Kaya is dead, long live OGS

[Uberdude]

Ah, I can see it now on my desktop (Win XP) but not my phone (Samsung S3 Android 4.3)

[Charlie]

I support keeping this board open as sort of a reminder if the Kaya creator ever visits here.

He took tens of thousands of dollars in donations, and he didn't even open the source before he disappeared. You would assume that everyone who supported him deserved at least that.

I am not surprised that Kaya went quietly to its death without so much as a source release. From the first, limited release it was clear that the developer(s) were building what they wanted to build without any care to what the community wanted.

The fact that they took money for it and gave us nothing is dishonest. I didn't give them money. In a way, they were quite ahead of their time; they pulled a Kickstarter before Kickstarter was a thing: disguising a blatant hand-out as an investment or pre-order.

[badukJr]

The worst part is I've heard he took a nice job in California making quite a lot of money, he could at least return some of it.

When our community is so small, something like that sucks us dry and sets us back in time as well.

[lemmata]

I sunk a couple hundred dollars into Kaya.gs despite the fact that I planned not to log on much until it gained a good user base.

I do not regret it. The lesson I re-learned was worth it: People who love go would not act the way that the founders of Kaya did. They probably loved what go does for their ego, not go itself. In the end, I was a sucker for not realizing that Kaya was just a vanity project. I don't hold ill will against the founders of Kaya. They didn't trick me at all. In retrospect, I ignored all the warning signs that were there (in blinking neon) from the beginning.

Have we been set back? Perhaps. By a lot? I think not. Programming a go server is not exactly rocket science. An intelligent comp sci undergraduate student could write a full-feature go server in one month. I doubt that we lost much because Kaya's founders did not release the source code. The lesson learned from the fiasco is more valuable than any bits of Kaya source code.

That said, I did find their illegal (at times) broadcasting of Baduk TV streams to be offensive. That was not cool when Gogameguru.com was working hard to get a legal BadukTV streaming business off the ground. That was the only thing that really pissed me off about Kaya.

[oren]

An intelligent comp sci undergraduate student could write a full-feature go server in one month.

Wow... this is REALLY not true...

[lemmata]

An intelligent comp sci undergraduate student could write a full-feature go server in one month.

Wow... this is REALLY not true...

They used to a lot harder things than that in one month back when I went to school. Before I switched full-time to more math-y stuff, I audited a graduate-level class devoted to just network programming (protocols, servers, peer-to-peer, blah-blah-blah). Half of the class consisted of undergrads who were actually enrolled (not auditing like me). For the course project, one guy made a somewhat primitive 3D FPS game+server. This took a little more than two months as I recall, but he had to take other classes while doing this so I doubt that he spent a full month on it.

So yeah, it's not as crazy as it might seem to think this.

EDIT: Of course, "full-feature" does not mean bug-free or even visually pleasing.

[dfunkt]

EDIT: Of course, "full-feature" does not mean bug-free or even visually pleasing.

It also depends on what you mean by full-feature. Some developers feel full featured means it uses the latest programming techniques as opposed to actually making go playing features that are useful to players.

[hyperpape]

http://www.yosefk.com/blog/low-level-is-easy.html

I'll just leave this here for y'all.

[Charlie]

I have absolutely no doubt that an intelligent undergrad could build 'a go server' in under a month. Heck, a brilliant 13-year-old could have one going in a week of all-nighters - complete with secret keyboard shortcut to change the mouse pointer into something obscene.

However, when embarking on a project, you need to build for longevity. This does not mean that you need to incorporate a million forward-thinking ideas like Facebook integration, live videos accompanying game broadcasts, Karma and revolutionary new challenge mechanisms. Similarly, longevity is not a magic byproduct of modern technologies and frameworks.

A grad. student or prodigy would NOT achieve longevity in a month. They might build a 'working' server and be able to demonstrate it in a University lab or on a projector but it would be a proof of concept only - it would illuminate the battlefield and reveal some of the 'unknown unknowns' that might be relevant to a properly planned Go server. I have my doubts that they'd be able to achieve true longevity in a whole year, let alone a month, simply because they would have little experience building long lived projects.

A more interesting question is this: could a brilliant, experienced, old-hand write a Go server in a month?

Full time? (Let's be mythical and redefine 'month' to be "160 man-hours"?)

Not a proof of concept? (That is, after the first month and the first release, he/she could then continue iterating and adding functionality & features?)

[badukJr]

I have visited a few websites where the layout was extremely similar to kaya, the elements and sliding things to bug you. Same size, font etc. They weren't related to go.
I think a lot of kaya was built from some template or website-in-a-box so you will never see the source released.

[Hong Ny]

Well, that's one of the reasons why I like OGS and chose it as the server of my choice. Development might be slow, but it is also steady. And I'm quite sure it will still be here in 10 years.

[Ryuukun]

I don't like OGS at all but I agree anyways. Should have a board here.