I think the biggest problem at the moment is stability. The server has gone through periods where it is unreachable, and this has severely cramped its style. Apparently this has been caused by some third-party add-on, which supports the general sentiment that it would be good to focus on basics before bells and whistles. On the other hand, it's clear to anyone who frequents Kaya, that Gabriel is hard at work to fix whatever problems arise, and to date has consistently managed to do so relatively quickly.

As to the basics - the quibbles seem minor to me. You can play and review on Kaya, and there is a functioning rating system. Sure, negotiating games and better review tools are necessary, but they don't seem like much of a hurdle - neither to people testing out the server nor to the developer. You are probably right that putting the cool stuff on ahead of making everything solid is putting the cart in front of the horse, but I think that most people who are watching Kaya's development think that its got a pretty nice cart, and the horse is growing stronger every day. The server is updated every single week. Also, some of these features such as built in translation and malkovitch comments are powerful tools for enhancing the online go experience. Imagine a pro making malkovitch comments in Korean and everyone can read them.

Don't forget also that what people are missing atm is not a place to play - KGS is still out there - but rather a place where somebody is interested in implementing interesting features that other servers apparently aren't interested in. By putting the cart before the horse, Gabriel has demonstrated his willingness and ability to do so.



Yeah, that is a bit lame.

_________________
Patience, grasshopper.

[quote="daal"]Don't forget also that what people are missing atm is not a place to play - KGS is still out there - but rather a place where somebody is interested in implementing interesting features that other servers apparently aren't interested in. By putting the cart before the horse, Gabriel has demonstrated his willingness and ability to do so.[quote="daal"]

Yeah, good point - except for the group of people who literally cannot play on KGS because of firewall or java problems, there's no urgent need for for kaya to focus on the playing basics if they believe its more important for their success to get some of the more advanced or unique features working.

Don't get me wrong, I think the bells and whistles that have been added are amazing (and bloody useful).

The problem is that there was a surge of excitement when Kaya went Beta and that surge could have been channeled better. The server was getting pretty close to critical mass for a while, and had some of these basics been implemented, I think we would see people playing numerous games on it today.

Instead, I feel Kaya has wasted a big opportunity, and while it may get another one, I feel that folks' excitement will be tempered now by some less-than-stellar impressions the first go-round, which may slow its growth.

_________________
Tactics yes, Tact no...

You have a very decent level of control over who you play. You can pick the specific time settings and handicap of the game that is going to be played.

We thought a lot on a way to avoid the negotiation system that has a few issues:

- It makes both players wait (the offerer and the challenger) meaning that it takes longer to start a game.
- User lockup: If the game offerer is afk, the challenger sends an offer and the game doesnt start, and he waits for acceptance. So now he is wasting time. Then, a third user could take that very same challenge, and wait. Hence 2 able players are waiting for someone afk.
- Extra user required actions: there are many user actions required to start the game.
- Multiple offering: as you wait and you are qualified to take several games, asking for a game on all of those is a common practice. This is cluttering, making many simultaneous dialog windows. Also tricky, the offerer can see people coming and going because of this efect.


The elimination of negotiation means that start games is faster and easier than ever which is evident on account releases, where people enter the server and start games within seconds.

Covering all the aspects that the negotiation system on KGS does

In the future, with blacklists you can avoid specific players you dont want to play, and also we plan to put an option to play an absolute even game(with whoever accepts).

That said, the Kaya system is not strictly better than the negotiation system: there is a specific consequence. A game offerer does not get to pick among a varied group of qualified challengers. The offerer doesnt get to see 2 players of his own rank and gets to pick who he wants to play.
I dont regard that benefit as a positive one. Being able to pick means that 2 players that could play each other are waiting the game offerer.


What I do regard as the biggest con is that regular online players are not used to it. Particularly what i see happen is that people make a game offer, and they dont expect the game to start, so they go make coffee, go afk. Then when they come back, they expected to pick challengers from a list, but the game had started, and the clock got running, so they get upset.
Maybe what we should add is info in some sort of way.


There is a first version of game reviewing, which is also high on the priority list for improvement. Adding markers, and saving the comments properly on the sgf, plus better navigation, are all listed as TODO's in the next few versions.

