how do I avoid repeated downloads when viewing PDFs etc?
When I click on a linked file to view it in an external application, Firefox always downloads a new copy (saving it with a numeric suffix in the /tmp/mozilla_USERNAME0 directory). This is daft - I don't always remember that I've looked at the file before recently, and Firefox ought to be able to reuse the existing download, or at least ask me whether it should do that. Is there a config setting somewhere I'm missing? (This is with 61.0.1 on Ubuntu, but I think it's always been the case.)
ప్రత్యుత్తరాలన్నీ (5)
The reason why Firefox handles it this way is because there is virtually no way to check if a file has been changed since the last download (or not) until it's fully downloaded.
Normally your Ubuntu system should clean out the /tmp directory regularly (if it is not a tmpfs, in the case of which files get discarded on reboot anyways).
What about Last-Modified and If-Modified-Since ?
That's the reason I said "virtually no way". Yes, those headers can be used to check if a redownload is necessary or not, but it's currently not the case. Feel free to create an issue on Bugzilla if you think it's a necessary feature!
p.s. not all server support those headers either ;)
న Gerd చే మార్చబడినది
So does this mean that Firefox no longer actually makes any use of all the stuff it dumps into my .cache directory? I can't see any logical difference between a PDF resource and an HTML resource.
There is a big logical difference. When viewing a webpage, the HTTP response is directly rendered, and also stored in the cache. When externally opening a document, the resource is treated like a download, and as downloads are generally one-time, they are not cached.