Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

1.8.0 release planning #5958

Closed
2 of 3 tasks
galad87 opened this issue Apr 16, 2024 · 46 comments
Closed
2 of 3 tasks

1.8.0 release planning #5958

galad87 opened this issue Apr 16, 2024 · 46 comments
Milestone

Comments

@galad87
Copy link
Contributor

galad87 commented Apr 16, 2024

There are some fairly large updated contribs that would be nice to have in an official release, plus the GTK4 GUI, hopefully in a few weeks.

Here's a (probably incomplete) list of major things still missing:

@Nomis101 @robxnano @sr55 @bradleysepos @jstebbins @hnsteyding

@jstebbins
Copy link
Contributor

jstebbins commented Apr 16, 2024

Can we also merge #4384 ?

I use it regularly. I think @bradleysepos was satisfied with my selection logic. It can be later tweaked if we find corner cases, but I think it's good as it is.

☑️ Done

@jstebbins
Copy link
Contributor

jstebbins commented Apr 16, 2024

This also would be really nice to have #5922

☑️ Done

@jstebbins
Copy link
Contributor

jstebbins commented Apr 16, 2024

And hopefully I can look into this issue before release #5948

☑️ Done

@bradleysepos
Copy link
Contributor

I think @bradleysepos was satisfied with my selection logic.

I believe so after re-reading the PR. Though I cannot recall testing it directly, I trust it does what you say it does.

@Nomis101
Copy link
Contributor

I can make PRs to update all translations, once it is sure that no more new strings will go into 1.8.0.

@sr55
Copy link
Contributor

sr55 commented Apr 16, 2024

Probably a good idea to make a note that this is coming on transifex again.

@Nomis101
Copy link
Contributor

Nomis101 commented Apr 16, 2024

Would somebody be able to update ghb.pot for gtk, so all the new strings introduced recently will be visible in Transifex, please? This does not work anymore for me, since this process moved to meson.

☑️ Done

@robxnano
Copy link
Contributor

@Nomis101 I was actually just about to do that. I made a pull request: #5959

@Nomis101
Copy link
Contributor

Probably a good idea to make a note that this is coming on transifex again.

I did so.

@hnsteyding
Copy link
Contributor

Maybe update harfbuzz to 8.4.0 ? #5965

@jstebbins
Copy link
Contributor

Would somebody be able to update ghb.pot for gtk, so all the new strings introduced recently will be visible in Transifex, please? This does not work anymore for me, since this process moved to meson.

I'm not sure how this is meant to be run. I attempted:

cd build/gtk
ninja ghb-pot

And this generated a new gtk/po/ghb.pot but something isn't correct. It substituted:

-# Copyright (C) 2024 HandBrake Team
+# Copyright (C) YEAR HandBrake Team

So I must be missing an option or configuration somewhere.
@robxnano what's the correct way to do this?

@robxnano
Copy link
Contributor

Oh, I just edited the year in myself. I don't think there's any way to set the year (or other variables) with xgettext, they're just supposed to be manually added to the .po files by the translators.

@robxnano
Copy link
Contributor

What about updating the Flatpak runtime to Gnome 46? It builds fine, just need to check that encoding is working.

@jstebbins
Copy link
Contributor

What about updating the Flatpak runtime to Gnome 46? It builds fine, just need to check that encoding is working.

Yes please.

@sr55
Copy link
Contributor

sr55 commented Apr 26, 2024

Shall we code freeze soon ? :)

@jstebbins
Copy link
Contributor

Whenever you say is fine by me. I've just been picking some low hanging fruit from our issues list to fix. I can stop at any time 😰 (says the addict).

@sr55
Copy link
Contributor

sr55 commented Apr 26, 2024

Pretend I said nothing ;)

@Nomis101
Copy link
Contributor

Thanks for merging all the localization PRs. This will give the localizers the opportunity to check how the strings look in the UI. I have noticed there is still some activity ongoing for at least one or two locales to finish the remaining strings, so I plan to do PRs for this new strings beginning or middle of next week (if thats OK).

@sr55
Copy link
Contributor

sr55 commented Apr 26, 2024

Sounds good.

@galad87
Copy link
Contributor Author

galad87 commented Apr 30, 2024

Is anything missing from the news?

@sr55
Copy link
Contributor

sr55 commented Apr 30, 2024

Looks like we got the bulk of it. Maybe if @bradleysepos has some time to review too?

@sr55 sr55 added this to the 1.8.0 milestone Apr 30, 2024
@bradleysepos
Copy link
Contributor

I should be able to make some time this week.

@Nomis101
Copy link
Contributor

Nomis101 commented May 3, 2024

There are some new (unapproved) patches for x265 regarding AArch64. I am not an expert in this, but at least this two might be useful for us?
https://mailman.videolan.org/pipermail/x265-devel/2024-May/013667.html
https://mailman.videolan.org/pipermail/x265-devel/2024-May/013672.html
For the other 10, I am not sure.

