Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Frank Murphy
On 08/08/10 20:19, Matt McCutchen wrote:

>
> My advice: stop doing that.

Taken


-- 
Regards,

Frank Murphy
UTF_8 Encoded
Friend of Fedora
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Matt McCutchen
On Sun, 2010-08-08 at 19:47 +0100, Frank Murphy wrote:
> On 08/08/10 19:28, Matt McCutchen wrote:
> 
> >> By that I mean the maintainer\Co-Maintainer\Good Citizen
> >> retelling the same fix\workaround posted earlier in the bug.
> >> To either enable\updates-testing or do xyz.
> >> Because the bug report hasn't been read by the "added to" person.
> >
> > As I said, I have never seen that happen.  Do you have an example?
> >
> 
> I do it constantly, on such, as the older version after
> workarounds would perhaps suggest,
> newer versions are not being attempted.
> https://bugzilla.redhat.com/show_bug.cgi?id=559484

My advice: stop doing that.  Remember that abrt adds the comment before
giving the user the link to the bug, so there is no way a user would
know to try the workaround until after their comment is added.  That
would be addressed by your proposed change to abrt.  But given the
current abrt design, when you see a "how to reproduce" comment appear,
the user may well go on to read the thread as their next step.  If they
don't, I would argue that it's their problem.  (Bug threads can get
long, and the useful comments hard to find, but that's a separate
issue.)  Hence, I do not believe there is a point in reposting the
workaround.

-- 
Matt

-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Matt McCutchen
On Sun, 2010-08-08 at 20:01 +0100, M A Young wrote:
> On Sun, 8 Aug 2010, Matt McCutchen wrote:
> 
> > Users should already be reading the existing information, because abrt
> > provides the link after it makes its changes (adding the user to CC and
> > adding a comment with the "how to reproduce" text, if it was nonempty).
> > The proposal would just be to reverse the order of the steps so that
> > users can, and hopefully will, avoid adding redundant comments.
> 
> It would also be good if this could happen before the debuginfo packages 
> are downloaded, which would mean that those with limited bandwidth could 
> contribute more easily if there was an existing bug.

That's not possible: the debuginfo packages are needed to generate the
stack trace, which is needed to find the existing bug.

-- 
Matt

-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread M A Young
On Sun, 8 Aug 2010, Matt McCutchen wrote:

> Users should already be reading the existing information, because abrt
> provides the link after it makes its changes (adding the user to CC and
> adding a comment with the "how to reproduce" text, if it was nonempty).
> The proposal would just be to reverse the order of the steps so that
> users can, and hopefully will, avoid adding redundant comments.

It would also be good if this could happen before the debuginfo packages 
are downloaded, which would mean that those with limited bandwidth could 
contribute more easily if there was an existing bug.

Michael Young
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Frank Murphy
On 08/08/10 19:28, Matt McCutchen wrote:

>> By that I mean the maintainer\Co-Maintainer\Good Citizen
>> retelling the same fix\workaround posted earlier in the bug.
>> To either enable\updates-testing or do xyz.
>> Because the bug report hasn't been read by the "added to" person.
>
> As I said, I have never seen that happen.  Do you have an example?
>

I do it constantly, on such, as the older version after
workarounds would perhaps suggest,
newer versions are not being attempted.
https://bugzilla.redhat.com/show_bug.cgi?id=559484


-- 
Regards,

Frank Murphy
UTF_8 Encoded
Friend of Fedora
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Matt McCutchen
On Sun, 2010-08-08 at 09:24 +0100, Frank Murphy wrote:
> On 08/08/10 03:25, Matt McCutchen wrote:
> 
> >>
> >> Would it be any benefit to the maintainers\bugzappers.
> >> If abrt opened the existing link, before it would report?
> >
> > And then what?  Encourage the user not to add a comment unless they have
> > new information?  If that is the proposal, I am in favor.
> >
> Yes that was the hope.
> But the main idea, was to get them to read the existing info.

Users should already be reading the existing information, because abrt
provides the link after it makes its changes (adding the user to CC and
adding a comment with the "how to reproduce" text, if it was nonempty).
The proposal would just be to reverse the order of the steps so that
users can, and hopefully will, avoid adding redundant comments.

> >> So there wouldn't need be be so many replies of
> >> please try xyz?, to help cut down on noise.
> >
> > I don't understand this part.  The noise I've seen consists of
> > abrt-added "me-too" comments with steps to reproduce.  I have never seen
> > someone repost the same workaround in response to such a comment.
> >
> By that I mean the maintainer\Co-Maintainer\Good Citizen
> retelling the same fix\workaround posted earlier in the bug.
> To either enable\updates-testing or do xyz.
> Because the bug report hasn't been read by the "added to" person.

