Closing this bug.

Status: NEW => RESOLVED
Resolution: (none) => FIXED

Sorry, dmorgan. I'm using your bug to test flags (well, to convince QA that flags are great and useful). You can click the "Ignore Bug Mail" checkbox at the top right of the page to not get spam for this bug. :)


Pseudo-comments:

Things look good on Mageia 6 i586.

@David and Akien: can you test on Mageia 6 x86_64 if the solution proposed in this bug fixes this issue? ;)

@neoclust: can you check with Mageia 5 and GNOME as I know you still use it? :-p

It would be nice if someone else could also test on Mageia 5.

Status: RESOLVED => UNCONFIRMED
Summary: test => test flags (a.k.a flags are great!)
Flags: (none) => test-mga5-32?, test-mga5-64?(mageia), test-mga6-64?(davidwhodgins), test-mga6-64?(rverschelde), test-mga6-32+
CC: (none) => davidwhodgins, LpSolit, mageia, rverschelde
Ever confirmed: 1 => 0
Resolution: FIXED => (none)

(In reply to Frédéric Buclin from comment #2)

> Pseudo-comments:
<snip>
> 
> It would be nice if someone else could also test on Mageia 5.

More pseudo-comments:

32 bit Mga5 tested fine here, using XFCE, but 64bit (again with XFCE) failed.

Was it OK to add a set of flags to be able to set another test-mga5-64 flag (since the existing one seemed meant for neoclust) and to set the already existing one for test-mga5-32?

How many sets of test-mga*-* flags can be created in one bug report?

Flags: test-mga5-32? => test-mga5-32+, test-mga5-64-
CC: (none) => marja11

In the set of flags I added only the one I had set to "-" was kept.... I think I had also set one to "?", with LpSolit in the field next to it... will test that again

Another pseudo-comment:

tested with Mate now, Mga5 64bit, and the test fails again.
Adding  "test-mga5-64" "-", but also adding 
"test-mga5-32" "?" "LpSolit@netscape.net, "
because I want to see that that gets saved :-)

Flags: (none) => test-mga5-32?(LpSolit), test-mga5-64-

(In reply to Marja van Waes from comment #4)
> but also adding 
> "test-mga5-32" "?" "LpSolit@netscape.net, "
> because I want to see that that gets saved :-)

it does :-D

(In reply to Marja van Waes from comment #3)
> Was it OK to add a set of flags to be able to set another test-mga5-64 flag
> (since the existing one seemed meant for neoclust) and to set the already
> existing one for test-mga5-32?

It is totally fine. :)

 
> How many sets of test-mga*-* flags can be created in one bug report?

Unlimited. You can even create several flags of the same type at once by listing users in the "requestee" field. For instance:

test-mga6-64?marja,lpsolit will create two requests:

test-mga6-64?marja
test-mga6-64?lpsolit

Cool, isn't it? :)


Pseudo-comment:

(In reply to Marja van Waes from comment #3)
> 32 bit Mga5 tested fine here, using XFCE

It fails when testing with LXQt. I get a blank screen. No idea why.

End of pseudo-comment.



Note that with flags, we now have:

marja11: test-mga5-32+
LpSolit: test-mga5-32-

So without reading comments, you already see that at least one tester had problems with Mageia 5 i586. So you know it would be inappropriate to add the MGA5-32-OK keyword.

@David: I think this is a good demonstration of why flags would be useful. Because each tester is free to set his own flag. This doesn't impact what other testers say, and you have an immediate global view of the situation.

Flags: test-mga5-32?(LpSolit) => test-mga5-32-

Testing flags...

pclx: test-mga5-64+

Flags: (none) => test-mga5-64+
CC: (none) => mageia

(In reply to Frédéric Buclin from comment #6)

In reply to my earlier:  Flags: (none) => test-mga5-32?(LpSolit@netscape.net),
you did:
Flags: test-mga5-32?(LpSolit@netscape.net) => test-mga5-32-

I received a new type of Bugzilla mail about that, one that has no footer, but has this at the top of the mail:


> Frédéric Buclin <LpSolit@netscape.net> has denied Marja van Waes
> <marja11@xs4all.nl>'s request for Frédéric Buclin <LpSolit@netscape.net>'s
> test-mga5-32:
> Bug 5: test flags (a.k.a  flags are great!)
> https://bugs.mageia.org/show_bug.cgi?id=5

I think for flags about testing that message is confusing: I asked you to test with mga5-32, that's what you did and the test failed. If you had denied my request, you wouldn't have tested ;-)

However, getting a separate mail about the flag-request is great!

Is a message like this one generic enough for all kinds of flags?:

> Frédéric Buclin <LpSolit@netscape.net> has responded to Marja van Waes
> <marja11@xs4all.nl>'s request for Frédéric Buclin <LpSolit@netscape.net>'s
> test-mga5-32:
> Flags: test-mga5-32?(LpSolit@netscape.net) => test-mga5-32-
> Bug 5: test flags (a.k.a  flags are great!)
> https://bugs.mageia.org/show_bug.cgi?id=5

(Or is it possible to have different messages for different flag types?)

(removing duplicated flags)

(In reply to Marja van Waes from comment #8)
> I received a new type of Bugzilla mail about that

Yes, they are called flag mails, which are different from the usual bug mails.


> I think for flags about testing that message is confusing: I asked you to
> test with mga5-32, that's what you did and the test failed. If you had
> denied my request, you wouldn't have tested ;-)

That's the default flag mail. :) Upstream, we had to be as generic as possible to match as many words for the flag name as possible.


> (Or is it possible to have different messages for different flag types?)

Not by default, but I could customize them if needed.


Anyway, per David Hodgins in bugsquad@, I have the feeling that he doesn't want flags in the QA process. :(

Flags: test-mga5-64-, test-mga5-64+, test-mga6-64?(davidwhodgins) => (none)

Created attachment 9607 [details]
ignore me