I think this is useful if there is an option to enable this.
Same information as Lizzie.

Ok, I've added something like this. Here's some sample output after leting two engines analyze a game:


If you're able to build it yourself, the code is now on github. But it's getting to the point where there is enough new stuff (the major bits this time being multi-engine analysis, slide export) to make a release, so that should happen some time this week after I've tested it a bit more.

I personally would like a different order:
Engine top choice: D5, 549 visits, winrate B: 47.14%
Game move: D6, choice #2, winrate B: 0.5495% (+7.81)
Analysis: Leela Zero 199-15x192 analysis @7.5 komi

The reason is I open these analyzed files on a mobile phone so I think the most important information should be on top (I don't want to nitpick). Like the extra choice# info.
Does it overwrite current comments? I wait for the compiled version!

Release 0.8 is now out - it includes two major new features, slide export (useful for slide shows or videos), and multi-engine analysis. Also, with engines like LZ, batch analysis can now produce reverse-komi evaluations by internally flipping the colors.
As usual, see the first post for more information.

I tested the Batch computer analysis mode V.08:
A) In which folder are the configuration files stored for the Engines? Because a new or deleted configuration file only (dis)appears when I restart Q5.
B1) The analysis function works fine with Leela 0.16. Including the Add evaluation info!
B2) But with Leela 0.17 beta (LZ_0.17-next-x64-28032019) it gets stuck in a loop at the first move (so it stays at 0%). See attachment for the log file.
C) It is not clear to me what the "If fixed engine komi" does or how it should be set-up. It needs a read me/help?

Which OS are you on? On Windows, the config file is in (I think) %APPDATA%/Local/q5go/, on Linux it should be ~/.config/q5go/.
I'm not sure what situation or actions you are describing. What exactly are you doing and what is the behaviour?

Hmm, I just tried with a build from the next branch on Linux, and it seems to work fine here. I'll try on Windows the next time I boot into it.

This is the reverse komi support. LZ only supports a fixed komi of 7.5, and you should put that in the configuration. The combo box then selects the behaviour - keep the 7.5 komi always, or switch to -7.5 if the analyzed game has reverse komi, or just analyze both.
That definitely ought to have a tooltip, which appears to be missing. I'll fix that.

This should be fixed now with version 0.8.1.

Many thanks to Adrian Petrescu, who has built the latest release on the Mac and provided a Mac binary package on his Github page.

I downloaded the binary, but when I unzipped it and tried to run it, after giving permission, I got the dreaded OSStatus=99999 error message. I have a MacBook pro with Mojave 10.14.4 and NVIDIA GeForce GT 750M 2GB.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Unfortunately it's not something I can help with - I've never owned a Mac. There are a few Google results, several of them suggesting a similar solution, but I can't tell whether those are safe to use.

Thanks, I did a search last night and nothing seemed to work better than using an earlier MacOS, or waiting for another update.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Hi!

When I tried to run q5Go in Windows 10 32bit, an alert box appear with the message: "Access Denied!"

Do you know why this error appear?

Thanks!

It's compiled for 64 bit, so I wouldn't expect this to work. I'm surprised to hear there even is a 32 bit Windows 10 version.

I've made a new 0.9 release and updated the first two posts. The major new feature is support for KataGo and its score report in analysis mode. But there is also support for additional variants (removing intersections from the board), IGS bug fixes, AI self-play and a few other new things.

A)
I would like to see a option to only analyze the first x number of moves. For example only analyse move 0-31 for Batch computer analysis.

B1)
I tried the computer vs. computer function. q5Go-0.9 stops play after the second move.
I would like to send a log file but there is no console (show messages option).
According to GPU-Z (and the coil whine ) the engine is still running in the background after the second move.
See attachment.
B2)
The noponder function is not enabled? Or do I have add this myself in the Arguments?

Thx for the update .

That sounds doable. I'll look into it.
Which engines were you trying to use? I'll admit I only tested this feature on Linux, not Windows. Does normal play against the engine work for you?
You can create a debug log by passing '-D logfilename" as an argument to the program. That might not include the necessary information, however. Maybe your engine can also create a log file? I'm pretty sure KataGo does.
That's up to you to configure in the engine set up; q5go can't really know what the engines want as arguments.

What is this?

In file included from ../src/preferences.cpp:16:
temp_/ui_preferences_gui.h: In member function ‘void Ui_PreferencesDialogGui::setupUi(QDialog*)’:
temp_/ui_preferences_gui.h:2357:54: error: ‘PlaceholderText’ is not a member of ‘QPalette’
palette.setBrush(QPalette::Active, QPalette::PlaceholderText, brush8);
^~~~~~~~~~~~~~~
temp_/ui_preferences_gui.h:2376:56: error: ‘PlaceholderText’ is not a member of ‘QPalette’
palette.setBrush(QPalette::Inactive, QPalette::PlaceholderText, brush9);
^~~~~~~~~~~~~~~
temp_/ui_preferences_gui.h:2399:56: error: ‘PlaceholderText’ is not a member of ‘QPalette’
palette.setBrush(QPalette::Disabled, QPalette::PlaceholderText, brush12);
^~~~~~~~~~~~~~~
make: *** [Makefile:2209: temp_/preferences.o] Error 1

I think I've seen this before and it seems to be some weird incompatibility between Qt versions. Some versions of Designer seem to add these things to .ui files and then it fails to compile on other versions. Looks like some instances crept in in commit cf309e41, they are quite easy to spot and those parts can safely be undone. If you don't want to try doing that yourself, I'll push out a fix as soon as possible (currently battling a dying hard disk).

Thanks for you info. I will wait for your next release. Not a big deal as I can use a previous version.

The master branch now has the update. As of now it's at the top, with no functional changes beyond 0.9. If you want the exact commit, it's 2fe4a0d.