"The exception breakpoint" error and 0 points on profile

[BrenoFariasdaSilva]

Hello, here i have two problems.
The first issue is related to the .bat file sometimes not saving the login, showing 0 channel points.
When I have all set up and a file for each channel using the same account, I open the file (which i had already configured, logged in, and used for a while just to make sure it is working just fine). Then, on another day, for example, the channel goes live, so i run the same .bat file as earlier. It opens but shows up having 0 points (which means it is not logged in anymore). It is happening every day. I have it all set up. The profiler, i always mark to "remember for 30 days" after solving the captchas and so on. Why is this happing?

The second issue is about the following screenshot.

When i have them opened and it's working (showing the right current points), sometimes it breaks as the images say. Tried to find any solution, but i had no success.
I verified the firefox version and it is 59.0.3. I deleted everything and set it up all over again many times and those two errors keep showing up.

Thank you for your time!

[Sleepydragn1]

For the first issue you listed, it's possible that the script is logged in, it's just that the points tracking system broke. The reason why I say this is because the point tracking code is a really poor way to obtain that information, and is potentially prone to breaking, but I couldn't think of a different way to do it. On top of this, the script should detect when it's not logged in. As such, the next time this happens, try running without the --headless flag to visually verify that you aren't logged in — that'll narrow down the issue for me. If you've already done this, reply back and we'll go from there.

As for the second issue, can you provide logs of the sessions in which this occurs? That would help me diagnose the issue. I will forewarn you, though, since SlimerJS runs on an older version of Firefox, it is fully possible that this is a Firefox-related bug that's isn't fixable on my end.

EDIT: For the second issue, it would also be helpful if you switched all of the options in twitchAFKConfig.js under the Debug header to true. It'll make the logs messier, but may provide more information.

[BrenoFariasdaSilva]

The logs seem fine. Nothing is wrong with them, but I will activate the Debug Header and answer you later about that. For the solution of the "0 points channel"...I tried taking out --headless and it clearly shows that it is not logged in.
The two following screenshots are for the attempt without headless mode on and for the .bat file text.

.
I used these commands for the previously used .bat file on the screenshot:

slimerjs -CreateProfile MainCorruptedG
slimerjs -P MainCorruptedG twitchAFK.js CorruptedG

As we can see, I don't think there is a mistake in my setup. I read everything carefully just to make sure I didn't mess up somewhere.

The script is really great and helped me a lot. But, I have already +90 .bat files. So we can imagine there are a lot of streams going on and off, and I found myself spending 2+ hours a day just because it suddenly is not logged in anymore.

[BrenoFariasdaSilva]

This screenshot is to show how the previous .bat file screenshot used to be, just to make sure it is right:
.
I'm also leaving a copy of my twitchAFKConfig as a txt file but without my credentials in case that would help:
twitchAFKConfig Example.txt

[BrenoFariasdaSilva]

The log txt for the files that aren't logged in(showing 0 channel points):
Datto_2021.01.12_17.35.txt

Another example but this one says "Network error" on the log file.
Arsenal_2021.01.12_18.00.txt

[BrenoFariasdaSilva]

The last log, related to the second error, which was the crash from the screenshot on the first issue message:
Datto_2021.01.12_18.26.txt
At the beginning, I thought it might be my PC, but I have a Ryzen 7 3800x, 16GB of Ram, and an RX 580 8GB, so I don't think that's really the problem

[BrenoFariasdaSilva]

Today, I found out that this error message appears only on the cmd, not on the log file.
It happened right before I launched Chrome (my main browser). It looks like that, whenever I open a browser (Chrome, Opera, or Edge), the cmd script crashes.
There you have the error message:

The CMD crashed after that error message, but my browser keeps running, but not responding until I restart the computer.

It says that is out of memory, but it wasn't

I'm trying to give you as much info I can, so it would be easier to find where is the problem.

This one might be useful too, IDK.

[Sleepydragn1]

I appreciate how thorough you're being, that helps a lot.

For the first issue: the messages prefaced with "[ERROR] [apollo]" are pretty suspicious. Apollo is a framework that deals with APIs and data, and can handle OAuth token-related stuff, which is involved in logging in and keeping a user logged in. It's hard to know exactly what's happening based on just that one error message, so here's a couple of theories:

It looks like it's attempting to revoke tokens, which could potentially mean that Twitch itself is intentionally logging you out. You mentioned running multiple profiles and instances, so it could be that Twitch has some sort of automation protection that's triggering because you're running a lot of different streams on different browser instances on the same account or IP.

Otherwise, it could be that it's revoking tokens in the process of refreshing your access token (i.e. it's "revoking" the old access token, or something), and that it's simply having trouble doing so, leading to you being stuck with an outdated access token. Note that it attributes the issue to a NetworkError, and that both times that the error occurs, the stream also fails to load. This could be some sort of networking issue on your end, but that's far from certain.

Either way, I haven't been able to personally recreate this in my testing, although I may need to create more profiles and simultaneous streams to see if that can force the issue on my end. If I were to recommend anything, I would try using just one profile temporarily to see if the issue continues to occur.

