Raising money to 'Open-Source' kaya.gs

[SpongeBob]

Hi guys,

I can't stop feeling disappointed about the current state that kaya.gs is in, with no real hope for improvement. Me and many others had such high hopes to finally get the Go server we always dreamed of. Surely this hope was kind of unrealistic, as everybody has a different taste, but - hey, what would the world be without dreams!?

Just as KGS is a one-man-show where wms has built a great server, but would not implement Fisher timing, no matter how many users wanted it so badly, now Gabriel takes his decisions, no matter how many users demand something different. I cannot blame him: I am a software developer myself and I also very much like to make my decisions.

So I see two issues:
1) Decisions taken by very few (one, maybe two individuals)
2) Limited manpower for development

Here's my proposal:

Raise money for the current kaya development team so that they step back and hand the project over and 'open-source' it. If (and this is a BIG if) they would agree to do this, they really deserve this money, given that they put in a lot of hard work and energy.


This thread has two purposes:
1. See how many people would want to make an investment
2. Discuss the feasibility

Disclaimer: I have no experience with Open Source development whatsoever. So I have no idea how this all could work out.

[kivi]

Parts of Kaya is already open source. Not everything, but for example there is already time system code available, so one can implement (and test properly) and then ask Gabriel to push it to the real server. Actually there is fischer-timing code in the repository (follow the "for geeks" link on the website). I don't know if it is functional or not in the actual server.

[snorri]

Projects of this size don't need funding to open-source. You just do it.

[Bantari]

Hi guys,

I can't stop feeling disappointed about the current state that kaya.gs is in, with no real hope for improvement. Me and many others had such high hopes to finally get the Go server we always dreamed of. Surely this hope was kind of unrealistic, as everybody has a different taste, but - hey, what would the world be without dreams!?

Just as KGS is a one-man-show where wms has built a great server, but would not implement Fisher timing, no matter how many users wanted it so badly, now Gabriel takes his decisions, no matter how many users demand something different. I cannot blame him: I am a software developer myself and I also very much like to make my decisions.

So I see two issues:
1) Decisions taken by very few (one, maybe two individuals)
2) Limited manpower for development

Here's my proposal:

Raise money for the current kaya development team so that they step back and hand the project over and 'open-source' it. If (and this is a BIG if) they would agree to do this, they really deserve this money, given that they put in a lot of hard work and energy.


This thread has two purposes:
1. See how many people would want to make an investment
2. Discuss the feasibility

Disclaimer: I have no experience with Open Source development whatsoever. So I have no idea how this all could work out.

Basically, what you are saying is that we need to BUY Kaya away from its developers.

I don't like this idea. I think that, if you are serious in what you say, it is much better to start a new open-source project from scratch. I know - some of the work will have to be duplicated, but then - it will give you/us the ability to do it right and exactly the way you/we want rather than possibly be forced down development paths we don't like.

Also, ideologically - people already put money into Kaya, to see it developed. To now ask them to put money again to see it abandoned, just does not sit well with me.

So - as they say on Shark Tank - I'm out.
But that's just my 2 cents.

PS>
Didn't they have some open-source server projects out there? I seem to remember something like that a while back... where did that go?

[oren]

I actually agree with Bantari for once, I think it's better to start from scratch than try to pay to open source kaya. Why not start a new one if you want to hack on it?

[Bantari]

I actually agree with Bantari for once

You say it like its a bad thing...

[SpongeBob]

Start from scratch? There is a huge asset right there, with hundreds of man-days of development and testing already done.

[Bantari]

Start from scratch? There is a huge asset right there, with hundreds of man-days of development and testing already done.

Man hours do not necessarily translate into quality.

There might possibly be hundreds of dead-end avenues within the code which will be more trouble then they're worth to unravel or circumvent. My impression of Kaya is that it concentrated and emphasized features which are not crucial and in some sense only bells-and-whistles to make it look pretty. This might very well be reflected in how the code was written, in which case you would be buying a lemon. I am not sure if this is the case, but I would certainly like to see the code before dishing out any money for it.

If you so dead-set on having a head start, I'd say it might be better to raise the money to pay a developer to give you the head start you think you need - to your specifications.

Anyways - I said I was out, so I should probably let others talk more and me less.

[Bantari]

Just one more thing...
What makes you think an open-source project would be any successful within this community?

I mean, a lot of Go players are programmers who could contribute, but still - all I see are single-motivated-man or small-dedicated-group projects... L19, Senseis, KGS, whatever... In all these years not a single open-source project that I can see was brought to any satisfactory conclusion... and I seem to remember there were some open source projects.

To me - this is another reason not to invest into such project up-front (as in buying Kaya) - chances are this will end up with disappointment just like the actual Kaya did. Better to start from scratch, with no up-front losses, and if it does not work out - at least we will have fun for a while.

But prove me wrong.

[oren]

What makes you think an open-source project would be any successful within this community?

NNGS!

[Bantari]

What makes you think an open-source project would be any successful within this community?

NNGS!

Ohh.... yeah - forgot about that!
Good catch. Was long time ago.

Still, NNGS was not really a full-fledged community-written software but only adaptation of FICS code to Go - and that was done mostly by Eric, with a few 'authorised' people doing minor tweaks after the fact, if I remember correctly. The NNGS source was open, but I am not sure if it was a truly open-source projects like the one proposed here.

[SpongeBob]

If there would be a group of people who wanted to start such a project from scratch, it would already have happened, right?

My assumption was that things are very different, with an up and running server already available. There would be less need for synchronization of the development efforts, as all the main architecture decisions have already been taken. Now bantari seems sceptical about the quality of those decisions. Myself, I would rather think that the server should have a modern and extensible architecture, given that it was started not so long ago by two experienced developers.

[Bantari]

If there would be a group of people who wanted to start such a project from scratch, it would already have happened, right?

My assumption was that things are very different, with an up and running server already available. There would be less need for synchronization of the development efforts, as all the main architecture decisions have already been taken. Now bantari seems sceptical about the quality of those decisions. Myself, I would rather think that the server should have a modern and extensible architecture, given that it was started not so long ago by two experienced developers.

Maybe.
Try contributing here:
viewtopic.php?f=18&t=566

[oren]

It would seem funny for the community to pay developers to develop a new go server where progress ends...

and then want to pay them again for unknown quality of code so you can play with it.

[SpongeBob]

Try contributing here:
viewtopic.php?f=18&t=566

Yeah, right - the last post in this thread is three years old and http://www.owgs.org/ points to nowhere.

See, it is very hard to get a project like this off the ground ...