<ppetrov^>
so, prtverify now reports redundant deps?
goodlings8 has quit [Quit: Okay bye...]
goodlings8 has joined #crux
groovy4shoes has quit [Ping timeout: 255 seconds]
groovy2shoes has joined #crux
farkuhar has joined #crux
lavaball has joined #crux
<farkuhar>
ppetrov^: stenur made the same observation in his latest mailing list message, and suggested that the severity level be downgraded from WARN to INFO. I did question whether WARN was appropriate, since the offense involves other ports beyond the one being verified.
<ppetrov^>
how does it involve other ports?
<ppetrov^>
actually, i like the idea very much. I am doing a big cleanup of my ports now
<ppetrov^>
i have another remark: prtverify reports an ERROR if any of the headers use tab instead of space, separating the from their contents
<farkuhar>
See findredundantdeps(1) for an explanation of the algorithm, which involves not just the other ports but also the active collections (inasmuch as it relies on prt-get).
<ppetrov^>
ок, тханкс
<ppetrov^>
ok, thanks
<ppetrov^>
so, the tab thing, should not be an ERROR, should it?
<farkuhar>
I'll look into the tab thing. I wonder if the preference for spaces instead of tabs was there from the beginning, or if it got added later.
<ppetrov^>
it never complained before about this
<ppetrov^>
i did use prtverify regularly
<ppetrov^>
well, i changed the tabs to spaces, anyway i like it more like this, but such cosmetical thing should not be an ERROR
<farkuhar>
ppetrov^: thanks for the feedback. Just to be clear, which error were you seeing: "header not found" or "empty header found"?
<ppetrov^>
let me check, since I already fixed these...
<ppetrov^>
e.g.
<ppetrov^>
ERROR ppetrov/p5-algorithm-munkres .............. header not found: Descriptio
<ppetrov^>
ppetrov/p5-algorithm-munkres .............. header not found: Maintainer
<ppetrov^>
ppetrov/p5-algorithm-munkres .............. header not found: URL
<ppetrov^>
having a tab on # Depends on: seems to be fine
<farkuhar>
That would make sense, because 20_pkgfile_headers.awk does not regard "Depends on" as a mandatory header. But you're right that the regexp on line 21 is not loose enough to accommodate tabs.
<farkuhar>
Anyway, it should be easy enough to fix. Just change the regexp to /^#[[:space:]]+(Description|URL|Maintainer):/
<ppetrov^>
ah, so simple :P I am not a bash guru such as yourself, man...
<farkuhar>
A complete fix would require also changing the regexp in the sub() command, two lines later. Otherwise the wrong key of the pkgfile_headers array will be initialized, and you'll get the same error as before.
<farkuhar>
Thanks for reporting. I'll test the proposed solution some more, and push any changes later today.
<tilman>
ppetrov^: starting an executable that has the set-uid bit set results in a process where uid=owner of the executable. example: your sudo executable is owned by root, so starting sudo as user ppetrov actually runs it with uid=0
_0bitcount has quit [Quit: Leaving]
<ppetrov^>
thanks tilman
<ukky>
ppetrov^: In the ideal Linux distro, there should be *no* executable files with suid bit set
<SiFuh>
Or Linux-PAM in core
<cruxbot>
[contrib/3.7]: python3-shtab: initial commit, version 1.7.1
<cruxbot>
[contrib/3.7]: power-profiles-daemon: initial commit, version 0.21
<cruxbot>
[contrib/3.7]: sbctl: initial commit, version 0.13
<cruxbot>
[contrib/3.7]: gpgme: allow building qt6 bindings
<cruxbot>
[contrib/3.7]: objfw: 1.0.12 -> 1.1
<cruxbridge>
<js> That was fast 😄
<cruxbridge>
<tim> I am doing ports anyway :) Might as well
<remiliascarlet>
ppetrov^: In the ideal Linux distro, static linking shall be the default, rather than a pain in the ass.
<ppetrov^>
static linking?!
<remiliascarlet>
Yes?!
<ppetrov^>
ok!?! :)
<ukky>
remiliascarlet: I do not agree with your statement, as I prefer non-static (shared) linking.
<remiliascarlet>
Static binaries are way more portable, not as prone to supplychain attacks, not prone to sudden breakages because it was once linked to a slightly older version of a library that just got updated, and more advantages. The only true advantage of dynamic/shared is that the binary sizes are smaller.
<remiliascarlet>
Shared linking is like communism; it sounds great on paper, but in practice it's the worst thing ever.
<remiliascarlet>
Go compiles to static by default, C to shared. As a developer in both languages, the differences pretty clear to me.
<ppetrov^>
hueheuhue, i just love analogies to Communism
<remiliascarlet>
Haha, get it? Anal, ogies!
braewoods_ has quit [Remote host closed the connection]