Arbiter development
Re: Arbiter development
My suggestion which I'm not sure whether is possible is make it possible to upk the game at a point of the blast, rather than replay the game in full.
Re: Arbiter development
I've fixed annoying .00 bug and added option to determine numer of digits after point.
- Attachments
-
- arbiter_0_48_4.zip
- (528.2 KiB) Downloaded 2256 times
Go IRC! (try mibbit)
Re: Arbiter development
Arbiter 0.48.5 (final) is ready.
- Attachments
-
- arbiter_0_48_5.zip
- (530.16 KiB) Downloaded 3166 times
Go IRC! (try mibbit)
Re: Arbiter development
I decided to release Arbiter sources. Archive below contains all sources except checksum writer.
- Attachments
-
- arbiter_0_48_5_source.zip
- (226.19 KiB) Downloaded 2590 times
Go IRC! (try mibbit)
Re: Arbiter development
Very occasionally, there seems to be a glitch where left-click is somehow registered as double-click. I'm not sure if it's me, my mouse, or Arbiter. It's only happened twice so far so I can't really say much yet about why it appears to happen.
Re: Arbiter development
I believe I have found the cause of the occurrence left-click becoming a double-click glitch. It happens when after right clicking and keeping the button held down the mouse is moved off the matrix of cells, and the button is released while the mouse is still off of the matrix. Upon putting the mouse back onto the board, the glitch stays until the clicking of the right mouse button. I think this glitch, although a major problem if one doesn't know it, can actually help make a better score if (ab)used the right way!
Re: Arbiter development
Do you have the latest version? I believed i've fixed this in 48.5.wylx wrote:I believe I have found the cause of the occurrence left-click becoming a double-click glitch. It happens when after right clicking and keeping the button held down the mouse is moved off the matrix of cells, and the button is released while the mouse is still off of the matrix.
Go IRC! (try mibbit)
Re: Arbiter development
I downloaded it from the link in the front page near the end of January...but it appears to be 48.4. Okay, so that's why—I think I'll download 48.5 now. Thanks.
Re: Arbiter development
The following is on 48.5:
Recently, I was playing Arbiter is Custom Mode when I minimized it. For some reason, when I restored Arbiter, the timer did not seem to move. When I ended the game, however, I realized that it was moving but was moving at a rate about 20x slower. Curious, I saved a replay file and played it to see what would happen, and it played the ending at the maniacal speed. This has only happened once so far.
EDIT: Apparently some part of the glitch stuck, because in all of my games in the same window after that one the timer increments with each click I make instead of each second. After I changed the board size, it stopped.
Recently, I was playing Arbiter is Custom Mode when I minimized it. For some reason, when I restored Arbiter, the timer did not seem to move. When I ended the game, however, I realized that it was moving but was moving at a rate about 20x slower. Curious, I saved a replay file and played it to see what would happen, and it played the ending at the maniacal speed. This has only happened once so far.
EDIT: Apparently some part of the glitch stuck, because in all of my games in the same window after that one the timer increments with each click I make instead of each second. After I changed the board size, it stopped.
-
- Posts: 419
- Joined: Sun Nov 30, 2008 11:50 pm
Re: Arbiter development
i'm using 48.3 but according to max this still occurs in later versions:
throughput and correctness aren't consistent if you watch a replay several time. if you right-click on the counters window to use "preview replay" and then watch the game several times the thorughput increases with every time you watch the game whereas the correctness decreases. it's not a serious bug, but those few of us who use throughput and correctness should be aware of it.
throughput and correctness aren't consistent if you watch a replay several time. if you right-click on the counters window to use "preview replay" and then watch the game several times the thorughput increases with every time you watch the game whereas the correctness decreases. it's not a serious bug, but those few of us who use throughput and correctness should be aware of it.
„Das perlt jetzt aber richtig über, ma sagn. Mach ma' noch'n Bier! Wie heißt das? Biddä! Bidddää! Biddddäää! Reiner Weltladen!“
-
- Site Admin
- Posts: 136
- Joined: Tue Jan 08, 2008 3:33 pm
- Location: UK, Scotland
- Contact:
Re: Arbiter development
Small bug? I just watched Michael Seyfried 90.00 video (find his profile and videos on the Germany ranking) and the Arbiter 48.5 video says 90.00 (90) instead of (91) on the Stats Box.
-
- Site Admin
- Posts: 136
- Joined: Tue Jan 08, 2008 3:33 pm
- Location: UK, Scotland
- Contact:
Re: Arbiter development
Just an idea I had today. It should be easy for Arbiter to notice if the mouse is on Left of Right hand setting and put 1 or 0 in the video file. It would be an easy way to collect this info about players.
Re: Arbiter development
A new version of Arbiter is released.
This is a huge security update, so you should switch to the new version as soon as possible.
This is a huge security update, so you should switch to the new version as soon as possible.
- Attachments
-
- arbiter_0_49_3.zip
- (529.72 KiB) Downloaded 2148 times
Last edited by Cryslon on Thu Mar 15, 2012 11:07 pm, edited 2 times in total.
Go IRC! (try mibbit)
Re: Arbiter development
And an update for avftools.
- Attachments
-
- avftools.zip
- (15.57 KiB) Downloaded 2204 times
Go IRC! (try mibbit)
Re: Arbiter development
Several Arbiter users experienced some lags. The most annoying lag reportedly occurred after blasts, and there were also small lags after clicking f2. A quick analysis showed that the feature causing lag is the so-called Debug Protocol, which is updated every time you win or blast a game, or hit f2. I decided to remove all interactions with the Debug Protocol occurring during normal playing.
