Improving the bug tracking workflow: starting document
Hi, Following the discussions, I have started to write a small document highlighting my current gripes with trac. I focus on some common scenario, and pin-point trac limitations. I mention possible new tools at the end, but that's not the main point: everybody who is also disatisfied with trac, and maybe even more importantly people who are currently satisfied and think their scenario is not covered should feel free to comment/modify it: http://scipy.org/scipy/numpy/wiki/ImprovingIssueWorkflow I put the initial version in svn as well: http://projects.scipy.org/scipy/numpy/browser/trunk/doc/neps/newbugtracker.r.... cheers, David
On Fri, Feb 27, 2009 at 3:28 AM, David Cournapeau <david@ar.media.kyoto-u.ac.jp> wrote:
Hi,
Following the discussions, I have started to write a small document highlighting my current gripes with trac. I focus on some common scenario, and pin-point trac limitations. I mention possible new tools at the end, but that's not the main point: everybody who is also disatisfied with trac, and maybe even more importantly people who are currently satisfied and think their scenario is not covered should feel free to comment/modify it:
http://scipy.org/scipy/numpy/wiki/ImprovingIssueWorkflow
I put the initial version in svn as well: http://projects.scipy.org/scipy/numpy/browser/trunk/doc/neps/newbugtracker.r....
cheers,
David
Just two quick comments: * I like the integration of the bug tracker and svn, browsing between old tickets and revisions is pretty easy. Similarly, integrated timeline for svn and issue tracker makes tracking new code and issues easy. * eclipse integration with trac issues works well with mylyn, but I haven't used it much and not for scipy, eclipse integration with svn is very good. Josef
On Fri, Feb 27, 2009 at 7:56 AM, <josef.pktd@gmail.com> wrote:
On Fri, Feb 27, 2009 at 3:28 AM, David Cournapeau <david@ar.media.kyoto-u.ac.jp> wrote:
Hi,
Following the discussions, I have started to write a small document highlighting my current gripes with trac. I focus on some common scenario, and pin-point trac limitations. I mention possible new tools at the end, but that's not the main point: everybody who is also disatisfied with trac, and maybe even more importantly people who are currently satisfied and think their scenario is not covered should feel free to comment/modify it:
http://scipy.org/scipy/numpy/wiki/ImprovingIssueWorkflow
I put the initial version in svn as well: http://projects.scipy.org/scipy/numpy/browser/trunk/doc/neps/newbugtracker.r....
cheers,
David
Just two quick comments:
* I like the integration of the bug tracker and svn, browsing between old tickets and revisions is pretty easy. Similarly, integrated timeline for svn and issue tracker makes tracking new code and issues easy.
* eclipse integration with trac issues works well with mylyn, but I haven't used it much and not for scipy, eclipse integration with svn is very good.
Josef
I connected my eclipse mylyn with the scipy trac tickets and have tickets that are assigned to me on my local computer where I can mark them as read or unread. But for now I like the web interface better. But this works after some trial and error. Also, since David mentioned sql queries in another thread, I set up a report that sorts tickets by change time. It helps to see which tickets where recently commented on. But since this is the first time, I do this, and I'm not very familiar with sql, this still needs improvements. However, this helps with the main problem, I had with the trac ticket listing, and maybe some additional specialized reports will make keeping an overview of tickets easier, even with the current trac version. Josef
josef.pktd@gmail.com wrote:
On Fri, Feb 27, 2009 at 7:56 AM, <josef.pktd@gmail.com> wrote:
<SNIP>
However, this helps with the main problem, I had with the trac ticket listing, and maybe some additional specialized reports will make keeping an overview of tickets easier, even with the current trac version.
