[Joda-interest] RE: joda API suggestion

[Al M. <al...@bo...>]

turns out i'm going to be working on some joda date code for the next week
or so. if there's a (hopefully small) self-contained unit of work that i can
help out with, let me know.

sounds like the test cases might be a priority. unfortunately, i don't think
i know enough to write good programmer tests.

since i'm still figuring out how all this works, another task i could work
on would be tutorial/documentation for the newbie. if this is a priority,
let me know.

regards,

al

-----Original Message-----
From: Stephen Colebourne [mailto:ste...@ho...]
Sent: Thursday, June 03, 2004 3:24 PM
To: Al Major
Cc: Joda Interest
Subject: Re: joda API suggestion


From: "Al Major" <al...@bo...>
> i downloaded the 0.95 zip file off sourceforge. i'm getting a number of
> compilation errors in the src/test directory. a junit include appears to
be
> missing.
The test folder in basically full of miscellaneous test files, most of which
fail. This is not a great siutuaton, and one of the things preventing a 1.0
release is lack of tests.

> so far, i like the looks of your work. the java date api is quite severely
> brain damaged in term of usage, with the simplest use cases requiring the
> use of deprecated methods. unfortunately, it's the standard. are you
seeing
> a lot of adoption for joda? any chance of folding it into jakarta lang for
> greater visibility?
This has come up before, and at the time jakarta was deep in arguments. I'd
prefer to get a 1.0 out, then move, but I try to keep an open mind.

> i'm also very interested in seeing functionality for recurrences, both
> systematic (for example, a lunar calendar or ephemeris) and ad-hoc (for
> example, user defined anniversary dates). i'd be happy to contribute code
> for this if you have design ideas on what such a facility should look
like.
We haven't tackled recurrences yet as there is just too little time to get
the main code done for 1.0.

At present, a new PartialInstant structure is being completed. Then some
more tests are needed. At that point, a 0.96 perhaps?

If you are interested in helping then let me know ;-)), see sourceforge for
the low traffic mailing list.

Stephen