Re: svn commit: r926912 - in /ofbiz/trunk/framework/base/src/org/ofbiz/base: concurrent/ concurrent/test/ conversion/ conversion/test/ json/ json/test/ lang/ lang/test/ util/ util/collections/ util/collections/test/ util/string/ util/string/test/ util/test/

[Scott Gray-2]

This seems important, could you explain how it works a little more?

Thanks
Scott

HotWax Media
http://www.hotwaxmedia.com


On 23/03/2010, at 9:17 PM, [hidden email] wrote:

> Author: doogie
> Date: Wed Mar 24 03:16:50 2010
> New Revision: 926912
>
> URL: http://svn.apache.org/viewvc?rev=926912&view=rev
> Log:
> Add new annotation, SourceMonitor, which provides a source-level
> documention of which people, on which mailing list, will be giving extra
> attention to changes to the annotated classes, constructors, and
> methods.
>
> Added:
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/lang/SourceMonitor.java
> Modified:
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/ExecutionPool.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/TTLCachedObject.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/TTLObject.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/test/AsyncTTLObjectTest.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/test/SyncTTLObjectTest.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/test/TTLCachedObjectTest.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/concurrent/test/TTLObjectTest.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/conversion/Converters.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/conversion/test/DateTimeTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/conversion/test/MiscTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/json/JSONWriter.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/json/test/JSONTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/lang/ComparableRange.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/lang/test/ComparableRangeTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/IndentingWriter.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/ObjectType.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/TimeDuration.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/UtilObject.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/collections/FlexibleMapAccessor.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/collections/test/FlexibleMapAccessorTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/string/FlexibleStringExpander.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/string/test/FlexibleStringExpanderTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/test/IndentingWriterTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/test/ObjectTypeTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/test/TimeDurationTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/test/UtilIOTests.java
>    ofbiz/trunk/framework/base/src/org/ofbiz/base/util/test/UtilObjectTests.java
>

[Adam Heath-2]

Scott Gray wrote:
> This seems important, could you explain how it works a little more?

There was a thread about this recently, where I announced which
classes had full coverage.  This annotation just means that the
person(s) listed in the annotation are monitoring the source.  That
means that those people might be who to turn too if there are
questions on how the code works.

[Scott Gray-2]

On 23/03/2010, at 9:37 PM, Adam Heath wrote:

> Scott Gray wrote:
>> This seems important, could you explain how it works a little more?
>
> There was a thread about this recently, where I announced which
> classes had full coverage.  This annotation just means that the
> person(s) listed in the annotation are monitoring the source.  That
> means that those people might be who to turn too if there are
> questions on how the code works.

But what does it do functionally?  How would it be different from:
// "Don't mess up my code buddy!" signed Adam Heath

Regards
Scott

[Adam Heath-2]

Scott Gray wrote:

> On 23/03/2010, at 9:37 PM, Adam Heath wrote:
>
>> Scott Gray wrote:
>>> This seems important, could you explain how it works a little more?
>> There was a thread about this recently, where I announced which
>> classes had full coverage.  This annotation just means that the
>> person(s) listed in the annotation are monitoring the source.  That
>> means that those people might be who to turn too if there are
>> questions on how the code works.
>
> But what does it do functionally?  How would it be different from:
> // "Don't mess up my code buddy!" signed Adam Heath

Shows up in javadoc.

The retention policy could be changed to to Runtime, so that a test
case that fails could report something automatically.

Remember, annotations are easier to access programatically.

[Scott Gray-2]

On 23/03/2010, at 9:46 PM, Adam Heath wrote:

> Scott Gray wrote:
>> On 23/03/2010, at 9:37 PM, Adam Heath wrote:
>>
>>> Scott Gray wrote:
>>>> This seems important, could you explain how it works a little more?
>>> There was a thread about this recently, where I announced which
>>> classes had full coverage.  This annotation just means that the
>>> person(s) listed in the annotation are monitoring the source.  That
>>> means that those people might be who to turn too if there are
>>> questions on how the code works.
>>
>> But what does it do functionally?  How would it be different from:
>> // "Don't mess up my code buddy!" signed Adam Heath
>
> Shows up in javadoc.

Got it

>
> The retention policy could be changed to to Runtime, so that a test
> case that fails could report something automatically.

Doesn't buildbot do that for us already?

> Remember, annotations are easier to access programatically.

My only concern is that we removed all author information when we moved to Apache and this feels very similar to it.  I understand you're not claiming authorship but it does feel kind of implied.

[Adrian Crum-2]

--- On Tue, 3/23/10, Adam Heath <[hidden email]> wrote:

> From: Adam Heath <[hidden email]>
> Subject: Re: svn commit: r926912 - in /ofbiz/trunk/framework/base/src/org/ofbiz/base: concurrent/ concurrent/test/ conversion/ conversion/test/ json/ json/test/ lang/ lang/test/ util/ util/collections/ util/collections/test/ util/string/ util/string/test/ util/test/
> To: [hidden email]
> Date: Tuesday, March 23, 2010, 8:46 PM
> Scott Gray wrote:
> > On 23/03/2010, at 9:37 PM, Adam Heath wrote:
> >
> >> Scott Gray wrote:
> >>> This seems important, could you explain how it
> works a little more?
> >> There was a thread about this recently, where I
> announced which
> >> classes had full coverage.  This annotation
> just means that the
> >> person(s) listed in the annotation are monitoring
> the source.  That
> >> means that those people might be who to turn too
> if there are
> >> questions on how the code works.
> >
> > But what does it do functionally?  How would it
> be different from:
> > // "Don't mess up my code buddy!" signed Adam Heath
>
> Shows up in javadoc.
>
> The retention policy could be changed to to Runtime, so
> that a test
> case that fails could report something automatically.
>
> Remember, annotations are easier to access
> programatically.

