Haha, wow. Did not know that was possible. I would vote to limit those, but guess that’s another topic 
Well fine, you talked me into it, if people want those for whatever reason, no more objections from me. But probably at least flag them as unusual if it happens (@Eugene )
I like the idea of having an unlimited max time fischer, but yeah would have to be either blitz, or correpondence. With normal ‘live’ increments it could turn into days long games, which might easily cause those games to end as timeouts.
But idea of having +1d/move without cap sounds great for a correpondence player like me, it would be really nice option to have.
Also for blitz games, gaining under 20 sec per move wouldn’t still cause overwhelmingly long games, so it might work quite well.
I have one misgiving where live (and blitz) games are concerned. I like that OGS has an approach of allowing a wide range of options but:
In a real life, over a goban game, I can see if my opponent decides to invest his precious reserve of time in making a particularly careful move and I can spend that time thinking about the game too. Online I have no way of knowing if he/she is thinking or rage quitting. Do I spend an hour patiently waiting only to have them timeout or do I quit (as many would) rewarding possibly malicious behaviour.
Online is different from real-life. The cap on live fischer time limits the degree to which people can be ‘stuffed around’ by others. I can’t see any significant advantage in increasing the current 1hour limit for live-speed games but I can certainly see problems and opportunities for abuse.
I am not asking to remove the time cap option entirely. I am asking to add the option of not having it in a game.
I am not talking about correspondence games. I don’t know what makes sense there, when the players are not at the board at the same time.
I’m talking about live games, for example: 60 minutes plus 20 sec per move. It is a slow game, but it cannot take much longer than 3:40 h. If I want it faster, I reduce the basic time and the bonus time. If I want it slower, I increase them. I see no reason to additionally introduce a rule that punishes players for playing fast, and I want the option of not having it.
I think that at the very least ogs should permit games at egf class A level using fisher time. Live tournaments in the UK seem to be moving to Fischer time and for a class A tournament this means time for 120 moves of 75 minutes. That is 45min main time plus 15 seconds per move, or 55m+10s or 65m+5s. There is no cap for time accrual.
What happens now? Is there some decision process going on in the background? Or will this just be left and forgotten?
Discussions like these do need a prompt along from time to time to be not forgotten, thanks for that.
There is another problem with “options” on game settings, which is that people frequently don’t notice them. This means that people will find themselves in these long live games by accident, and Kosh’s concern becomes even more relevant.
That being said, if the discussion is largely done, and concensus hasn’t been reached that it’s a great idea, but the proponent wants to pursue it, then next things to do would be
EuG
I find it a bit irritating that the idea that a Fischer game without a cap might take a lot longer still seems to exist.
It doesn’t work that way. If I now play a 20/10 game with 60 min cap, the cap is irrelevant, as it will never ever be reached. If I play a 60/20 game, however, the site forces me to have a 60 min cap as well, and if I keep playing at a 20 s/move pace, it will punish me on average every other move. This doesn’t make sense, and it certainly doesn’t make the game faster.
If you want a faster game, use shorter time controls. If you fear that people might not realize the duration of the game, display it to them. You can calculate a rough upper bound very easily with Fischer time: for XX/YY (where XX is the basic time in minutes and YY the bonus time in seconds), the time for 300 moves in minutes is 2×XX + 5×YY. For example, a 20/10 game takes at most around 90 min =1:30 h. A cap only disturbs that calculation.
I’ll look into github.
And again, I’m not talking about correspondence games.
And Chess sites manage to have Fischer time without causing undo confusion. It’s actually easier to explain to a new player, in my opinion, than byo-yomi or Fischer with cap. You get x minutes, and every time you move, you get y more seconds.
I have raised an issue on github here: https://github.com/online-go/online-go.com/issues/839
Anoek is requesting community pressure to at least always allow longer caps than the basic time (the option of disabling the cap seems to be surprisingly hard to implement and roll out). So, please get your voice in there.
Best wishes.
I think it seems reasonable that if main and increment time can both be selected then cap should be able to be selected as well. But even if that is too much, I don’t think it’s too much to ask that cap is always at least > main time.
I don’t play Fischer controls all that often, but off the top of my head I would say the cap should be a minimum of 2*(main+increment) though I’d be fine with keeping 1 hour as the lower bound.
However, if the problem is that for functionality the cap needs to be an integer, it could always be something functionally infinite like 1000 hours?
You can start a corr game with infinite time per side. How much longer does your game need to last?
If you really wanted a reasonable pace, you wouldn’t play Fischer corr. 
Again, this is not about correspondence games, and it is not about the overall length of the game. All I actually want is a standard Fischer clock, or, as it seems the case, if the cap cannot be made optional, at least that it can be set to a value that makes it irrelevant.
I signed up to GitHub especially to support this! Hope it helps! Current setting limits don’t really make sense.
I just want to point out that one of the benefits of Fischer time is that it is better for dealing with escapers and rage-quitters. For instance, if you play a Fischer time game with a 5 minute cap, you have to wait a maximum of 5 minutes to get your victory if the opponent quits.
If I were to change anyting about Fischer time, it would be the default values (including Automatch). Right now Fischer time (2 min + 30 sec/move) already results in significantly longer games than Boyoyomi (20 min + 3x30sec), so maybe having +25 or even +20 sec/move as the default would result in better time parity.
That’s right. We have (in Europe) settled for a rule of thumb: that the bonus time for a 19×19 game should not exceed about 1/120 of the basic time. For example, typical weekend tournament settings are 40/15 (that is 40 min basic/15 s bonus time, about 2:45 h ). A reasonably fast evening game (1.5 h) can be played with 20/10.
So, I think default Blitz setting should be 5/2, default live maybe 15/7 or 20/10.
I support @Harleqin’s suggestion. Why not give people more flexibility when setting up a game? Maybe some people won’t want to use it, but what if others do?
Even if adding this option is not easy due to technical reasons as @anoek mentioned, a similar change would just be to allow a much larger maximum cap.
Personally, I could see some utility for this feature in setting up ultra-fast correspondence games. Currently, I think there is a bit of a gap between correspondence and live games that could be useful to open for some people that can maintain the pace.
I think @Harleqin was talking about a 10 second increment to get that figure. Note that even if the cap was only 10 minutes, a game could still theoretically drag on for days.
Would 4 hours be enough? That’s 1h base time + 180 * 60s.
And 35min for blitz?
btw the server doesn’t care high the max time is. One can create live games with arbitrary max time via API.
KoBa still likes the idea of having only +1d/move without any caps