https://www.compuram.biz/memory/fujitsu-siemens/notebook/lifebook/s-series/s7210/8gb-dual-channel-memory-kit-2-x-s26391-f681-l40-x25aa.htm is not a common type of RAM, it appears08:15
brolin_empeyOksana: As far as I know, usually Intel/AMD/VIA/whoever do not requalify their old hardware designs for the maximum amount of main memory after the hardware design is released.  The official maximum is only the maximum that was qualified by Intel/AMD/VIA/whoever and possibly also the computer/motherboard maker.08:17
OksanaAs in, 110AUD to equip one laptop with two such modules08:17
brolin_empeyafter the hardware design is released = after the product containing the RAM controller is released, such as the north bridge or the CPU for an x86 computer.08:20
* brolin_empey was born in 1987 so he is younger than the 80286 (released in 1982) but not this 80286-based computer from 1989.08:24
brolin_empeyAlso, interestingly Wikipedia says that the original XBox, which uses a Pentium III CPU, uses DDR SDRAM but I have never knowingly seen a Pentium III computer that uses DDR SDRAM.08:27
OksanaDisassemble an XBox?08:29
brolin_empeyOksana: No, I do not have one.  If I really wanted to check that claim it would probably be easier to find a photograph of an Xbox motherboard to try to read the label/marking on the DRAM IC(s).08:34
brolin_empeys/that uses DDR SDRAM/that uses DDR SDRAM for the main memory/08:38
brolin_empeyHaving a peripheral, such as a video card, that uses DDR SDRAM does not count.08:38
brolin_empeyOK, I searched the Internet and found this article:08:53
brolin_empeyhttp://www.vanshardware.com/articles/2001/november/011116_Xbox/011116_Xbox.htm08:53
brolin_empeywhich says that the DRAM IC is a Samsung K4D263238M-QC50 , which apparently truly is DDR SDRAM instead of SDR SDRAM but probably only because the Xbox uses an nVidia chipset instead of an Intel chipset.  Other than the Xbox, I do not think I have knowingly seen a Pentium III computer or, more generally, a P6 microarchitecture computer with an nVidia chipset.  As far as I know, the Intel chipsets used in P6 microarchitecture computers support only SDR SDRAM,08:53
brolin_empey not DDR SDRAM.08:53
brolin_empeyThis article also talks about an Intel computer using HyperTransport, though;  I thought HyperTransport is an AMD technology.08:54
brolin_empeyThis article also says that the Coppermine CPU used in the 2001 Xbox motherboard is a Mobile Celeron, not a Pentium III, but I guess the only difference is the amount of cache.08:57
brolin_empeyOr at least the primary difference, maybe not the only difference.08:57
houkimeIf there is no progress on this matter and no consensus reached this Sunday I will write some minimal contribution guide myself. One can still edit it later. I willl write it in contributor-centric manner, because contributors for stagnating project is the most precious thing in the world.15:41
houkimeStrict real-time quality control is useless if contributors are in pain and can't move forward.15:48
houkimeif such situation occurs new ways should be invented to provide quality15:49
houkimeThe volume of tasks and contributions needed to the toolchain doesn't really matter because it is really THE only way out.15:50
houkimeotherwise the project is not fun to participate and is dead15:51
houkimethis is how community projects work. They absolutely must be interesting and fun to participate. It is not even #1 it is #0.15:54
houkimeend result doesn't even matter as much as fun does15:54
houkimeand even "chinese copycats" don't matter if fear of them deters contributor fun16:02
houkimechomwitt says that the current apparent conflict between joerg's ideas of how things should be done and me trying to ease the flow makes him uneasy.16:16
houkimeand  if a community member chomwitt feels uneasy we have a problem, esp so that it is likely that others will feel it too.16:19
houkimehowever I don't know how to end conflicts with joerg's vision without deterring contributors flow and leaving neo900 die.16:22
houkimeI'm open to your suggestions though and will be glad to at least try to discuss solutions in a constructive manner.16:25
Joerg-Neo900do you have anything to sugest yourself, instead of just spreading FUD?16:55
Joerg-Neo900even on github you have to register before you can contrinute16:56
Joerg-Neo900contribute*16:56
houkimeRegistration on github is an automated and extremely quick process. And it is the single thing you need in order to contribute. no cryptosigned ndas.17:01
houkimeBesides, getting the sources (which in my suggestion should just hang somewhere in the open) is not the only problem with contribution flow right now.17:04
houkimekicad4 is very painful and buggy (try opening current layout with legacy toolset enabled) and no matter what theoretical bonuses lie in using old software it doesn't really make for it17:07
Joerg-Neo900I'm fed up with this constant rant about the project being structured in a way so it can't find contributors. Please point us to a prior project of yours where you did it in a better way and had better success with layouting ccomplex circuitry17:07
houkimeI peronally don't have such projects because i'm very young. But Sparkfun for example does.17:08