And automatically fix bad commits?

;)

[Adrian Crum-2]

--- On Tue, 3/23/10, Scott Gray <[hidden email]> wrote:

> From: Scott Gray <[hidden email]>
> Subject: Re: svn commit: r926912 - in /ofbiz/trunk/framework/base/src/org/ofbiz/base: concurrent/ concurrent/test/ conversion/ conversion/test/ json/ json/test/ lang/ lang/test/ util/ util/collections/ util/collections/test/ util/string/ util/string/test/ util/test/
> To: [hidden email]
> Date: Tuesday, March 23, 2010, 8:55 PM
> On 23/03/2010, at 9:46 PM, Adam Heath
> wrote:
>
> > Scott Gray wrote:
> >> On 23/03/2010, at 9:37 PM, Adam Heath wrote:
> >>
> >>> Scott Gray wrote:
> >>>> This seems important, could you explain
> how it works a little more?
> >>> There was a thread about this recently, where
> I announced which
> >>> classes had full coverage.  This
> annotation just means that the
> >>> person(s) listed in the annotation are
> monitoring the source.  That
> >>> means that those people might be who to turn
> too if there are
> >>> questions on how the code works.
> >>
> >> But what does it do functionally?  How would
> it be different from:
> >> // "Don't mess up my code buddy!" signed Adam
> Heath
> >
> > Shows up in javadoc.
>
> Got it
>
> >
> > The retention policy could be changed to to Runtime,
> so that a test
> > case that fails could report something automatically.
>
> Doesn't buildbot do that for us already?
>
> > Remember, annotations are easier to access
> programatically.
>
> My only concern is that we removed all author information
> when we moved to Apache and this feels very similar to
> it.  I understand you're not claiming authorship but it
> does feel kind of implied.

Maybe there could be some indirection where the annotation could contain a mailing list group or some such.

[Adam Heath-2]

Adrian Crum wrote:
>> Remember, annotations are easier to access
>> programatically.
>
> And automatically fix bad commits?

What are you saying?

[Adam Heath-2]

Adrian Crum wrote:
> Maybe there could be some indirection where the annotation could contain a mailing list group or some such.

It already does.  The default address is [hidden email].  Go
read the commit again.

You can have a single email address, but multiple names.

[Adam Heath-2]

Scott Gray wrote:

> On 23/03/2010, at 9:46 PM, Adam Heath wrote:
>
>> Scott Gray wrote:
>>> On 23/03/2010, at 9:37 PM, Adam Heath wrote:
>>>
>>>> Scott Gray wrote:
>>>>> This seems important, could you explain how it works a little more?
>>>> There was a thread about this recently, where I announced which
>>>> classes had full coverage.  This annotation just means that the
>>>> person(s) listed in the annotation are monitoring the source.  That
>>>> means that those people might be who to turn too if there are
>>>> questions on how the code works.
>>> But what does it do functionally?  How would it be different from:
>>> // "Don't mess up my code buddy!" signed Adam Heath
>> Shows up in javadoc.
>
> Got it
>
>> The retention policy could be changed to to Runtime, so that a test
>> case that fails could report something automatically.
>
> Doesn't buildbot do that for us already?

The person who breaks the build is different then the person who might
know a bit more about the code in question.

>> Remember, annotations are easier to access programatically.
>
> My only concern is that we removed all author information when we moved to Apache and this feels very similar to it.  I understand you're not claiming authorship but it does feel kind of implied.

Good point.  But, this isn't an author tag.  Author tags imply that
they will grow without bound, for everyone who ever modifies
something, would need to add their author tag.

[Adrian Crum-2]

--- On Tue, 3/23/10, Adam Heath <[hidden email]> wrote:
> Adrian Crum wrote:
> >> Remember, annotations are easier to access
> >> programatically.
> >
> > And automatically fix bad commits?
>
> What are you saying?

I was saying something along the lines of "Gee Adam, an annotation that emails a list when something in the source code changes is nice, but... couldn't you make it fix bad commits automatically too?"

In other words...

It was a joke. It must be late.

Or my humor sucks.



     

[Adam Heath-2]

Adrian Crum wrote:

> --- On Tue, 3/23/10, Adam Heath <[hidden email]> wrote:
>> Adrian Crum wrote:
>>>> Remember, annotations are easier to access
>>>> programatically.
>>> And automatically fix bad commits?
>> What are you saying?
>
> I was saying something along the lines of "Gee Adam, an annotation that emails a list when something in the source code changes is nice, but... couldn't you make it fix bad commits automatically too?"
>
> In other words...
>
> It was a joke. It must be late.
>
> Or my humor sucks.

ooh, ooh, ooh, I know the answer to this one.  wait, its, its ...

all of the above!

[Jacopo Cappellato-4]

On Mar 24, 2010, at 4:55 AM, Scott Gray wrote:

>
> My only concern is that we removed all author information when we moved to Apache and this feels very similar to it.  I understand you're not claiming authorship but it does feel kind of implied.

I agree with Scott, I don't like the idea of (re) adding names of committers to the source files... even if it is not an "author" tag it implies some sort of (light) ownership that I don't like. It is also bad from a purely aesthetic point of view.

I'd suggest to move this information to a Wiki page.

Jacopo