| amarsh04 | still going on with a very slow update from i386 to amd64 - I would like to automate the downloading of amd64 packages matching the installed i386 packages, many of which may only be on snapshot.debian.org - has anyone already come up with script to do this? | 11:18 |
|---|---|---|
| JackFrost | Are you asking if someone wrote crossgrader? | 11:20 |
| amarsh04 | JackFrost - I've read a few bits of information on cross-grading and am partway through. Apart from the prospect of manually installing amd64 packages at the same version as the i386 versions, I was just looking at automating that part | 11:35 |
| JackFrost | https://salsa.debian.org/crossgrading-team/debian-crossgrading/-/blob/master/INSTRUCTIONS.md like that? | 11:36 |
| amarsh04 | Thanks, I hadn't seen that page, but had seen a different one on wiki.debian.org | 11:38 |
| freem | Hi. I do not know if this is widely known, I personnally just found out about it, because I decided I should search if, really, in 2025, linux is still too stupid to get that "Foo" and "foo" should point to the same file. So, I found this: https://www.collabora.com/news-and-blog/blog/2020/08/27/using-the-linux-kernel-case-insensitive-feature-in-ext4/ in short, it is posible (if kernel is built with support, which is the case for me, using stock kernel) | 17:07 |
| freem | to configure the system to ignore case when trying to refer to a file. I thuoght it would be a good idea to share the information, as this is something that I have always considered to be vastly superior in the windows world, compared to linux. That notably helps having duplicated files because one failed to use the exact damn stupid case, which can happen fairly easily, notably when exchanging code with people who work with user-friendly operating | 17:07 |
| freem | systems. | 17:07 |
| brocashelm | got xfce 4.12 running on one devuan stable machine; hail gtk2/x11 | 20:41 |
| gnarface | freem: it's a huge security hole, don't do that. | 23:36 |
| freem | security hole? Can you tell me more? | 23:36 |
| gnarface | nope. | 23:36 |
| freem | I see. | 23:36 |
| gnarface | well, i can tell you a bit more, but only a tiny bit; based on your initial statement, you'll simply disagree, but the security hole is obvious enough that if you're thinking about it right, you'll get it, and if you're not, it won't matter what i say. | 23:37 |
| gnarface | it's much better to just change your own brain to pay closer attention to such things, that way when files are changing contexts between systems you control and systems you don't, nobody can play any funny games on you | 23:38 |
| gnarface | the minor inconvenience of accidentally having a "Foo" and "foo" file (something someone rational might actually want to do on purpose, mind you) pales by comparison | 23:39 |
| gnarface | and don't even get me started on weird contexts where Windows will accept a value as "foo" and store it as "foo" but still display it to you as "Foo" ... leading huge swathes of users to not even know what their actual username is. this is bad! | 23:40 |
| gnarface | or having a long and short form username | 23:41 |
| gnarface | or then on mac where it does "case folding" where it actually stores case-sensitively but compares case-insensitively... all bad. | 23:41 |
| gnarface | any time you have a lossy abstraction you have a pattern vulnerable to abuse | 23:42 |
| gnarface | anyway, that's just why it's still like that on linux even though it seems more advanced the way windows does it | 23:43 |
| gnarface | not all progress is an improvement | 23:43 |
| gnarface | and the most insidious types of progressive regressions are the ones that make your brain simpler | 23:44 |
| gnarface | you can do what you want with your own systems, but this is my earnest advice to you: dno't. | 23:46 |
| gnarface | *don't | 23:46 |
| freem | ok, thanks for your input. Unlike what you supposed, I am open to actually think over people's arguments. | 23:48 |
| gnarface | well i apologize for that then | 23:53 |
| freem | np | 23:54 |
| gnarface | you can make it more like Windows by just chmodding everything to 777 (rwxrwxrwx) too, but the risks with that should be more obvious at least, i hope | 23:54 |
| freem | those years too many people are unable to change their mind, too many polarisation | 23:54 |
| gnarface | i had a boss once who thought passwords were even stored case-insensitively on windows. i discovered this was why he was always having trouble remembering his laptop password, shortly before the company went out of business | 23:57 |
| freem | lol what | 23:57 |
| gnarface | for real dude | 23:57 |
| freem | at this point, it's not a technical problem, though | 23:57 |
| gnarface | true, but he got that mindset from the way the filesystem handled case, and he had basically no other frame of reference for it | 23:58 |
| freem | well, the difference between a file name and a password is that a password needs to be hard to guess | 23:59 |
| freem | while you want filenames to be easy to guess (and remember) | 23:59 |
Generated by irclog2html.py 2.17.0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!