[web-devel] [Yesod][Persistent] FooIdEq,FooIdIn

[Michael Snoyman]

Well, let's discuss the motivations for requiring the annotations at
all. I *think* no one but me really knows why I did this, because I
have neither documented my intentions nor taken advantage of them code
wise.

While it's true that it lowers the number of constructors created at
compile time, that was *not* my primary concern. The idea was to allow
for optimizations: for a SQL backend, the SQL migration code could
produce indices automatically. For other backends, it could have an
even more dramatic effect on how the data is stored.

But I think that argument doesn't really apply to an IdIn constructor.
Greg, I like your proposal of including it by default. If there are no
objections, let's try to get it in for the next release.

One other note about the next Persistent release: we're going to be
splitting the package up into a core and a Template Haskell set. Some
people are using Persistent in an iPhone app, and the iPhone compiler
cannot support TH. Just a little heads-up.

Michael

On Sat, Mar 12, 2011 at 9:27 PM, Greg Weber <greg at gregweber.info> wrote:
> It might be better to save the space after the table line just for future
> table annotations.
> Then to annotate id, it would be added as a field.
> This brings up another point though- do we need to add "In" annotation- why
> not just enable it? Does it make the compiles take really long, etc? We
> should at least document the rational in the persistent section.
> On Sat, Mar 12, 2011 at 10:46 AM, Michael Snoyman <michael at snoyman.com>
> wrote:
>>
>> On Sat, Mar 12, 2011 at 3:07 AM, Katsutoshi Itoh <cutsea110 at gmail.com>
>> wrote:
>> > Hi
>> >
>> > I had often wanted FooIdIn, FooIdEq on Foo Persistent type.
>> > Would you support this by default?
>> >
>> > from japan.
>>
>> As Max said, glad to see that you're safe.
>>
>> For FooIdEq: I don't think there is any need, since it can be
>> expressed more succinctly with the "get" function. As for FooIdIn:
>> this *was* a recently requested feature. Maybe we can get it in for
>> the next release. I think it would make sense to add the In attribute
>> to the entity itself, eg:
>>
>> mkPersist [$persist|
>> Person In
>>    name String
>>    age Int
>> |]
>>
>> Any objections?
>>
>> Michael
>>
>> _______________________________________________
>> web-devel mailing list
>> web-devel at haskell.org
>> http://www.haskell.org/mailman/listinfo/web-devel
>
>