| Summary: | aspell / emacs / gpm circular dependency | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | mike <mikezero1> |
| Component: | New packages | Assignee: | Gentoo's Team for Core System packages <base-system> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | CC: | app-dicts+disabled, emacs, galtgendo, gentoobugzilla, thif |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
| Bug Depends on: | 33440 | ||
| Bug Blocks: | |||
| Attachments: | gpm-1.20.1-installdirs-ELISP.patch | ||
|
Description
mike
2005-07-19 07:59:33 UTC
May I ask how is this related to Gnome? Reassigning. Are you installing Gentoo? If you, then the correct way is 'emerge --emptytree system'. I put it as Gnome, becuase I was under the impression that it was a gnome dependency. I did start with an "emerge --emptytree system" that was where aspell failed the first time. In my subsequent efforts to try to emerge it I have just been using emerge system or emerge aspell. (In reply to comment #3) > I did start with an "emerge --emptytree system" that was where aspell failed the > first time. In my subsequent efforts to try to emerge it I have just been using > emerge system or emerge aspell. You did not answer my question at all. Is this a fresh Gentoo install? Either way, try to emerge sys-libs/gpm first. yes, this is a fresh install. gpm has aspell as a dependency, should i try to emerge it --nodeps? (In reply to comment #5) > yes, this is a fresh install. > > gpm has aspell as a dependency, should i try to emerge it --nodeps? No, gpm does not have aspell as a dependency, (In reply to comment #5) > yes, this is a fresh install. > > gpm has aspell as a dependency, should i try to emerge it --nodeps? Oh, now it finally makes sense... app-text/aspell with USE="gpm" depends on sys-libs/gpm which with USE="emacs" depends on virtual/emacs which with USE="spell" depends on app-text/aspell which... Yes, try with --nodeps... ok, it's compiling now. You were right. Circular dependencies. Thank you for the help. -mike (In reply to comment #7) > ok, it's compiling now. You were right. Circular dependencies. Thank you for > the help. Good that it works now. @usata: Seems that the fix for Bug 80217 has broken aspell now, instead of gpm. ;) CCing the herds maintaining the ebuilds that are causing this dependency problem, perhaps someone will find a solution. we can always just punt emacs support from gpm, i dont mind Since the emacs support in gpm just seems to be from a contrib directory anyhow, that sounds good to me. Another idea is to have a separate emacs-gpm package that installs the emacs support for those who want it. waste of space i say ... doubt anyone actually uses that ;) I just ran into this bug doing an emerge --emptytree system. It confused me for a bit because I didn't realize that the gpm USE flag was, by default, enabled...had to explicitly disable it in my make.conf. One possibility for avoiding it then would be to make gpm not enabled by default. Maybe this would be bad though...? Why not make emacs an RDEP of gpm instead of DEP? wouldnt help, portage treats them the same besides, gpm needs emacs in order to compile support for it i'll just punt it from gpm gpm no longer supports emacs *** Bug 104296 has been marked as a duplicate of this bug. *** *** Bug 110213 has been marked as a duplicate of this bug. *** Please note that Emacs 22 and later include t-mouse.el, which was formerly installed by sys-libs/gpm. (So the commented lines for Emacs support in the gpm ebuild should probably be removed now.) *** Bug 202946 has been marked as a duplicate of this bug. *** (In reply to comment #19) > *** Bug 202946 has been marked as a duplicate of this bug. *** ... but addresses a valid issue, since gpm-1.20.1-r6 installs a spurious /usr/share/emacs/site-lisp directory. Reopening, patch follows. Created attachment 139061 [details, diff]
gpm-1.20.1-installdirs-ELISP.patch
This fixes the issue for me. It should be included into the patchset.
then you re-open the original bug report ... you dont re-use an old bug |
