You are not logged in.
I have a package whose version used to be 1.02b-4.
Now the new package is is version 1.1-2 in my repo.
Can someone explain to me why I get the following from pacman? How is it doing version counting?
:: itext-1.02b-4: local version is newer
Offline
pygtk 2.4.1-2
this means the "normal" behaviour of "major-minor-revision" and the digit after "-" is arch' package version, this means that this is the second time pygtk is packaged and put in arch' repos,
arch + gentoo + initng + python = enlisy
Offline
You can use the vercmp utility to test pacman's version comparisons, if you want. It returns an integer as a strcmp() call would.
[jvinet@mars misc]$ vercmp 1.02b 1.1
1
So pacman should see 1.1 as greater than 1.02b
[jvinet@mars misc]$ vercmp 1.12a 1.12
-1
Pacman considers trailing letters as pre-release versions, so 1.12a is less than 1.12.
Offline
Just to make it really obvious:
1.02b-4 is the same as 1.2b-4.
Offline
Just to make it really obvious:
1.02b-4 is the same as 1.2b-4.
yep: strcmp() says 0 ;-) ... but this is not common sense (a trailing 0 is never common sense)
The impossible missions are the only ones which succeed.
Offline
yep: strcmp() says 0 ;-) ... but this is not common sense (a trailing 0 is never common sense)
If that the case with your strcmp() function then that's a bug. It should only be the case with strverscmp().
Offline
isn't apeiro mixing the numbers in his explanation?
pjmattals package shouldn't been replaced with 1.1, should it!?
arch + gentoo + initng + python = enlisy
Offline
the vercmp program uses the rpmvercmp routines (check the pacman source) which, though I haven't looked at the details, I would assume uses BCD style versioning.
so 1.02 translates:
atoi("1")=1, atoi("02") =2
just a guess.....
Offline
so 1.02 translates:
atoi("1")=1, atoi("02") =2
I've said that two days ago already...
Offline