This sort of direction is much more effective and less likely to be lost when it is done in
bugzilla, as I've explained to Jan and other developers many, many times.
Below for educational purposes, since apparently this isn't fully understood, I've listed the
approximate qa translations of Jan's input, the sort of which needs to take form in bugzilla (and
which I've already done in bugzilla unless noted otherwise).
--- msevior@physics.unimelb.edu.au wrote:
>
> Hi Jan,
> Thanks very much. I need just this ort direction right now. I'm not
> ------------------------------------------------------------------------
>
> From Jan:
>
> I have tried to make some kind of, "what is important" report - and I
> would like to know, what you think - the concept probably needs some
> finetuning. Should I continue, looking at other components - or is useless
It most certainly should continue, but as I've said, in bugzilla.
> and I should let it be. The purpose is to have a look at a given
> component, extract the most important bugs and try to relate and comment
> them. Minor bugs and requests are not listed. Please notice ">>>" marks.
> Either the qa people need some idea of how to continue work, which is the
> same as, please make a statement / comment ...we need input - or some bugs
> should have a coders attention right away. - Jan
>
> ---------- Crash
[keyword crash, severity implication]
> When we paste large tables into smaller tables, we're not sure what to do
> http://bugzilla.abisource.com/show_bug.cgi?id=5762
> >>> your comment please: Should be do it your way or like Word does? They
> look related.
KS already triaged this one, actually, and well so. It's P2 (the qa translation of Jan's "what is
important"), its severity has been upgraded, it is targeted at 2.0.x because the possibility of
backport has not been ruled out and we're still depending on 2.0 to be production for the next
couple of months. All I had to add was the crash keyword because, as can be seen in the comments
on the bug, it can cause a crash.
>
> When we have nested tables, some operations do crash
> http://bugzilla.abisource.com/show_bug.cgi?id=6983
> http://bugzilla.abisource.com/show_bug.cgi?id=6920
> http://bugzilla.abisource.com/show_bug.cgi?id=6960
For each of these three,
Again, some of the triaging has already been done. Being an easy crasher, it is already of
critical severity, and it already has the crash keyword, thanks to sum1. They are confirmed.
Now, Jan's indication that this is an important bug translates to upgraded priority (P2) and,
since this takes place in the context of the upcoming 2.2 release series, an initial target
setting of 2.2. Of course if it turned out to be an easy backport, it still could be, if it
turned out to require a huge amount of manhours of coding, it could of course be postponed to
2.2.x or 2.4 (with discussion on the bug).
>
> - if we do some testing first, maybe we'll notice a pattern. I'll ask sum1
> to have a look at it.
>
This comment could be made on the bugs (especially if citing specifics), and/or a discussion could
be opened up on the qa list about "patterns" in nested table bugs and any bz ramifications of that
(including what we draw martin's attention to).
> ---------- 5 minute fix
[the severity may not be very high, but the target for example may reflect that it is (afawk) a
quick fix by being set accordingly]
> Validation of fixed size value
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6741
> >>> you should just fix it, so we can resolve and forget it.
This comment should be made on the bug, that it should be fixed for being quick, the target
initially set to 2.2 for aforementioned reasons, the priority upgraded as a translation of whoever
is setting it (Jan in this case) feeling that it is important within the context of 2.2 bugs.
Of course this specific example is now resolved fixed, so I'll forgo actually making the changes
for you.
> ---------- Selection
>
> Please note comment #11 in bug 5534.This is not a table only bug, but I
> notice it very often, when reading bugreports.If we can fix this one,
> we'll solve several other "non table" bugs at the same time.
>
> http://bugzilla.abisource.com/show_bug.cgi?id=4987
The above comment should have been put here. The priority would be upgraded (if only to P3) as a
translation of Jan's evaluation ("what's important"), although since Ryan had previously
downgraded it within the same context (Future), some dialog should take place between them
(possibly on list or on bug, and including martin since he's already evaluated and adjusted as
well) to determine just where the priority should agreeably lie. The target was set to Future,
and would _not_ be upgraded to 2.2 due to martin's statement that it would require substantial
time and change, which translates into release postponement. Again, with discussion it might be
determined to promote to 2.4, if it's determined worthy enough and martin expects to do the
refactoring in that time frame (if not, it stays at future and can wait for another release, like
2.6/3.0).
> http://bugzilla.abisource.com/show_bug.cgi?id=5534
The target and priority were again set and upgraded accordingly. Jan's feeling that this is
important, including the fact that it may lead to other fixes, upgrades the priority. Also, the
version field had not been updated despite the fact that the bug was found to be in cvs-head (see
comments), so that was updated accordingly.
