fund features

Comments, questions, bug reports, etc.
cadgis
Posts: 4
Joined: Wed Dec 19, 2018 1:45 pm

fund features

Post by cadgis » Wed Dec 19, 2018 1:47 pm

Hi,
is it possible to pay you to implement features we need?
Kind Regards,
the CAD/GIS-Labor der HfWU

synchronicity
Site Admin
Posts: 277
Joined: Sun Aug 02, 2015 10:31 pm

Re: fund features

Post by synchronicity » Thu Dec 20, 2018 3:25 am

No. You're free to do so yourself, of course; wimlib is open source...

cadgis
Posts: 4
Joined: Wed Dec 19, 2018 1:45 pm

Re: fund features

Post by cadgis » Thu Dec 20, 2018 1:17 pm

ok, sad, let us know if you change your mind. In theory yes, in reality, coding is a profession and it's not our job here, we just administrate the computer rooms for the students and are not allowed to do anything that is not described in our tasks descriptions... Anyway, Wimlib is already a very good program, so thanks for sharing it :)

zipmagic
Posts: 31
Joined: Thu Aug 06, 2015 7:09 am

Re: fund features

Post by zipmagic » Fri Dec 21, 2018 12:20 pm

I think what we need to do is get every interested party together in funding new features, and set up a foundation.

This is similar to the Outercurve Foundation that runs WiX, for example.

I myself am paying for an arm64 port, but the work is very slow coming, and one developer has bailed out already.

Clearly there's incredible potential in this open source library that more than one person agrees upon.

I would gladly chair the foundation :ugeek:

cadgis
Posts: 4
Joined: Wed Dec 19, 2018 1:45 pm

Re: fund features

Post by cadgis » Fri Dec 21, 2018 1:47 pm

If I understand synchronicity good, the problem is not money but will/time or whatever.

Paying another dev to implement the features is too risky:
- we want it to be of the same high quality as the rest of wimlib (fast, lightweight and reliable).
- It needs to be well tested = used widely = added to master.
How can we guarantee all this with an external dev. He/she also will need time to learn the codebase, which adds an enormous upfront cost.
From what I can see, there is only synchronicity who has commits in the repo. I don't know the state of the docs or if there are tutorials to get started with coding with wimlib. If there where any, I guess more people would be contributing as it's a great lib. So it mean an external dev would have to learn by mostly reading the code directly?
We are open to suggestion, but it should be really well thought.

zipmagic
Posts: 31
Joined: Thu Aug 06, 2015 7:09 am

Re: fund features

Post by zipmagic » Sat Dec 22, 2018 2:30 am

Yes there are risks.

If you have any other ideas, let me know?

synchronicity
Site Admin
Posts: 277
Joined: Sun Aug 02, 2015 10:31 pm

Re: fund features

Post by synchronicity » Sat Dec 22, 2018 4:43 am

Yes, I just don't have as much time for wimlib anymore, what with a full time job and more important open source projects to work on. I'd like for more developers to get involved with wimlib. But indeed, for someone to make meaningful contributions they need to actually be qualified and not just some random person hired on a freelance programmers website.

There isn't documentation for developing wimlib per se, beyond the instructions for how to build it. But I think an actually good developer, especially one experienced with C programming and filesystems, could figure most things out pretty well. I've tried to make things clean and logical, follow established conventions, and leave helpful comments.

Logistics-wise, I don't currently list an official way to submit patches for review. But if someone were actually to be doing substantial, genuinely good work I'd be glad to arrange to review and/or apply patches via email, Github, or any other way that makes sense.

Unfortunately I don't think wimlib is widely used enough for a foundation to provide adequate funding. Even open source projects that are much more essential, e.g. zlib and OpenSSL, tend to be very underfunded.

So I'm afraid that if you can't convince me to implement your feature soon enough (considering that due to limited time I have to prioritize based on how many people want the feature, how easy it is including the future maintenance cost, whether reasonable arguments exist that the feature request is misguided, and other factors), the best hope is probably to take it up yourself. And if you have questions I'd encourage you to post them here on the forums, where you might be able to get some other people involved too.

synchronicity
Site Admin
Posts: 277
Joined: Sun Aug 02, 2015 10:31 pm

Re: fund features

Post by synchronicity » Sat Dec 22, 2018 4:49 am

Also I don't believe cadgis ever stated what feature he is requesting, specifically :-)

If you explain it, it could turn out that other people are interested in it too.

Or perhaps it even already exists, just not in the way you thought...

zipmagic
Posts: 31
Joined: Thu Aug 06, 2015 7:09 am

Re: fund features

Post by zipmagic » Sat Dec 22, 2018 12:24 pm

synchronicity wrote:
Sat Dec 22, 2018 4:49 am
Also I don't believe cadgis ever stated what feature he is requesting, specifically :-)

If you explain it, it could turn out that other people are interested in it too.

Or perhaps it even already exists, just not in the way you thought...
True that!

zipmagic
Posts: 31
Joined: Thu Aug 06, 2015 7:09 am

Re: fund features

Post by zipmagic » Sat Dec 22, 2018 12:27 pm

synchronicity wrote:
Sat Dec 22, 2018 4:43 am
Unfortunately I don't think wimlib is widely used enough for a foundation to provide adequate funding. Even open source projects that are much more essential, e.g. zlib and OpenSSL, tend to be very underfunded.
Can you give us a sense of what would not be considered underfunding, for a project like wimlib? It would be great to have an idea of the range of investment required...

Post Reply