[Bug 1018162] Re: [MIR] mksh
Hrm, ok. As long as it’s indeed kept in sync by someone… I’d have preferred the pseudo-package way to have a zero delta, but that’s your decision. Thanks. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
Right, so, it passes my review, given the above security history, and the packaging itself, while esoteric, does the job, and modulo build- depends, we'd intend to keep it in sync with Debian anyway. I'm promoting it now, and will upload my diet-free version in ~30m. ** Changed in: mksh (Ubuntu) Status: New => Fix Released -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
Indeed, I'm going to do an upload that drops the dietlibc-dev build-dep, and then look at approving this MIR, which should be a no-brainer for the pdksh->mksh switch, but due diligence and all that. Or, rather, I'll review and promote it, then upload the new one, so it's actually built in main. :P -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
Re: [Bug 1018162] Re: [MIR] mksh
On Fri, Jun 29, 2012 at 08:18:43PM -, Thorsten Glaser wrote: > OK. With dietlibc, with dietlibc defused by depending on some empty > pseudo-package that only exists in Ubuntu, or with a diff against the > Debian package (not recommended)? A delta against the Debian package to drop the dietlibc build-dep would be the preferred solution here. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
OK. With dietlibc, with dietlibc defused by depending on some empty pseudo-package that only exists in Ubuntu, or with a diff against the Debian package (not recommended)? If pseudo-package, I’ll probably do another mksh upload before the freeze for some minor string and documentation fixes and can squeeze that in if it’s here in time. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
oh, but of course graphviz and shunit2 aren't referencing it as pdksh, they're referencing it as ksh. So these are not frivolous dependencies after all, and it looks like mksh should be MIRed into main. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
Removing the pdksh source package from quantal; this is not a bug in that package, so marking invalid. As for mksh being promoted to Ubuntu main, I can't see any reason why pdksh was needed in the first place. The two build-dependencies on the package look quite frivolous to me. I think it would be better to remove these build dependencies (in Ubuntu and in Debian) instead of promoting mksh to main. ** Changed in: pdksh (Ubuntu) Status: New => Invalid -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
[Bug 1018162] Re: [MIR] mksh
OK, it’s happened, mksh has been synched, pdksh is now obsolete. Availability: mksh has until now built on all Ubuntu platforms and should continue to do so; if not, an active maintainer (mirabilos) will review build logs and take actions. Rationale: The package replaces another package we currently support (pdksh) and promises higher quality (it does!) and/or better features (it does!), so that we can drop the old package from the supported set. Security: yep Quality assurance: all points met UI standards: no translation, but that’s not a regression relative to pdksh. Also, this is a shell intended for people knowing how to use a command line, or for running scripts, so it’s not likely to face a user unable to understand it. Furthermore, many system errors (from calls to strerror() and friends) are internationalised by libc. Dependencies: only src:dietlibc is in universe, all other B-D, Depends and Recommends are in main; dietlibc is an optional build-dependency (meaning the package compiles without it installed without any patching required), and klibc is normally used in preference nowadays anyway (which is in main), so dropping dietlibc would not change the package. (However, unless Ubuntu adds a pseudo-package, the Build-Depends line cannot/will not be changed in Debian to exclude dietlibc-dev. Build- Depending on "dietlibc-dev | some-ubuntu-pseudopackage" would be acceptable.) Adding dietlibc to main is subject to your deliberation, you could think about it. Standards compliance: ok Maintenance: upstream is maintainer in Debian, very active; pdksh was orphaned Background information: see above mksh had one CVE for an esoteric feature, which was immediately fixed. Code is regularily reviewed by tools (llvm+clang scan-build, Coverity scan/prevent) and people (e.g. during the Android inclusion process). No suid, sbin, dæmons or network. No plugins. Please advice on whether an MIR for dietlibc would be considered. -- You received this bug notification because you are a member of Ubuntu Bugs, which is subscribed to Ubuntu. https://bugs.launchpad.net/bugs/1018162 Title: [MIR] mksh To manage notifications about this bug go to: https://bugs.launchpad.net/ubuntu/+source/mksh/+bug/1018162/+subscriptions -- ubuntu-bugs mailing list ubuntu-bugs@lists.ubuntu.com https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs
