Summary: | net-print/hplip-3.9.4b hp-setup not working | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Petteri Räty (RETIRED) <betelgeuse> |
Component: | [OLD] Printing | Assignee: | Denis Dupeyron (RETIRED) <calchan> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | a0401241, billie, cruzki123, flameeyes, junk4me46806, kirikaza, leio, printing |
Priority: | High | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
URL: | https://bugs.launchpad.net/hplip/+bug/399522 | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Attachments: |
Error message of hp-setup -i with USE=-qt4
fix for hplip problem |
Description
Petteri Räty (RETIRED)
2009-06-17 19:08:20 UTC
Same problem here. I suspect the hplip-3.9.4b-plugin.run has changed, to lose the -i option, or it was not tested during the bump and the newer version has changed. Upstream report: https://bugs.launchpad.net/hplip/+bug/399522 Does the solution from the upstream bug work? (In reply to comment #3) > Does the solution from the upstream bug work? > yes it works fine Can anybody try this with hplip-3.9.10. Running "hp-plugin -i" works fine here, but I can not check with hp-setup because my printer does not require a plugin. Whith 3.9.12 did not work. Making the change sugested in the upstream AND instaling pygobject (as sugested by this thread https://answers.launchpad.net/hplip/+question/68660) I can finally install my printer. (In reply to comment #6) > Whith 3.9.12 did not work. > > Making the change sugested in the upstream AND instaling pygobject (as sugested > by this thread https://answers.launchpad.net/hplip/+question/68660) > I can finally install my printer. > The link you posted above seems unrelated to our problem here. The launchpad question is about missing dependencies which should be installed when installing hplip. From Petteri's output there must be a problem after hp-setup is trying to run the the downloaded run script. From hp-setup it is called with: "sh hplip-3.9.12-plugin.run -- -i" or "-u" depending on how hp-setup is called. When I run this command the plugin installs successfully. I tested with the -u and with the -i switch. As mentioned above I can not run it from hp-setup as this part is only executed if the printer in question needs a plugin. Please can anybody experiencing this problem run "sh hplip-3.9.12-plugin.run -- -ig". This should give extra debug output. I misscopy the link. The correct one was: https://bugs.launchpad.net/hplip/+bug/399522 But after doing this I hit the problem that the other link said. (In reply to comment #8) > I misscopy the link. The correct one was: > > https://bugs.launchpad.net/hplip/+bug/399522 > > But after doing this I hit the problem that the other link said. > I guess you have the qt4 use flag disabled for hplip as enabling it pulls in the required dependencies. Probably the python dbus bindings and pygobject are also needed when using the non gui mode but this would be another issue. Can you please download the plugin from http://www.openprinting.org/download/printdriver/auxfiles/HP/plugins/ and run the command I mentioned above. I forgot to said that after intaling pygobject I can finally run hp-plugin and the command you provide. And I haven't enable de qt4 flag, so I think this would be the problem. Thanks for the help ;) *** Bug 309013 has been marked as a duplicate of this bug. *** +*hplip-3.10.2-r1 (20 Mar 2010) + + 20 Mar 2010; Daniel Pielmeier <billie@gentoo.org> + -files/hplip-3.9.4+glibc-2.10.patch, +hplip-3.10.2-r1.ebuild, + +files/hplip-3.10.2-plugin.patch: + Revision bump. Update ebuild to use newer python eclass related functions. + Add a patch which fixes plugin installation from bug #274538. Remove + unused patch. Please test if this version fixes the plugin installation. *** Bug 316777 has been marked as a duplicate of this bug. *** + 04 May 2010; Daniel Pielmeier <billie@gentoo.org> + files/hplip-3.10.2-plugin.patch: + Another try to fix plugin installation. This should fix bug #274538 and + #316777. Apparently the upstream patch did not fix the problem so I replaced it with a new one. See also bug #316777. The user confirmed that the issue is fixed with the new patch, so I will close this for now. If there are still issues please reopen. (In reply to comment #14) > + 04 May 2010; Daniel Pielmeier <billie@gentoo.org> > + files/hplip-3.10.2-plugin.patch: > + Another try to fix plugin installation. This should fix bug #274538 and > + #316777. > > Apparently the upstream patch did not fix the problem so I replaced it with a > new one. See also bug #316777. The user confirmed that the issue is fixed with > the new patch, so I will close this for now. If there are still issues please > reopen. > I just had the same problem with 3.10.6. It appears the plugin patch was dropped between the 3.10.5 and 3.10.6 ebuilds. I applied a similar patch (solved the issue on my own before I found this bug) and it fixed the issue. (In reply to comment #15) > I just had the same problem