Upgrade Storagemtime is not saved
zotero disregards the "Last-Modified" http header, and does not set the modified date of downloaded files accordingly. This causes, for example, pdfs to show the download date under "Modified" instead of the actual date the file was last modified.
If that's important to you, you can explain why and make it a feature request, but this is working as intended.
If you want it changed, provide a rational beyond "that's how it should be" and we can discuss this productively. My argument for leaving this as is is that it allows you to check whether you've edited a file since importing it in Zotero". What's your use case for changing it?
I think this is pretty fairly called a feature request, since even Firefox doesn't seem to override the mod time of downloaded files, but I don't have any objection to following Last-Modified if it's available.
adamsmith: You'd still be able to do this just by seeing if the file's mod time was more recent than the Date Added/Accessed of the parent item, no?
As for least astonishment - you're the first person to mention this in 7 years of Zotero, so it's not like this is a constant source of consternation and I'd argue that changing 7 year old behavior creates astonishment in itself so there'd really have to be a good reason for it. I'm not convinced. I'm not going to make the call in the end, and maybe Dan sees this differently, but changing the current behavior would seem like a mistake to me.
FWIW, I don't see why we shouldn't maintain the mtime if we can unless this currently interferes with sync.
On the other hand, I don't really see much use for this behavior either. IMO, last modified time as supplied by the remote web server is quite useless. I don't think rtc1 has presented a real use case either. We do preserve mtime on import from RDF, RIS, etc., which is much more important.
I for one consider my local copy of a file to be separate from the server's copy unless it is being synced in some way rather than simply downloaded. I find it quite astonishing when the modified date is earlier than the created date in reference to the same file.
http://support.microsoft.com/kb/299648
fcheslack: "I for one consider my local copy of a file to be separate from the server's copy unless it is being synced in some way rather than simply downloaded." But what is the difference? The entire discussion thread at http://lists.w3.org/Archives/Public/www-tag/2011Nov/0025.html is very unfortunate, the simple argument that in doubt, to preserve is better than not to preserve, causes many replies with empty and meaningless academic statements.
aurimas: "consistent behavior at least with copying files on Windows" I'd consider zotero's download to be file archiving, and cp --archive from GNU coreutils preserves the mtime. wget, by default, does so, too. It is unfortunate that firefox doesn't seem to offer an option, only as an addon. Last word on this seems still not to have been spoken, see https://bugzilla.mozilla.org/show_bug.cgi?id=178506#c249 which refers to the mailing list disucssion mentioned by fcheslack. ext4fs has crtime which tells you when you downloaded the file. The mtime seems not a place to rely on for this information anyway.
As I said before, I don't really see very much use in looking at mtime of files downloaded by Zotero, but clearly you have a reason. I also don't see any down side to preserving the mtime where possible as long as it doesn't interfere with sync (and I think Dan has implied that it doesn't).
It may help move this feature request faster if you had a more solid use case than which Dan addressed with As it stands now, it doesn't look like it would be very high on the todo list.