Yeah, it’s a real pain, sadly. Tbh, I don’t think we will ever find a major CAD company support Linux again - even Siemens, who supported NX on Linux for ages have stopped.

From my POV we have two choices: Either we make FreeCAD a viable alternative that beats the competition or at least is on the same page as them - which I find highly unlikely with the current system, so a fork+someone who finances it would be needed- or we find ways to optimise/enable Windows based CAD on Linux*. The former worked for the other tool we regularly use: QGIS. That has become the de facto standard in a lot of fields and has sometimes even pushed out commercial competition.

The later is imho the better way for CAD as it is really really hard for companies to change their CAD (even within windows and with a commercial product) - I have a business estimate for an medical product company who estimated 30k € per employee under ideal conditions, possibly more if something goes wrong(Training, loss of production, licencing, converting of files, integration of external databases,etc.). We have done it for games (tbf,with a lot of help from valve) and surely can do it with CAD (which in theory should be easier).

The last option is a bad one: In theory we could use FreeCAD as a backengine and develop themes that replicate the workflow of other products. But for that FreeCAD would need to improve on so many points beforehand…

Photoshop is a professional level software that is used by hobbyists as well - we compare affinity to this level as well and that’s okay.

So we should compare FreeCAD on this level as well. And from that perspective it’s sadly exactly what I called it.

The roughness from a commercial perspective is an issue as it costs money - because it takes people much more time to do things,even when they work.

And there are still way too many issues with it that sometimes are a result of infighting within the development community and exist for5+ years. To name a few:

• More complex imports are basically a nightmare especially with more complex facets

• Large file handling is unstable as f***. Our CAD files are commercial building size or “complex medical product” sized and despite having more than enough resources allocated FreeCAD crashes frequently without even proving any hints to the user why. The issue behind it is known for years, though.

• We had multiple issues with using older files that were saved on different OSes - really great if you can’t access files that are 16 months old. Also a known issue.

• Standardised rollout is still basically impossible.

Just to name a few… It’s simply not on the level even Solidworks has in that regards (which has it’s own issues,yes, I am on the same page with you there). While I don’t really like Siemens NC (or Solidedge for that matter) it’s indeed a reasonably good software - but me disliking them might be the result of them dropping Linux support more or less unannounced. AutoCAD and it’s sister products are imho worse than Creo,but again: More of a personal thing. In the end they sadly (!) beat FreeCAD in all aspects. By far. Which is pretty much a catastrophe as FreeCAD is the only Linux alternative atm.

But we are talking about a commercial level here - Adobe Photoshop is primarily a professional software that is also used by prosumers/hobbyists,not vice versa. We all judge e.g. Affinity on that level (rightfully).

And seen from that level FreeCAD is,well, what I said. Sure,it might do for some hobbyists and even some small companies, but even then it shows it’s massive structural flaws. Which partly, and this is why I am so openly critical of it, exist for 5+ years and are there due to the ongoing infighting in the development community.

The problem with is roughness is also a problem in terms of commercial use. When I do things as a hobbyist it’s just my time that is consumed. Not ideal,but it is what it is. In a commercial setting my staff takes more time due to this roughness and that costs money - much more money than commercial solutions cost. Which is bad - especially as it forces people to stick with Windows as there are no properly working alternatives on Linux.

And yes, onshape and fusion are horrible to hobbyists in that regard, but Solidedge(free) and to some extend Solidworks(cheap) are decent.

Sadly FreeCAD is absolutely shit compared to what commercial CAD products offer - and sadly even 1.0 didn’t change their problems.

Yeah, that would be favourable.

But: Blackberry has acquired multiple companies that deliver government grade android devices that are fully degooglefied and heavily secured. They are the de facto market leader.

Even if they bring out a properly secured and degooglefied Android that would be a huge step in the right direction.

Yeah, Blender is one of the few points where it works. QGIS is the other.

Yeah. Gaming isn’t the issue for a long time. Productivity is. Rantmode

Proper CAD for Linux? Nonexistent, even worse, some manufacturers intentionally make sure you can’t use a VM either until you massively pay extra.(Looking at you Dassault) FreeCAD is a shitshow (and that is entirely the communities fault) and no professional competitor has shown any incentive - even though there is a increasing market for Linux in some professional capacities. And the current projects to get bottles/wine/etc. to work are maintained by a single guy (bless him) who tried to do it for multiple systems at once and seems to have given up mostly.

