Re: [Haskelldb-users] Latest patch to HaskellDB

[Justin B. <jgb...@gm...>]

I should also mention that the latest patch to HaskellDB breaks
backward comparability in a big way. If you don't want to go through a
painful update, I suggest staying with the current version of
HaskellDB (v 0.12) for now. The UPGRADE file will contain directions
on whats needed to update existing programs, but it is very raw at the
moment.

Some of the major stuff:

  * Regenerate all DBDirect-generated modules - they will no longer compile.

  * Operators which have changed:

    (!) becomes (.!.)
    (<<) becomes (.=.)
    (#) becomes (.*.)
    (field <<- value) becomes (field .=. constant value)

  * emptyRecord needs to be added to the end of all projections:

      project $ TInt.f02 << count(s ! TInt.f02)

  becomes

      project $ TInt.f02 .=. count(s .!. TInt.f02) .*. emptyRecord



On Wed, Sep 16, 2009 at 9:17 AM, Justin Bailey <jgb...@gm...> wrote:
> I've just pushed an update to HaskellDB to the repo and I'm hoping to
> get feedback and/or patches:
>
>  * Thanks to Brian Bloniarz, HaskellDB now uses the HList library to
> represent queries. An upgrade guide will be provided. The file UPGRADE
> in the root directory is very rough but describes changes necessary.
>
>  * All cabal files have been updated to v 1.0.
>
>  * HUnit tests have been updated to reflect new changes.
>
>  * A PrintQuery module, for debugging intermediate query forms, has been added.
>
>  * HDBC driver support has been updated to support HDBC 2.0.
>
>  * Versioning updates to require the Haskell Platform, 2009.2, have been made.
>
> HaskellDB isn't quite ready for another release yet, but more to come
> on that soon.
>
> Justin
>