CERN's contribution to KiCAD

[djsb]

Just thought I'd share this link to information on CERN's contribution to the development of KiCAD

http://www.ohwr.org/projects/cern-kicad/wiki

Not much there yet but but there a BIG plans (subject to donations-Page for donations to be added later)

http://www.ohwr.org/projects/cern-kicad/wiki/WorkPackages

I hope it is of some interest.

David.

[rolycat]

Very ambitious - it looks as though this might eventually provide a really credible (and free!) alternative to the likes of Altium.

I'm particularly fond of this work package objective:

• Study ergonomics of various commercial/proprietary PCB applications (when in doubt about any particular UI solution, check how it has been done in a certain proprietary app that is very popular among OSHW folks and do exactly opposite).

[BravoV]

 

[andersm]

Something's finally happening? That's great! The project went so long without any visible activity I assumed it was abandoned before even starting.

[Bloch]

Very ambitious - it looks as though this might eventually provide a really credible (and free!) alternative to the likes of Altium.

Yes this is great news. At least for the open source hardware movement.

[free_electron]

Maybe there is hope after all.
There is one element that is blocking though. They want to standardize the library on 2d.... That should be 3d... Especially if you want to i tegrate 3d models and complex stackups like blind and buried vias.....

[amspire]

I think the 2d is only referring for the geometry used by the push & shove router.

I think they plan to improve the 3d functionality as well.

[c4757p]

Oh hell yes.   

[codeboy2k]

Good that they forked it.  kicad needs to improve. The devs weren't willing.

[Bored@Work]

Good that they forked it.  kicad needs to improve. The devs weren't willing.

I don't read the announcement as an announcement of a fork. They write

do it in close collaboration with the current main Kicad developers.

And they talk about contributing.

[Smokey]

I'm all for the concept, but I wonder what their motivation is for doing this.  It's not like they can just make it a grad student thesis project or something and get free labor like a pure university sponsored project.  It seems like it would take a ton of work to get kicad up to the level of a professional EDA package.  That's time their programmers could have been using to write particle physics code (or whatever they would have been doing) as opposed to just buying a tool they can get off the shelf.  I would think that you would want coders also familiar with PCB layout so they understand what they are doing, which further limits their programmer pool to work on this.  Is there some feature they can't get from a commercial EDA?  I find that hard to believe, and even if commercial EDA packages are missing something they need, they have to build up all the other basic features they are missing from kicad in addition to the special one to make it useable.

[Bored@Work]

The basic argument is that true open hardware needs to be done with open source tools, not proprietary tools.

This leads to the question why that particular CERN section (decoded the group's name is beams/controls/hardware&timing) is into open hardware? Their official mandate https://espace.cern.ch/be-dep/CO/Hardware_Timing/default.aspx doesn't mention it. So they must have made a case that OH is benefitial to their work. This looks like how they made their case https://indico.cern.ch/conferenceDisplay.py?confId=190126

[ve7xen]

Perhaps they are committed to openness in academia, and aren't focusing purely on short-term results? CERN is large enough that they can probably spare a couple of developers to work on tools that will improve their ability to collaborate and share with other academics in the long run.

IMO academia is one place where using open source tools makes a lot of sense in the long run.

[codeboy2k]

I don't read the announcement as an announcement of a fork. They write

do it in close collaboration with the current main Kicad developers.

And they talk about contributing.

OK. We'll see if it turns out to be true.  Kicad's main developers are pompous. They resist change, and they have shown an unwillingness to receive contributions in the past.

I expect that the cern-kicad tree in git will progress faster, as they will want to do real designs with Kicad, and share their designs with other labs around the world, who may also contribute back to CERN and/or request CERN to add features to Kicad that they need.

[HackedFridgeMagnet]

I'm all for people putting in some more effort in Kicad but this sounds a bit strange.
What do the Kicad developers say about the CERN roadmap? Sounds like a fork.

If they are going to add/change a plugin architecture then it seems like a complete rewrite would be easier.

Goal: Conflict-less way of developing tools. Script-driven tools. Refactoring towards clean MVC model.

Scientists writing MVC in python sounds like a project that will never finish. I hope I am wrong.

[AlfBaz]

I don't know anything about kicad, but if its truly open source and the existing kicad developers remain obstinate, there's nothing stopping cern from forking down their own development path.

As for cern allocating resources for this, it's not inconceivable that they have a multinational funding regime and budgets to develop tools. Looks like somebody in cern thinks this would be a good thing

[Bored@Work]

OK. We'll see if it turns out to be true.  Kicad's main developers are pompous. They resist change, and they have shown an unwillingness to receive contributions in the past.

Yes. From some previous interaction I know they don't even get what the problems are. But maybe a big name like CERN helps to kind of kick in a few doors.

Until now it looks as if the kicad developers will accept CERN code.

