CERN's contribution to KiCAD

[HackedFridgeMagnet]

Design rules -> Global design rules, add the width you want to the list of track widths.

Well that was easy enough, I can't really explain how I never found that before.
Thanks

I do like the look of the Open GL though and I can live with the blurry fonts.
Looks like the CERN team is actually helping. I must admit I was doubtful.

Only thing I need now is an export to Altium because some manufacturers seem to want that instead of gerbers. <kidding>

[hikariuk]

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.

Bear in mind that CERN "students" are generally PhD candidates working on the blunt end of research.

[madworm]

The KiCad developers have just merged 1 year of CERN development on KiCad.

The push & shove router is in there as well. Not perfect, but very impressive. The code compiles and I've shot a short video ;-)

[Dago]

Does anybody know a link where I could dload the testing binaries for kicad for windows with the CERN additions? I'm at work so I don't have the opportunity to compile them right now.

[madworm]

* https://code.launchpad.net/~kicad-testing-committers/kicad/testing
* http://www.ohwr.org/projects/cern-kicad/wiki/WorkPackages
* http://www.ohwr.org/projects/cern-kicad/wiki/Kicad-gal
* http://www.ohwr.org/projects/cern-kicad/files/
* https://launchpad.net/kicad-winbuilder
* http://escalalibre.com/edwt/

[minisystem]

This thread has been a fascinating read for me, an EDA newb.  I was using EagleCAD lite for small, simple boards but due to a combination of inexperience and fear, always relied on the autorouter to route boards. I graduated to KiCad because I needed to design larger boards but fell into the same trap of the autorouter (Freerouter). Finally, I got to a project where the autorouter choked and I had to start connecting traces by hand. This is a task I'm inherently not suited for and was getting ready to give up ( ) when the CERN push and shove router arrived on the scene. It's made all the difference. It's interesting to read about it's development history and common gripes among experienced users, but for an inexperienced user like me, KiCad (with CERN involvement) is looking better and better all the time. 

[dandumit]

I'm an old Kicad fan and just recently found out about push an shove.
Unfortunately I have an old laptop and it only supports OpenGL 2.1.
On Cairo , it's moving pretty slow.
Do you know what's the minimum OpenGL needed ?
BR,
Daniel

[madworm]

Currently a minimum of 2.1 is required. It does run with Intel GM945 graphics chips. Usable but not super fast.

Cairo is no joy, even with a decent NVidia card + dual core CPU.

[dandumit]

Hi,
I have an intel 965 Cipset video .
When I try to switch to OpenGL Canvas it tells me that OpenGl version 2.1 it's not supported.
I am running a precompiled version of PCBNew 4495.
Maybe newer versions are more "tolerant" to OpenGL version ?
Br,
Daniel

[madworm]

4495 is quite recent. Very unlikely.

Best guess is to check the graphics driver. I don't know what OS you run, but openGL 2.1 support was "unlocked" on my linux laptop just recently by an upgraded mesa driver. That machine is getting a bit old as far as computing power is concerned.

[dandumit]

I am running win 7 on a old  Dell Latitude  D630 .
I have just compiled Kicad 4578.
Again it's complaining that cannot run on a Open GL 2.1 .
I have the latest version of Intel Video Driver.

I'm confused. Do I need a new laptop just to have push and shove?

DAniel

[madworm]

As of now, it seems to be the case. You could try a recent (released 2nd half of this year) linux distro instead of winblows. Any serious gaming is certainly out of the question on that machine, so you might not really need the MS stuff on there.

I don't know how old your laptop is, mine is about 7-8 years old. Good enough for 'standard' kicad, barely usable for openGL. It is showing its age more and more. Unfortunately graphics chips can't be upgraded.

[Twistx77]

Hi,

Yesterday I was listening to the Amp Hour and Chris Gammel said that he have been talking to one of the KiCad developers from CERN and it seems that the KiCad "project" hasn't been approved by the CERN yet.

I suppose that it is a question of money, right?

Then you can see in the CERN's KiCad page this:

"All donations will be put to immediate use in developing Kicad. Any amount is welcome and will be gratefully received. This page will be updated regularly to show how your gifts are contributing towards our targets.
A gift of...   Would provide...
€50   1 hour of expert Kicad development
€400   1 day of expert Kicad development
Combining the hours and days of development supported by your gifts will bring the project ever closer to reaching its targets.
Reaching a target of...   Would allow developers to...
€150,000   Make Kicad usable for very complex Printed Circuit Board designs with acceptable productivity.
€300,000   Bring Kicad in line with the best proprietary tools, and outclass them in some respects. "

Anyone else think that they might be able to pull off a KickStarter campaign to get the founds needed?

I would gladly pay up to 100$ if they commit to the task.

They could offer maybe help routing boards or maybe to make components for KiCad or anything else.

What do you guys think?

 

[Frost]

What do you guys think?

I'm not sure what to think about it.
CERN is one of the worlds largest research centers
with a billion euro budget per year.

