The future of KGS

[Kirby]

There is a pretty substantial difference between "no KGS"/"KGS" and "HTML5 KGS"/"Java KGS"

There is also a pretty substantial difference between "HTML 5 KGS" and "no HTML 5 KGS". I'm not saying that adding a web client will be as big of a difference as making KGS in the first place, but the original argument was that he might put effort into something, and:

... And for what, exactly?

The gain from doing the web client may be less than the gain of making KGS in the first place, but that does not imply that there is no gain at all. (Unless one values the java client greater than or equal to the web client, for which I argue against.)

[Kirby]

If I install a new JRE, I have two versions on my machine that are in different directories. This is annoying to me as an end-user.

as an developer and an end-user, I find this its best feature

Why? If you have the latest JRE, supporting bytecode from earlier of the JRE, why not simply use it for all of your development and/or java use?

The only argument I can see is to allow for people using older versions of the JRE not to have to upgrade. But in that case, just use functionality supported by that JRE.

[billywoods]

three, yes three

Huh? Are you expecting me to have heard of "nova.gs" or "funnode.com"? Kaya is doing a decent job with the arbitrary bunch of features it has, and when it fills out (in basic features and in users) and people can actually find games on there, it may begin to rival KGS, but that time has not yet come. I look forward to it.

perhaps he will even welcome a credible english language alternative and can retire from the stress of running kgs.

Is that likely? He could have said "you know what, I'll just go right ahead and be awesome and make KGS open-source" and everyone would have loved him for it.

My prediction is that there's really not much to be said for KGS's long-term future. Either wms overhauls it or it will die out when any serious competition comes along. It's great for what it is, but it's complicated and grey and clunky and riddled with compatibility issues, and in a battle between two servers it will lose to anything created this decade.

[quantumf]

The gain from doing the web client may be less than the gain of making KGS in the first place, but that does not imply that there is no gain at all. (Unless one values the java client greater than or equal to the web client, for which I argue against.)

Perhaps, and presumably wms agrees with you, if javaness is to be believed.

Another way of looking at it is the consider the opportunity cost of writing a new client vs. adding much-wanted features and bug fixes to the existing client. So its a gain tradeoff.

The inability of all the html5 based go servers to dent kgs in any meaningful way is a pretty clear indication (to me) that actual users really don't care at all about the technology of the client.

[Kirby]

...

The inability of all the html5 based go servers to dent kgs in any meaningful way is a pretty clear indication (to me) that actual users really don't care at all about the technology of the client.

I think it's difficult to say that we have a "clear indication" that users don't care about the technology of the client, but we can certainly hypothesize about it. And if I were to make a hypothesis about the situation, it would be based mostly on my personal experience, and why I currently use KGS more than, for example, Kaya.

And why is that? Kaya's more accessible when I'm using my iPad, for example. And from this discussion, I think it's clear that I *do* care about the technology of the client, and it *does* affect my experience.

The answer, I feel, is the same reason that I use Facebook a lot more than Google+... It's not that I don't like the technology used, or that I don't think that the interface is good. It's because of the users. KGS has an established community analogous, in a way, to all of the friends I already had on Facebook by the time I had even heard of Google+.

Is this to say that I don't care at all about the technology of the client? No, I certainly find the user experience more pleasing on a newer platform.

The real reason I use KGS often is that many people I know are still there - TheCaptain, twoeye, BigBadWolf, and many of the people here on this forum. So that's why it's difficult to move even when the technology seems obsolete.

But that's just the magic of an html client. We can get the benefits of modern technology, and maintain the community that's already established there.

That's why I think it's a great venture to pursue.

[Charlie]

Kaya.gs failed when they opted for this interminable limited BETA phase that they are still in and failed again when they ignored the user base who were jumping up and down shouting about how cool Kaya was but how it was essentially useless because you could never get a game against another human.

They chose to focus on arbitrary features of dubious demand like correspondence Go, BadukTV streaming, score estimators and Facebook integration. All of these are cool features but none of them solved the real problem: you couldn't get proper games on the server.

Had they followed up their initial 200-account release with a desperate push to improve stability, open up account registrations and build a community, I think Kaya would be buzzing, today, and this thread would be about how people who are tired of KGS' issues are leaving for Kaya.

It wouldn't matter whether Kaya was competing with KGS or not - they would be a viable alternative and that would be good enough. In time, we would learn whether KGS could continue to compete.

[Javaness2]

there was Alex Selby's HTML server 5 years before any of these came out.

Is this GoShrine? or is that another to add to the list?

No, this is a server Alex made for the British Go Association. It was pretty nice, but it only received occasional use, I have the feeling it might vanish soon.

[Javaness2]

KGS competes with other servers (Tygem, WBaduk), which have probably taken more of the user base than Kaya (or any of the other HTML 5 projects) (Why do people on this forum always focus on Kaya?). IGS still survives because of the user base it created initially. NNGS died, as we can predict other servers will die, because of a lack of players. I don't remember if it was feature richer than IGS, but I feel it wasn't. Dragon is still there, despite other correspondence servers having better functionality.
I think most people use a Go server to play Go, to watch Go, to study Go, and to chit chat (about Go). You can argue over the order of importance of those 4 essentials, but you have to have them all in my opinion. The evidence shows that the fine details are of lesser importance.

It wouldn't matter whether Kaya was competing with KGS or not - they would be a viable alternative and that would be good enough. In time, we would learn whether KGS could continue to compete.

[PaperTiger]

Huh? Are you expecting me to have heard of "nova.gs" or "funnode.com"?

If you were looking for a KGS alternative and read either the Announcements sub-board or all new posts, then there's a good chance you could have heard of them.

