The new version of KiCAD is a fact.

[Bassman59]

In the real world, there are alternative workflows and quality processes that are equally effective. You know, just in case God/Bassman59 is not one of the employees.

Yes, there are alternative workflows. There are some workflows where you do some work once, up front, and reuse that work across many designs. And there are some workflows where the Very Smart Engineers who post on internet forums like to re-invent the wheel for every design. I suppose they like to re-do the work again and again because they can either bill the client for that work, or they can appear to their bosses that they're very busy indeed.

[macstr1k3r]

Hey!

been reading this thread and have gotten the impression that this info will add value to the discussion.

Most of the stuff everyone has been asking for or was not happy with is planned to be implemented/fixed

here is a link detailing the efforts and commitments by CERN
http://www.ohwr.org/projects/cern-kicad/wiki/WorkPackages

[firewalker]

BTW... 4.0.1 is out (fixes an Os X crash).

I believe that only crash fixes will merge to the 4.x series.

Alexander.

[ElektroQuark]

And Spanish translation is about 75% now

[AF6LJ]

Anyone like to talk about KiCAD?

I was thinking the same thing...
I need to get back on that microphone preamp I am designing using KiCad.

[EEVblog]

MODERATOR: I've gone and deleted and/or edited a bunch of off-topic mud slinging. Stick to the topic please.

[donotdespisethesnake]

Most of the stuff everyone has been asking for or was not happy with is planned to be implemented/fixed

here is a link detailing the efforts and commitments by CERN
http://www.ohwr.org/projects/cern-kicad/wiki/WorkPackages

The CERN roadmap is a really positive move for Kicad, it needs that sort of direction. I'm not sure how far these are commitments from CERN, more of a wishlist. I am hoping Item 10, "Improve the UI" happens sooner rather than later. The #1 bugbear of Kicad is the awkward and inconsistent user interface.

4.0.x is really more of a 3.5 beta, there are several features which are half-finished - and create even more confusion for users (e.g. you need to be in the right canvas before certain features are enabled. Of course there is no hint to the user about that). The github libraries are causing a lot of confusion for many people, which could have been easily avoided.

I am still plugging away with 4.0.x, trying to get it configured correctly and a sample project migrated but it is slow going. To get things done I am still using the "old stable" BZR 4022.

[kripton2035]

yes UI must be improved ...  no copy-paste thing in many places.
the zoom with the scrollwhell of the mouse is too fast and happening when you don't want to (at least in os x)
there are 2700 or so items in the library, that is way too small to find what you want
I'm not getting further in this version 4, I'll go back to diptrace for now.

[c4757p]

there are 2700 or so items in the library, that is way too small to find what you want

I strongly recommend not judging any EDA package by its library. Since when can you trust them anyway? I've never seen an EDA package whose library wasn't terrible - yes, KiCad included, as well as Altium, Eagle, DipTrace...

If you can't find a part you need, make it. Making parts is part of the deal with any EDA.

[tggzzz]

there are 2700 or so items in the library, that is way too small to find what you want

I strongly recommend not judging any EDA package by its library. Since when can you trust them anyway? I've never seen an EDA package whose library wasn't terrible - yes, KiCad included, as well as Altium, Eagle, DipTrace...

If you can't find a part you need, make it. Making parts is part of the deal with any EDA.

Over 3 decades ago a potential client asked my company to provide a set of simulation models for TTL/LSTTL logic. Upon asking what their acceptance criteria was, they didn't have any and would automatically accept anything we delivered. Why? Because they had a sales deal hanging on the presence of those models, and it was a case of th famous rag-trade phrase  "never mind the quality, feel the width". I declined to quote.

I've never got over the concept that many libraries are like that - unless you are paying significant money to a company dedicated to proving library parts.

[kripton2035]

I was hoping the kicad community was a rather large one.
2700 items is what I found online ready to install in kicad.
I hopped (too much may be) that a long ago installed community would have made many more items
 and that you could find the essentials parts in it.

[nctnico]

Standard libraries are highly overrated. With every CAD package I have used to far I have created my own. Standard footprints are nice and some symbols as well but in the 25 years of using CAD packages every single PCB needed at least 1 new symbol and or footprint. IOW: to get the most use out of a CAD package you have to learn how to make new symbols & footprints.

[Warhawk]

Standard libraries are highly overrated. With every CAD package I have used to far I have created my own. Standard footprints are nice and some symbols as well but in the 25 years of using CAD packages every single PCB needed at least 1 new symbol and or footprint. IOW: to get the most use out of a CAD package you have to learn how to make new symbols & footprints.

very well said 

[jancumps]

What isn't overrated though is Dave Vandenbout's KiPart utility:

http://www.xess.com/blog/giving-back-to-the-community/

I've been raving about it before. It's a great script to add components to your library - certainly when the pin count is high and the pin description can be snooped from the pdf datasheet of your component.
Took a few minutes to copy the pin info for the TDC1000 from datasheet to a spreadsheet and generate a component.

[timofonic]

It would be cool to have that kind of functionality into KiCad, by plug-in or itself.

I did read gEDA people even experimented with OCR and vectorizing to automate footprint creation. I don't have the link, but I can find it if there's some interest.