@jstebbins
Copy link
Contributor

FYI, I'm traveling until next Monday. My time will be limited until then.

@galad87
Copy link
Contributor Author

galad87 commented May 6, 2024

@Nomis101 the first looks like only for arm64e, but we compile for the normal arm64 arch, the second probably should be useful, but maybe we can integrate the whole set after they merge it, in the next minor release.

SVT-AV1 2.1 should be out in a week or so, let's wait a bit so we don't ship with an out of date version.

@Nomis101
Copy link
Contributor

Nomis101 commented May 6, 2024

@Nomis101 the first looks like only for arm64e, but we compile for the normal arm64 arch, the second probably should be useful, but maybe we can integrate the whole set after they merge it, in the next minor release.

Thanks for looking into it. This is of course also a good idea. It might just take a while until there is a new x265 release, 3.6 has been 3 years in the making.

@galinart
Copy link
Contributor

galinart commented May 8, 2024

Sorry folks,

  • Nightly build GUI binary crashes on Windows with 2 Intel GPUs: iGPU + dGPU after selection of H.265 QSV 2160p or AV1 QSV HW presets.
  • Nightly CLI binary works fine.
  • 1.7.3 GUI also works.

Seems this is regression since we have not contribute anything since 1.7.3 release. I will try to know more details tomorrow.

@sr55 any thoughts?

@galinart
Copy link
Contributor

galinart commented May 9, 2024

Found breaking change 069b785

Reproducible only via GUI
Call stack:
image

@galad87 are you able to reproduce the crash? Preview function was modified 4366cd3#diff-574ebf6c5d95e912886b84976f946f09a74d7c04c72be275182bafff202f14eaR920

@galad87
Copy link
Contributor Author

galad87 commented May 9, 2024

I don't any have QSV hardware ready to use at the moment. I wonder if it's trying to use vpp in the preview format filter for some reason? It would be useful to see where it's crashing in libhb.

@galad87
Copy link
Contributor Author

galad87 commented May 9, 2024

Does it happen with a specific source file or with everything?

@sr55
Copy link
Contributor

sr55 commented May 9, 2024

I'll try reproduce it shortly to and see if I can't get a stack trace

@sr55
Copy link
Contributor

sr55 commented May 9, 2024

Confirmed. It happens specifically when QSV decode is turned on. No Crash if QSV decode is disabled in preferences.

@sr55
Copy link
Contributor

sr55 commented May 9, 2024

No joy with GDB :(

@galad87
Copy link
Contributor Author

galad87 commented May 9, 2024

It would be nice to switch to the clang toolchain to be able to debug libhb inside visual studio.
Anyway, try to disable qsv decode in with something like job->qsv.decode = 0 in hb_get_preview3. Before there was an option to set the format in cropscale.c to convert the previews to RGB, now it uses the format filter, and VPP might get enabled by mistake.

@sr55
Copy link
Contributor

sr55 commented May 9, 2024

I've disabled all hardware decoding during previews (including other vendors). Builds running. Lets see if this helps.

If I disable static previews on the screen in the mean time, it appears to work, fill path, hyperencode enabled so it looks like the encode path is fine.

@galinart
Copy link
Contributor

No joy with GDB :(

yes, it is challenging to get underlying hb.dll call stack

@galinart
Copy link
Contributor

galinart commented May 10, 2024

I've disabled all hardware decoding during previews (including other vendors). Builds running. Lets see if this helps.

If I disable static previews on the screen in the mean time, it appears to work, fill path, hyperencode enabled so it looks like the encode path is fine.

The new build works fine. No crashes have been observed. Thanks for the fix.

@galinart
Copy link
Contributor

Another issue, not related to QSV, has been found

On some machines, not everywhere, I see connecting issue to the Worker process. Disabling process isolation fixes the issue.

HandBrake Nightly 20240509181016-90173af8c-master (2024051001)
OS: Microsoft Windows NT 10.0.22631.0
CPU: 13th Gen Intel(R) Core(TM) i9-13900K
Ram: 15949 MB, 
GPU Information:
  Intel(R) UHD Graphics 770 - 31.0.101.5445
  Intel(R) Arc(TM) A770 Graphics - 31.0.101.5445
Screen: 1920x1080
Temp Dir: C:\Users\intel\AppData\Local\Temp\
Install Dir: C:\Program Files\HandBrake Nightly\
Data Dir: C:\Users\intel\AppData\Roaming\HandBrake\Nightly

-------------------------------------------

 # Starting Encode ...

[12:33:57] base preset: H.265 MKV 2160p60 4K
[12:33:57] Remote Process started with Process ID: 14648 using port: 8044. Max Allowed Instances: 1
[12:33:57] Worker: Starting HandBrake Engine ...
[12:33:57] Worker: Parent Process Id 18408
[12:33:57] Worker: Starting Web Server on port 8044 ...
[12:33:57] Worker: Starting Listener: 1
[12:33:57] Worker: Server Started
Unable to connect to the HandBrake Worker instance after 10 attempts. Try disabling this option in Tools -> Preferences -> Advanced.