The scipy trac seems to be version 0.10.2 which has a number of known security issues. Asides from that trac release prior to 0.11 leak memory when using Apache, i.e. http://trac.edgewall.org/ticket/6614 which was a well known and often complained about bug, so many of your performance problems will likely go away once you upgrade (it as my impression you use trac+Apache)
Josef
Cheers, Michael
_______________________________________________ Scipy-dev mailing list Scipy-dev@scipy.org http://projects.scipy.org/mailman/listinfo/scipy-dev
2009/2/27 Michael Abshoff <michael.abshoff@googlemail.com>:
The scipy trac seems to be version 0.10.2 which has a number of known security issues.
That's about to change! We've already set up a new NumPy Trac for testing. Hopefully we can soon switch over to http://new.scipy.org/trac/numpy/timeline You'll notice that the new server and 0.11 is much more responsive. Cheers Stéfan
Stéfan van der Walt wrote:
2009/2/27 Michael Abshoff <michael.abshoff@googlemail.com>:
The scipy trac seems to be version 0.10.2 which has a number of known security issues.
That's about to change! We've already set up a new NumPy Trac for testing. Hopefully we can soon switch over to
http://new.scipy.org/trac/numpy/timeline
You'll notice that the new server and 0.11 is much more responsive.
Well, I rarely look at numpy or scipy's trac since I have plenty of things to do, too, and I am not the guy dealing with numpy/scipy problems in Sage at the moment :) I was just surprised that such an obvious problem with trac+apache or other crippling issues with your trac install weren't just fixed. It took William a couple hours to migrate four trac installs, one of them with a 400 MB database. If Sage's trac had the performance of the ones you had to deal with I would not have rested until the issue were fixed since no working trac means standstill for Sage.
Cheers Stéfan
Cheers, Michael
_______________________________________________ Scipy-dev mailing list Scipy-dev@scipy.org http://projects.scipy.org/mailman/listinfo/scipy-dev
On Fri, Feb 27, 2009 at 11:25 PM, <josef.pktd@gmail.com> wrote:
I connected my eclipse mylyn with the scipy trac tickets and have tickets that are assigned to me on my local computer where I can mark them as read or unread. But for now I like the web interface better.
Different people, different workflows :) Web-UI is of course a must, that's definitely easier for newcomers anyway (no need to install anything). Some things could be improved in the WEB-UI as well: batch editing, for example, is something I really like with redmine - reassigning things in batch has to be done through SQL I believe.
Also, since David mentioned sql queries in another thread, I set up a report that sorts tickets by change time. It helps to see which tickets where recently commented on. But since this is the first time, I do this, and I'm not very familiar with sql, this still needs improvements.
With trac 0.11, there will be no much need for SQL: there is finally a user-friendly interface for advanced queries with trac macro systems. This does not help much for batch editing, though. Maybe we will have to implement our own plugin to trac if we stay with trac. David
On Sat, Feb 28, 2009 at 12:01 PM, David Cournapeau <cournape@gmail.com> wrote:
On Fri, Feb 27, 2009 at 11:25 PM, <josef.pktd@gmail.com> wrote:
I connected my eclipse mylyn with the scipy trac tickets and have tickets that are assigned to me on my local computer where I can mark them as read or unread. But for now I like the web interface better.
Different people, different workflows :) Web-UI is of course a must, that's definitely easier for newcomers anyway (no need to install anything). Some things could be improved in the WEB-UI as well: batch editing, for example, is something I really like with redmine - reassigning things in batch has to be done through SQL I believe.
Also, since David mentioned sql queries in another thread, I set up a report that sorts tickets by change time. It helps to see which tickets where recently commented on. But since this is the first time, I do this, and I'm not very familiar with sql, this still needs improvements.
With trac 0.11, there will be no much need for SQL: there is finally a user-friendly interface for advanced queries with trac macro systems. This does not help much for batch editing, though. Maybe we will have to implement our own plugin to trac if we stay with trac.