[Bloch]

Open letter fron kicad mailing list

[Kicad-developers] Introducing CERN BE-CO-HT
Javier Serrano Thu, 11 Apr 2013 06:22:11 -0700
Dear all, As you know, the BE-CO-HT section at CERN has been contributing to Kicad for some time now. We see this involvement as a key part of our activities regarding Open Source Hardware [1]. We are also contributing to the adoption of VHDL and SystemVerilog in the Icarus Verilog simulator, so as to enable cross-language simulation. Other related activities include the operation on the Open Hardware Repository [2] and the development of the CERN Open Hardware Licence [3].
Regarding our contribution to Kicad, it took us some time to put our act together, but I believe we are now ready to announce that development has started in earnest. We have come up with a preliminary roadmap [4], which has been discussed with the main Kicad developers [5]. We have agreed that this strategy is at least a good enough base to start development work.
The main actors involved on our side will be:
- Maciej Sumi?ski, aka "Orson". He is going to be with us for a year within CERN's Technical Student Programme, working full time on Kicad. He started in March. He has already started working on a new View component for Kicad as you saw yesterday. - Tomasz W?ostowski, aka "Tom" or "Tomek", Orson's supervisor. Tom works on hardware design most of the time, but is also developing a Push & Shove router for Kicad.
We are very excited about the opportunity to contribute to Kicad, and I would like to thank all Kicad developers (in particular the main developers: Jean-Pierre, Dick and Wayne) for bringing Kicad up to a level of quality and features where we would judge it worthwhile to invest in. We very much look forward to collaborating with you all.
In addition to the manpower I described, we have a small budget we sometimes use to get work done through companies. If browsing Kicad code you see files with a CERN copyright statement and an author either from among the main Kicad developers or another company (only Igalia [6] so far), that's the way that code was contributed. We believe one way of making Kicad sustainable and reliable for professional use is to have companies sell support contracts for Kicad, and companies which contribute code to Kicad will of course be ideally placed to sell such contracts. This should sound familiar to people who know how many of the major Linux distributions work from a financial point of view. There are probably other ways to reach these goals, and I am very interested in this topic, so please don't hesitate to contribute ideas in this domain.
Our budget for Kicad development is as I said very modest. If you think what we are doing is useful and you would like to help us work better and faster, please consider donating. CERN's Knowledge Transfer Group has set up a donations page [7], as part of a global fund-raising effort for projects which present an interest to society, even if they are not in the core mandate of CERN (Particle Physics research). There is also a link to the donations page from the CERN Kicad wiki [4]. I would also like to ask you to forward this link to whoever you think might be interested. The donations page is in a very primitive state right now. It does not allow payment through Paypal, credit cards or other methods yet. You are requested to send a message to cern.and.soci...@cern.ch and then you get bank account details which allow you to make a money transfer. This is a bit inconvenient but I can tell you any contribution at this point in time will be very useful to go faster through the roadmap. Thanks in advance for your help. We are really looking forward to working with all of you. Cheers, Javier
[1] https://indico.cern.ch/conferenceDisplay.py?confId=190126
[2] http://www.ohwr.org/
[3] http://www.ohwr.org/cernohl
[4] http://www.ohwr.org/projects/cern-kicad/wiki
[5] Referred to as "Main Authors" in file AUTHORS.txt of the Kicad sources.
[6] http://www.igalia.com/
[7] http://cernandsociety.web.cern.ch/technology/kicad-development

[ve7xen]

Great news!

[elgonzo]

Great news!

Positive, sir! 

[Bored@Work]

Now that the cat is out of the bag we have a realistic view:

CERN has one student who'll work on it for one year. Oh dear, what KiCad needs is professionals with experience, not yet more student hacks.

Further, the student's supervisor, a hardware guy, is hacking a router for it on the side. Well, a hardware guy who wants to do some software. Oh dear. And now it becomes clear why their roadmap contains such rubbish like a Python interface.

For everything else they need to find funding.

To sum it up, hardly the breakthrough that can turn KiCad around.

[marshallh]

Gonna take more than 2 programmers with spare time to fix the trainwreck of Kicad... No way I'd touch it

[firewalker]

Is it so badly "coded"?

Alexander.

[c4757p]

Is it so badly "coded"?

Alexander.

Yes, including the well-deserved quotation marks around "coded". By "programmers".

[HackedFridgeMagnet]

Gonna take more than 2 programmers with spare time to fix the trainwreck of Kicad... No way I'd touch it

I use Kicad quite happily, it is not a train wreck.

Speaking of train wrecks, last night I opened up a project in  Altium reader, I was missing some if not most of the library files required. What I got was 15 or so Altium reader processes each with about 30 modal dialogues saying missing library files. 
How annoying. I was tempted to take a screen shot, now I wish I had of.
I couldn't close it without process explorer and a lot of button pressing.

At least that sort of stuff doesn't happen with Kicad.