As I was kinda bored after work today, I had a closer look at what I saw during my fuckup in the morning. Well, Steve said, that when he looked at metadata.xml it’d be “really common” .. still that isn’t making it right ..
There is a reason we do have a herds.xml (exactly for the reason to associate people with packages, and that’s what the <herd> tag is for in metadata.xml) file. So after a preliminary look through the repository, here are the winners:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 |
700 : perl 126 : xemacs 63 : haskell 47 : sound 32 : ha-cluster 31 : crypto 19 : desktop-misc 16 : netmon 15 : forensics 13 : web-apps 8 : mips 8 : app-backup 7 : kde 6 : tcltk 6 : net-im 6 : media-tv 6 : dev-embedded 5 : voip 5 : theology 5 : samba 5 : net-p2p 4 : sparc 4 : java 4 : graphics 2 : net-mail 2 : kernel 2 : fonts 2 : embedded 2 : cpp 1 : x11 1 : wxwidgets 1 : www-servers 1 : tex 1 : shell-tools 1 : sh 1 : sgml 1 : sci 1 : python 1 : proaudio 1 : php 1 : media-optical 1 : kerberos 1 : hp-cluster 1 : gentopia 1 : amd64 |
Don’t know how accurate that list is, but you can check it for yourself. The commands I’ve used are these:
|
1 2 3 4 5 |
for i in $( < ../herds.list ); do grep --exclude=eclass --exclude=CVS --exclude=profiles --exclude=skel.* -R "$i@gentoo.org" /cvs/gentoo-x86/* ; done > redundant-metadata-xml.list |
|
1 2 3 4 5 6 |
for i in $( < herds.list ); do echo -e " $( grep "$i@gentoo.org" ~/public_html/redundant-metadata-xml.list | wc -l )t: $i"; done | grep -v "^ 0" | sort -nr > public_html/redundant-metadata-xml.overview |
While herds.list holds a list (separated by n) with all the herds there are. The raw files are here and here and here. Knock yourself out!
> 13 : web-apps
oh yes, i already fixed tons of these … 13 to go, yay! *sigh*
We use the maintainer tag for herds when the bugzilla alias is not just the name of the herd.
This is the case for perl (perl-bugs@gentoo.org) which explain it being top of the list.
Me and Jakub (and others) will hate you if you revert that.
> We use the maintainer tag for herds when the bugzilla alias is not just the name of the herd.
I know that pretty well Diego, but then again; there’s tools to look it up. Sure it’s faster, if you put the data again into metadata.xml, but they *really* belong into herds.xml.
There is a simple solution to fix this mess once and for all: get rid of the herd concept and use mail aliases directly. It’s just an IMO useless layer of indirection that has been causing confusion since it was implemented.
I agree with genone. Or at least make the herd have the _same_ address of the bugzilla account.