[KE5FX]

It would be cool to have that kind of functionality into KiCad, by plug-in or itself.

I did read gEDA people even experimented with OCR and vectorizing to automate footprint creation. I don't have the link, but I can find it if there's some interest.

You might be thinking of AutoBGA.  It looks pretty nifty.

[timofonic]

It would be cool to have that kind of functionality into KiCad, by plug-in or itself.

I did read gEDA people even experimented with OCR and vectorizing to automate footprint creation. I don't have the link, but I can find it if there's some interest.

You might be thinking of AutoBGA.  It looks pretty nifty.

That's not what I did see, but it seems interesting too. Unfortunately, it seems not updated since 2011 (I checked GitHub mirrors, no forks).

I meant these utils...

http://wiki.geda-project.org/geda:tragesym_tutorial
http://www.penguin.cz/~utx/pstoedit-pcb/

[bson]

KiPart is nice, an excellent time saver!  I particularly like that it can easily break the component up into blocks (power, bus, USB, analog, etc) according to schematic needs.  It also permits labeling pins with the function they will actually perform in this particular application rather than all the various functions they could potentially be assigned but aren't.

[timofonic]

what about merging KiPart into KiCad?

What about KiCad 5.0?
?

[s8548a]

But here's the thing: LM317 in a TO-220 has a different manufacturer part number from LM317 in SOT-223. So, thinking about how you'd change from the TO-220 to the SOT-223, isn't it easier to have both LM317T and LM317EMP in your library? Each component in the library knows the correct footprint AND has a part number that the purchasing people can use to buy the correct parts. Changing from one to the other, at the design level, is a simple matter of deleting the LM317T from the schematic and replacing it with LM317EMP. 

Here's a better, yet still trivial example.  There are far more complex chips that have multiple packages with different pinouts.



On the soic, the inputs are at the top and the outputs at the bottom.  On the ssop the outputs are on the right, inputs on the left and there are four nc's.  (Or, I should say uC-facing vs external-facing.)  You really want the freedom to pick one or the other during layout, because you won't know until then if there's any benefit to using the soic over the ssop.  Without context switching to the schematic.  No tool handles this adequately, but KiCAD and its manual netlist management at least permits multiple exploratory layouts: I can run two copies of pcbnew on Linux side by side, making two exploratory layouts in parallel, for the same schematic.  And it's not just pinouts, I can experimentally design for 2 and 4 layer and then see if the benefit of the latter outweighs the cost, or conversely use it to prove my assumptions.  All I need to do is generate appropriately named netlists or subdirectories for the different variants, and go to town.  In the documentation I then do a writeup of the different variants and why one was chosen over the other (more quantitatively, as in halving the number of vias, or made a 2-layer board possible within the dimensional constraints).

I am curious how is that possible? because if you're changing between SO16 and SSOP20 you will need to use 2 schematics right? I mean both IC's pin configurations are different.

To try different layouts for a schematic option,  I just renamed the finished layout by the Save as option ( a project file .pro creates automatically in that new name)

did't got a clue to run multiple Pcbnew side by side, how did you do that? am using Win7.

[Wilksey]

That is one of the things I liked about Eagle other packages, is that you can assign pins to pads separately, I think KiCAD automatically assigns pin1 of the symbol to pin1 on the footprint, not sure if they changed this on newer versions?

[PCB.Wiz]

That is one of the things I liked about Eagle other packages, is that you can assign pins to pads separately, I think KiCAD automatically assigns pin1 of the symbol to pin1 on the footprint, not sure if they changed this on newer versions?

Not sure what you mean here ? - all PCB tools I have seen, communicate via netlists, and they have the general form of RefDes.PinID
That means the PinID and RefDes have to both agree/equate between the SCH and PCB sides.
PinID can be either Number, or Alpha like E.B.C, or A1..G9 on BGA

[Wilksey]

I mean KiCAD assigns symbol "pin 1" to footprint "pin 1".  Other packages allow you to link symbol "pin1" to footprint "pin2", or if you have multiple footprints (say DIP and QFN), allows you to assign symbol "pin 1" to DIP footprint "pin 1" but symbol "pin 1" to QFN footprint "pin 2", without changing the names, so the associations change within the library not within the project.

[HackedFridgeMagnet]

IIRC Kicad does force a schematic part pin ID to a footprints ID.
But these don't necessarily have to be numbers.
eg. A Mosfet Part (component?) can have pins marked G, D, S. (instead of 1,2,3)
then the Footprints can have G, D, S too. As long as these match. Otherwise the pad doesn't get a net.

This is useful if you haven't decided whether to go SM or TH. as I think the pin numberings are different on FETs between these packages.

Obviously you cant match a component marked 1,2,3 with a footprint marked G,D,S.
Hope that makes sense.

[technotronix]

It may be 'out' but it's not actually 'out'. The download still pages still give you 4.0.0 RC2, while the documentation date back to April. Who knows when they will actually finish releasing it?

I guess it's called a release when the final commits and tags are made to the git repo.
Compiling and packaging binaries for different platforms is another process. For unix variants (e.g. Ubuntu) this can be more or less easily automated and the binaries should be out real soon I think. Windows binaries are not so straightforward I suspect and the solution is to just stop using Windows 

Stop using windows?