with 3.10.6. It appears the plugin patch was > dropped between the 3.10.5 and 3.10.6 ebuilds. I applied a similar patch > (solved the issue on my own before I found this bug) and it fixed the issue. > The patch was dropped intentionally as upstream added a fix for this problem. They did not apply the patch [1] I provided in [2] but this [3] one instead. This is also why you experience bug #331163. [1] http://launchpadlibrarian.net/47809379/hplip-3.10.2-plugin.patch [2] https://bugs.launchpad.net/hplip/+bug/399522 [3] http://launchpadlibrarian.net/49192341/core_install.patch I ran some tests that are documented in Bug 331163 Comment 2 [1] that shows that my system does not work with the new patch (Listed as [3] above). [1] http://bugs.gentoo.org/show_bug.cgi?id=331163#c2 [3] http://launchpadlibrarian.net/49192341/core_install.patch I think that part of the problem is that I don't have the qt4 use flag enabled. The documentation states that -u is for gui (normal default) and -i is for interactive. If I run hp-setup -i, it gives an error indicating an unrecognized option. Running hp-setup with no arguments works and run in interactive mode. This is also true of hp-plugin. To me, the correct solution is to find out why -i is not accepted when USE=-qt4 and fix that (then the old patch is not needed). The error message from hp-setup -i is attached. The funny part is that the error message shows usage info for -i and even includes examples of using -i and then complains that -i is invalid. Here are my use flags for hplip: qwerty ~ $ emerge -pv =net-print/hplip-3.10.6 These are the packages that would be merged, in order: Calculating dependencies... done! [ebuild UD] net-print/hplip-3.10.6 [3.10.6-r1] USE="X hpcups libnotify scanner -doc -fax -hpijs -minimal -parport -policykit -qt4 -snmp -static-ppds -udev-acl" 0 kB [1=>0] Total: 1 package (1 downgrade), Size of downloads: 0 kB Portage tree and overlays: [0] /usr/portage [1] /usr/local/overlay/ Note that I currently have a version my a local overlay installed that requires dbus-python and applies the "old" patch (I want printing to work while this bug is being processed). I had the portage version installed when I ran the tests in bug 331163. Created attachment 241515 [details]
Error message of hp-setup -i with USE=-qt4
(In reply to comment #17) > I think that part of the problem is that I don't have the qt4 use flag enabled. > The documentation states that -u is for gui (normal default) and -i is for > interactive. If I run hp-setup -i, it gives an error indicating an > unrecognized option. Running hp-setup with no arguments works and run in > interactive mode. This is also true of hp-plugin. To me, the correct solution > is to find out why -i is not accepted when USE=-qt4 and fix that (then the old > patch is not needed). The error message from hp-setup -i is attached. The > funny part is that the error message shows usage info for -i and even includes > examples of using -i and then complains that -i is invalid. If you build hplip without GUI (USE="-qt4") there is no -i or -u flag available because only interactive mode is possible. The confusing thing is that hp-setup lists the options as available. Created attachment 292713 [details, diff]
fix for hplip problem
It seems post #8 at the following site has a working patch for this problem: https://bugs.launchpad.net/hplip/+bug/399522 From the post: Here is a patch that seems to fix this bug. It works on my machine. The issue, as I understand it is: A) Python scripts pass -i to other python scripts when running in interactive mode. B) When built without the gt4 gui_build option, -i is not a valid option. The previous patch from comment #4 attempted to fix issue part A. This patch was rejected because it affected the behavior when built with gui_build (hp-setup -i will still launch the gui because hp-plugin is called with no arguments and its default is to launch the gui.). The patch from comment #6 did not affect code execution, but instead printed a message to the user to install python dbus and python gobject. The patch operated under the assumption that installing these packages would fix the issue. I have both of these packages installed (gentoo packages dev-python/dbus-python-0.83.0-r1 and dev-python/pygobject-2.20.0) and am still seeing this bug. Comment #7 reiterates this point. The attached patch address issue part B, which I believe is the correct place to fix this bug. The attached patch always allows the -i option as long as interactive mode is available. Without this patch, if interactive mode was the only version, -i would receive an error even though the program could run in interactive mode just fine (in fact that is what the user desires). This patch has the advantage of not changing the current behavior of the code. When built with gui_build, this patch will have no effect since -i was already accepted in that case. When built without gui_build, this patch will fix this bug and allow proper execution. |