Jump to content
Read the Funtoo Newsletter: Summer 2023 ×

Oleg Vinichenko

Members
  • Posts

    619
  • Joined

  • Last visited

  • Days Won

    53

Everything posted by Oleg Vinichenko

  1. i believe the 2.3.8-r8 portage version supposed to fix this. please, update.
  2. do not use backtrack, its an artificial hack withing a portage, which is -r6 is attempting to fix. is any of mentioned ebuilds are present in /var/lib/portage/world? thats boost, ffmpeg, libcd1394, libcdio-paranoia, boost-build
  3. firstly update to portage-2.3.8-r6, then do not use backtrack option at all and see if it changes
  4. yes, there is a package.provided feature. adding category such as: x11-drivers/nvidia-drivers-381.22 into /etc/portage/package.provided will make portage think that this exact version installed on users box, when in reality it's not. notice a syntax, if you need specific version (381.22) you don't put atom ("=" in this example) in front of category, unlike it's needed for masks or unmasks or keywords.
  5. its not possible to get a clue about this failure and why echo='echo' fixes. paste build.log of libxml2 failure.
  6. this is pretty much happens often with portage unable to solve conflicts. Daniel is working on improving this, notice that during many years portage developers didn't actually tried to fix it. For this particular problem: akonadi cannot work with maridab-10.2, and 10.2 is the version you have installed on your box. if you put maridab-10.2 into package.mask, it will solve.
  7. this is quite unclear what happened on your box. updating old box or migrating to meta-repo? setting of python_targets in make.conf is safe but it shouldn't be actually needed. so far it's hard to track the exact steps were on your box, can you elaborate?
  8. downgrades are expected. put package.mask for dev-util/pkgconfig for now. i'll look for this issue in details.
  9. echo ">xfce-extra/xfce4-xkb-plugin-0.7.1-r1" >> /etc/portage/package.mask this is what i have on my laptop too.
  10. if you are not comfortable with frequent vim updates, just put the mask from the version you have installed and then control when you need to update :) there could be prime branch for editors, but it's uncertain atm.
  11. you can do reinstall but now, it's maybe a good idea is to try using unmasked versions. so, now i guess it's python-3.6 is now the default python?
  12. i'm not able to understand this failures based on giving outputs. something went broken on your system due to non default settings, like unmasked pythons, etc.
  13. this is due to the portage and ego both are built for different python versions. show output of eselect python list
  14. so, the reason of failing is that your toolchain (gcc, glibc, binutils) built without multilib support but your profile is multilib, did you by chance disabled it? maybe you installed from pure64 stage and then changed profile? to fix this you might resurrect your toolchain packages from stage3 (make sure it's multilib, i.e do not use pure64 for that), follow this guide: http://www.funtoo.org/System_resurrection tbz2 way will work for you, so get the fresh new stage, mount it somewhere, prepare bin packages for glibc, gcc, binutils, linux-headers and then extract into your system as described.
  15. alright, there seems edit performed to add new information. you should not use that gentoo file in /etc/portage/repos.conf. ports-2017 was alpha version of meta-repo, please follow: https://github.com/funtoo/meta-repo/blob/master/README.rst and notice the overlay wiki page i posted above and pay attention to the symlink creation, so that you will be able to use meta-repo and sage overlay.
  16. this looks like your toolchain has disabled support for multilib. paste output of epro show-json
  17. the idea would be to use science overlay as you had before apply the instruction from: http://www.funtoo.org/Local_Overlay
  18. this is fixed already, you need to update your portage tree.
  19. currently, you can fix this by applying the mask for mariadb-10.2
  20. it could happen that you didn't rebuild the ebuilds which affected by ABI changes between gcc-4 and gcc-5. Follow http://www.funtoo.org/News:GCC_Update
  21. 2.23-r4 would be glibc, not glib. Show output of gcc-config -l
  22. http://www.funtoo.org/News:Python_Multiplexing(And_another_Ego_update)
×
×
  • Create New...