I think you havent visited the server in a while to see the new things, shapenaji.

Lastly i agree with daal. The switch to https was rocky, and a few issues from third party software have arised frequently lately. It is my top priority for Kaya to run super smooth, as it was running 2 months ago and even better. We are wrapping up ways to dissipate the effects of the bugs on the third party libraries.

Let me just say that we have a long way to go to increase the user base. Everyone that finds something they don't like naturally think that the lack of this or that feature is the root cause.
If you compare the basic rate of accounts registered vs the number of people online in any server you will see that less than 2% of registered accounts are online at any given moment.

All feedback is appreciated and i had a great time reading all these posts. Bear in mind that Kaya improves a lot each week, its a very live project and the donations and contributions of people supporting this project has helped enormously in the development of it.

_________________
Founder of Kaya.gs

That makes no sense at all. It sounds like you have 0 control whatsoever over *who* you play.

Quick match and such is all fine, but no way to confirm an opponent is just a terrible idea.

Regarding averell's question, do you intend to offer the ability to do direct challenges, that, say, Tygem allows? Because there are definitely situations where I want to play a certain person, e.g. a friend, or a tournament opponent. The KGS system is pretty annoying, but admittedly, the Tygem situation is also annoying (getting spammed by endless strangers). I'm not clear that there is a perfect solution...

Direct challenges already exist. Still, the current system is equivalent to Automatch on KGS (for the one making the game). Selling the lack of custom game offers as a feature is just silly.

If direct offers are available, as are automatch offers, and blacklisting/censor lists, what other situations are covered by KGS-style custom games? i.e. what problems does the lengthy negotiation system of KGS solve?

You have no individual control over your opponent. For example I like to discriminate users who's name starts with vowels. Other people only want to play people with pics. FYI censoring on KGS does block those people from being automatch opponents. There is no confirmation when the game starts. If you've ever used KGS, it's just like automatch (for the one offering the game). What is being sold as a new feature, is the lack of opponent/game start confirmation.
Of course then the game starts immediately but let me just say: what is this, i don't even...

Hmm. At first I thought you were joking, but on reflection, I think you've just been a bit opaque, perhaps as a test to me I presume what you really mean is that you want to be able to avoid playing people who have (to you) offensive names or images. OK, fair enough. To me this would not be a high priority item, but I suppose for some it would be.



Agreed, the blacklist would need to be apply in all situations. Censoring and blacklists are not identical things, tho.

To be honest actually I only play automatch on KGS because I'm lazy. But it's still very simple.

Some kind of confirmation before the game starts is a good thing: Removing this does not make the game start faster. If one player isn't there to click, all that happens is that he starts with less time.

With that being said, once the confirmation button is there, if you only have "OK" and no "Cancel", then that does not represent greater freedom of choice.

I think averell tries to say, that if for example Player X 5k sets a game up, but there is just Player Y 5k who could play with him because of Player X set the hc on 0 each way, and all other 5k's are offline, that if Player Y 5k doesn't like the time settings, he can't click on Play to negotiate them and make a counter offer but has to do it in the Chat. Ofc thats a bit exaggerated, but I think that is the point, or am I on the wrong track?

My biggest issue is selecting someone of a particular rank. I like the option, in KGS, of putting up a game and when someone challenges, getting to decide "Hmm, do I want to play a 4 stone game?"

There is a little increased time in starting a game (pretty minimal, on KGS I usually get a challenger in under 30 seconds), but that is vastly better than having to play a game that I don't want to play.

That risk associated with just having a game offer up and having to play whomever shows up first makes it untenable as a place for me to play right now. (I recognize I haven't been there in a while, but this is the big reason, I'm not going to log in if I'm not going to play)

A blacklist doesn't solve my problem, because it may very well be that I would like to play a 4 stone game with that person later, but I'm not really in the mood right then.

Also, the type of game I want to play may depend on the quantity of people at a certain rank online. I may put up a game for a few minutes and have no takers at my rank, and at that point, I'll start playing handicap games. But that doesn't mean I want to play a handicap game right out of the gate.

Automatch and Direct challenges are both great features, and if people want to play that way, it should be there for them. But the default should be a game where each person gets to say whether they want to play with the other person.

