I did not know that the intel driver was borked, i updated my and installed fresh the system from my daddy in the beta phase of ubuntu, i did not know that there are special isues with the intel driver. And i could not imagine that the driver stays in this state till release.
I did not have such problems with as example the free ati driver ever. The progress is not always that fastk, so you wait sometimes very long for new features especialy for new hardware.
But I never had more problems with a newer driver than with an older one.
But i agree thats not only problem of xorg-people (especialy intel-devs), ubuntu-devs did not had to use this bloody driver version. But its also politics of the intel devs.
To use the short way and stop support for old driver-code while the new isn't without big isues.
The problem by feature-driven releases its a serial development-process what is not really needed with git and other distributed systems, if you have devoloped 10 from 12 features in 4 months and the last 2 ones take to finish 12 months all who dont want to use unstable releases have to wait 16 months for the 10 features who were already done after 4 months. Thats not good (for the users)
Also there is a clock ticking if the last release is long ago the developer wants to release, because much users want the feature/fixes or the developer need more testers. But if the last 2 features are not really ready the developer tests the stuff less or release sub-stable code to get it finaly out.
So have the choise between realy long release-cycles which are not even calculatable (moving of release-dates) or unstable releases, i think both stinks
I did not have such problems with as example the free ati driver ever. The progress is not always that fastk, so you wait sometimes very long for new features especialy for new hardware.
But I never had more problems with a newer driver than with an older one.
But i agree thats not only problem of xorg-people (especialy intel-devs), ubuntu-devs did not had to use this bloody driver version. But its also politics of the intel devs.
To use the short way and stop support for old driver-code while the new isn't without big isues.
The problem by feature-driven releases its a serial development-process what is not really needed with git and other distributed systems, if you have devoloped 10 from 12 features in 4 months and the last 2 ones take to finish 12 months all who dont want to use unstable releases have to wait 16 months for the 10 features who were already done after 4 months. Thats not good (for the users)
Also there is a clock ticking if the last release is long ago the developer wants to release, because much users want the feature/fixes or the developer need more testers. But if the last 2 features are not really ready the developer tests the stuff less or release sub-stable code to get it finaly out.
So have the choise between realy long release-cycles which are not even calculatable (moving of release-dates) or unstable releases, i think both stinks
Comment