DISCUSS: deprecating a TON of Delegator/GenericDelegator calls(huge)

[Adam Heath-2]

Deprecate:
* GenericValue.getRelatedDummyPK(String)
* Delegator.getRelatedOne and Delegator.getRelatedOneCache
* Delegator.findByAnd and Delegator.findByAndCache
* Delegator.findByPrimaryKey and Delegator.findByPrimaryKeyCache

Attached is the git log of the set of changes.  It's about 2M of diff.
 I haven't done full testing of all of it, some is in code that hardly
ever gets run, and we all know that the test cases we have don't cover
much.

As you can see from the log, I've been holding on to these for quite a
while.  I'd like to get them off my plate.  However, if I commit them
to trunk, it'll make it more difficult to backport changes to the just
released branch.  So, I'm asking if I can commit these to both trunk
and 12.04.  If not, we'd have to wait even longer to fully remove
these methods(at least another year).

[Jacopo Cappellato-4]

Hi Adam,

On May 3, 2012, at 5:20 PM, Adam Heath wrote:

> So, I'm asking if I can commit these to both trunk
> and 12.04.  If not, we'd have to wait even longer to fully remove
> these methods(at least another year).

I didn't review your patch yet so please consider this as a general statement; but I don't see big issues in backporting this kind of work (the same holds for the work that Adrian is doing) if this makes sense, will make the refactoring easier and quicker and there are no big objections.

Jacopo

[Adam Heath-2]

On 05/03/2012 10:44 AM, Jacopo Cappellato wrote:
> Hi Adam,
>
> On May 3, 2012, at 5:20 PM, Adam Heath wrote:
>
>> So, I'm asking if I can commit these to both trunk
>> and 12.04.  If not, we'd have to wait even longer to fully remove
>> these methods(at least another year).
>
> I didn't review your patch yet so please consider this as a general statement; but I don't see big issues in backporting this kind of work (the same holds for the work that Adrian is doing) if this makes sense, will make the refactoring easier and quicker and there are no big objections.

It's not possible to review the patch; what was attached was just the
changelog and diffstat(git log --stat=120,120).

Such a large set of changes has a fairly good chance of causing
problems.  Whoever does such a wide-reaching search/replace won't
nescessarily understand everything that the code is doing.

Btw, I did not just do a blind programatic change.  I used grep to
find each occurrence, then manually typed in the correct new call.  So
you can be assured that at least one pair of eyeballs looked at each line.

[Adrian Crum-3]

On 5/3/2012 4:44 PM, Jacopo Cappellato wrote:
> Hi Adam,
>
> On May 3, 2012, at 5:20 PM, Adam Heath wrote:
>
>> So, I'm asking if I can commit these to both trunk
>> and 12.04.  If not, we'd have to wait even longer to fully remove
>> these methods(at least another year).
> I didn't review your patch yet so please consider this as a general statement; but I don't see big issues in backporting this kind of work (the same holds for the work that Adrian is doing) if this makes sense, will make the refactoring easier and quicker and there are no big objections.

Thanks Jacopo. There are some Mini-language API changes that I would
like to back-port (to make the branch forward-portable), and it's good
to know that the process is worth considering.

-Adrian