Re: [OpenJade-devel] progress and plans

[Terje B. <li...@po...>]

Ian Castle <ian...@op...> wrote:

>I think that opensp 1.5 is probably ready to go (I've given up worrying
>about a Windows build - if people are interested in Windows, they're mor=
e
>likely to try with a "final" release, rather than a "pre-release" so a
>1.5.1 may well pick up any changes for Windows).

As a data point; I finally got around to installing Red Hat 8.0 on a spar=
e
box and OpenSP 1.5pre8 builds out of the box there. That's GCC 3.2,
autoconf 2.53, automake 1.6.3, and libtool 1.4.2.

If you want access to it -- to build RPMs or smoketest changes etc. -- I
can set you up an account (IIRC you said you didn't have recent Red Hat
around these days).


We're gearing up for a new major release of the W3C MarkUp Validator --
which now _requires_ OpenSP 1.5pre8 or newer -- so I, at least=92 would v=
ery
much like to see a final release of OpenSP 1.5 that we can point people a=
t
(and preferably also RPMs and .debs etc.).


>However, there also needs to be a stable version of openjade which can
>coexist nicely with opensp... and it seems a bit silly to have to have
>two different versions of libosp...

Yes, especially since I suspect distributors are mainly looking for the
jade bit and only taking OpenSP along as part of the package. To get
OpenSP 1.5 into distros it needs to work with a good version of OpenJade
(or the other way around, rather ;D).


>OpenJade and OpenSP are new getting pretty old (1996?). A lot of the C++
>features that were "new" when James Clark first started on the project
>(namespaces, templates etc) are now pretty standard in most compilers on
>the various platforms. I'm thinking of removing some of the conditional
>compiliation that wraps these features to make things a bit easier to
>understand (and move towards ANSI Standard C++ as much as possible)... a
>lot of these #ifdefs just serve to obfusticate the source somewhat. Any
>thoughts on this?

My -- totally uninformed -- opinion on the matter is that anything that
makes the code easier to read and understand can only be good; and pretty
much essential to enable bringing new blood (developers) into the project.