Realthunders work

Have some feature requests, feedback, cool stuff to share, or want to know where FreeCAD is going? This is the place.
Forum rules
Be nice to others! Read the FreeCAD code of conduct!
freedman
Veteran
Posts: 3436
Joined: Thu Mar 22, 2018 3:02 am
Location: Washington State, USA

Realthunders work

Post by freedman »

I know there are some strong opinions about realthunders fork of FreeCAD but that's not my target, but it is open discussion after all :) .
I was kind of hoping we could find some middle ground here on the forum.

To start with could we agree that assembly3 (asm3) has taken FreeCAD to another level, it appears there are big improvements in file structure, topological error handling, linkstage, multi-object bodies, tree options, and I'm sure more that I haven't seen yet?

I have been here (on the forum) for a couple years and haven't used asm3 mostly because it was moving so fast that I didn't trust it on designs, but it appears that maybe that's changed and it has settled down. Isn't there some way we can recognize realthunders work and get it on the forum in a major way so we can see how it compares to Std-FreeCAD. I would like to see something placed at the top level, right under the Help section "FreeCADLink" and then a couple sub groups "Real problems" and "All other". I say that because I presume realthunder went off on his own so he could code stuff the way he wants. If I had his skills I would have done the same thing. I feel he deserves this, he has been working on improving FreeCAD for a few years, and he has done just that. I can see how the main members don't want to incorporate something and have the programmer disappear, he seems pretty committed.

I'm new to this code sharing thing so I though that the process would be the most powerful code always prevails or someone just branches it and makes it more powerful on their own. Standard FreeCAD doesn't seem to be moving much, at least in big improvements. No disrespect to developers, I'm sure your working hard.
I presume at some point if realthunder gets so far ahead of Std-FreeCAD that the forum will fracture and his new powerful branch will percolate to the top, maybe somewhere else. I don't want to see that happen. Is there room on this forum for a 1st and 2nd generation? Can we use asm3 and still use other assembly programs. I personally like assembly4, because I think it will do better on large projects, I don't know how that plays out.
Thanks, my 2 cents....
User avatar
yorik
Founder
Posts: 13640
Joined: Tue Feb 17, 2009 9:16 pm
Location: Brussels
Contact:

Re: Realthunders work

Post by yorik »

We all value Realthunder's work a lot and if you look at the commit history you'll see a lot of efforts from everybody (above all Werner) to integrate everything he does. So far I don't think there has ever been something he did that was rejected. But this is a really huge amount of work because most of Realthunder's changes touch dozens of files scattered all over FreeCAD, it affects a lot of hidden things, and they are often very complex for who is not an experienced coder. What's sorely missing here is people willing to help to review and thoroughly test Realthunder's pull requests. Everybody just keeps asking "why hasn't it been done already" and nobody really does anything to help (sorry for the rant :| )

Also as a note, Werner's assessment of the FreeCAD code is uttermost precious to the project, as he single-handedly basically maintains the clean structure and stability of FreeCAD. This is a blessing for FreeCAD, it is the strongest thing that permits FreeCAD to evolve well. We really should value that a lot, and try our best to remove weight from his shoulders ;)
User avatar
bernd
Veteran
Posts: 12849
Joined: Sun Sep 08, 2013 8:07 pm
Location: Zürich, Switzerland
Contact:

Re: Realthunders work

Post by bernd »

yorik wrote: Wed Nov 25, 2020 10:30 am Also as a note, Werner's assessment of the FreeCAD code is uttermost precious to the project, as he single-handedly basically maintains the clean structure and stability of FreeCAD. This is a blessing for FreeCAD, it is the strongest thing that permits FreeCAD to evolve well. We really should value that a lot, and try our best to remove weight from his shoulders
+1

Any of the bigger merges of realthunder brought a lot of improvements to FreeCAD, but they brought regressions too. These needs to be found and needs to be fixed. This is what has happened with App::Link. A great work has been done in this regard. BTW: still some minor known regressions exist.

If all improvements of realthunder would be merged fast, people would no longer complain, why his work is not merged. BUT a lot of more people would complain about a lot of problems and regressions soon.

As Yorik says, we need more people who support Werner.
User avatar
dulouie
Posts: 95
Joined: Tue Oct 03, 2017 4:14 pm

Re: Realthunders work

Post by dulouie »

I think this is the visible difference for the end user between working on new features and maintaining old code and stability. For most developers it is more exciting to implement new features than to care about the stability and security of old code.
Of course I'm happy about every single new feature in realthunders LinkStage3 branch and I use this branch actively but security and stability are also very important.

