julian-klode

9 ms·

61.
by julian-klode 7y ago
Last evaluation has shown that the size increase of zstd compared to xz warrants an introduction of delta debs first, so that people don't have to download more. But then with deltas, mirror sizes grows even more, which is a hard sell.
62.
by julian-klode 7y ago
APT does in fact check if the root of the repo changed and if so does not download anything else and will then notice that the input files have not changed when it's checking if the cache needs rebuilding. What you're seeing here
63.
by julian-klode 7y ago
Wow, that's a good example I did not have on my mind, but it's entirely the reason why I made that change.
64.
by julian-klode 7y ago
Pidgin works fine if you use one device. If you have multiple devices or want to receive messages while you're offline, it does not support that.
65.
by julian-klode 7y ago
Gajim's an unstable mess that half of the time is not even connected to the server because it did not manage to reconnect after resume, or it just crashes. It is fairly useless on dark theme, rendering grey on grey for the backlog. Apa
66.
by julian-klode 7y ago
So why do you use a browser made by an ad company, that is all about analysing your browser history and targetting ads at you?
67.
by julian-klode 8y ago
If you've seen today's bug yesterday, or carefully looked at the previous CVEs, you would have seen that https would have significantly reduced the probability of exploitability.
68.
by julian-klode 8y ago
Of course you'd get more security. Analogy: It's like your hanging on a rope and also add a safety net. If the rope breaks, you only fall on the safety net instead of the ground. All software has bugs, so if you add two buggy secu
69.
by julian-klode 8y ago
Well, defining the expiration date is up to the server. Debian picks a week. Ubuntu does not use it, I have started a launchpad branch last year, but um, I don't know launchpad, so it might take some more years ;)
70.
by julian-klode 8y ago
No you would not be able to use HTTP/2. HTTP/2 is not implemented in apt, and probably won't be for a very long time, as it directly clashes with apt internals. That said, pipelining over https is surely possible to, and redu
71.
by julian-klode 8y ago
CoCs are not silly, they solve very important and real problems.
72.
by julian-klode 8y ago
The blogger's bug report he linked to was about a package in universe. Universe is essentially the community playing ground - packages there are not supported, and most of them are synced unmodified from Debian and not there because an
73.
by julian-klode 8y ago
Hey craftyguy. I assume you are talking about networkd-dispatcher? I understand if you feel like your responses are being ignored. I was focused on other things the past weeks, I'd love to be more responsive. WRT patching things in wei
74.
by julian-klode 9y ago
This is impossible. Debian packages are stateful: They have maintainer scripts that can do stuff, they are inherently non-transactional by design. Even if you can rollback updates, a lot of the time that's no help - your user data form
75.
by julian-klode 10y ago
The JSON license does not prevent usage in weapons. Only use for evil. Weapons are not necessarily evil.
76.
by julian-klode 11y ago
bash supports that, just do shopt -s globstar to enable it.
77.
by julian-klode 11y ago
Just use apt. apt install apt remove apt search apt update etc.
78.
by julian-klode 11y ago
Well, just use apt(8): apt install - Install packages apt remove - Remove packages apt upgrade - Upgrade packages without removing apt full-upgrade - Upgrade packages, allowing removals apt update - Refresh sources apt show - Show a package
79.
by julian-klode 11y ago
Now, I was wrong, apt is indeed already in the handbook. "6.2. aptitude, apt-get, and apt Commands" https://www.debian.org/doc/manuals/debian-handbook/sect.apt-...
80.
by julian-klode 11y ago
There is a manual page. Nobody updated those fancy doc things. http://manpages.debian.org/cgi-bin/man.cgi?query=apt
81.
by julian-klode 11y ago
Your variable my_string is a pointer to a constant string. char * my_string It must be a const char * my_string You can still assign a different string to that, as my_string is not const. For my_string to be const, you'd have to use c
82.
by julian-klode 11y ago
You could just store using offsets instead. You don't have to do sign bit. 2 bit example: 00 => -1 01 => 0 10 => 1 11 => 2 or: 2 bit example: 00 => -2 01 => -1 10 => 0 11 => 1 This even wrapsaround naturally.
83.
by julian-klode 11y ago
Microsoft did not implement them in Visual C and wanted to claim C11 conformity without doing the work, so they bullied it out.
84.
by julian-klode 11y ago
You could have types __builtin_type_like_char_that_is_not_char and __builtin_type_like_uchar_that_is_not_uchar
85.
by julian-klode 11y ago
I have to take that back. For Contents-Source, it's ugly. The rest is fine though.
86.
by julian-klode 11y ago
So, we have basically rewritten the tornado patch idea using the C++17 string_view API (subset implemented in APT::StringView). This improved performance by about 10% on my test systems. Here's the commit: http://anonscm.deb
87.
by julian-klode 11y ago
Where do you want LMDB? apt has a super fast binary cache, the repositories and dpkg store their package lists as text files. RPM systems have the advantage IIRC that they can basically just download a bunch of sqlite3 databases, which mean
88.
by julian-klode 11y ago
Not all our files are uncompressed, so directly using mmap does not make sense for us, because we also need to be able to use files compressed with gzip, bzip2, xz (and soonish probably lz4).
89.
by julian-klode 11y ago
You don't really need cumulative diffs anymore. Sure, they would be nice to have, but with pipelining and client-side merging of the patches before applying, things are not that much slower.
90.
by julian-klode 11y ago
> Do you want me to try to train the APT developers in some way? That won't work Try it out, we're always listening and around in #debian-apt on OFTC.

More ›