Talk:Comparison of file systems

The limits in this article are not the same as in the ReFS article. The limit is 35PB. Can someone verify it and fix? ויקיטכני (talk) 19:11, 26 March 2024 (UTC)

It unfortunately seems ReFS isn't well-documented so far. I couldn't even find a source for its time granularity.

Microsoft seems to prefer keeping ReFS in the Windows server world only. The opposite happened to exFAT, where Microsoft realized in 2019 that its proprietary status limits its adoption, famously in stock Android OS (although some third-party vendors like Samsung already supported exFAT long before), so Microsoft made it open. CDVDBD 💿 📀 09:21, 1 June 2025 (UTC)

There is no column for "last modified time stamps" in the table at Comparison of file systems#Metadata. I assume it was left out because pretty much every file system has it. Is there any file system that doesn't support it?

If all do, I suggest writing "Every file system listed in this table supports the last modified time stamp" or something like that. CDVDBD 💿 📀 08:59, 1 June 2025 (UTC)

I added a note at the bottom, worded in a way that does not exclude the possibility of such a file system existing out there, but if it does, it is very obscure at least. CDVDBD 💿 📀 09:46, 1 June 2025 (UTC)

Whether TRIM is supported is a question of how the driver / operating system implements the file system, not of the file system. Technologically, nothing would stop the 1978 Unix File System from being used with TRIMs emitted for disused blocks (because, as almost all file systems on this page, it's block based, and as soon as the last user of a block has stopped using it, it could be marked as unused).

That has absolutely nothing to do with "allocation and layout policies" (aside, again, from almost all file systems being block-based, because they're used on block devices; this is, as far as I can tell, even true for "absurd" things like thinking about trim on magnetic tape file systems).

Therefore, the "Trim support" column needs to be moved to a different table (in this table, it would need to be "yes (green)" from top to bottom). It might make sense to make it a symbol in entries of the "OS support" table. MüllerMarcus (talk) 11:57, 22 November 2025 (UTC)

I don't know if it's necessarily true, that "No filesystem has ever allowed NUL". It's a very broad sweeping statement, and although it may be true for the filesystems listed, I feel like it should at least have a citation. Thoughts? JMVR1 (talk) 12:12, 9 June 2026 (UTC)

It's a very broad sweeping statement, possibly made by people whose first experience with file names and pathnames was with an operating system primarily programmed in languages in which character strings are represented as NUL-terminated rather than, say, counted. It may well be true, but, yes, it needs a citation (and one that doesn't assert it based on "since strings are NUL-terminated, NUL can't be used as a character in a file name", as that's based on the broad sweeping assumption that they're NUL-terminated in all operating systems). Guy Harris (talk) 00:35, 7 July 2026 (UTC)

https://en.wikipedia.org/w/index.php?title=Comparison_of_file_systems&diff=1364812261&oldid=1358552310 (a Wikipedian away from his usual account)

The left column of all the tables is completely unreadable in dark mode. ScienceD90 (she/her) (talk) 02:59, 14 July 2026 (UTC)