Hi, On 17-03-2019 03:47, Axel Beckert wrote: > Hi Paul, > > Paul Gevers wrote: >> I am going to raise my question again, if synaptic can't be run on the >> default desktop in the default login mode, isn't the average user better >> of if we don't ship synaptic with buster? > > Huh? What kind of logic is that? So any X program which doesn't run > under Wayland but does work under X should not be shipped at all? WTF? > > Of course with these issues, it doesn't make sense to install Synaptic > by default. But it's still useful for anyone who doesn't use Wayland, > which is AFAIK anyone who has no properly accelerated GPU.
I am not contesting that. However, synaptic is a relative popular package with quite some votes, which is working on the default desktop with the default login mode in stretch. It now *fails* in buster in a very user unfriendly way (on x86 as you mention below, but that *is* our largest arch). If the failure mode would be better, I wouldn't be asking this question and would probably have set this bug as buster-ignore, but with the current situation I'd hate to waste the time of the current users of synaptic (running default GNOME on x86) in stretch debugging the situation when they upgrade. Also, I wasn't aware of the "accelerate GPU" thing, so thanks for bringing that up. Can anybody guess how much x86 users would have those? That is already a thing for quite some time, no? If somebody improves synaptic to give a proper error message on Wayland, then I'll unblock that improvement on the condition that synaptic is still in buster and we're not too close to releasing buster. > And besides: Isn't GNOME/Wayland only on x86 architectures the > default, but on any other architecture XFCE (with X I assume) is the > default desktop because GNOME doesn't properly work on them? Following > the same logic, shouldn't we stop shipping GNOME? </cynicism> For the reasoning I set out above, it's not the failure I consider the worst issue here, but the regression without proper error message. Paul
signature.asc
Description: OpenPGP digital signature