 # Job Failed (-10)

@sr55
Copy link
Contributor

sr55 commented May 10, 2024

I've pushed another build with some debug logging.
It looks like the worker process is started up ok and bound to the port so it appears that we just can't talk to it.

Two most likely possibilities is security software / firewall interfering with it or potentially a broken system dll file.

Hopefully the new logging will give us a hint.

Snapshot builds running now.

@galinart
Copy link
Contributor

I have tested new build. No detailed logs appear.

HandBrake Nightly 20240510112631-2860a8ebd-master (2024051001)
OS: Microsoft Windows NT 10.0.22631.0
CPU: 13th Gen Intel(R) Core(TM) i9-13900K
Ram: 15949 MB, 
GPU Information:
  Intel(R) UHD Graphics 770 - 31.0.101.5445
  Intel(R) Arc(TM) A770 Graphics - 31.0.101.5445
Screen: 1920x1080
Temp Dir: C:\Users\intel\AppData\Local\Temp\
Install Dir: C:\Program Files\HandBrake Nightly\
Data Dir: C:\Users\intel\AppData\Roaming\HandBrake\Nightly

-------------------------------------------

 # Starting Encode ...

[14:21:33] base preset: Fast 1080p30
[14:21:33] Remote Process started with Process ID: 10748 using port: 8038. Max Allowed Instances: 1
[14:21:33] Worker: Starting HandBrake Engine ...
[14:21:33] Worker: Parent Process Id 14568
[14:21:33] Worker: Starting Web Server on port 8038 ...
[14:21:33] Worker: Starting Listener: 1
[14:21:33] Worker: Server Started
Unable to connect to the HandBrake Worker instance after 10 attempts. Try disabling this option in Tools -> Preferences -> Advanced.

 # Job Failed (-10)

@galinart
Copy link
Contributor

Another observation, in case if worker failed and connection is not successful, it starts to be zombie process (Task Manager -> Background processes) even if I manually delete the task in Queue.

image

@sr55
Copy link
Contributor

sr55 commented May 10, 2024

There is an idle shutdown in the worker, but it won't activate until the GUI connects. I guess I may need to extend that to shutdown if we don't get a connection after a minute or so.

@sr55
Copy link
Contributor

sr55 commented May 10, 2024

I've added more logging since it appears to be a non-fatal error that's occurring which blows my theories out the water I suspect.

I've also hopefully fixed the stale process issue.

New snapshot will be available in about an hour.

@galinart
Copy link
Contributor

Stale process is fixed.

The detailed log is available now.

HandBrake Nightly 20240510142627-887494df5-master (2024051001)
OS: Microsoft Windows NT 10.0.22631.0
CPU: 13th Gen Intel(R) Core(TM) i9-13900K
Ram: 15949 MB, 
GPU Information:
  Intel(R) UHD Graphics 770 - 31.0.101.5445
  Intel(R) Arc(TM) A770 Graphics - 31.0.101.5445
Screen: 1920x1080
Temp Dir: C:\Users\intel\AppData\Local\Temp\
Install Dir: C:\Program Files\HandBrake Nightly\
Data Dir: C:\Users\intel\AppData\Roaming\HandBrake\Nightly

-------------------------------------------

 # Starting Encode ...

[23:40:18] base preset: Fast 1080p30
[23:40:18] Remote Process started with Process ID: 19876 using port: 8037. Max Allowed Instances: 1
[23:40:18] Worker: Starting HandBrake Engine ...
[23:40:18] Worker: Parent Process Id 18152
[23:40:18] Worker: Starting Web Server on port 8037 ...
[23:40:18] Worker: Starting Listener: 1
[23:40:18] Worker: Server Started
Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unexpected Response: State: False, StatusCode: Forbidden, Response: <!doctype html>
<html>
<head>
<meta http-equiv="refresh" content="0;url=https://7rx90003.ibosscloud.com/ibreports/ibp/bp.html?bu=http://127.0.0.1:8037/IsTokenSet&bc=The+requested+URL+cannot+be+accessed.&ip=172.28.235.22&er=ERR_ACCESS_DENIED"/>
</head>
<body>
</body>
</html>

Unable to connect to the HandBrake Worker instance after 10 attempts. Try disabling this option in Tools -> Preferences -> Advanced.

 # Job Failed (-10)


@sr55
Copy link
Contributor

sr55 commented May 10, 2024

Yup, it's some kind of firewall/network/VPN appliance on your end. It's Man in the Middling the connection between the UI process and the Worker.

It's pretty aggressive network filtering to block localhost loopback.

All you can do in these cases is turn off process isolation unless your corporate security team can add an exception.

@galinart
Copy link
Contributor

@sr55 thanks for confirmation that this is not HandBrake issue. In meantime, I will clarify machine configuration details.

@galad87 galad87 closed this as completed May 20, 2024
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Development

No branches or pull requests

8 participants