[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: future of LedgerSMB

 From a crossplatform point of view I would have thought that Perl was 
actually about the best supported tool out there...?
No actually Python or Java is going to give us the best capability there.

It's not really relevant, but I still reckon that Perl is more widely supported than Java.  For example Perl runs on Linux...


I don't really know Python that well, but it seems like a nice tool.

Can't really see why the language matters all that much though - personally I like Perl a lot, I'm only using Ruby for my latest project because I liked the Rails framework for that project.

Also of course Perl has such a huge library of functions already available through CPAN....

  I like the idea of
a more MVC architecture, (such as catalyst or whatever the perl 
equivalent is called?).  However, it's just a coding style issue

MVC doesn't require a toolkit :).

True, but Django, catalyst, rails, etc all come with a huge bunch of pre-written functionality for web apps.  So you could almost say that in fact MVC-web *does* require a toolkit...

I agree with your point though

I guarantee you we won't do rails. Rails has a dependency that enforces
bad database design, basically the requirement that a primary key be

I wasn't suggesting that Rails was used for a moment!

However.... Just to take up your point, you *can* use any primary key column that you like with rails.  It's slightly more complicated if you insist on a compound key, but certainly you have few problems using arbitrarily named primary keys.  Essentially rails has a default name, but you can set the actual primary or foreign key col names in your model and everything works fine.

Rails and Django, etc, all have some very interesting features which are worth exploring and trying to fit into whatever MVC architecture is actually in use for a given project.  Since we were previously talking about how to have external API's interfacing into SMB I thought it relevant to mention the Rails REST features of the new 1.2 rails.  There are some nice ideas there which may be helpful (when implementing the solution in Perl....)

In summary, has anyone got any experience of Catalyst, and any reason why the whole product shouldn't slowly crawl into a Catalyst MVC framework (iteratively) over the next versions...?

Just an idea...

Ed W