David
From what I have seen from a quick look at the new trac 0.11.3, it still doesn't allow custom queries sorted by changedate. So it will still be useful to set up a set of reports by changedate and milestone or version or component as in http://trac.edgewall.org/report
Do you mean the ticket query macro for the wiki or is there some other way for doing advanced queries? I didn't find the ticket query macro very useful for my own use since it doesn't provide the nice table. The XML-RPC interface to trac might work quite well for scripting, this is also how eclipse/mylyn synchronizes with the trac tickets. To eclipse usage: I used it, together with svn integration, quite a bit for pure python packages. Since I build and install scipy into new directories each time, I find the workspace concept of eclipse a bit cumbersome, and if I'm not sure how thinks work, I just work with the Idle shell. But since eclipse is distributed with pythonxy, there might be more users of eclipse. Josef
On Sun, Mar 1, 2009 at 5:44 AM, <josef.pktd@gmail.com> wrote:
On Sat, Feb 28, 2009 at 12:01 PM, David Cournapeau <cournape@gmail.com> wrote:
On Fri, Feb 27, 2009 at 11:25 PM, <josef.pktd@gmail.com> wrote:
I connected my eclipse mylyn with the scipy trac tickets and have tickets that are assigned to me on my local computer where I can mark them as read or unread. But for now I like the web interface better.
Different people, different workflows :) Web-UI is of course a must, that's definitely easier for newcomers anyway (no need to install anything). Some things could be improved in the WEB-UI as well: batch editing, for example, is something I really like with redmine - reassigning things in batch has to be done through SQL I believe.
Also, since David mentioned sql queries in another thread, I set up a report that sorts tickets by change time. It helps to see which tickets where recently commented on. But since this is the first time, I do this, and I'm not very familiar with sql, this still needs improvements.
With trac 0.11, there will be no much need for SQL: there is finally a user-friendly interface for advanced queries with trac macro systems. This does not help much for batch editing, though. Maybe we will have to implement our own plugin to trac if we stay with trac.
David
From what I have seen from a quick look at the new trac 0.11.3, it still doesn't allow custom queries sorted by changedate. So it will still be useful to set up a set of reports by changedate and milestone or version or component as in http://trac.edgewall.org/report
It looks like I read a bit too quickly, and it will be for 0.12: http://trac.edgewall.org/milestone/0.12
Do you mean the ticket query macro for the wiki or is there some other way for doing advanced queries? I didn't find the ticket query macro very useful for my own use since it doesn't provide the nice table.
I don't care too much about the presentation - this can be improved. As you mentioned, the xml-rpc might be very useful - I may be able to program something for my own workflow. After all, that's what python is for :) cheers, David
On Fri, Feb 27, 2009 at 9:56 PM, <josef.pktd@gmail.com> wrote:
On Fri, Feb 27, 2009 at 3:28 AM, David Cournapeau <david@ar.media.kyoto-u.ac.jp> wrote:
Hi,
Following the discussions, I have started to write a small document highlighting my current gripes with trac. I focus on some common scenario, and pin-point trac limitations. I mention possible new tools at the end, but that's not the main point: everybody who is also disatisfied with trac, and maybe even more importantly people who are currently satisfied and think their scenario is not covered should feel free to comment/modify it:
http://scipy.org/scipy/numpy/wiki/ImprovingIssueWorkflow
I put the initial version in svn as well: http://projects.scipy.org/scipy/numpy/browser/trunk/doc/neps/newbugtracker.r....
cheers,
David
Just two quick comments:
* I like the integration of the bug tracker and svn, browsing between old tickets and revisions is pretty easy. Similarly, integrated timeline for svn and issue tracker makes tracking new code and issues easy.
Yes, I agree this is a useful functionality. I think most integrated solutions ala trac/redmine and co have this feature.
* eclipse integration with trac issues works well with mylyn, but I haven't used it much and not for scipy, eclipse integration with svn is very good.
I note that you care about eclipse integration. This is an important point I think for other people as well, and something I can't really document myself, as I don't use IDE. David
participants (5)
-
David Cournapeau -
David Cournapeau -
josef.pktd＠gmail.com -
Michael Abshoff -
Stéfan van der Walt