[Pharo-project] Call for help for reviewing code
Hi guys there are plenty of bugs fixes that are waiting for review in the bug tracker. If each of us take two we will get really done fast. So can you give a hand and have a look the fixes. Stef
Stéphane Ducasse wrote:
Hi guys
there are plenty of bugs fixes that are waiting for review in the bug tracker. If each of us take two we will get really done fast. So can you give a hand and have a look the fixes.
Stef
Thinking ahead to a time when I might contribute to bug fixing, such a call to arms makes me wonder if several people might rush to look at the same bug I would choose, and fix it while I'm half way through looking at it - thus wasting my time. A bit defeatist I know, but there you go - btw, how people actually dived in? I am reminded of what I experience in my busy work setting. An email "To:" multiple people often does not get actioned, while "To:" a single person gets actioned more often. It is like people can't help but think that someone else is going to do it. This sparks a thought that something I would sign up is to receive an occasional mail out of a random issue to individuals. I might sign up for say one bug a week. The alternative of searching through the issue tracker for suitable material to attack is a bit overwhelming. Getting one occasional email is much less so, with nothing lost by ignoring the assigned issue it if it exceeds my skill level. Then that one time I can do a little makes me feel more involved. just some random ideas, cheers, BenComan
On 10 November 2011 16:42, Ben Coman <btc@openinworld.com> wrote:
Stéphane Ducasse wrote:
Hi guys there are plenty of bugs fixes that are waiting for review in the bug tracker. If each of us take two we will get really done fast. So can you give a hand and have a look the fixes.
Stef
Thinking ahead to a time when I might contribute to bug fixing, such a call to arms makes me wonder if several people might rush to look at the same bug I would choose, and fix it while I'm half way through looking at it - thus wasting my time. Â A bit defeatist I know, but there you go - Â btw, how people actually dived in?
you can always, before actually diving in, leave a comment inside a particular issue that your are working on it. you're not forced to put only fixes on issue tracker :)
I am reminded of what I experience in my busy work setting. Â An email "To:" multiple people often does not get actioned, while "To:" a single person gets actioned more often. It is like people can't help but think that someone else is going to do it. Â This sparks a thought that something I would sign up is to receive an occasional mail out of a random issue to individuals. Â I might sign up for say one bug a week. Â The alternative of searching through the issue tracker for suitable material to attack is a bit overwhelming. Â Getting one occasional email is much less so, with nothing lost by ignoring the assigned issue it if it exceeds my skill level. Â Then that one time I can do a little makes me feel more involved.
this doesn't seems realistic. I certainly not such kind of person, who would do such job: picking random issue and sending an individual mail to some random person taken from 'bugfixers list' :) I think you can do such job pretty easily by youself: open an issue tracker , close your eyes , randomly move the mouse and then click. Repeat until you find something, which you can deal with.
just some random ideas, cheers, BenComan
In practice, if some bug are not fixed within a couple of weeks then it is not so critical and people can live without it being fixed. Actually it would be nice to have a bot which automatically closing unattended issues after month of inactivity, and mark them as 'closed (unresolved due to inactivity)'. Because if there's nobody who wants to fix it, why we should keep it listed? -- Best regards, Igor Stasenko.
On 10 November 2011 17:09, Igor Stasenko <siguctua@gmail.com> wrote:
Because if there's nobody who wants to fix it, why we should keep it listed?
Eh, because it's a bug. If you accept it as a bug, then it is a bug. "Nobody bothers to fix it" is not a valid issue resolution. It should stay there forever until you can think of a good enough reason not to fix it, or you actually fix it. -- Milan Mimica http://sparklet.sf.net
mmimica wrote:
"Nobody bothers to fix it" is not a valid issue resolution.
I agree. I sometimes find bugs that I can't fix because I don't understand enough or just don't have time right then, even though the bug is very annoying or disruptive. Over time, I gather info through using the system that I can collect as comments in the issue. Also, every once in a while (e.g. ESUG), I'll have a block of time that I can sit down and take down a big issue, but months may go by before such a block comes up. Maybe a status called "old" or "stale" or something else so these can be filtered? Sean -- View this message in context: http://forum.world.st/Call-for-help-for-reviewing-code-tp3992746p4031712.htm... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
+1 On Nov 11, 2011, at 4:49 PM, Sean P. DeNigris wrote:
mmimica wrote:
"Nobody bothers to fix it" is not a valid issue resolution.
I agree. I sometimes find bugs that I can't fix because I don't understand enough or just don't have time right then, even though the bug is very annoying or disruptive. Over time, I gather info through using the system that I can collect as comments in the issue. Also, every once in a while (e.g. ESUG), I'll have a block of time that I can sit down and take down a big issue, but months may go by before such a block comes up.
Maybe a status called "old" or "stale" or something else so these can be filtered?
Sean
-- View this message in context: http://forum.world.st/Call-for-help-for-reviewing-code-tp3992746p4031712.htm... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
Stéphane Ducasse wrote
So can you give a hand and have a look the fixes.
I just looked through about 6 issues marked FixReviewNeeded. I loaded the fixes and requested more info, and learned a few things: * it's hard to deal with issues that do not accurately explain the issue, preferably in a reproducible way * this is especially important when dealing with system critical things, e.g. the Debugger, that lead to image freezes and emergency evaluators. * including tests is key because it's not always clear what the fix is even supposed to do HTH, Sean -- View this message in context: http://forum.world.st/Call-for-help-for-reviewing-code-tp3992746p4092471.htm... Sent from the Pharo Smalltalk mailing list archive at Nabble.com.
I just looked through about 6 issues marked FixReviewNeeded. I loaded the fixes and requested more info, and learned a few things: * it's hard to deal with issues that do not accurately explain the issue, preferably in a reproducible way * this is especially important when dealing with system critical things, e.g. the Debugger, that lead to image freezes and emergency evaluators. * including tests is key because it's not always clear what the fix is even supposed to do
:) You see what we face :) Stef
participants (5)
-
Ben Coman -
Igor Stasenko -
Milan Mimica -
Sean P. DeNigris -
Stéphane Ducasse