houkimebasically sparkfun is all structured in a way to comfort contribs17:08
Joerg-Neo900probably that's the core problem here. I'm very old and have seen lots of shit happening17:08
Joerg-Neo900and this isn't the first project I'm contributing to17:09
houkimevery nice of you. in 2018 we can do stuff in a fun open way and don't suffer from old shit that you've seen as much17:10
Joerg-Neo900furthermore I am financially liable for this prohect17:10
houkimeand thereby you will let it die?17:12
Joerg-Neo900anyway, where are your suggestions to *improve* stuff. I'm really fed up with 90% of what you say is "this isn't good, this sucks, this is buggy" (with the latter usually turning out to be a lack of understanding in the complete picture on your side)17:13
^7heo< houkime> In 2018 we can do stuff in a fun open way and don't suffer from old shit that you've seen as much17:14
^7heoThat is SO besides the point that I don't even know how to answer it.17:15
^7heoDoing hardware and doing software are roughtly equivalent to respectively going to war and playing CS:GO.17:15
^7heos/tly/ly/17:16
houkimeOk. give me a sec to write a nice list of suggestions to improve contrib flow.17:20
houkimebtw ^7heo: DO have a look at how a hardware org sparkfun operates. I was not speaking about software.17:21
^7heoBut make it nice alright? It's much more important than it being useful. ;)17:21
^7heoyou're comparing things that have nothing in common.17:22
^7heosparkfun is a retailer.17:22
Joerg-Neo900oh c'mon we obviously simply put out fingers in our ears and sung LALALALA when you came up with that truckload of issues you're just running into, in the past17:22
^7heoJoerg-Neo900: wie alt bist du btw?17:23
Joerg-Neo9005817:23
^7heodas ist nicht so alt ;)17:23
Joerg-Neo900"zu alt fuer so nen scheiss"17:24
^7heo:D17:24
houkimeok, here is it18:51
houkimemaybe not perfect but sth to discuss without "please propose sth constructive"18:52
houkimehttp://susepaste.org/6223641918:52
houkimeI will go and check in in several hours so you can uninterruptedly write whatever criticism you want.18:57
houkimebe back later.18:58
Joerg-Neo900sorry, I won't stuff that contains accusations like "done because it is NOT blocked by joerg"21:47
Joerg-Neo900if you think I'm your enemy then this project isn't for you21:48
Joerg-Neo900complete 1) (no I'm dicussing it nevertheless) is *again* moot since it moans about implementations while not listing the goal. WTF is your problem with "complicated scripts" you never even seen in your life so far?21:49
Joerg-Neo900your points been answered weeks ago, you're repeating yourself. See https://irclog.whitequark.org/neo900/2018-08-08#2278087721:54
Joerg-Neo900there's this one guy in that other project wrtitten in python, who constantly pushes violently for switching the complete project to golang. Guess about his success with his idea, when he doesn't deliver sound arguments why the project needs such a radical switch21:57
Joerg-Neo900his arguments are like "*I* can write much faster in go than in python" and "go is a nice languange" and "python is lame"21:58
Joerg-Neo900I'm very sorry we don't use gitlab like PRs. We simply don't. Get over it, or come up with an evidence where we need this massive complication and redactoring of what's there. "Because you like PRs better than sending patches" is NOT such evidence22:00
Joerg-Neo900>>(however right now even formalities may take several days (and they TOOK in my case) I consider this inacceptable contributor experience)<<  Metacollin's "contributor experience" was 5h to understand what we do **and provide his first contribution: a Mac patch for eeshow** - how long did it take until you even took a LOOK at the easily accessable info about WHAT EESHOW IS?22:05
Joerg-Neo900and by "provide contribution" I actually refer to a git commit&push22:08
Joerg-Neo900re 3:  congrats, that's exactly what we have, afaik22:10
Joerg-Neo900>>4. Make kicad5 and each stable release hereafter an official tool to work with files:<< you start to annoy me. Show me the "stable kicad 5 release" of 2 months ago!!22:12
Joerg-Neo900honestly, *everything* is in the CC-BY:NC-CA and GPL, when you think we're doing everything wrong, pretty please YOU START DOING IT BETTER. Stop blaming us and trying to force US changing stuff22:14
Joerg-Neo900you know what's killing the project for sure? contributers looking at the channel log and seeing your endless rant about *everything* in this project being absolutely unbearable22:16
Joerg-Neo900and a basic consideration: I don't think layout (pcbnew project files) lend themselves well to management with git, and generally layout doesn't lend itself to cooperative work at all. So I suggest you simply upload the layout stuff to regular dir using your sftp account. Since it seems you can't be bothered to learn how to "git commit && git push"22:34
Joerg-Neo900and schematics are not in the realm of layouter, only for reference. Any changes un schematics need to get done by the schenmatics folks who know how to use kicad4 to keep compatibility for that22:36

Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!