arch1t3cht / Aegisub

Cross-platform advanced subtitle editor, with new feature branches. Read the README on the feature branch.
http://www.aegisub.org
Other
772 stars 33 forks source link

[Bug] Opening multiple instances of Aegisub at the same time generates unexpected bugs #7

Closed VegethB closed 1 year ago

VegethB commented 2 years ago

As the title suggests, if several Aegisubs are opened at the same time, "random" bugs are generated, the most frequent one is shown in the video (since sometimes it doesn't just happen that specific bug but others).

Bug1)

https://user-images.githubusercontent.com/17975351/192144347-c440d402-3519-4b56-bca3-9a8a170836bf.mp4

Bugs screen: Bug2) ![Immagine 2022-09-25 153427](https://user-images.githubusercontent.com/17975351/192146356-93fe4bd2-15ae-427f-9c99-1af979e57841.png) ![Immagine 2022-09-25 153501](https://user-images.githubusercontent.com/17975351/192146363-d511c701-0bab-4887-a3ec-f78cde9b797c.png) Bug3) ![image](https://user-images.githubusercontent.com/17975351/192155023-fc877015-05b3-47bf-8c53-b80961de19f2.png) ![image](https://user-images.githubusercontent.com/17975351/192155057-12e84af7-7dce-4172-affd-e432ac849503.png) ![image](https://user-images.githubusercontent.com/17975351/192155078-dc933bd6-4c56-4ed8-ae97-fa631678528e.png) ![image](https://user-images.githubusercontent.com/17975351/192155190-baf3e1db-7c48-4ad9-9087-37bfdc4efe39.png)

The most obvious thing to think about is that it is due to the program's inability to create unique logs for opened files. However, this extends to prevent certain scripts from being loaded (generating an automation script not loaded error).

Why open multiple instances like this instead of opening them manually (thus allowing at least 1 second to pass): I have to mark the times in ms of the RAWs. As long as you do 12 ep a day it's not a problem... Different speech if I have to open series of 300+ episodes. Those 3 seconds of doing it by hand change the whole day (also because at a certain you break to repeat the same action 300 times).

Steps:

To do what I do in the video (open aegisub from the context menu): Create keys in the registry like this image image "C:\SoftWarez\arch1t3cht's Aegisub fork\aegisub-Arch1t3chts.exe" "%1"

HW Specs: I7-4790k 4ghz; Kingston sata3 SSD; Network Paths 1gb / s (sata3 network drive); 20GB DDR3 1600mhz Nvidia GTX1070ti 8gb

NOTE!

This bug happens with any aegisub mentioned in the readme (code), AegisubDC, AegisubTypesetTools and the official Aegisub. In the case of this video, the Aegi of this Github was used in portable mode (because AegisubDC is still the main one). In the particular case of this build, I tried with Avisynth + (the latest to the day of this post), as source video (instead of the usual ffmpeg), nothing has changed. In AegisubDC, vsfiltermod (instead of libass) is used to view subs

logs.zip crashlog.txt

Thanks in advance.

Long and complete video of: What is all this mess for 🙃 https://mega.nz/file/x4EViaTL#5WwRS_XH3tjThV4cJ-onZNEtn4joLFQ05Gw8prmeXOs

VegethB commented 1 year ago

🖕

arch1t3cht commented 1 year ago

...It's not like I ignore issues on purpose. But bugs like these, which are actually collections of multiple issues, some of which are very hard to debug or work around will just take time to fix.

Also, please read the repository's README:

I'll gladly take any bug reports, but if you encounter an issue, please check first if it occurs only on my fork, or also on earlier TSTools builds. If it wasn't introduced by my fork, I can still take a look, but I can't promise anything.

Bugs that weren't introduced by my own changes will get a lower priority when there's more important things to fix. Especially when it's bugs like this one that come up in niche situations and have simple workarounds (e.g. making a wrapper script that starts multiple instances of Aegisub with some delay in between)

But, since you closed the issue, it's probably not important any more :)

VegethB commented 1 year ago

...It's not like I ignore issues on purpose. But bugs like these, which are actually collections of multiple issues, some of which are very hard to debug or work around will just take time to fix.

