[sage-devel] Re: Rules for closing tickets in trac
On Monday 22 October 2007, mabshoff wrote: Hello, since this has come up repeatedly I would like to clarify: Do not close any tickets in trac unless you have been explicitly told to do so by either malb, was, cwitty or mabshoff. This is to avoid having issues slip through the cracks. Once a ticket is closed and off the top of the time line chances are nobody will ever look at it again unless you stumble across it by accident. Just like the [with patch] byline we should come up with something to indicate that a ticket should be looked at like [should be closed], [is invalid] or [is won'tfix]. In addition you should give a reason why you think the action you requested should be taken (the more precise the better) and retag the ticket against the current target, i.e. 2.8.9 at the moment. If you leave it in 2.9 for now it is unlikely to be found and looked at because there are another 140 tickets open against that one. William can configure trac so that closing tickets is limited to a few, but so far he has not done so. But as we have to deal with an ever increasing number of tickets we need to follow the process in order to avoid losing issues and also keep the confusion and the work to deal with tickets to a minimum. I got a clarification for the clarification: Michael mentioned me (malb) and Carl (cwitty) because William (was) asked us to prepare the next release. So for 2.8.9 the three of us are the release managers and consequently we are supposed to close tickets. and a question: Take ticket 729 as an example ( http://trac.sagemath.org/sage_trac/ticket/729 ): It was closed by Robert (rml) because the bugfix/feature request was invalid. Michael (mabshoff) reopened it due to the rule state above. But there is nothing for the release manager to do and I feel perfectly comfortable with rml closing invalid tickets for the Graph subsystem. So I think in those cases it makes perfectly sense to just close tickets. IMHO this rule could be relaxed if we had the e-mail subsystem working (I know it is being worked on). Because in that case, the submitter and the owner of a ticket get an e-mail and can complain if it isn't actually fixed. Martin -- name: Martin Albrecht _pgp: http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0x8EF0DC99 _www: http://www.informatik.uni-bremen.de/~malb _jab: [EMAIL PROTECTED] --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Rules for closing tickets in trac
On Oct 22, 10:40 am, Martin Albrecht [EMAIL PROTECTED] wrote: On Monday 22 October 2007, mabshoff wrote: Hello, Hello, since this has come up repeatedly I would like to clarify: Do not close any tickets in trac unless you have been explicitly told to do so by either malb, was, cwitty or mabshoff. This is to avoid having issues slip through the cracks. Once a ticket is closed and off the top of the time line chances are nobody will ever look at it again unless you stumble across it by accident. Just like the [with patch] byline we should come up with something to indicate that a ticket should be looked at like [should be closed], [is invalid] or [is won'tfix]. In addition you should give a reason why you think the action you requested should be taken (the more precise the better) and retag the ticket against the current target, i.e. 2.8.9 at the moment. If you leave it in 2.9 for now it is unlikely to be found and looked at because there are another 140 tickets open against that one. William can configure trac so that closing tickets is limited to a few, but so far he has not done so. But as we have to deal with an ever increasing number of tickets we need to follow the process in order to avoid losing issues and also keep the confusion and the work to deal with tickets to a minimum. I got a clarification for the clarification: Michael mentioned me (malb) and Carl (cwitty) because William (was) asked us to prepare the next release. So for 2.8.9 the three of us are the release managers and consequently we are supposed to close tickets. and a question: Take ticket 729 as an example (http://trac.sagemath.org/sage_trac/ticket/729): It was closed by Robert (rml) because the bugfix/feature request was invalid. Michael (mabshoff) reopened it due to the rule state above. But there is nothing for the release manager to do and I feel perfectly comfortable with rml closing invalid tickets for the Graph subsystem. So I think in those cases it makes perfectly sense to just close tickets. Well, I see it the same way: rml is responsible for the graph subsystem, so he is the person with the expertise to determine that the ticket is invalid. But he closed that ticket against 2.9, while it is customary to close invalid tickets against the sage-duplicate/ invalid milestone. The same applies for #731. It is not so much the marking the ticket invalid, it just ended up against the wrong milestone. What caused me to actually raise the issue is #656: That one is clearly not a duplicate. #968 is an enhancement relative to #656. During Bug Day 4 William also told rlm not to close tickets until the issue had been officially resolved. IMHO this rule could be relaxed if we had the e-mail subsystem working (I know it is being worked on). Because in that case, the submitter and the owner of a ticket get an e-mail and can complain if it isn't actually fixed. Yep, that has been requested for quite some time. William has enabled smtp support for trac IIRC, but it doesn't work (yet). Martin Cheers, Michael -- name: Martin Albrecht _pgp:http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0x8EF0DC99 _www:http://www.informatik.uni-bremen.de/~malb _jab: [EMAIL PROTECTED] --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Rules for closing tickets in trac