But they are not able to spend 250.000 Euro per year,
lets say for two years, to make this thing usable?
That amount of money will be peanuts compared
to their total budget.

So I think in reality it seams like there's no real interest
within the organisation to further develop this thing 

[Bored@Work]

Yesterday I was listening to the Amp Hour and Chris Gammel said that he have been talking to one of the KiCad developers from CERN and it seems that the KiCad "project" hasn't been approved by the CERN yet.

I suppose that it is a question of money, right?

Right from the beginning they mentioned they have a very tight budget for KiCad work, and that they are looking for donations. They were probably making a bit too much noise about their potential contributions, compared to the budged they have.

What do you guys think?

Without checking CERN's bylaws I would think that PCB layout software development is not part of CERN's core mission. Either they manage to make a very strong case towards CERN decision makers why KiCAD work is needed for CERN, get enough donations, or will have to cut down their involvement.

[dandumit]

Well, Cern has his buget's for different projects...

For now I am very pleased with push and spring functionality of Kicad.

If I would know how to lead this development I would offer myself for administrative parts.

I wonder if would work a Kickstarter campaign ?
It would be needed also pretty much publicity ...

Kind REgards,
Daniel

[Twistx77]

$50? get Autotrax dex
$500? get diptrace
$5K? get Altium
$1M? teach me how to open kickstarter account, i'll quit my current job
another jk.
[/quote]

Yes you can pay that for those softwares but first then you would have to pay for every update, second you can't run those in Windows, Linux and OSX and third, they are not open source which maybe for most people won't matter but hey between something close and something open, I stay with the open .

Anyway, I think that if money is the problem KickStarter could be at least a bit helpful.

I would like to work on KiCad but I've never worked in a software project so big. Does anyone know what would be the best way to start learning all the things needed to work in this kind of projects? I know C and Java but I only know how to work in small projects where there is not many people involved.

[johansen]

My advice is figure out how to get better understanding between the users and the developers.
They recently pushed an update, i forget which BZR it was but it fixed some issue where the libraries are now auto populated in the library manager... like they should be.
Only problem is I can't figure out how to include my custom libraries, which, are a mess btw, and i need a library manager that will properly organize the individual parts inside the various libraries they are bundled in.
I googled the issues (getting the new library manager to load the libraries) and came across a bunch of politics in a bug report surrounding someone else who had the same problem and "couldn't read the directions"

a friend of mine and coworker, and my boss were actually interested in using it for a few projects, rather than  Altium, for various reasons. It really does have that potential but folks; the developers are French, that's all you need to know. 

[dandumit]

... were actually interested in using it for a few projects, rather than  Altium, for various reasons. It really does have that potential but folks; the developers are French, that's all you need to know. 

Well (despite the fact that how you say it French would be a bit offending) personally I admire those French guys ! They have managed to make an open source product that's better than many paid products.  No one else in this world have done this yet.
More than that development team it's composed from people around the world.

I have never passed the step learning curve of Eagle. Personally I think that Kicad it's a far better alternative.

[johansen]

I have never passed the step learning curve of Eagle. Personally I think that Kicad it's a far better alternative.

neither did I, and there's nothing other than gEDA that's not crippled.
and yes that was a joke.

[poorchava]

I believe the CERN developer working on KiCAD is actually Polish, not French. His name's Tomasz Wlostowski I think.

Anyway - Kicad is still light years behind DipTrace, not to mention Altium.

[dandumit]

I believe the CERN developer working on KiCAD is actually Polish, not French. His name's Tomasz Wlostowski I think.

Anyway - Kicad is still light years behind DipTrace, not to mention Altium.

DipTrace Lite500 pins, 2 signal layers - $ 145

Also I didn't find push & shove on dip trace ...

[ve7xen]

My advice is figure out how to get better understanding between the users and the developers.
They recently pushed an update, i forget which BZR it was but it fixed some issue where the libraries are now auto populated in the library manager... like they should be.
Only problem is I can't figure out how to include my custom libraries, which, are a mess btw, and i need a library manager that will properly organize the individual parts inside the various libraries they are bundled in.
I googled the issues (getting the new library manager to load the libraries) and came across a bunch of politics in a bug report surrounding someone else who had the same problem and "couldn't read the directions"

a friend of mine and coworker, and my boss were actually interested in using it for a few projects, rather than  Altium, for various reasons. It really does have that potential but folks; the developers are French, that's all you need to know. 

I encountered this recently too. If you go to Pcbnew->Prefs->Library Tables you can do Append Row and put in the path to your .mod file. Set the type to Legacy. It'll then appear in the library selector, and your existing layouts should be able to find their modules.

It'll take some time and effort but this new library manager looks promising.

[dandumit]

It'll take some time and effort but this new library manager looks promising.

From which release has appeared this library manager ?

Daniel

[ve7xen]

It'll take some time and effort but this new library manager looks promising.

From which release has appeared this library manager ?

Daniel

I'm not sure in which commit it appeared. I compiled BZR 4631 a few days ago and it has it.