Also, please read the repository's README:

Bugs that weren't introduced by my own changes will get a lower priority when there's more important things to fix. Especially when it's bugs like this one that come up in niche situations and have simple workarounds (e.g. making a wrapper script that starts multiple instances of Aegisub with some delay in between)

But, since you closed the issue, it's probably not important any more :)

I'm sorry, don't take it as a direct insult to your person... but if you wrote me this 1 year ago I would have said "ok" and that was it. Did we have to get to this for a feedback? Instead, you release 2 versions with 40 fixes and after 1 year and 3 months not even the shadow of an answer... sorry, but I'm breaking my balls.

I'll gladly take any bug reports, but if you encounter an issue, please check first if it occurs only on my fork, or also on earlier TSTools builds. If it wasn't introduced by my fork, I can still take a look, but I can't promise anything.

it is your fork that I wanted to use ( which is why I report the problems to you and not to whoever wouldn't answer me... too bad that in the end I didn't receive answers anyway).

I can still take a look, but I can't promise anything.

Precisely! I read the readme and decided anyway to write the issue to you who seemed active and willing... but instead...

Since any of my interactions with the typesetT and DC went unanswered why would I write a bug to them where (if they ever responded) they would write me the same thing as you? That is: go write it to guy and then guy says to me: go write it to that other guy... In the end, it ends up that you have to write it on the original github that there are no more releases and the devs since 1700. Since you were detached from all of this, I simply gambled your fork (losing badly).

Anyway, I closed it here and here's my last answer. Have a nice day or evening.

seproDev commented 1 year ago

1 year and 3 months

You might want to double-check your maths here. But I am sure arch1t3cht is happy to issue a full refund to you for all the pain and suffering he caused by not spending his time on your very niche use case.

Also, looking at your video, I don't understand why you want to open a new Aegisub instance for each video. Can't you just drag and drop the next episode into Aegisub once you have noted down the timestamp? Alternatively, as it doesn't seem like you need all episodes open simultaneously, you could just write a simple script to open every video in the folder one after another, waiting for the previous Aegisub instance to close.

VegethB commented 1 year ago

image image

I don't remember ever writing a right date... maybe you weren't taught "irony" (and I even wrote "3 months" after "1 year" to make sure you understood... 🤷‍♂️... or do you prefer 4 since we are over 25?)

your very niche use case.

In fact, opening a second instance of a software is unique in the world 👏.

Alternatively, as it doesn't seem like you need all episodes open simultaneously

Try registering a 5760x2160 canv to OBS and see if you understand anything seen on a 1920x1080 (assuming you have 4 1080p monitors with 3 in series and 1 above). You can't take timestamps so lightly if you don't COMPARE the two videos (or more than 2...). I also specified that simultaneous opening is a MUST and not a SHOULD. First of all reasons, I have to wait seconds for aegisub to load the next video (and the seconds change from 230mb to 1.4gb to 24 minutes to 1:30 hours... but maybe it's too hard to understand when you don't have the hours counted in your life...) Knowing how you think, you will answer me: "Aegisub was not meant for that use". And it makes me laugh that it all ends in this sentence of thought. Because it proves once again that 90% of Github devs have no idea of the difference between the terms "REPORT A PROBLEM" and "FIX A PROBLEM".

Those like you: I reported a problem = give alternative solutions to work around the bug. I wonder if the problem is the education that schools provide you on text analysis in your country… But then I'm fucking wasting my time explaining to you. It's clear that you don't have the skills to solve this problem, you're telling me to go elsewhere because: 1) you don't give a shit; 2) you don't have the skills to do it. Thanks for the enlightenment. Go on your way, I don't give a shit about you like you do me. I'm now removing the notification to this shit of issues and 0 respect from people who think they are gods (...because respect has to be earned first). ByeBye shitty people 👋!

arch1t3cht commented 1 year ago

@juice_squad

arch1t3cht commented 1 year ago

They're blocked now, but anyone else reading this can refer to this as the perfect example of what not to do in open source repositories.