Summary: | tetex.eclass pointless mysql check | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Robert Buchholz (RETIRED) <rbu> |
Component: | Eclasses | Assignee: | TeX project <tex> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | jakub, qa |
Priority: | High | Keywords: | QAcanfix |
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- |
Description
Robert Buchholz (RETIRED)
2006-11-25 06:01:01 UTC
I fail to see your problem, virtual/mysql just depends on mysql or mysql-community, don't see how any revbump would solve anything. Run emerge --sync; emerge -uDav world. The problem is that I do not have virtual/mysql installed, because the old libwww did not depend on it. I do however have the actual mysql package installed, because the old libwww depended on that. Now the new ebuild of tetex checks for the existence of virtual/mysql because the new version of libwww should have it installed -- and it fails because the virtual is not there. This did happen after emerge --sync && emerge -uDN world and could be fixed by manually doing emerge virtual/mysql But if libwww was revbumped, it would not need manual action for being fixed. You don't need virtual/mysql installed, you need mysql or mysql-community. The virtual installs _nothing_. This plain doesn't make any sense, sorry. Oh cool, this comes from the "hundreds of bugs reporting "cannot find -lmysqlclient"" check in tetex.eclass. I have found exactly 3 - 14709, 67336 and 69308. How about that we stop checking for PEBKAC cases and remove the check? Ping! (QA) Although Robert was right about the revbump as virtual/mysql is a new style virtual the check still doesn't make much sense, hence removed. |