Why open hardware needs open software | Hacker News Reader

Why open hardware needs open software

blog.snapeda.com

126 pointsnatashabaker29 comments

As a mechanical engineer, I'm envious of polished open source tools that exist in other fields. For parametric modeling and engineering drawing of mechanical parts, the free software tools are not nearly as mature and user friendly as commercial equivalents. When creating engineering drawings and models for public use, I would much prefer to use free software such as FreeCAD, SolveSpace, and OpenSCAD, because I want designs to be modifiable by anyone.

Unfortunately, using these free tools feels like a throwback to a less developed era in desktop computer aided design. Commercial CAD software with no-cost hobbyist licenses are much more usable, but tie you into their licensing structure and cloud data system. There's no telling whether you'll be able to open your Fusion360 or Onshape projects 10 years from now.

I have a lot of hope that eventually free software will overtake commercial software in a lot of fields, and establish a regime where everyone has access to "professional grade" tools. Something like that exists in the information technology field, and it has unlocked a lot of human potential over the last few decades. Just imagine every school child having access to advanced, professional grade free software tools in every sort of field of human creativity.

Whenever a free CAD software exists, AutoCAD or similar buys them out.

I'm not sure the solution.

That's only really possible when there's a company behind it. It's much harder to buy out a whole OSS project, the only real option is to buy out the main contributors.

Especially like in the case of KiCad, some of the main contributors are CERN :D

This is exactly the situation in the realm of ASICs. Patent situation is a fucking minefield, so whenever a startup arises, the oligopoly smashes it to pieces and then buys out. This was actually the same for compiler technology (remember shitheads Borland?) and it was only pushed back by a large community and social traction. Unfortunately the realm of engineers (electrical, mechanical) is MUCH smaller to gain such traction to change things. So the situation is quite hopeless.

If there's some hope, it's China. Hundreds and hundreds of small shops don't care about patents and build their RISC-V based chips.

Incidentally, one of the no-cost commercial tools just got bought out by one of the big players: https://www.ptc.com/en/about/onshape

Our tools are still not that good.

I strongly agree! I often wonder why a big alliance of industrial companies (say Ford, Toyota, Airbus, Mitsubishi...) hasn't formed to create a high-performance open-source mechanical CAD, modelling software. Surely they each pay a fortune in proprietary licenses each year? It seems like it would save them a lot in the long-run to back an open-source standard and then just pay a small amount each year to support its maintenance.

"I often wonder"... because fucking "nobody was fired for buying IBM". For your average section manager of an engineering unit, CAD costs are just plain old "resources", just as engineers. As long as these costs are in check, NO ONE would risk their warm and fluffy high paid positions in undertaking such risky initiative. Even proposing one. Remember, initiative is punishable. This craeates a culuture of hairless (or grey-haired) high-paid impotents (in a figurative and literal sense) that would do as little as possible to risk their position, including taking initiatives with high rewards (like developing open source CAD tools).

Cost aside, it would do wonders for productivity if everyone was using the same tool. Right now, collaborating with customers and suppliers on design work is often difficult due to mutually incompatible data formats (even between old and new versions of the same tool).

Apart from well known KiCAD, I hope more open-source electronic design and engineering software projects will get attention from the industry. Software like LibreCAD[1] and FreeCAD[2], Qucs[3], gEDA[4], Yosys[5] and Symbiflow[6], Chisel/FIRRTL[7], OpenROAD initiative[8], Degate[9], and many others.

[1] https://librecad.org/

[2] https://www.freecadweb.org/

[3] https://github.com/Qucs

[4] http://www.geda-project.org/

[5] http://www.clifford.at/yosys/

[6] https://symbiflow.github.io/

[7] https://www.chisel-lang.org/

[8] https://theopenroadproject.org/

[9] https://github.com/nitram2342/degate

LibrePCB is an up-and-coming contender that aims to do things right from the library-management perspective: https://librepcb.org/

