I really hope kicad gets the blender treatment some day. It has been a dream of mine to actually not have to use altium(or eagle, or orcad) when working for an EE company.
As someone who used/attempted to use Kicad in 2009 (it was introduced as 'another option' in my electrical engineering coursework alongside Eagle - long before the latter was acquired by Autodesk), it has received the Blender treatment. It has trivial all-in-one installers for most platforms, a decent standard library, solid documentation, very few bugs... it's as easy to use as Eagle today, which was not the case in 2009.
Being taken seriously by companies, getting the big investments to increase developement speed by lots. Actually overtaking its competitors in terms of features. A big usability overhaul. Fun stuff like that.
Having worked in the EDA industry, though more on the electronics simulation side, there's a surprising amount of complexity in these tools. I think the complexity is such that it needs more than a single entity supporting it otherwise only the features that CERN directly uses will get enhanced.
Can confirm from the plugin simulation side. Integrations and import/export are important. Drafting by itself is useless to many companies without facilitating the corresponding engineering work.
Agreed! The data challenges are the biggest blocker IMO. It's so hard to know the quality of models even when received directly from manufacturers. Hopefully companies like SnapEDA can help with a lot of it though I think there needs to be some kind of model interchange standard
SnapEDA is a pox upon the industry attempting to be yet another middleman in between you and parts libraries.
Unfortunately, companies like SnapEDA bring negative value to the industry because they produce crappy libraries that you can't trust.
So, they reduce pressure on the parts vendor to produce a decent symbol/model. At the same time, I can't trust the symbol/model until I check it personally and run it through a project. And, finally, some contract terms generally prohibit me from redistributing that symbol (go check the terms you agreed to when you signed up).
I had some really bad issues using vendor supplied (TI even!) and snap parts with misnumbered pins or the wrong width package... and many bad experiences with eagle built in libraries.
I much prefer using a formula based tool to make packages for a specific part's mechanical drawing, I just don't trust what other people have made yet. I'd like to be able to but the trust isn't there yet.
Maybe kicad could list parts as verified against a standard footprint based on user feedback so that I know that whatever random part number actually fits this particular soic wide or whatever bga pattern. Maybe a part number has no notes but it's nice to know that no one has checked it manually yet too. Like the wine database.
That is great and all, but I am an EE student, and last time I looked for an internship, there are a grand total of zero actual serious EE companies that use kicad. It is pretty much all Altium rn (and god I hate using altium). So ye, what I mean by being taken seriously by companies, I am talking about more than one or two companies.
On top of the other answers: a much more friendly UI. Blender used to have a horrendously steep learning curve because the UI was just alien and ignored most agree-upon convention used by other 2D and 3D packages.
Not the case anymore, Blender is much more user-friendly and beginner-friendly.
The huge amount on online video tutorial also helps tremendously.
It was really not that bad, it just looked scary, once you got past the first steps, it was fine. But I guess those first steps made a big difference in adoption
What I tend to recommend folk do in the synth-diy community is to find existing designs like Yves Usson's yusynth stuff, and make a "cover version" of it. Draw the circuit diagram, and from that draw the PCB, trying to keep them as close as possible to the originals.
That way you're climbing the "using the tools" learning curve separately from the "laying out designs" learning curve, which makes it all a lot easier.
A lot of small things, for example editing multiple parts at the same time (select multiple parts, E, change, lets say the value of multiple resistors, enter). Or how the power ports are horrible to use (you have to create a new symbol for each new net (these should really just be some sort of global label, with a power icon). Or a better, iso standard footprint generator. Being able to have multiple sheets open at the same time. Just small things like that.
As for the UI, actually, just steal the blender UI concept, that modular window concept is great, though, if you define "user friendly" as, "The user immediately knows what to do when they open the program" idk what would help with that, it seems pretty on par with other pcb design tools in that aspect.
Altium is like the Solidworks of ECAD. Intuitive enough to pick up pretty quickly and you can get pretty far with it. It's also frequently buggy and the company is not responsive at all, if you file a bug report you're lucky if they fix it several releases down the line. Hope you paid the thousands of dollars for the support subscription.
Orcad is, I dunno the Catia? Was the "serious" software back in the day so preferred more by greybeards who scoff at using amateur stuff like Altium. Interface is much more dated and less intuitive. A company I worked for had a mild mutiny because they hired an Apple guy that insisted that everyone learn and switch to Orcad and the EE team just balked at it.
If Kicad can catch Altium on usability and handle dense, complex boards then it could be the best package out there. Just the fact that they actually fix bugs puts them at an advantage. They may already be there, I haven't played with it in a while since I'm at an Altium shop.
