Hacker News Reader

Comment by bayesian_horse | Hacker News Reader

Just came to say FreeCAD is also really good.

Last time I tried FreeCAD it seemed like a bunch of really good tools that had kind of been thrown in a bag and shaken around a bit. All the bits were there but it just seemed so clunky. Then I tried Fusion 360 and it's like the exact same Lego pieces had been rearranged into something that made sense, with the timeline and nonlinear editing tying it all together.

Is FreeCAD different now, or was I expecting it to be something it's not, or was I just using it wrong?

>> Last time I tried FreeCAD it seemed like a bunch of really good tools that had kind of been thrown in a bag and shaken around a bit. All the bits were there but it just seemed so clunky.

That's my impression too - and it tends to crash. If you look at how FreeCAD imports DWG files for example, it runs an external instance of a DWG->DXF converter built from (I think) LibreDWG, and then it uses their other python DXF importer to bring it in. In a way it makes sense because LibreDWG is GPLv3 only (not LGPL) so they can't directly incorporate it. But the awkward way it gets used feels like the way a lot of FreeCAD is put together. OTOH it can do a LOT and it is Free Software.

It is, I also have a big soft spot for SolveSpace[1] - the feature to speed/binary size ratio is fantastic.

1: https://solvespace.com

You didn't read the link, it's a Blender plugin to use SolveSpace inside of it.

It uses the same python sketch constrainting module derived from a SolveSpace internals fork that FreeCAD also uses [1]. I'll give the parent project a go anyway, it's a cool idea.

1: https://github.com/realthunder/slvs_py

And it's also worth noting one of the SolveSpace folks is in Ukraine and his people could use our support.

https://twitter.com/app4soft/status/1499029843689852934?s=20...

That guy is... Well, he was a large contributing factor to the previous Solvespace lead maintainer quitting. I find him both useful for bug reports and some feature ideas, but he can also be one to overstep (to put it kindly). He goes under multiple user names, and it's not entirely clear if he Ukrainian or Russian - he did contribute one of those language translations, but I forget which one.

Hi, Paul. Lets clear what You just said about me:

> He goes under multiple user names.

Not true. I'm using `app4soft` almost everywhere now: https://git.io/DailySolveSpace

On GitHub my account name is "Symbian9", but nickname is `app4soft` too.

> I find him both useful for bug reports and some feature ideas, but he can also be one to overstep (to put it kindly).

Its just because I'm a human living in free country, where each citizen has a right to being criticize something in same time trying to help fix & improve it.

> it's not entirely clear if he Ukrainian or Russian.

I'm Ukrainian living in Ukraine. Here is my statement for HN:[0]

I has nothing to connect me with Russia, so do not do that anymore.

> he did contribute one of those language translations, but I forget which one.

Common man, You know that I contributing to Ukrainian locale:[1]

> Well, he was a large contributing factor to the previous SolveSpace lead maintainer quitting.

Not true: `whitequark`, being on HRT therapie, completely broke communication between bug reports & feature request contributors and main SolveSpace development branch.

Also, `whitequark` resigned from SolveSpace project itself, just after `Evil-Spirit` (Russian living in Russia) mistakenly call they "him", instead of "she/they" — `Evil-Spirit` did not know that `whiteaqurk` was on HRT and transited gender from men to girl.[2]

Also, `whitequark` is the man who near converted SolveSpace to doble-licensed open-source, far away from its inital GPLv3.[3]

Thanks to `jwesthues` (SolveSpace founder), we rolled to GPLv3, removing CLA & double-licensing.

Since then SolveSpace is FLOSS, fully Free/Libre Open-Source Software again.

Any questions?

[0] https://news.ycombinator.com/item?id=30395897

[1] https://github.com/solvespace/solvespace/commits/master/res/...

[2] http://web.archive.org/web/20201013183045/https://nitter.net...

[3] https://github.com/solvespace/solvespace/issues/1098

A few things:

> he was a large contributing factor to the previous SolveSpace lead maintainer quitting

If you look at https://github.com/solvespace/solvespace/issues/714, it refers to "a few people [who] behave in a way that is unconstructive and directly leads to maintainer burnout [...] have not encountered this level of entitlement and persistence when working on any other OSS, of which I maintain plenty, and I do not wish to encounter it ever again.". That's this guy. At the time I chose not to name him publicly nor to ban him from the organization; the latter was clearly a mistake.

(The unfortunate incident with Aleksey was the proverbial straw. I regret that it happened, since Aleksey has done a lot of good work for the project; if memory serves, at the time I was looking for other ways to fund his work.)

> who near converted SolveSpace to doble-licensed open-source, far away from its inital GPLv3.[3]

This happened after I spent several months paying Aleksey Egorov (Evil-Spirit), who did some of the best and most important work on SolveSpace, a competitive salary out of my own pocket, and it has been done in coordination with Jonathan Westhues. (You'll see his name on the CLA.) SolveSpace had a commercial licensing option before the initial open-source release, which provided a (small, but useful) revenue stream, and that seemed like a good option to continue using.