_________________
Tactics yes, Tact no...

Yeah, this is another reason for a negotiating step, it gives some wiggle room between players who want to play at different time controls, without the negotiating step, they won't play at all, but it's possible that there's a time control that they both might agree on.

_________________
Tactics yes, Tact no...

_________________
PLAY GO EVERY DAY

But see, I get to decide if that's actually what I want, WHEN they challenge. I don't want to decide that before I start the game. Who knows who's on,

Maybe there are 40 5d's with itchy trigger fingers and there's no problem. On the other hand,

Maybe there's 1 septuagenarian 5d that doesn't react quickly, and 39 1d's itching for some handicap go.

Or Maybe there just aren't many players of any kind on there, and after waiting a bit I realize I'll just have to take what I can get.

In all of those cases, I'd rather be able to decide, and the rank slider won't help me.


I want the possibility of having a game with whomever is on, but I want to be able to agree to it.
Hell, I might play a 6 stone handicap game, it really depends on who challenges.

Taking away my ability to accept my challenger is a non-starter.

_________________
Tactics yes, Tact no...

Yes, thats an advantage the typical game offerer has on a negotiation system.
Of course, it is a dis-service to the challengers, that have to wait for him to make a decision.
The negative waiting is multiplied by the number of challengers that wait for a decision.

You can get exactly what you describe here if you just look at the game list. You can pick exactly who you play at your very liking that way at the moment of your choosing.

Its a benefit-balance: the game offerer doesnt get to pick the challenger exactly, but gets to pick the time settings or rules. The game challenger gets to pick the game offerer but doesnt get to pick the exact time settings he wants.

The negotiation system clearly gives the game offerer a higher advantage: maybe it would induce more players to play games, but maybe not. The certainty is that it slows and complicates the process for both offerers and challengers.

Long time ago I worked on doing the negotiation step, I've spent a lot of time building it and getting it functional, and as soon as i tried it, I felt immediately that negotiable games would be set aside. The game challenger has a lot less friction on non-negotiable games and those with the extra step would get easily neglected.
Knowing you click and play makes you avoid negotiable games that require you to wait for a possibly negative response.

Hence i looked for an alternative.

I think your concern over getting the exact game you want will dissipate as the player base grows and you have a wealth of challengers of your own rank. Negotiation will be reconsidered if after having a larger player base, the need for it is still felt.


That said, there are definite features in the works to make playing more fun and engaging.

_________________
Founder of Kaya.gs

@Kaya.gs.: Like averall said, isn't that basically the same as the automatch function? (Except slightly slower because it removes the automation.) The whole point of a custom game is the customisability. Personally, I do prefer to start playing as quickly as possible, but it seems to me that when both automatch and custom game options already exist, making one more like the other is just removing functionality.

While Gabriel`s reasoning for not allowing the picking of challengers is understandable, maybe he has overlooked a major loophole: people can still reject unwanted challengers manually. If a 5d sets up a game, and an unwanted opponent takes it, all the 5d has to do is resign without playing, as such a game won`t have any value in the ranking system (or am I mistaken?). That would cause annoyance to the challenger, who thought she or he had a game, and would take the 5d a longer time to execute than simply being able to reject a challenge straight away.

_________________
Learn the "tea-stealing" tesuji! Cho Chikun demonstrates here:

If I am going to dedicate an hour, two hours to a game, I don't mind spending a few minutes getting it right. It seems like you are making a big issue out of waiting for a negative response. My solution is to let someone challenge multiple game offerers at the same time - when someone accepts the negotiation the game starts and all other challenges are automatically cancelled. This gives the best worlds of being able to negotiate and not wasting time. Why not implement it?



This can even become worse: The challenger is annoyed so much, that when the 5d relists the game, the previous challenger immediately forces the 5d back into the game.

To be honest, I care more about WHO I play than the rules I choose. Of course if I put up a 30m main with 30s byo, I don't want a 0m main with 10s byo, but if someone suggests 20m main I will accept it. But on KGS, I always, always look at the challenger's game history to see if they have something suspicious going on. The 'no negotiation' will work as long as the beta accounts are limited and people will forgive playing whoever because they want the server to succeed. After that it will be a mess when the abusers show up.