You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@drill.apache.org by Muhammad Gelbana <m....@gmail.com> on 2017/05/25 13:34:27 UTC
Issues categorization suggestion
Hi,
I suggest to categorize issues according the level of expertise needed to
solve each one. This will encourage want-to-help\learn but
not-so-experienced-with-drill developers to take a thorough look into
issues not requiring high level of experience with Drill.
What do you think ?
*---------------------*
*Muhammad Gelbana*
http://www.linkedin.com/in/mgelbana
Re: Issues categorization suggestion
Posted by Abhishek Girish <ag...@apache.org>.
I've seen the "newbie" flag in Drill too. I think Jacques tagged a couple
of issues with that label. We could continue to use that.
On Thu, May 25, 2017 at 11:41 AM, Julian Hyde <jh...@apache.org> wrote:
> In Calcite we assign a "newbie" flag to some issues. A more detailed
> categorization takes significant effort for the person triaging the
> bugs, so isn't worth it.
>
> On Thu, May 25, 2017 at 9:23 AM, Paul Rogers <pr...@mapr.com> wrote:
> > Great suggestion.
> >
> > What I’ve learned over the last year, however, is that if something in
> Drill seems simple, it is only because I’ve missed understanding some of
> the details. It is quite hard to determine the effort level to fix a bug
> without first doing a detailed analysis of the bug. If the bug ends up
> being easy to fix, that is usually only because I’ve previously learned an
> area.
> >
> > That said, if the goal is to find good “starter bugs”, we can tackle
> that directly. Perhaps as the more experienced folks see bugs arrive in
> areas of our expertise, we can determine if the bug could be fixed by a new
> contributor. That is, the bug is isolated, seems to be pretty obvious, and
> the fix is likely to have few side-effects.
> >
> > Maybe use labels to identify such issues. Then community folks can grab
> those as a way to get started with Drill development. And, that experienced
> developer can provide guidance to help the new folks get started.
> >
> > Let’s take a concrete example. Thanks for filing DRILL-5539:
> "drillbit.sh script breaks if the working directory contains spaces".
> >
> > On the surface, this looks pretty easy: just put quotes where needed.
> But, you’ll find that drillbit.sh calls drill-config.sh to do all the heavy
> lifting, and that drill-config.sh is called by many of our scripts. And, it
> does lots of path work to find the config files, find directories, find
> Java and so on. Further, we have a unit test for the scripts that should be
> modified to test for the case you found.
> >
> > So, if you wanted to tackle this bug, I can explain this background in
> the JIRA ticket and help you identify all the places to fix, then show you
> how to use the unit test to verify that everything still works.
> >
> > Thoughts?
> >
> > Thanks,
> >
> > - Paul
> >
> >
> >> On May 25, 2017, at 6:58 AM, Charles Givre <cg...@gmail.com> wrote:
> >>
> >> +1
> >>
> >> Sent from my iPhone
> >>
> >>> On May 25, 2017, at 09:34, Muhammad Gelbana <m....@gmail.com>
> wrote:
> >>>
> >>> Hi,
> >>>
> >>> I suggest to categorize issues according the level of expertise needed
> to
> >>> solve each one. This will encourage want-to-help\learn but
> >>> not-so-experienced-with-drill developers to take a thorough look into
> >>> issues not requiring high level of experience with Drill.
> >>>
> >>> What do you think ?
> >>>
> >>> *---------------------*
> >>> *Muhammad Gelbana*
> >>> http://www.linkedin.com/in/mgelbana
> >
>
Re: Issues categorization suggestion
Posted by Julian Hyde <jh...@apache.org>.
In Calcite we assign a "newbie" flag to some issues. A more detailed
categorization takes significant effort for the person triaging the
bugs, so isn't worth it.
On Thu, May 25, 2017 at 9:23 AM, Paul Rogers <pr...@mapr.com> wrote:
> Great suggestion.
>
> What I’ve learned over the last year, however, is that if something in Drill seems simple, it is only because I’ve missed understanding some of the details. It is quite hard to determine the effort level to fix a bug without first doing a detailed analysis of the bug. If the bug ends up being easy to fix, that is usually only because I’ve previously learned an area.
>
> That said, if the goal is to find good “starter bugs”, we can tackle that directly. Perhaps as the more experienced folks see bugs arrive in areas of our expertise, we can determine if the bug could be fixed by a new contributor. That is, the bug is isolated, seems to be pretty obvious, and the fix is likely to have few side-effects.