When I look at Github, I see wmayer working on Freecad every day, improving and rebuilding things. I think this is the work that many end users don't see, they only have eyes for new features.
Jee-Bee
Veteran
Posts: 2566
Joined: Tue Jun 16, 2015 10:32 am
Location: Netherlands

Re: Realthunders work

Post by Jee-Bee »

I don't think the problem that users don't see it... I think that the problem is that users read on the forum about new features that exist and are waiting to be merged. but the merging takes more than a half year (https://github.com/FreeCAD/FreeCAD/pull/2723 this one almost a year).
Specially when there are screenshots around of the assembly 3 branch that is a huge difference graphically but also what is possible.

I agree that there should be ways the regression is minimized but i also thanks that we can ask that kind of stuf to the one who created it...
The problem with this is that how longer it takes before it is added the bigger the chance that the creater is busy with other things...
On the other hand not all regression is bad... I remember that there was something with line and linesegment. i don't remember exactly what the function was... but after some time it was known and finished....
drmacro
Veteran
Posts: 8862
Joined: Sun Mar 02, 2014 4:35 pm

Re: Realthunders work

Post by drmacro »

The current stats from github for the last 30 days:

80 active pull requests.

Excluding merges, 25 authors have pushed 248 commits to master and 250 commits to all branches.
On master, 2,345 files have changed and there have been 154,683 additions and 78,459 deletions.

We should definitely throw all of realthunders code into the mix ASAP. :mrgreen:
Star Trek II: The Wrath of Khan: Spock: "...His pattern indicates two-dimensional thinking."
freedman
Veteran
Posts: 3436
Joined: Thu Mar 22, 2018 3:02 am
Location: Washington State, USA

Re: Realthunders work

Post by freedman »

yorik, What's sorely missing here is people willing to help to review and thoroughly test Realthunder's pull requests.
I think this is actually my target for this post, even though I didn't have a total plan. FreeCADLink is an unknown item and needs to be tested. The forum is where we get feedback but there is no place to give that feedback currently. A new major heading in the forum "FreeCADLink (Experimental)", and a few sub sections for help and requests could solve this unknown. Even if we only keep it in the main section for a few months, I think this would answer the question. Right now on the forum if I go into "Assembly" and find "Assembly 3 preview" there is 1764 replies, all inline. I can't learn anything from a post like that.

I think this is what would need to occur:
1) Add a specific section to the forum with sub topics.
2) Have some kind of standard release page and a new release section. This way we can all download the same release for testing and realthunder can update the new release section as necessary.

I don't know how hard it is to modify the forum layout, maybe someone will comment.

If we find that FreeCADLink isn't too bad then maybe just change the target commits to FreeCADLink if they apply.
Werner's assessment of the FreeCAD code is uttermost precious to the project, as he single-handedly basically maintains the clean structure and stability
I wish I could help, maybe realthunders code could be helpful, the only way to know is get it out front in the main stream and let folks try it and respond, I'm sure there will be much response.
Thanks
drmacro
Veteran
Posts: 8862
Joined: Sun Mar 02, 2014 4:35 pm

Re: Realthunders work

Post by drmacro »

freedman wrote: Wed Nov 25, 2020 6:48 pm ...the only way to know is get it out front in the main stream and let folks try it and respond, I'm sure there will be much response.
But, it IS readily available and there is much discussion about it here and elsewhere (gitter, IRC, discord, Facebook, Raspberry PI forums...)
(in fact realthunder is on discord daily answering questions.)

Not to mention the subject of Youtube videos.
Star Trek II: The Wrath of Khan: Spock: "...His pattern indicates two-dimensional thinking."
freedman
Veteran
Posts: 3436
Joined: Thu Mar 22, 2018 3:02 am
Location: Washington State, USA

Re: Realthunders work

Post by freedman »

I plan to do a couple designs using FreeCADLink, I don't feel like I have a place to put useful comments, like in our current Help section.

Well, I've had my say. Thanks all.
chrisb
Veteran
Posts: 53919
Joined: Tue Mar 17, 2015 9:14 am

Re: Realthunders work

Post by chrisb »

freedman wrote: Wed Nov 25, 2020 6:48 pm ...the only way to know is get it out front in the main stream and let folks try it and respond, I'm sure there will be much response.
This is definitely something where we can support Werner. But there is more he keeps an eye on, and I am glad he does: It is the quality of the code itself. It is not sufficient that it works and no errors were found by the users. It must be maintainable for years to come. And thus it has to fit into the software landscape as well as it has to follow other programmiing standards, such as proper reuse of existing classes etc. It also can be the other way round, that existing stuff has to be adapted to the new code.
A Sketcher Lecture with in-depth information is available in English, auf Deutsch, en français, en español.
Post Reply