Kaya is doing a decent job with the arbitrary bunch of features it has, and when it fills out (in basic features and in users) and people can actually find games on there, it may begin to rival KGS, but that time has not yet come. I look forward to it.

Anybody truly sick of KGS's lack of maintenance has the ability to play there or on any of the other servers now. There was mention in another post about Kaya being in closed beta for too long back when they had buzz, and I agree, but they recently opened up another 700 accounts and there are still plenty available.

[Pippen]

I did chose KGS because

1. it had the best looking board and stones,
2. it had a simple interface with not too many features to get lost in,
3. it provided the opportunity to chat with people on and off the game and therefore brought together Go players beyond playing their beloved game and
4. because I liked the friggin rank graph plus the rank system is not too soft or too hard (I don't like to call me a 2d when in almost all other organizations I'd be a 2k...like in AGA where a 2d is like a 1k in EGF/KGS).

As long as KGS keeps these points alive, I'll stick to it.

[jts]

Some of the discussion here verges on the edge of eccentricity. Multiple people seem to be taking the view that the primary asset of a Go server is some eclectic array of bells and whistles, and that building a better Go server is a technical problem of selecting and programming the best set of bells and whistles. The primary asset of a Go server is people. More people mean that it's easier to find games; with truly large servers you can even be picky about your time settings, rules, rank range, and so on. More people means it's easier to find people who want to chat, or who have answers to specific questions, or who want to join your Vegan Rengo Long-Cycle Ko room.

Go players are not going to weigh the timing options, graphics, and administrative policies of various servers against one another and choose the fairest of them all, they're going to go where the people are. I seriously doubt that KGS will be dethroned by a technically superior server (if the concept of an absolutely "better" server even makes sense - I'm dubious) unless and until the admins manage to alienate enough people to generate a self-perpetuating cycle of falling interest. Unless something like that happens, new servers can only make a modest contribution to the decline of KGS by poaching 50 fanatical players at a time.

[emeraldemon]

The primary asset of a Go server is people.

100% agree. This is a classic network effect.

[billywoods]

Anybody truly sick of KGS's lack of maintenance has the ability to play there or on any of the other servers now.

In theory, yes. I would love to play on kaya, but basically every time I've logged on there's been a 6d and a 19k and a 23k and everyone else is idle. (And there are no games in progress.) I've spent a reasonable supply of patience sitting on kaya waiting for games, and not found any.

Some of the discussion here verges on the edge of eccentricity.

I think you have some sort of point, but your post sounds pretty eccentric too. I'm not sure the first piece of software on the market is always the one that remains number 1 forever. (How many social networks were there before Facebook? How many millions of users did they have before they were all stolen away?) KGS has some stability because nothing else competes, but new (hence server-agnostic) western go players are still increasing steadily in number.

Let's be honest: nobody needs however many squillions of players KGS has. A couple of dozen of players of each rank is enough, and once you can play equally satisfying games equally easily on two servers, people will float around between both and eventually settle on the one that annoys them less. I'd have started lurking on go9dan if it had delivered on its promises, or kaya if there were any players. And KGS is old and grey and clunky and buggy, new users can't understand how the interface even works, the stones don't click, the SGF editing tools are primitive... you can find plenty of reasons, big or small, to give it up if there's a better option. The move is so easy because nobody (except the few hundred frequent chatters in the EGR) has any attachment to KGS itself whatsoever - they're purely there because it's the best place to play go.

[Bantari]

Let's be honest: nobody needs however many squillions of players KGS has. A couple of dozen of players of each rank is enough

This is not true.
In order to get a couple of dozen of players of each rank at any time you log in - you pretty much need a squillions of people! Especially considering the pyramide effect of the rating scheme - i.e. there is always more people of lower rank than there is of higher rank.

Think about it.

[Bantari]

Some of the discussion here verges on the edge of eccentricity. Multiple people seem to be taking the view that the primary asset of a Go server is some eclectic array of bells and whistles, and that building a better Go server is a technical problem of selecting and programming the best set of bells and whistles. The primary asset of a Go server is people. More people mean that it's easier to find games; with truly large servers you can even be picky about your time settings, rules, rank range, and so on. More people means it's easier to find people who want to chat, or who have answers to specific questions, or who want to join your Vegan Rengo Long-Cycle Ko room.

Go players are not going to weigh the timing options, graphics, and administrative policies of various servers against one another and choose the fairest of them all, they're going to go where the people are. I seriously doubt that KGS will be dethroned by a technically superior server (if the concept of an absolutely "better" server even makes sense - I'm dubious) unless and until the admins manage to alienate enough people to generate a self-perpetuating cycle of falling interest. Unless something like that happens, new servers can only make a modest contribution to the decline of KGS by poaching 50 fanatical players at a time.

I agree with the above.
The primary reason NNGS (and later KGS) had a chance to blossom was that IGS pushed away too many players due to draconian admin behavior. And still there are thousands of people playing on IGS, even though it is inferior and not always free (players in japan still have to pay, yes?)

The few complaints notwithstanding, KGS is a pretty decent server, it was good to players and it helped build communities for many years, and I seriously do not see people just abandoning it because a server with HTML client came along. Don't get me wrong - I am VERY curious about Kaya, and have big hopes for it, but... KGS is perfectly adequate to what 99% of players need. What I see in the future is some kind of balance, with people playing on KGS and Kaya, and other places, but not really abandoning nothing. Like now - people play on IGS, KGS, Tygem, and whatnot... and all servers thrive!

From the business perspective, I think people get is ass-backwards.
Its not Kaya that is the competition KGS needs to worry about, but the other way around.
KGS does not live or die by the amount of people who use it. Can the same be said about Kaya?

Which brings me to Kaya's business model. Anybody knows anything about it?