As I said, I have never seen that happen.  Do you have an example?

-- 
Matt

-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-08 Thread Frank Murphy
On 08/08/10 03:25, Matt McCutchen wrote:

>>
>> Would it be any benefit to the maintainers\bugzappers.
>> If abrt opened the existing link, before it would report?
>
> And then what?  Encourage the user not to add a comment unless they have
> new information?  If that is the proposal, I am in favor.
>
Yes that was the hope.
But the main idea, was to get them to read the existing info.

> Adding the user to CC is still correct.

Agreed on that

Other users are responsible for
> disabling the "email me when someone else is added/removed to CC"
> preference if they so desire.
>
>> So there wouldn't need be be so many replies of
>> please try xyz?, to help cut down on noise.
>
> I don't understand this part.  The noise I've seen consists of
> abrt-added "me-too" comments with steps to reproduce.  I have never seen
> someone repost the same workaround in response to such a comment.
>
By that I mean the maintainer\Co-Maintainer\Good Citizen
retelling the same fix\workaround posted earlier in the bug.
To either enable\updates-testing or do xyz.
Because the bug report hasn't been read by the "added to" person.

-- 
Regards,

Frank Murphy
UTF_8 Encoded
Friend of Fedora
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-07 Thread Matt McCutchen
On Sat, 2010-08-07 at 11:51 +0100, Frank Murphy wrote:
> Thougth I ask for some feedback.
> 
> Currently when abrt finds an existing bug,
> it goes
> " similar bug found, (shows a link) adding you to cc"
> 
> looking at some of the bugs, that I have reported or cc'd on.
> Where someone earlier in the thread had sent a workaround/fix etc..
> and cc's just being added automatically.
> 
> Would it be any benefit to the maintainers\bugzappers.
> If abrt opened the existing link, before it would report?

And then what?  Encourage the user not to add a comment unless they have
new information?  If that is the proposal, I am in favor.

Adding the user to CC is still correct.  Other users are responsible for
disabling the "email me when someone else is added/removed to CC"
preference if they so desire.

> So there wouldn't need be be so many replies of
> please try xyz?, to help cut down on noise.

I don't understand this part.  The noise I've seen consists of
abrt-added "me-too" comments with steps to reproduce.  I have never seen
someone repost the same workaround in response to such a comment.

-- 
Matt

-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


Re: abrt thoughts pre-rfe q?

2010-08-07 Thread Michael Schwendt
On Sat, 07 Aug 2010 11:51:07 +0100, Frank wrote:

> Thougth I ask for some feedback.
> 
> Currently when abrt finds an existing bug,
> it goes
> " similar bug found, (shows a link) adding you to cc"
> 
> looking at some of the bugs, that I have reported or cc'd on.
> Where someone earlier in the thread had sent a workaround/fix etc..
> and cc's just being added automatically.
> 
> Would it be any benefit to the maintainers\bugzappers.
> If abrt opened the existing link, before it would report?
> So there wouldn't need be be so many replies of
> please try xyz?, to help cut down on noise.
> 
> I could be talking out of my ear.
> Just a thought.

Not sure that will be fruitful. What makes you think the reporter would
not add another comment?

[The same can be observed in bodhi, where subsequently added comments (and
votes) ignore everything that has been said before. Sometimes an update
submitter writes "Please don't vote +1 on this", but anonymous and
registered users add their votes and comments nevertheless. Notifications
or instructions from the packager get lost in all the noise.]


Somewhat related: I ABRT I'd prefer if I could edit the "How to reproduce?"
field _before_ waiting for ABRT to fetch debuginfo packages. Currently, I
enter the notes in normal text editor, because ABRT insists on downloading
packages first, and the through-put monitor tells me the access to the
repo isn't good.
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel


abrt thoughts pre-rfe q?

2010-08-07 Thread Frank Murphy
Thougth I ask for some feedback.

Currently when abrt finds an existing bug,
it goes
" similar bug found, (shows a link) adding you to cc"

looking at some of the bugs, that I have reported or cc'd on.
Where someone earlier in the thread had sent a workaround/fix etc..
and cc's just being added automatically.

Would it be any benefit to the maintainers\bugzappers.
If abrt opened the existing link, before it would report?
So there wouldn't need be be so many replies of
please try xyz?, to help cut down on noise.

I could be talking out of my ear.
Just a thought.
-- 
Regards,

Frank Murphy
UTF_8 Encoded
Friend of Fedora
-- 
devel mailing list
devel@lists.fedoraproject.org
https://admin.fedoraproject.org/mailman/listinfo/devel