Graphic design? While the situation is a little bit better,it’s still a shitshow. No, GIMP and Inkscape are not sufficient replacements for Adobe or even Affinity. They are “good enough” for most things,but they are not nearly ready for production use in any professional capacity.

Office? Yeah. Sadly equally bad. I really really really hate Microsoft and Office. But: They are inherently good at what they do. Not because people get used to it - but because they work. I used LibreOffice since back when it was still StarOffice. (And have used Lotus before that) But we as the open source community still rather fight about ribbons (even though they became the standard everywhere) than get LibreCalc halfway production ready or make proper collaborative working possible. Or get a proper fucking search into thunderbird.

And this is the problem: OSS is so damn up its own ass, that it does not see the bigger picture. We can fight about the kernel allowing Rust, having Ribbons, which is the proper workbench in FreeCAD or about packet managers, distro flavours,etc. In the end what will happen is that the other side will be alienated, excuse themselves from further contributions and, and this is even worse, a lot of possible future contributors will also not contribute. And wow, someone was right and can think he (and it’s almost always a he) thinks he knows the only truth.

While the actual truth is held by the others. The ones that don’t even are bothered by the whole fucking discussing because they make the money, they influence millions and they are the ones setting de facto standards. And yes, that will mean we will need to adapt.

Including adapting market standards. When 95% of the world does a thing “that way”, it’s simply preposterous to claim “your way” is the right way, even it’s for historical reasons. (Easy example: CTRL C / CTRL V)

Same goes for adapting software. If 10% of the development power of Libre Office,GIMP, etc. would have been used to further Wine/Proton to get people to be able to use their industrial standard software we would have seen much much much larger adoption rates,both professionally and for private users.

Because that is literally what happened in gaming. Once Valve basically put massive efforts into allowing Windows games to be played on Linux - and not into developing native Linux games all of a sudden Linux gaming went ahead. Because it is a advantage for your game to work natively and well on a steam deck.

This is even more relevant for production software. If a CEO/CIO has reached a point where his main production software runs on Linux and he has deployed Linux in his company his next software contract for other software will go towards the company who runs better in their environment.

Rant out

(Nothing personal,mate, I just spent the last two days to get fucking CAD to work on Fedora…)

Or if the US fucks up enough for the rest of the world to put a UNO reverse card on the US. If China and the EU do that, the US is fucked within a few months. A “you can’t trade with either of these markets when you trade with the US” would be interesting.

And tbh, from what I gather it’s absolutely a option that is being discussed in diplomatic circles. The main reason it’s not on the table is the huge amount of debts the USA has in China. And the EU will use it as a backup arrow for “further escalation”. Maybe someone should tell Trump who actually delivers the machines for the factories he wants to “bring back”. Hint: It’s not the US.

Yes,you understand how sanctions work.

A person living in a sanctioned country can also no longer buy certain things, travel to certain countries or use certain services. You couldn’t buy a Boeing Plane during WW2 in Germany as well…

And considering that Russia is waging a fucking genocidal war and a hybrid war in Europe and a majority (according to relatively independent statistics) of the population stil supports that shit and has done so for a long time (when they still could have changed course) it’s god damn right these sanctions exist.

BTW: Cuba is being sanctioned by the US for simply nationalising US held companies (Fidel Castro wasn’t that much of a communist in the beginning) since 1960 and basically none gave or gives a rats ass.

Yes, exactly. That’s why it’s preferable to use them self-hosted as the E2E of course ends on the server.

They have been,well, complicated and uncomfortable, a few years back but gradually improved. I use both and they are alright. The WhatsApp Bridge works flawlessly for me, no issues at all, the signal bridge has occasional issues that require a restart of the container (as in “once every one to three months”),but that’s more on the signal end of things. While they are not ideal they are the best option at the moment.

Matrix is also an option and heavily audited+ federated. And unlike Signal not based in the US.

And the best? You can easily selfhost a bridge to signal and WhatsApp.

And it has some options to interpret data following strict W3C standards. Which was incredibly helpful when learning, as it encouraged me (and a lot of others) to don’t go down the IE/Netscape and later Chrome “specialities” road. (Yes,I am that old…I still remember MS fucking FrontPage)