Knowing this, you can also see why Aleksey's unfortunate statement about gender felt so painful. All I wanted was to keep developing a great FOSS CAD!

Thanks for the clarifications.

Thank You, Ricardo, for understanding!

To be clear I had nothing personal against `whiteqark`[0,1] and LGBTQ in whole - they (`whitequark`) did a lot of hard work on cleaning SolveSpace core code.

But in same time they almost broke SolveSpace development management (due to their own mental health issues increased with HRT side effects[2]), in which development I'm joined in as QA since 2013.

BTW, I should clarify that `whitequark` is ex-Russian boy Peter Zotov (rus. Петр Зотов), who is actually trying blame me, Ukrainian living in Ukraine, while Russian Army kills Ukrainian kids in Ukraine, bombing cities, schools, maternity hospitals, shooting whole families in cars when they tried to escape Russia's occupation...[2,3]

[0] https://news.ycombinator.com/user?id=whitequark

[1] http://web.archive.org/web/20200216155543/http://ritconf.ru/...

[2] https://twitter.com/SolveSpace/status/1502321643229794310

[3] https://twitter.com/app4soft/status/1502340942682066946

[4] https://twitter.com/MFA_Ukraine/status/1502251070709456900

Interesting.

Just to nitpick: this is not used by the normal Sketcher workspace in FreeCAD for constraint solving. According to the link it is used for constraint solving in the assembly3 workbench, which is much more niche than Sketcher.

Sorry yes, should have clarified!

Exact package that got me back into Blender again, and discovered right here on HN too ;)

Compared to what? I can't think of a CAD program I could compare it to, and say this about!

It's really bad compared to every commercial parametric CAD program I've used (SOLIDWORKS, Pro/E, NX, Fusion 360).

It's pretty bad compared to Solvespace too.

One of those FOSS programs like KiCAD where it is full of features but they've done absolutely zero work on UX, which makes using it a frustrating "ok you don't do it the obvious way, let me spend 15 minutes googling how to extrude a square" experience.

Extruding a square is exactly the same as SolidWorks though?

You sketch a square on any plane, then extrude it (they call it the "pad" tool instead of "extrude" but otherwise it's identical).

I think the only thing FreeCAD users have to know, which SolidWorks users don't, is that there is a "Part Design" workbench where this SolidWorks-style workflow is located. There are also many other workbenches for other things like drafting, architecture, or CSG, and it's not immediately obvious which workbench you want to choose for the SolidWorks-style experience.

I've only just begun using FreeCAD (and CAD in general, especially parametric CAD), but one of the problems I've seen when sketching on a plane has been the topological naming problem[0]. I don't know how it's changed since 0.19, but basically, making changes to your model after sketching on a plane can cause the sketch to anchor to another plane instead.

I was fist introduced to this problem in a video[1] made by Flowwie, who makes some fantastic introductory videos for FreeCAD. I'd strongly suggest watching it, because this isn't something the tutorial on the wiki[2] warns you about. Flowwie goes on to suggest an alternate method of working around this problem which works by using reference measurements (measurements within a sketch that aren't specified, but calculated for use in other sketches).

[0] https://wiki.freecadweb.org/Topological_naming_problem

[1] https://youtube.com/watch?v=FVKhejma69U

[2] https://wiki.freecadweb.org/Creating_a_simple_part_with_Part...

[edit: I accidentally linked to a video I didn't mean to link to! This one is the one I intended.]

A fix is in progress -- developer realthunder has submitted a patch last year to solve the topological naming problem, and you can try it in their development branch.

>You sketch a square on any plane, then extrude it (they call it the "pad" tool instead of "extrude" but otherwise it's identical).

yeah , but that's exactly the complaint; a lexicon of terms is required to use FreeCAD whenever you're accustomed to (any) of the industry standards.

I hop between F360 and SolidWorks routinely with little issue, FreeCAD is always the odd one that requires searching.

I want to like FreeCAD; I keep trying it every year. I hate the anti-user tactic of Autodesk and I hate the cost of SolidWorks -- but it hasn't reached parity yet for my uses.

> they call it the "pad" tool instead of "extrude" but otherwise it's identical

They call it something Illogic and non-standard which means you have to Google it but otherwise it's identical!

To be honest extruding a square was meant to be an exaggerated example but it sounds like it wasn't such an exaggeration!

You don't have to google it. First of all, "pad" is not illogical, and it's arguably a more familiar word to non-native speakers than "extrude". Secondly, the tool's icon looks just like an extruded cube, so it's what you're likely to reach for first if you're just experimenting for the first time. It's also right next to the revolve tool, just like in SolidWorks. When you click on it you'll immediately see a preview of the extruded cube that you expected, so you did the right thing. And lastly, if the name of the pad tool bugs you a lot, you can submit a patch.

