<remiliascarlet>
farkuhar: All a website is supposed to be is interesting content to read. Feels like nowadays websites are anything but interesting content to read.
<remiliascarlet>
SiFuh_: "This Spongue Bob thing is popular? How the fsck did it get popular?" Children worldwide having watched it across multiple generations.
<remiliascarlet>
Same reason why Disney and Loony Toones are very popular, having recognition across 4 generations (boomers, gen X, millenials, and zoomers), but I don't think gen alpha and onward will recognize them.
serpente has quit [Remote host closed the connection]
<farkuhar>
modular kernel for a laptop because "you never know what you will connect"? Heh, might be true for some laptops, but the netbooks they're pushing these days have almost no expansion slots or USB ports. On those models it's much easier to anticipate the range of hardware that you'll ever connect.
<farkuhar>
Slower startup after bumping the ff version to 136 is the regression that serpente noticed. Here's another one (supposedly fixed in 136.0.2): data loss on shutdown if the user hadn't already opted into "clear history and site data" at versions <= 135. https://bugzilla.mozilla.org/show_bug.cgi?id=1952564
<zorz>
farkuhar: it depends, for example, nvidia drivers no need
<zorz>
if you dont use nvidia
<zorz>
likewise amd. or broadcom
<SiFuh>
zorz: Or firefox, no need if you don't use it
<farkuhar>
serpente wasn't even using firefox when he noticed the regression; he was using librewolf instead. I should have asked whether the slowdown was noticeable even for users of opt/firefox-bin.
<farkuhar>
I wonder if the bug that automatically opted in to "clear cookies and site data on shutdown" could have had as a side effect the slower startup that serpente noticed. One would think that a cleared-out folder of cookies and site data would lead to faster startup times, not slower.
<zorz>
SiFuh: exactly no need for firefox.
<zorz>
i use firefox, because of metamask and rabby wallet. otherwise no need for me to use firefox.
<SiFuh>
Who disturbs me from my slumber?
<zorz>
heh
<SiFuh>
Just don't use a browser at all. It isn't needed
<zorz>
actually most of my surfing is with tor-browser, from my city to frankfurt usually i am 3-4 hops and exit nodes are not thaat slow.
<zorz>
SiFuh: how i am going to gamble?
<SiFuh>
zorz: Easy, put a blind fold on and walk around the highway.
<zorz>
fuckin lenovo..... who the fuck buys lenovo
<SiFuh>
zorz: Why are you fuckin it?
<zorz>
i did on the server xbps hold kernel
<zorz>
and then reboot
<zorz>
and nothing was working
<zorz>
and the problem was time
<zorz>
heh
farkuhar has quit [Quit: leaving]
zorz has quit [Quit: leaving]
zorz has joined #crux-social
serpente has joined #crux-social
<ppetrov^>
so... about 3.8... emm... are we there yet? or at least rc5
<ppetrov^>
:P
<ppetrov^>
i am not updating my ports for 3.7 anymore, waiting for 3.8, when i will cut ppetrov to 1/4
<SiFuh>
So ppetrov^ will become pp ?
<ppetrov^>
ba-dum-tss
<ppetrov^>
comedy genius SiFuh strikes again
<SiFuh>
ppetrov^: Reminds me of a joke I heard in the 80's "I can say the alphabet, abcdefghijklmnoqrstuvwxyz" "Where is the p?" "It's running down your leg"
<SiFuh>
blur blur blur. I told serpente that joke today abou the alien in the bar.
<ppetrov^>
the p... hilarious
<SiFuh>
For kids
<ppetrov^>
it was sarcasm
<SiFuh>
I can't beleive that I have to pay money to search court records for cases in Australia
farkuhar has joined #crux-social
<farkuhar>
There's no good way to write a flexible rust-bin port that will work for both CRUX-glibc and CRUX-musl. Even if I redefine make_footprint() to narrow the output and print only the files that the two builds would have in common, the make_signature() and check_signature() functions (and a source array that depends on the owner of /usr/lib/libc.so) cannot easily accommodate different upstream tarballs.
<farkuhar>
Maybe the core team is following ppetrov^s example and leaving the opt/rust port unchanged in the CRUX 3.7 branch. The rust version bump 1.85.0 -> 1.85.1 might have been committed in the 3.8 branch, which wouldn't have triggered any cruxbot message (at least while 3.7 is still considered the current CRUX).
<farkuhar>
Nope, no rust update yet in the 3.8 branch of the opt repo. It seems I gave beerman too much work by misinterpreting his initial request to "move docbook-xsl down to opt", thinking he needed it only for gtk4 in the 3.8 branch. But in fact he wanted it for the 3.7 branch too, and I committed a slightly different port there, resulting in a botched merge.
<farkuhar>
ppetrov^: Would you like to propose in #crux a moratorium on any further commits to the 3.7 branches (except for critical security updates)? That way they can stop running the automatic merges into the 3.8 branches. The only downside is if a CVE needs immediate patching, it would involve two separate commits (one in 3.7 and another in 3.8).
<ppetrov^>
me? to propose a moratorium? come on, man
<farkuhar>
One way to endow rust-bin with enough flexibility for both CRUX-glibc and CRUX-musl is to leave the source array empty, and just redefine download_source() to fetch the appropriate upstream tarball based on the owner of /usr/lib/libc.so; that way it's possible to create a signed manifest without downloading multiple tarballs. But then there's no source integrity verification at all, which probably isn't a wise idea.
<farkuhar>
It's much easier to endow cbindgen with the flexibility for both CRUX-glibc and CRUX-musl, because the downloading step happens during build(), in the line `cargo fetch --target x86-64-unknown-linux-gnu`. Change the argument of --target based on the owner of libc.so, and we have a flexible port!
<farkuhar>
Ideally there shouldn't be anything in the build() function that requires an Internet connection; all the sources should be present on disk by the time download_source() has run. But the rust ecosystem has stretched packaging practices a little bit, and redefining download_source() to hold a single `cargo fetch` command is an ugly way to respect historic packaging conventions.
<farkuhar>
ppetrov^: why not you? You're the one who first declared a personal moratorium on updating your repo for CRUX 3.7
<zorz>
SiFuh: SiFuh: is for the moratoriums!
<zorz>
:p
<farkuhar>
It looks like SiFuh was right about zorz not knowing how to read. Somehow it escaped zorz's attention who was the actual author of this post. "19:17 <ppetrov^> i am not updating my ports for 3.7 anymore, waiting for 3.8"
<ppetrov^>
farkuhar, the real reason is i am quite busy. And when I have some free time, I drink
<zorz>
< ppetrov^> me? to propose a moratorium? come on, man
<zorz>
SiFuh: is for moratoriums
<ppetrov^>
I actually, do not understand what you mean by "SiFuh: is for the moratoriums!"
<ppetrov^>
you mean, he supports moratoriums?
<ppetrov^>
because the way you wrote it if word by word translated to bulgarian means exactly this
<zorz>
he likes to piss them off
<zorz>
hahahaha
<ppetrov^>
man i am so tired
ppetrov^ has quit [Quit: Leaving]
<zorz>
ppetrov^: did you see what happened in skopje