Here's the updated version:
Here's the updated version:
- Attachments
-
- arbiter_0_49_4.zip
- Arbiter 0.49.4
- (529.68 KiB) Downloaded 2234 times
Go IRC! (try mibbit)
Re: Arbiter development
1.The zini&h.zini stats of exp games are lost in stats.csv
2.I defined QG=time^1.7/3bv ,howerver it seems "^" can't be used in counters
2.I defined QG=time^1.7/3bv ,howerver it seems "^" can't be used in counters
Re: Arbiter development
OH SHITgacord wrote:1.The zini&h.zini stats of exp games are lost in stats.csv
I don't think it is useful, but since i'm updating it, i added ^. It has the same priority as multiplication and division (not higher, because i'm too lazy to add a priority level).2.I defined QG=time^1.7/3bv ,howerver it seems "^" can't be used in counters
- Attachments
-
- arbiter_0_49_5.zip
- (529.22 KiB) Downloaded 2332 times
Last edited by Cryslon on Mon Jun 04, 2012 4:21 am, edited 1 time in total.
Go IRC! (try mibbit)
Re: Arbiter development
Great!,I just need to add "()" to ensure my counters works correctly H.Zini stats are also very useful
Re: Arbiter development
Since RQP = (Time+1)/BBBVS = (Time+1)*Time/BBBV ≈ Time^2/BBBV
I made a reserch on 48000+ exp games video data from saolei.wang
My point is to find out "Time^?/BBBV" can best show the quality of a game
And after several tests I confirmed QG = Time^1.7/BBBV can best show the quality of a game
I add this to the counters as the follows:
QG|TIMEESTREAL^1.7/BBBV{3}
If the same QG comes from a 120 BBBV game,it is easy to calculate the Time and BBBVS:
MaybeE1|(TIMEESTREAL^1.7/BBBV*120)^(1/1.7)
#|/
@|120/((TIMEESTREAL^1.7/BBBV*120)^(1/1.7))
If the same QG comes from a 210 BBBV game,it is easy to calculate the Time and BBBVS:
MaybeE2|(TIMEESTREAL^1.7/BBBV*210)^(1/1.7)
#|/
@|210/((TIMEESTREAL^1.7/BBBV*210)^(1/1.7))
I made a reserch on 48000+ exp games video data from saolei.wang
My point is to find out "Time^?/BBBV" can best show the quality of a game
And after several tests I confirmed QG = Time^1.7/BBBV can best show the quality of a game
I add this to the counters as the follows:
QG|TIMEESTREAL^1.7/BBBV{3}
If the same QG comes from a 120 BBBV game,it is easy to calculate the Time and BBBVS:
MaybeE1|(TIMEESTREAL^1.7/BBBV*120)^(1/1.7)
#|/
@|120/((TIMEESTREAL^1.7/BBBV*120)^(1/1.7))
If the same QG comes from a 210 BBBV game,it is easy to calculate the Time and BBBVS:
MaybeE2|(TIMEESTREAL^1.7/BBBV*210)^(1/1.7)
#|/
@|210/((TIMEESTREAL^1.7/BBBV*210)^(1/1.7))
Re: Arbiter development
Arbiter is the best! But... it would be really nice to check flagging/NF in the autosave config and include it in the recording file name.
-
- Site Admin
- Posts: 136
- Joined: Tue Jan 08, 2008 3:33 pm
- Location: UK, Scotland
- Contact:
Re: Arbiter development
Hi Cryslon, found a small bug maybe. In really long videos (thanks JoshK for your 960+s custom video) the Counter decimals do not always match the exact time.
For example, Anthony Clark (4372) has a 35 mines on 9x9 video of 158.13 seconds but my counter says 159.10 (the video file and the replay bar both give the correct time). In the IRC not all of us got the same time error but we all were off by a few hundreths.
Thanks for all your work!
For example, Anthony Clark (4372) has a 35 mines on 9x9 video of 158.13 seconds but my counter says 159.10 (the video file and the replay bar both give the correct time). In the IRC not all of us got the same time error but we all were off by a few hundreths.
Thanks for all your work!
Re: Arbiter development
I can confirm problems with really long videos. I was watching the same 900+s and the video was really jumpy, and the mouse movements on the screen seemed to update roughly every 1s.
Re: Arbiter development
I have 8000 games on arbiter now
A small bug : Top25 NF IOE of BEG INT EXP are infact Top25 NF IOE of EXP INT BEG(I find it as my 25EXPIOE is bigger than my EXPIOE record)
A small bug : Top25 NF IOE of BEG INT EXP are infact Top25 NF IOE of EXP INT BEG(I find it as my 25EXPIOE is bigger than my EXPIOE record)
Re: Arbiter development
fixed; wrong parameters in float to string conversion functionCounter decimals do not always match the exact time.
Really? The bug i found causes the hundredth digit to equal 0 when time exceeds 100. So everybody should have seen 158.10/159.10 on that video.not all of us got the same time error but we all were off by a few hundreths.
Funnily, the problem was caused by GUI, more specifically, by the progress bar. I removed the thing on videos longer than 200 seconds, now it seems to work nicely.I was watching the same 900+s and the video was really jumpy, and the mouse movements on the screen seemed to update roughly every 1s.
No, actually there remained an ancient piece of code which showed correctness instead of IOE.Top25 NF IOE of BEG INT EXP are infact Top25 NF IOE of EXP INT BEG
- Attachments
-
- arbiter_0_49_6.zip
- Arbiter 0.49.6
- (529.24 KiB) Downloaded 2491 times
Go IRC! (try mibbit)
Re: Arbiter development
Using 0.49.6 now and the 25NFIOE works fine!