| lotuspsychj3 | good morning | 01:41 |
|---|---|---|
| JanC | stuffandthings: I'm not sure that was a conscious decision for ntfs3 vs. ntfs-3g | 14:05 |
| stuffandthings | it just occured to me, maybe 3g is left in there for people upgrading from a system that used to use it | 14:06 |
| JanC | that would be handled on upgrade, so it wouldn't have to be in new installs really | 14:07 |
| JanC | but if I had to guess: it's possible there are some things that are supported or not in one vs. the other | 14:08 |
| ravage | When 3g came around the kernel driver was read only | 14:10 |
| JanC | e.g. I'm not sure whether (& if so: how) ntfs3 supports user mapping? | 14:11 |
| ravage | I think it now has write support but the default driver for everything you would do on an NTFS filesystem is still 3g I think | 14:12 |
| ravage | Have not followed the kernel development side really | 14:12 |
| stuffandthings | i think so, ive been using 3 for a while just as a user | 14:12 |
| JanC | well, the old 'ntfs' driver could write, but that was strongly non-recommended, as it would likely mess up any modern NTFS filesystem IIRC :) | 14:13 |
| stuffandthings | by default, ntfs3 was used to mount ntfs partitions | 14:13 |
| JanC | the 'ntfs3' driver should be good | 14:13 |
| stuffandthings | i never had a prob with 3g, other than it being slow | 14:13 |
| stuffandthings | ntfs3 is a bit problematic but nothing showstopper | 14:13 |
| stuffandthings | and kernel 6.11 is going to fix lots of ntfs3 stuff | 14:14 |
| JanC | looking at the kernel documentation for ntfs3, I don't see anything about user mapping, something ntfs-3g does support, so maybe it's still there for those who need that? | 14:14 |
| JanC | (it's used to map between Windows & linux users) | 14:16 |
| stuffandthings | well all i know is that when i mount the drive with dolphin, and then go to the properties, it says ntfs3 | 14:16 |
| stuffandthings | and i can write to it | 14:16 |
| stuffandthings | all while never getting asked to elevate | 14:17 |
| JanC | that works through udisks & PolicyKit | 14:20 |
| stuffandthings | whats the command to see if its 3 or 3g? | 14:21 |
| stuffandthings | well its most certainly uses ntfs3 by default | 14:30 |
| JanC | mount should tell you, but I assume the Dolphin info is right too | 14:32 |
| stuffandthings | i did a grep to confirm | 14:33 |
| JanC | this is one of the features that I think aren't supported by ntfs3: https://github.com/tuxera/ntfs-3g/wiki/File-Ownership-and-Permissions#mapping-the-users | 14:38 |
| JanC | most people probably don't need it | 14:38 |
| stuffandthings | this is whats coming in 6.11: https://www.phoronix.com/news/Linux-6.11-NTFS3 | 14:40 |
| stuffandthings | pull request link is in there | 14:40 |
| stuffandthings | nothing about whats in the link unfortunately | 14:41 |
| JanC | oh, now that I think about it, I think there are no tools for ntfs3, so no way to create an NTFS file system or check it for errors, so for that the ntfs-3g tools are used | 14:41 |
| stuffandthings | oh yeah? i never realized that | 14:42 |
| JanC | *that* might be why it's still installed :) | 14:42 |
| stuffandthings | probably | 14:42 |
| stuffandthings | thats a big one | 14:42 |
| stuffandthings | is it just me or should ubuntu enable windows_names flag by default | 14:44 |
| stuffandthings | seems like an easy way to prevent data loss for newcomers | 14:44 |
| JanC | maybe it's set by default? did you try? | 14:46 |
| stuffandthings | its not, i enabled it though | 14:47 |
| stuffandthings | but by default, its not on | 14:48 |
| stuffandthings | i can imagine a scenario in which someone copies a file with an unknown illegal character, runs windows and wonders where the file went | 14:50 |
| stuffandthings | that actually happened to me, fortunately, i know about looking at scandisk logs and the found.000 dir | 14:50 |
| JanC | the worst would be a file that has a special meaning in Windows :) | 14:53 |
| JanC | maybe file a bug about that | 14:53 |
| stuffandthings | should i? i was going to but people were talking me out of it, pointing to bugs filed from 2009, in which the consensus was 'boo ms bad, ms do bad things, me angry for hearing about this!' | 14:55 |
| JanC | ugh? | 14:55 |
| stuffandthings | basically seems like no one wanted to hear that anything might be wrong, both in the chan and on the posts from '09 | 14:56 |
| JanC | do you have a link for that? | 14:56 |
| stuffandthings | lets see if i can find it again | 14:56 |
| JanC | I mean, I can see that happening on a forum, but it shouldn't happen in a bug tracker | 14:56 |
| stuffandthings | ...and on the main ubuntu chan | 14:57 |
| JanC | I mean, it would be best if you can mark a disk somehow, so that the ntfs mount utility can know whether you want that enabled or not for that particular disk | 14:58 |
| stuffandthings | thats the thing, i raised it with them and got crap for it so i kind of didnt bother any more | 14:58 |
| JanC | but without that "marking" it should default to what is least likely to cause issues | 14:58 |
| stuffandthings | right? | 14:59 |
| stuffandthings | thats what i thought too | 14:59 |
| JanC | I mean, if it isn't used to exchange data with Windows why would you use NTFS at all? | 14:59 |
| JanC | there might be some other reasons, but those are probably going to be fairly niche | 15:00 |
| stuffandthings | in my cause because i have a few big, full drives as relics from a time when i mained windows, but yeah, i agree with the sentiment | 15:01 |
| stuffandthings | the main reason to have an ntfs drive is to interact with data that works on windows, if it defaults to not do that properly, its kinda bad | 15:02 |
| stuffandthings | in my cause i rebuilt a kubuntu iso with that flag on | 15:03 |
Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!