>
> >>> we need yours and the other coders attention. It requires insight
> knowledge...
I'm not sure which bug(s) this was intended for, but it should be added as a comment to that
bug(s), and cc whoever's attention needs to be had if they aren't already on the bug.
>
> ---------- Pixeldirt Border remains, when deleting table.
>
[keyword pixeldirt]
> http://bugzilla.abisource.com/show_bug.cgi?id=6963
For reasons given above, target set initially at 2.2, priority upgraded to P2.
> http://bugzilla.abisource.com/show_bug.cgi?id=6808
Setting aside possible dupiness between the two bugs (there are two bugs herein, one of which is
overlapped by 6963, but 6963 and 6808 are separate bugs at least for now), the pixeldirt keyword
should be set, martin has already targeted at 2.2, and again Jan's correspondance with martin
translates to a priority upgrade as before.
>
> >>> maybe one of these "1 pixel" bugs. You should have a look at it, to
> locate if it's as simple as that. Something about coordinates. Strange
> color leftovers.
>
Again, that's a comment that belongs on a bug, where it can be kept track of (I can personally
testify that martin is known to lose or forget mail about bugs).
> http://bugzilla.abisource.com/show_bug.cgi?id=6509- not a major problem,
> but as it's pixeldirt, it disturbs the overall impression. That's why I
> mentioned it.
"not a major problem" -> severity is not high
"as it's pixeldirt" -> pixeldirt keyword, although I don't follow the reasoning behind that here.
"what's important", "disturbs the overall impression" -> may translate to priority upgrade,
initial target might be 2.2.x or 2.2 depending on martin's evaluation of how much time/change it
might take. It isn't a killer, so it doesn't have to be in 2.2.0 at all costs, it's a matter of
evaluation.
>
> ---------- Environment Paragraph spacing (above & below) within tables
> doesn't work
> http://bugzilla.abisource.com/show_bug.cgi?id=5725
>
Again, feelings of importance translate to priority, target depends on evaluation of what can and
can't be done without significantly delaying 2.2 and other bugs.
Also, if you have reproduced this lately, it would be nice to have the version field updated.
> Some elements cannot be inserted directly above tables
> http://bugzilla.abisource.com/show_bug.cgi?id=6777
> >>> Your comment please: It looks like, that there are some weird things
> going on, just outside tables. Do you need more details?
Again, comment/discussion should be made on the bug, and feelings of importance translate to
priority upgrades within the context of 2.2, where martin has already targeted it.
>
> Position and size Tables can be positioned or imported outside boudaries
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6214
> http://bugzilla.abisource.com/show_bug.cgi?id=5912
>
> >>> your comment please: Should we follow Word or the more logical
> approach to keep tables inside (we could reposition, when importing).
> Column resizer causes weird table resizes or even make it dissappear
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6215
>
> When deleting last row (with merge cells) - we made a gap
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6742
>
Again, comments belong on the bug(s), and raise the priority if it needs to be. They are already
targeted for 2.2.
> Deletion with nested tables causes rendering confusion
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6961
>
Again, if you feel it's important or the need to remind the coder, say so on the bug (and
potentially upgrade the priority or set a tentative target). Even if someone disagrees about the
priority/target set, it'll at least get the bug more attention.
> Table incorrectly displayed when a merged cell spans two pages
>
> http://bugzilla.abisource.com/show_bug.cgi?id=6748
>
See above.
> - These are major bugs, hence they need attention.
>
If the coders aren't paying attention because they aren't on the bug, cc them so they become aware
of what's going on. Otherwise, these comments belong on the bugs, and the priority can be
upgraded to reflect importance, as well as the target set or adjusted to reflect the necessity and
feasibility of a fix before a given release if need be.
NB: when initially setting target milestone of a bug, if in doubt (ie between 2.2 and 2.2.x),
choose the later (at least while we're already behind schedule for the earlier target). It can
always be discussed and upgraded if need be.
I hope this helps you understand how to reflect your evaluations in bugzilla, so that everyone can
see and benefit from them, and they aren't lost/useless in the mail shuffle. That's why it's
there.
When this work is not in bugzilla (for example, high priority bugs aren't marked as such), there
isn't much we can do about it (for example, a query for high priority bugs doesn't show those that
aren't marked as such).
-MG
__________________________________
Do you Yahoo!?
Yahoo! Mail is new and improved - Check it out!
http://promotions.yahoo.com/new_mail
Received on Fri Jun 18 20:58:43 2004
This archive was generated by hypermail 2.1.8 : Fri Jun 18 2004 - 20:58:44 CEST