>> It's pretty bad compared to Solvespace too.

That sounds like a partial dig toward Solvespace. Aside from lack of features and the 90's UI look, what don't you like about it? Asking as one of the people trying to make it better.

I didn't mean to give that impression - I do really like Solvespace! It has a slightly unusual but pretty intuitive UI and the modelling actually works!

But it's still clearly in another class to the commercial software I listed. Both in terms of UI polish and features. Though in fairness it's way ahead of them in terms of being light and quick to run.

The main issues I ran into were no bevels/fillets/drafts, and the fact that small circles become octagons. Last I looked into it the latter was fixable, it just hasn't been fixed, and the former was too difficult to implement with the surface representation it uses.

I got the impression that only the guy who originally wrote it understands the actual CAD stuff, and he doesn't work on it any more so the fundamental CAD operations are probably fixed now.

>> I got the impression that only the guy who originally wrote it understands the actual CAD stuff, and he doesn't work on it any more so the fundamental CAD operations are probably fixed now.

That was true for a few years, but I've got a pretty good handle on it, as does one of the other maintainers who implemented boolean intersection. I added the helix and revolve (vs lathe) groups, and have fixed a lot of bugs in triangulation and NURBS code.

>> The main issues I ran into were no bevels/fillets/drafts, and the fact that small circles become octagons.

The octagon issue is fixed - they become 16-gons minimum, and the chord tolerance can be lowered to get more edges. If you're exporting NURBS that doesn't really mater because the actual curves are exported in STEP.

As for the other features, I had been working on that here: ]https://github.com/solvespace/solvespace/issues/453#issuecom...

Although my implementation will be strictly for the ends of extrusions and not a general purpose edge modifier tool. I also don't know when it will land.

We just keep plugging away at whatever pace we have time for ;-)

There have been some new contributors too and some of them even dig in pretty deep.

Nice! Keep up the good work.

The person you're responding to has done some great work on the fundamental CAD stuff, moving some functions to use eigan for significant performance increases. They seem to me like they've got a pretty solid grasp of the underlying CAD kernel.

I think it would be great if some other people were able to step up and help modernizing other parts of it, making distribution easier, that kind of stuff. I think phkahler has the skills to revive solvespace, but there needs to be some other core contributors who can help out with the build, UI, on-disk format, and the like.

I think the edge solvespace has over the closed-source competition is that it's actually enjoyable to use. Every time I think about sitting down to make something with solidworks, my heart sinks. On the other hand, I actually enjoy using solvespace.

Obviously, there are some lacking features, occasional jank, and (much less these days) weird pathological performance edge cases, but it's like all open source software: as it gets more polished, as more features get added, you have less and less reason to fire up the horrible UX closed source programs.

>> I think the edge solvespace has over the closed-source competition is that it's actually enjoyable to use.

That is IMHO exactly why people use it, myself included. It really encourages a sort of "play" with the model which seems to come from heavy use of the constraint solver. You never enter dimensions in a property window somewhere, you enter them right on the sketch which means your focus is almost always right there. The interactivity is very high for a CAD tool.

For me it's the orthogonality - sometimes I hit some weird thing that's really tricky to do, or surprising, but mostly, solvespace gives you a very elegant set of tools for building shapes.

That means that instead of stopping every five minutes to fiddle with all the little switches and toggles on some inane popup window you're in a flow state a lot of the time. It's like geometry, where you build complex forms through a straight edge and compass. For this reason, I think that even if solvespace was completely abandoned tomorrow, it would still be of interest to future programmers making future CAD programs.

I use Freecad everyday and I love it. Version 0.20 is very stable. I build relatively complex robot assemblies for 3D printing in it. Reminds me of solidworks.

Are you using Assembly3 or 4 or dodo? Do you ever use spreadsheet to control the whole assembly or just each part?

Assembly 3! I've never used 4 or dodo. Do you recommend?

I assume this project is about making it easier to make renderable models in Blender, not about using Blender as a true CAD/CAM tool, so it wouldn't compete with FreeCAD.

Does it have the option to lock the camera rotation around the Z axis yet like any other sane 3D viewer? Otherwise hard pass, because it's impossible to rotate the view properly.

Also consider OpenSCAD: https://files.openscad.org/examples/Basics/union.html

difference(); union(); def my_obj(p1, p2, p3) { for(...p1...) { sphere(p2); cube(p3); } }; ...etc...

All the reviews I hear about it are generally poor. That said, I hear a lot of poor reviews on open-source software, which I generally have a good experience with. I don't do any CAD, and haven't used FreeCAD personally, so I can't say for certain.

Tranlasted: "I have no knowledge of the subject you speak, but here is some hearsay information for you that I can't personally vouche for or dispell."

Love it

What was the point of this? Yes, that's what I said. And?

Reply on news.ycombinator.com