For the second issue: 0x80000003 seems to be a graphics-related crash. Some possible solutions that were mentioned in some threads I found:

1. Update your graphics drivers. Otherwise, it could be that your drivers are too new, and the older version of Firefox has incompatibility issues with them for whatever reason.
2. Try disabling hardware acceleration. In Firefox 59.0.3, this is under Options > General > Performance > Use recommended performance settings (uncheck), and you'll need to uncheck "Use hardware acceleration when available".
3. Try installing an older version of Firefox. SlimerJS supports Firefox 53-59, so I would try installing Firefox 53.0.3, which you can find here. Remember that you'll need to change SLIMERJSLAUNCHER in your environmental variables to reflect this install's location to get SlimerJS to actually use it, though. I've verified that SlimerJS/twitchAFK does work with this version, FYI. Also, as a personal anecdote, I found that although SlimerJS seems to work fine with 59.0.3 on my end, for whatever reason opening the browser manually and navigating to Twitch (or some other websites) causes a crash, albeit one with an unrelated error to yours. On the other hand, the older 53.0.3 does not have this issue for me.

The other possible issue I can think of, based on you mentioning that launching a browser causes issues, and that screenshot of Opera claiming that it's out of memory, is that you're running out of VRAM. I think that's kind of unlikely based your card having 8GB of it, but given enough instances of twitchAFK and other programs using VRAM, it's possible. If you click on the GPU in Task Manager's performance tab, it has a "Dedicated GPU memory" graph and readout towards the bottom that should make this fairly easy to verify.

[BrenoFariasdaSilva]

I thought about the network possibility and did some testing with Youtube. I have a 240mb/s network that I use the ethernet cable on it. I do open a lot of lives, but I reached out to never have more than 10. That's the current max number of lives that are running on "MIN" settings, which (probably) is 160p. So, I opened 10 Youtube videos at the same time, also using as many different browsers as possible. Also, I left open the software I'm used to having when I have the Twitch Script on. It ran ok. One or two out of the 10 videos sometimes got stuck loading for less than 5s. So, it doesn't look exactly to be a network issue, but it could also be some issue inherent in Twitch's code, as I found out by searching that are many people with really great PCs with multiple streams opened and it crashes (so let's count that, but I really do think it is something related to Twitch's decoding script).

As for the VRAM, the first thought I had was my Computer Resources. So, I always used to check the task manager and see if the problem was my PC. I never had much use. My RAM was the one most used, but it got like 10GB out of 16GB. I also already thought about the VRAM. It also never got to use more than 2GB of VRAM (out of my 8GB). I could tell that it could not be exactly the problem as sometimes I had only 2 or 3 streams and it crashed.

For today, I will try disabling hardware acceleration. I might not get you an answer today about it. I will try it for at least 24h straight. I also already have the last Graphics Update. It could as well be an incompatibility with the latest version as you mentioned.
At the weekend, in case the problem persists, I'm gonna try the Firefox version you told me, but even if the "crash issue" is solved by just disabling hardware acceleration, I will try with the Firefox version you told to see if it also have any impact on the "[ERROR] [apollo]". By Monday I might already have done it all and have the answers.

If there is anything I can do to help, please let me know :).

[BrenoFariasdaSilva]

This error also happened:

I made a quick google search for: Firefox the exception illegal instruction 0xc000001d.
Got some answers, but none of them ever mentioned this: 0x00007FFEAAE247EC.
Most of the answers I got talked about Update Firefox and Send Crash Report.
I already had disabled Hardware Acceleration, but I can guarantee that it happened before, but at the time the PC froze, so I could get the screenshot.

[BrenoFariasdaSilva]

My Computer Task Manager when one tab crashed:

With that in mind, I think it's safe to say that isn't any Computer Resources issue, but also let's not forget that Twitch's Decoder/Encoder might not the most efficient. I haven't tried out Firefox 53.0.3, but Hardware Acceleration might not be the solution (or at least, not the only one). As you said, it could exactly be any Twitch's code that tries to avoid people with multiple streams from the same IP. Also, in terms of performance, Twitch's code might not be good with scalability (Like, using multiple Threads when having multiple streams opened. I got to this conclusion as I changed the Task Manager CPU view to Logical Processors. It showed me that there are a lot of cores having more than 80% usage and some lower than 10%.

It is my guess. You better not blindly believe in me, as I'm not exactly a Programmer. I'm learning the basics of C Coding on my own. All that I said (here and on the previous messages) it's basically common sense. I'm not an expert or something like that, just some random guy that likes Tech in general.

[BrenoFariasdaSilva]

After uninstalling Firefox 59.0.3, installing Firefox 53.0.3, changing the SLIMERJSLAUNCHER path to the new Firefox Folder, I tried to run but the .bat file crashes in a fraction of time. I spammer the PrintScreen Key and got the error it showed.

It says that it does not recognize "--headless" command. So, I ran without that command on the .bat file. It worked for about an hour, then I was watching some Youtube videos when it crashed. But it looks to have solved that "Apollo Error" as I haven't need to re-login or anything like that. I read again the original project instruction in order to try to guess why the headless command isn't being recognized, but I haven't got anywhere.

Do you have any idea why it is not recognizing the headless command?