I must say I'm glad you did respond as it means you take it seriously.

While you take it as a personal attack, it was merely a description of what I 
have seen and experienced.
Attacking is something I associate with name-calling with nothing backing it up.
A performance-review from a boss that describes problems with an employee's 
results is not something I would consider an attack.

I know 1st-tier phone support where the person I'm speaking with has no idea 
regarding the issue and is reading from a script that is not applicable to the 
issue, when all that is needed is to advance me to someone who would be able to 
provide deeper support (test a patch).
It's the kind of support that would be equal to "reboot your computer 3 times" 
when Excel is selecting multiple cells with a single click. It would be more 
appropriate to see if the mouse button has bounce, but even then it could still 
be a software bug.
That is the impression I get when I'm asked to test a newer kernel or OS to see 
if it would fix the issue.
It was never said "why" it was thought it would fix the issue (eg, reference a 
pull-request that addresses an issue that might be remotely applicable to the 
current issue).
I can see the value of 1st-tier in order to filter out "newbs" for the 
developers when a configuration/setting change is all that is needed, but as 
much as the tier can be helpful for the developers (and newbs), they can also 
be headaches for those with issues that need deeper triage.

"... why one is asked... documented, both in those comments..."
I must be missing those comments you are referencing that provides the "why".
My "why" questions:
https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1516531/comments/29  "What 
is it about Xenial..."
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1465862/comments/11  
"...wondering what this 'testing' is supposed to be doing..."
Other's bug's that ask for a test with an upgrade, but no "why" given:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1523108/comments/4
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1506914/comments/4 "Can 
someone please explain what..."

With the wiki, I think you are probably referencing "Please do not complain 
because someone sent what one perceives to be a automated or "canned" response."
I also see this issue as "Please check to see if you(r) problem is a 
regression." so I'd say bisecting should be done rather than about testing 
upstream. But I know by saying just that goes against "Please avoid arguing 
with triagers and developers".

For me, testing a newer kernel is less complicated than testing a whole new OS 
that is only in the early testing stages when I'm dealing with a "production" 
(day-to-day use) computer.
It would involve:
1) backing things up and hoping that backup worked.
2) upgrading (in my case, I always start with a fresh install)
3) restore, which never goes well with an encrypted home. something I have to 
do from live-cd.
4) installing programs, which also means building some from scratch like kodi 
which also means building the addons.
Even if I got there after hours (day(s)?) of work, the testing OS may not fix 
the issue and may also have other unusable problems.

Note how I actually do the footwork that I consider reasonable. I'll do
research before submitting a report. I do research to back up my
comments. This isn't off-the-cuff stuff that I am doing.

Also I WON'T be marking this report as "Invalid" as it's VERY valid considering 
how it affected Kodi and Chrome (haven't tested totem or games in fullscreen).
But YOU may mark it as incomplete since I am unlikely to test xenial, 
especially since the break happened in 15.10 (worked in 15.04).
I may put xenial on a VM to get a feel for it's current state, but based on my 
experience with UX bugs in 15.04 and 15.10, if I'm going to have to clean house 
on my system, I'd just as soon avoid 16.04 (not even alpha yet) and see if 
other dists come through the pipeline clean(er/ly).
I'll have to back up my data anyway, so I may just do an in-place upgrade to 
xenial and if it fails with fullscreen or some other notable bug, I'll go ahead 
an move on to a different dist.

I may look at Arch since the documentation I've seen online, when searching for 
system-stuff, has been better than Ubuntu. Ubuntu does have plenty of 
documentation, but they are all over the place with multiple versions that are 
about the same exact thing.
Again, I wouldn't call that an attack, but merely pointing out a problem. I'd 
provide links to examples, but this comment is already getting long-winded and 
this paragraph doesn't relate specifically to the "help" issue (though it is 
about docs which was mentioned in your last comment).

I won't be offended or hurt if you want to reference these comments as what to 
(do/not do).
If I'm in the wrong, so be it. I currently don't see it. I'm trying to do good 
by submitting a bug and providing good data and in turn I'm only asked to test 
an upstream (unstable/pre-alpha) OS on my personal system with no basis 
provided on why it is thought it might work.

similar/side note: this reminds me of an article I've seen on
reddit/slashdot referencing sjw vs linus.

-- 
You received this bug notification because you are a member of Ubuntu-X,
which is subscribed to xorg in Ubuntu.
https://bugs.launchpad.net/bugs/1516531

Title:
  Fullscreen Freeze.

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/xorg/+bug/1516531/+subscriptions

_______________________________________________
Mailing list: https://launchpad.net/~ubuntu-x-swat
Post to     : ubuntu-x-swat@lists.launchpad.net
Unsubscribe : https://launchpad.net/~ubuntu-x-swat
More help   : https://help.launchpad.net/ListHelp

Reply via email to