>
> Maybe use labels to identify such issues. Then community folks can grab those as a way to get started with Drill development. And, that experienced developer can provide guidance to help the new folks get started.
>
> Let’s take a concrete example. Thanks for filing DRILL-5539: "drillbit.sh script breaks if the working directory contains spaces".
>
> On the surface, this looks pretty easy: just put quotes where needed. But, you’ll find that drillbit.sh calls drill-config.sh to do all the heavy lifting, and that drill-config.sh is called by many of our scripts. And, it does lots of path work to find the config files, find directories, find Java and so on. Further, we have a unit test for the scripts that should be modified to test for the case you found.
>
> So, if you wanted to tackle this bug, I can explain this background in the JIRA ticket and help you identify all the places to fix, then show you how to use the unit test to verify that everything still works.
>
> Thoughts?
>
> Thanks,
>
> - Paul
>
>
>> On May 25, 2017, at 6:58 AM, Charles Givre <cg...@gmail.com> wrote:
>>
>> +1
>>
>> Sent from my iPhone
>>
>>> On May 25, 2017, at 09:34, Muhammad Gelbana <m....@gmail.com> wrote:
>>>
>>> Hi,
>>>
>>> I suggest to categorize issues according the level of expertise needed to
>>> solve each one. This will encourage want-to-help\learn but
>>> not-so-experienced-with-drill developers to take a thorough look into
>>> issues not requiring high level of experience with Drill.
>>>
>>> What do you think ?
>>>
>>> *---------------------*
>>> *Muhammad Gelbana*
>>> http://www.linkedin.com/in/mgelbana
>
Re: Issues categorization suggestion
Posted by Paul Rogers <pr...@mapr.com>.
Great suggestion.
What I’ve learned over the last year, however, is that if something in Drill seems simple, it is only because I’ve missed understanding some of the details. It is quite hard to determine the effort level to fix a bug without first doing a detailed analysis of the bug. If the bug ends up being easy to fix, that is usually only because I’ve previously learned an area.
That said, if the goal is to find good “starter bugs”, we can tackle that directly. Perhaps as the more experienced folks see bugs arrive in areas of our expertise, we can determine if the bug could be fixed by a new contributor. That is, the bug is isolated, seems to be pretty obvious, and the fix is likely to have few side-effects.
Maybe use labels to identify such issues. Then community folks can grab those as a way to get started with Drill development. And, that experienced developer can provide guidance to help the new folks get started.
Let’s take a concrete example. Thanks for filing DRILL-5539: "drillbit.sh script breaks if the working directory contains spaces".
On the surface, this looks pretty easy: just put quotes where needed. But, you’ll find that drillbit.sh calls drill-config.sh to do all the heavy lifting, and that drill-config.sh is called by many of our scripts. And, it does lots of path work to find the config files, find directories, find Java and so on. Further, we have a unit test for the scripts that should be modified to test for the case you found.
So, if you wanted to tackle this bug, I can explain this background in the JIRA ticket and help you identify all the places to fix, then show you how to use the unit test to verify that everything still works.
Thoughts?
Thanks,
- Paul
> On May 25, 2017, at 6:58 AM, Charles Givre <cg...@gmail.com> wrote:
>
> +1
>
> Sent from my iPhone
>
>> On May 25, 2017, at 09:34, Muhammad Gelbana <m....@gmail.com> wrote:
>>
>> Hi,
>>
>> I suggest to categorize issues according the level of expertise needed to
>> solve each one. This will encourage want-to-help\learn but
>> not-so-experienced-with-drill developers to take a thorough look into
>> issues not requiring high level of experience with Drill.
>>
>> What do you think ?
>>
>> *---------------------*
>> *Muhammad Gelbana*
>> http://www.linkedin.com/in/mgelbana
Re: Issues categorization suggestion
Posted by Charles Givre <cg...@gmail.com>.
+1
Sent from my iPhone
> On May 25, 2017, at 09:34, Muhammad Gelbana <m....@gmail.com> wrote:
>
> Hi,
>
> I suggest to categorize issues according the level of expertise needed to
> solve each one. This will encourage want-to-help\learn but
> not-so-experienced-with-drill developers to take a thorough look into
> issues not requiring high level of experience with Drill.
>
> What do you think ?
>
> *---------------------*
> *Muhammad Gelbana*
> http://www.linkedin.com/in/mgelbana