Fault reports and bug fixes and expectations

Alan Coopersmith alan.coopersmith at oracle.com
Tue Jun 18 16:29:50 UTC 2019


On 6/17/19 3:31 AM, Wheatley, Martin R wrote:
> Engineers in our control room really would like the action they request to 
> complete rather than suffer from a random failure (as identified in the bug) so 
> my question is:
> 
> If Oracle will only take the Xorg source rather (as opposed to fixing it - or 
> even discussing a fix with Xorg - and pushing the fix back to Xorg)

Oracle does contribute fixes to Xorg for issues as necessary - but the time we
have for that is very limited, and we prioritize issues based on how many people
are affected.   (And we absolutely prioritize work on Solaris 11.4 over releases
in extended support, nearing their end of life, like Solaris 10.)

This code in xauth has been in place for 25 years now and it works for millions 
of users as they only write to their .Xauthority file at the start of the
session.  (And in modern desktops, they have a separate file per session, set
via the XAUTHORITY environment variable, so even in the rare scenarios where
they start a second session with the same home directory, they don't clash.)

If you are regularly hitting this, I'd ask what you are doing that requires
writing to the .Xauthority file so much more often than everyone else in the
world.  Do you need to write new xauth records so often?  Do you even need to
write them at all?   Why not just use "xhost +si:localuser:...." to grant the
user access instead of relying on the old shared secret method?

>  who and when 
> will the issue be fixed so that I can put pressure on Oracle to generate a patch 
> from the new Xorg code thus improving  the reliability of my control room 
> operation? 

You are demanding a team of volunteers provide you an exact date for when
they'll provide you free help?   You're lucky that someone has taken interest
in your bug and decided to work on it - many bugs linger for years because
there's simply far too few people contributing to X.Org to handle them all,
and most of the ones who do contribute have more interesting things to work on.

> > After all, do you want to be on a plane when the pilot says “we 
> pushed the button to lower the undercarriage – most times the wheels do come 
> down but sometime they might now!”???

If the wheels on a plane fail to lower at landing time, the manufacturer will
investigate.   If they find that they work reliably on every plane in the fleet
where they are only lowered at landing time, and not on the one plane where they
are constantly raised and lowered during flight, they are not going to issue a
recall and change the landing gear design across the entire fleet, but ask the
pilots of that plane to reconsider why they are doing something so unique.

	-alan-

[Absolutely speaking solely for myself, not anyone else.]


More information about the xorg mailing list