Take ticket 729 as an example (http://trac.sagemath.org/sage_trac/ticket/729): It was closed by Robert (rml) because the bugfix/feature request was invalid. Michael (mabshoff) reopened it due to the rule state above. But there is nothing for the release manager to do and I feel perfectly comfortable with rml closing invalid tickets for the Graph subsystem. So I think in those cases it makes perfectly sense to just close tickets. Well, I see it the same way: rml is responsible for the graph subsystem, so he is the person with the expertise to determine that the ticket is invalid. But he closed that ticket against 2.9, while it is customary to close invalid tickets against the sage-duplicate/ invalid milestone. The same applies for #731. It is not so much the marking the ticket invalid, it just ended up against the wrong milestone. Why do we need to change the milestone for invalid tickets? Martin -- name: Martin Albrecht _pgp: http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0x8EF0DC99 _www: http://www.informatik.uni-bremen.de/~malb _jab: [EMAIL PROTECTED] --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Rules for closing tickets in trac
On Oct 22, 11:13 am, Martin Albrecht [EMAIL PROTECTED] wrote: Take ticket 729 as an example (http://trac.sagemath.org/sage_trac/ticket/729):It was closed by Robert (rml) because the bugfix/feature request was invalid. Michael (mabshoff) reopened it due to the rule state above. But there is nothing for the release manager to do and I feel perfectly comfortable with rml closing invalid tickets for the Graph subsystem. So I think in those cases it makes perfectly sense to just close tickets. Well, I see it the same way: rml is responsible for the graph subsystem, so he is the person with the expertise to determine that the ticket is invalid. But he closed that ticket against 2.9, while it is customary to close invalid tickets against the sage-duplicate/ invalid milestone. The same applies for #731. It is not so much the marking the ticket invalid, it just ended up against the wrong milestone. Why do we need to change the milestone for invalid tickets? The idea is to collect all invalid tickets in one milestone so that they can be found by accessing one milestone. You can pull all of them via custom query, but that is not as transparent. I only recently started doing that, so there are invalid tickets attached to older milestones, but I had a discussion with William about that in IRC before creating that special milestone. The discussion happened about 3 weeks ago. Martin Cheers, Michael -- name: Martin Albrecht _pgp:http://pgp.mit.edu:11371/pks/lookup?op=getsearch=0x8EF0DC99 _www:http://www.informatik.uni-bremen.de/~malb _jab: [EMAIL PROTECTED] --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Rules for closing tickets in trac
What caused me to actually raise the issue is #656: That one is clearly not a duplicate. #968 is an enhancement relative to #656. During Bug Day 4 William also told rlm not to close tickets until the issue had been officially resolved. I'm assuming you're talking about 956. The reason I closed 956 and deferred to 968 was because the patches are both to the same area of code, and if you apply the patch in #956, then the ones in 968, funny stuff might happen. The recommended patches in #968 will cover both, so... --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---
[sage-devel] Re: Rules for closing tickets in trac
On Oct 22, 5:17 pm, Robert Miller [EMAIL PROTECTED] wrote: What caused me to actually raise the issue is #656: That one is clearly not a duplicate. #968 is an enhancement relative to #656. During Bug Day 4 William also told rlm not to close tickets until the issue had been officially resolved. Hello Robert, I'm assuming you're talking about 956. You are right, sorry for the type. The reason I closed 956 and deferred to 968 was because the patches are both to the same area of code, and if you apply the patch in #956, then the ones in 968, funny stuff might happen. The recommended patches in #968 will cover both, so... Well, #968 has two patches, one of which is idential to the one attached to #656 except that it is rebased against the tree after your fixes went in. I would have suggested to attach the updated version for #656 to that ticket and add a comment not to apply the first, but the second version due to the rewrite attached to #968. Cheers, Michael --~--~-~--~~~---~--~~ To post to this group, send email to sage-devel@googlegroups.com To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/sage-devel URLs: http://sage.scipy.org/sage/ and http://modular.math.washington.edu/sage/ -~--~~~~--~~--~--~---