[Koha] Emilda and Koha: how to choose between them?

[Pierrick LE GALL]

On Wed, 17 May 2006 20:48:16 +1000
Paul Gear <paul at gear.dyndns.org> wrote:

> Hi folks,

Hi Paul,

I don't know Emilda, but as a Koha developer, let me answer about your
Pros and Cons :-)

> * [Emilda cons] Developer/user communities not so active

That's a really bad point in my opinion (see my conclusion below).

> * [Koha pros] Pretty (the suggested demo sites look awesome!)

I suppose you've seen Liblime [1] templates. They are not the default
template but they are always up to date when new releases are
available. The important is that you can easily switch to other
templates and even create your own. We've planned to open an online
extension manager where anyone will be able to contribute by adding a
template. Hopefully, it should open by the end of this month (May 2006).

> * [Koha pros] Seems to have quite an active developer and user
> community

You're right, the developer community is really active and worldwide
(mainly USA, France and New-Zealand). I'm sure we could improve the
size of the user community by using more user oriented tools (that's
another debate).

> * [Koha cons] Install not so simple - no Debian package; requires
> messing about with CPAN

I totally agree, installing Koha is not simple. We have to make an
effort to simplify the installation procedure. Nevertheless, in my
opinion, Koha will hardly be as simple to install as a Php/MySQL
software. Simply because of Perl additionnal modules.

Furthermore, with Koha 3.0 will come a new step in the technical
installation: zebra server. I think Emilda also uses one (I confirm
after visiting Emilda website), so if you've found Emilda installation
easy, it means Koha can make zebra installation easy.

> * [Koha pros] I hate Perl [...]

Of course Perl is more complexe than PHP. As a PHP and Perl developer,
I appreciate the simplicity of PHP and the possibilities of Perl data
structure. If the Koha team had to make a technology choice today to
build a new application from scratch, I'm not sure Perl would still be
used. At the beginning of the project in 1999, when Chris (from Katipo)
decided which technology to use, PHP was not as popular as today and
Perl was quite the most obvious choice for a web based application.

The advantage of Perl versus PHP is CPAN, even with PHP having Pear.
In some way, the drawback of Perl is also CPAN, because you need to
install many dependencies from CPAN, check the compatibility, be sure
the module is maintained.

Anyway, the most important is that Perl is a free software oriented
programming language. Such as Python, Ruby or PHP (even if PHP is
driven by a single private company, in the same way of thinking
PostgreSQL is more "free" than MySQL). I would understand the complaint
if Koha was written in .Net or Java for example.

> Are there any other considerations that i should be looking at? [...]

On a free software project, one of the most important thing to consider
is the activity of the project: how many developers, are they active,
how open is the project (easy to contribute, features discussed
between users and developers to reach the best solution, availability of
a roadmap in time and features). I would even say that maturity is less
important than the activity (even if Koha is quite mature). With a good
activity, you can be sure the project won't be abandonned in a year.

[1] http://liblime.com

Cheers,

-- 
Pierrick LE GALL
INEO media system