[DBIx-Class-Devel] people/abraxxa/doc_plain_value_removal

[Peter Rabbitson]

On Fri, Apr 05, 2013 at 05:39:05PM +0200, Alexander Hartmaier wrote:
> On Wed, Apr 3, 2013 at 3:45 PM, Peter Rabbitson <rabbit+dbic at rabbit.us>wrote:
> 
> > On Wed, Apr 03, 2013 at 03:32:15PM +0200, Alexander Hartmaier wrote:
> > >
> > > The new syntax is nice and concise.
> > > Do we want to separate the 'backward compatible' and 'convenience'
> > > shortcuts?
> >
> > I am not sure which one you mean "backwards compatible". They all have
> > their
> > place.
> >
> This wording is from the DBIC BIND VALUES docs in ResultSet
> 

That wording is correct. In the context in which it appears. Please 
engage brain. FWIW I would welcome patches improving the wording in this 
section. Note - improving. This is not as subjective as you think, there
is a baseline of "A is better than B" that appeals to humans rather
universally. Attempt to find this baseline.

> ... I've already changed the cookbook examples in 4e68a35.
> 
> ... I've rebased people/abraxxa/doc_plain_value_removal.

Yet you did not throw away parts that you were asked to throw away [1] 
and you did not squash relevant changes but piled more stuff on top of 
your old work which was half-rejected, and left someone (currently me) 
to figure out how to reign in this entire mess to extract the work that 
is actually relevant. These same commits came from someone who 
publically accused me of making it hard to contribute to DBIC. Reality 
check much?

> Did you forget to include the links? I couldn't find [4] or the others.

Yes I did, sorry about that. The links for message
http://lists.scsys.co.uk/pipermail/dbix-class-devel/2013-April/000409.html
are as follows:
 1: https://github.com/dbsrgits/dbix-class/commit/0e773352
 2: https://github.com/dbsrgits/dbix-class/commit/00c12490
 3: https://metacpan.org/module/DBIx::Class::ResultSet#DBIC-BIND-VALUES
 4: https://metacpan.org/module/DBD::SQLite#Functions-And-Bind-Parameters

> Can we document when the simplest syntax is not enough for DBIC to figure
> out the correct bind datatype?

*You* can document it. Or add a TODO point somewhere or something. If 
others "could" do it it would have ben done already. If you are asking 
"Can I (abraxxa) document this" - I am not sure what to answer.

Bottom line:
Out of d2cac29c, c4984b86 and 4e68a35a I was able to salvage ac09a390
which is now at the head of topic/constructor_rewrite. The changes are
as follows (you can also trivially see them by executing
`git diff --color 4e68a35a ac09a390c`)

 d2cac29c - partially taken (with a correct link - pod linking != URL linking)
            This chunk was dropped because I can not understand what does it
            actually say. Explain/add in a separate commit.

 +If you need to search for more than just the left-hand-side function result
 +you have to wrap the search arguments in an -and arrayref:            

 c4984b86 - thrown away as per [1]
 
 4e68a35a - taken as-is

Please delete adjust your branch accordingly (I can not delete it myself, 
because I can not know if you have it locally or not).

And in conclusion - while I am very efficient at this kind of 
cherry-picking, it still took me good 20 minutes alongside writing this 
reply. I strongly believe none of this would have been necessary if you 
applied the minimal amount of care when submitting your work for a 
second review. In light of what resource-drain it is to get you to 
produce a useful patch, I would recommend being very careful when you 
accuse others of being unhelpful and/or obstructionist.

Cheers

[1] http://lists.scsys.co.uk/pipermail/dbix-class-devel/2013-April/000403.html