-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Alec Warner wrote: >>>> Zmedico did a lot of things with usage of global variables, however I >>>> think that getting all that tested ( especially in scripts that we don't >>>> keep track of ) is detremental to getting portage stable. I agree that >>>> it's an important step; however it's just code cleanup. It is not >>>> necessary for 2.1. >>>> >>>> I'm looking at the diff from pre9 and pre10, and I will backport any >>>> bugfixes if that makes everyone happy. >>>> > > In my cleanup of globals I took special care to maintain backward > compatibility. I have mr_bones_ and halcyon doing profiling to > track down the cause of the repoman performance issue. It should be > a simple fix when we find the problem. > > Zac > >> As most people would expect you to, however there are no regression >> tests to prove you succeeded other than just having people test it.
Well, it's been the tree for 2 days now we'll surely get bug reports as soon as people run into these hypothetical issues (though I expect very few, if any regressions). I think the globals cleanup is worth having in 2.1 because it makes the code more maintainable. If you want to move back to a more stable revision, I'd suggest 2.1_pre7 (before manifest2). I believe manifest2 introduced more potential for regressions than the globals cleanup did. Zac -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (GNU/Linux) iD8DBQFEVpLl/ejvha5XGaMRAjH4AJ9lezus6+SGKyg3KzKddwEE1ADwOwCfbAxv Rcduq3huWTa7Dlw5Nbb+r0c= =JXxC -----END PGP SIGNATURE----- -- gentoo-portage-dev@gentoo.org mailing list
