Re: Interface names in IndexedDB (and WebSQLDatabase)

(Are these comments going into someone's queue somewhere, or should I be
concerned there was no further response?  I ask because I'd kind of like to
start checking .idl files into WebKit.  :-)

On Fri, Jan 22, 2010 at 9:53 AM, Jeremy Orlow <jorlow@chromium.org> wrote:

> In general, sounds good to me.  Note that there already is an
> IndexedDatabase interface in your spec though.
>
> I'd also suggest renaming at least the following:
>
> ObjectStore
> KeyRange
> Environment
> DatabaseError
>
> At which point, there's not too many interfaces left without the IDB prefix
> (mostly synchronous variants of these interfaces) so maybe we should just
> prefix everything?
>
> Thanks!
> J
>
> On Fri, Jan 22, 2010 at 8:16 AM, Nikunj Mehta <nikunj@o-micron.com> wrote:
>
>>
>> On Jan 22, 2010, at 12:01 AM, Jeremy Orlow wrote:
>>
>>  The interface names in IndexedDB (and to an extent, WebSQLDatabase) are
>>> very generic.  Surprisingly, the specs only collide via the "Database"
>>> interface (which is why I bring this up), but I'm concerned that names like
>>> Cursor, Transaction, and Index (from IndexedDB) are so generic that they're
>>> bound to conflict with other specs down the road.
>>>
>>> Note that all but 5 interfaces in the WebSQLDatabase spec are prefixed
>>> with SQL (for example, SQLTransaction) which helps a lot.  It seems as
>>> though the remaining could also be prefixed by SQL to solve the problem.
>>>
>>
>> That will help.
>>
>>
>>
>>> I'm wondering if the majority of the IndexedDB interfaces should also
>>> have some prefix (like IDB?) as well since many of its terms are quite
>>> generic.
>>>
>>
>> I am fine with the following renaming:
>>
>> Database -> IndexedDatabase
>> Cursor -> IDBCursor
>> Transaction -> IDBTransaction
>> Index -> IDBIndex
>>
>> Nikunj
>>
>
>

Received on Wednesday, 27 January 2010 02:42:29 UTC