Myracle GUI
Moderator: Ras
-
mar
- Posts: 3002
- Joined: Fri Nov 26, 2010 2:00 pm
- Location: Czech Republic
- Full name: Martin Sedlak
Re: Myracle GUI
hmm, I'll investigate. I don't like the idea of changing the schedule compared to default to maintain color balance
-
vencenda
- Posts: 47
- Joined: Thu Dec 09, 2021 8:18 pm
- Full name: Michael Titherick
Re: Myracle GUI
Only point 1 changes the schedule, right? Maybe you could make it the default, not just for "cycle colors" feature. It makes more sense to me for the BYE to be the fixed "participant".
Or you could get rid of the non-"cycle colors" behavior entirely for Round Robin, and just have one type of scheduling, that keeps the colors balanced. For Gauntlet it's a different story of course.
-
mar
- Posts: 3002
- Joined: Fri Nov 26, 2010 2:00 pm
- Location: Czech Republic
- Full name: Martin Sedlak
Re: Myracle GUI
build 131 is up:
- fix color cycling balance in RR (as was proposed here)
I will not change the default behavior, because myracle doesn't save but rather rebuild the schedule (so the change would break existing saved tournaments)
plus it's sort of standard to have bye being the last one
the schedule is only changed in RR tour with cycle colors enabled in the initial cycle pairings
two more things:
1) nobody here is entitled to insist on anything, especially not those who haven't donated
2) nobody is entitled to ask for any more features or improvements - I don't care who you (think) you are
- fix color cycling balance in RR (as was proposed here)
I will not change the default behavior, because myracle doesn't save but rather rebuild the schedule (so the change would break existing saved tournaments)
plus it's sort of standard to have bye being the last one
the schedule is only changed in RR tour with cycle colors enabled in the initial cycle pairings
two more things:
1) nobody here is entitled to insist on anything, especially not those who haven't donated
2) nobody is entitled to ask for any more features or improvements - I don't care who you (think) you are
-
mar
- Posts: 3002
- Joined: Fri Nov 26, 2010 2:00 pm
- Location: Czech Republic
- Full name: Martin Sedlak
Re: Myracle GUI
build 132 is up:
- workaround: UCI_Chess960 is not sent if it doesn't differ from default value (old Fatalii plays fine now)
- localizations (czech, spanish, french and german) including help files, pdf manual has to be regenerated from scratch too
note that the crosstable text remains english as well as unit suffixes
- new feature: if engine default options change (except for UCI_Elo which is ignored) compared to what is stored in engine config, the engine will fail to load in tournament mode and an error message is displayed.
it's necessary to go to engine settings and click detect to fix the values to actual, check hash and other things and resume the tournament.
- detect engine options behavior has changed a bit: if old value was old default, it's set to new default if default has changed
what's needed to update for this build:
- help folder
- gui_data.zip
- pdf manuals
- and of course the binary
now a bit more information about what happened and why detecting the "default option change" is important:
people often update engine by replacing the newly released engine binary in the same folder. what happened (real world example):
princhess changed some defaults for time management UCI options from version 0.20 to 0.21; this means that 0.21
(where the binary was replaced, not reinstalled) used the saved values coming from old 0.20, resulting in much faster play
and therefore losing elo. so from now on Myracle detects such changes and prevents to start an engine that has "inconsistent"
default options so that this never happens again.
Myracle is now very close to being finished, apart from a couple of minor things it's time to focus solely on bugs from now on.
also the localizations are a big maintenance burden, so one more reason to not add any new features
Thanks again to everyone who participated in this thread by providing valuable feedback, suggestions and bug reports
and also huge thanks to the few donors.
- workaround: UCI_Chess960 is not sent if it doesn't differ from default value (old Fatalii plays fine now)
- localizations (czech, spanish, french and german) including help files, pdf manual has to be regenerated from scratch too
note that the crosstable text remains english as well as unit suffixes
- new feature: if engine default options change (except for UCI_Elo which is ignored) compared to what is stored in engine config, the engine will fail to load in tournament mode and an error message is displayed.
it's necessary to go to engine settings and click detect to fix the values to actual, check hash and other things and resume the tournament.
- detect engine options behavior has changed a bit: if old value was old default, it's set to new default if default has changed
what's needed to update for this build:
- help folder
- gui_data.zip
- pdf manuals
- and of course the binary
now a bit more information about what happened and why detecting the "default option change" is important:
people often update engine by replacing the newly released engine binary in the same folder. what happened (real world example):
princhess changed some defaults for time management UCI options from version 0.20 to 0.21; this means that 0.21
(where the binary was replaced, not reinstalled) used the saved values coming from old 0.20, resulting in much faster play
and therefore losing elo. so from now on Myracle detects such changes and prevents to start an engine that has "inconsistent"
default options so that this never happens again.
Myracle is now very close to being finished, apart from a couple of minor things it's time to focus solely on bugs from now on.
also the localizations are a big maintenance burden, so one more reason to not add any new features
Thanks again to everyone who participated in this thread by providing valuable feedback, suggestions and bug reports
and also huge thanks to the few donors.