hu.dwim.rdbms and hu.dwim.perec tests

[Attila Lendvai]

> Is the cl-perec mailing list still the canonical place for questions ? I
> can't subscribe via Mailman.


i'm not sure about the mailing list situation since the cl.net
migration.

but we haven't set up anything new since then, so i guess the answer
is yes (granted that it works, i'm CC'ing this mail to see).


> I'm trying to use perec(useing the LIVE repo) on sbcl-1.1.14/x86_64
> & postgres 9.2 and I see 3 test failing in hu.dwim.rdbms and a huge
> number in perec(which I haven't investigated yet).


don't use the HEAD, use the LIVE repos. they are better tested and i
don't pull there unbaked patches.

perec HEAD is in a half-done cleanup state which breaks things.


> To test only the Postgres backend in hu.dwim.rdbms I use funcall
> hu.dwim.rdbms.test::test/postgresql.
>
> The attached patch fixes two of them, I think mainly because
> hu.dwim.rdbms::expand-sql-ast-into-lambda-form-cached is supposed to be
> used at macroexpansion time to output an expansion and not at runtime,
> so eval more or less works there but I have the impression that the test
> would need to be rewritten.
>
> The third failing test is TYPE/RATIO and it's not clear to me what the
> test is trying to check.


no surprise. the comment is bitrotten... sorry, i'll push this new
comment:

;; make sure that postmodern is configured in a way that doesn't
;; silently distort ratios into floats when inserting them into the
;; database. note that this is not the case with the official
;; postmodern codebase.


i'll move on now to fork postmodern on github and rebase our patches.
it also moved to git... :/


> As for perec, I'll try to see what's happening. I think that part of the
> errors are caused by the sqlite and oracle backends which I'd rather
> exclude. I'll keep you informed of my progress.


as seen from this mail, do ask before you invest too much of your time
into any troubles!

-- 
• attila lendvai
• PGP: 963F 5D5F 45C7 DFCD 0A39
--
“Think you can, or think you can't ― either way, you'll be right.”
— Henry Ford (1863–1947)