I am one of the FreeCAD developper (vejmarie). I believe some good innovations are coming up within the next release. I recently added some Cloud features with storage capabilities directly to s3, opening doors to build a collaborative platform (I am working on a mockup). Waevfront OBJ exporter has been improved allowing some direct export to webGL format and many other things. Regarding the features, an assembly workbench is under work, and some fancy stuff have been designed with FreeCAD (there is some Renault Twizy design files available). The FeM workbench with OpenFoam bridge allows also to initiate some simulation model. FreeCAD has in its backend a lot of capabilities that we can expose to end user. We are lacking from good feedback on the UI other than we are not yet reaching Solidworks, but FreeCAD is today a solution which works on Mac, Windows, Linux, has an open file format (Brep shape and XML representation), and is fully open source. We made tremendous progress since version 0.16 and there is a strong wish to bring in innovation regarding design workflow. Just tell us what might be your best tools and we can try to do it !

As someone who uses FreeCAD on and off, I thank you very much for your efforts! There are a lot of features.

It is definitely not lack of features that holds FreeCAD back, though. It is the UI/usability/stability. In fact, I'd prefer a solution WITHOUT cloud features as it's a liability in my profession, and in general ends up putting an expiration date on the program (as eventually servers get turned off, etc).

Again, thank you SO much for your efforts.

You are welcome. The Cloud implementation I made, is fully compatible with the minio server which is an open source implementation of s3 written in Go, meansing that you can keep everything in house and this was something super important to me. FreeCAD is a huge piece of software. Thanks for your support ! I agree with you we change too much things but since 0.17 we moved to something more modular with new object which were reuired to implement a descent assembly workbench. This is done and I do not think this is going to change that much anymore (but who knows) ;) We will do it the best as we can for sure.

I wonder if a joint hardware-software development project could benefit KiCad in a similar manner that the Open Movie Projects have helped Blender’s development.

For those who don’t know, Blender has developed a few complete animations/shorts with artists and developers working jointly, thus bringing in real world experience to the project.

I wonder if something similar could be done with an open hardware project.

A few days ago wayne did post that his position was terminated : http://kicad-pcb.org/blog/2019/10/KiCad-Lead-Developer-Annou...

This is referring to WIT's funding. As per the post:

In the short term, I will continue working on KiCad while pursuing as many avenues as possible to generate enough revenue to allow me to continue to work full time on KiCad.

Who are WIT, anyway? Their site is a placeholder: https://www.wit.com/

Archive.org can help in cases like this, e.g.: https://web.archive.org/web/20190517010832/https://www.wit.c...

That was rather abrupt, the position started in May or so?

Which also means that probably the negative effects of this on KiCAD will be rather minor, as its more of a return to how things were than anything.

The announcement on the site is dated April 30.

That's unfortunate, KiCAD is a fantastically useful piece of software and Wayne has been great for it, from the little I know about it.

I love kicad and use it daily.

One of my most wanted feature would me to be able to select a trace and get simulated electrical characteristics, like resistance, capacitance, inductance (for example to design a loop antenna, I integrated a few of them in my lastest PCB, for NFC, and had to calculate by hand)...

Of course this is not a small feature but wishing for that says a bit on how good kicad has become, there is very little small feature to wish for as most are here.

At first sight the title seems to wrong, that software is easier to be open compared to hardware.

agreed; open hardware generally guarantees the possibility of open software, even if the product ships with closed software.

however; the interview is really talking about using open software to create your open hardware designs. if you've developed your schematic using proprietary software, then only people who have the same proprietary (paid) software can contribute.

another interesting point they touched on is open instruction sets (e.g. RISC-V). people are developing open hardware platforms with open software solutions but in the middle there is ARM or some other proprietary processor instruction set.

I've been out of the EE world for a couple of years now, but looking at the screenshots of KiCAD schematic capture tool, it looks remarcably simular to Eagle (same colors, graphics, etc). Are KiCAD and Eagle related in some way?

Nope, they're completely separate. Codebase, people, timeline.

Eagle predates KiCad by a wide margin, iirc. KiCad started in France, Eagle was German (iirc). I think modern Eagle is on Qt, KiCad uses wxWidgets for it's GUI library.

Reply on news.ycombinator.com