ungoogled-chromium: Ungoogled layer breaks Chromium DRM reading

@networkException asked me to post that I have widevine working in Version 98.0.4758.102 elosten unportable. I run this script after each upgrade and it grabs and installs the widevine directory from chrome into my ungoogle chromium install. This is a bash script for linux only with ungoogled chromium installed at /usr/bin/chromium and /usr/lib/chromium/WidevineCdm but could be adapted.

https://gist.github.com/dkebler/b90ca57ac481a428dcb6cbbd1e36553d

netflix doesn’t work either nor does spotify

I don’t have a Netflix subscription. However, I confirm Disney+ and Spotify (free account) keep working, with the procedure I indicated here above: https://github.com/Eloston/ungoogled-chromium/issues/1831#issuecomment-1094061018

It takes about 2 hours to build the package. It is going to be quite a job to test them one by one!

You could mimic the bisect strategy: divide amount of patches roughly in half, apply them. Test if anything changed. No changes → divide the remaining number of patches in half and repeat. Something changed → divide the number of applied patches in half and un-apply them. Repeat until one or several patches remain. There is a downside to it: if two or more patches are to blame, they could get into separate batches 😦

… nobody is able to help me to narrow down …

I’m afraid not many of us use widevine to begin with. Let alone with ungoogled-chromium.

I’m curious though, because the error quotes specifically, “contact Disney” and ask what is “error code 83” - maybe if they say what they’re expecting on their end? Are you able to do that? It would narrow the problem down. The disney log looks like it’s looping? Maybe because an expected response is missing and it retries?

Widevine had always been a problem in the past and I’ve never been able to get it to work personally. I know for some vendors they have a separate contract with Google so they are provided with a particular (I guess signed, haven’t look into it too much) version of widevine, but for us we cannot do that.

Hi. New update.

I tried to restore MEI_preload by:

sed -i "/disable-mei-preload.patch/d" patches/series
sed -i "/mei_preload/d" pruning.list
sed -i "/mei_preload/d" domain_substitution.list

Unsuccessful…

I am willing to keep testing, but I am in a kind of shooting-in-the-dark mode. I don’t know enough what is required for the DRM validation with the server.

Would anyone be able to give me some hints on where to look at, to narrow down the scope?

I’ll look into this in a few hours, until then you could maybe try building without domain substitution

I have tried already to un-substitute google and chromium items from the regex list and to remove any libdrm items from the substitution list. It didn’t help. I can try to remove everything if you think it can help. I will do it in 2 hours, except if you have another idea since then.