In short there are a few limitations I noticed for the Discord specific module while purging. The app cannot handle a purge that goes past a certain amount. I tried purging a DM chat that had about 73k messages and after a certain point the app seems to fail especially when it’s minimized.
Essentially either the entire GUI will freeze or it will go black and also freeze. It might unstick but more often than not nothing happens. I’ve left it on overnight into next day and come back to it pure black. It also sometimes will be using a lot of RAM before it gets to that point, I think I saw it using 7 GB of it in Task Manager and I’m not sure why other than that it’s keeping all the messages it deleted in memory for some reason? Only reason I can think of. Tl;dr cannot handle massive amounts of DM purging without failing at some point.
People who I think have the account blocked on Discord don’t show up in the DM list even though it has DMs with that person. The app seems unable to index these people when the account should still be able to access those DMs to delete. Tl;dr cannot access and delete DMs with people who are blocked and vice versa.
Re: Blocked Users - can you see your DMs with these people inside the Discord app “direct messages” section? If the DM is ‘closed’ you may not be able to find it (and Redact may struggle to load it). If you search for the user in the Discord DMs, and open your conversation with them, it should be added back to the list - in Discord, and Redact (probably after a refresh).
Give the above a try if you haven’t yet, it could clear up the issues. If you’ve tried these steps already, please get in touch with us directly via email/opening a ticket in discord & we’ll investigate and help out!
+1 confirming the memory issue from a different angle.
What I’ve seen on Windows 11 so far
Free build, fresh launch, ~250 MB RAM
After 4 hours of running on a 12k DM cleanup, climbs to 2.1 GB
GUI stays responsive but progress slows considerably toward the end
Forced restart and resume from where it left off worked fine
Workaround that helped me was toi pause every 30-45 minutes, fully close the app, reopen, resume. Tedious but it never crashed using that pattern.
The data package import method wooly had mentioned before is genuinely the cleaner approach for big jobs. The live API path keeps state in memory in ways the import path doesn’t, which I’m guessing is why memory grows so fast but I Don;t know?