[JDEV] Time to develop a client ?

[Jens Alfke]

On Thursday, August 16, 2001, at 04:45 AM, Mathieu Nouzareth wrote:

> I am relatively new to Jabber and was wondering how much time it takes 
> to
> develop a decent client with the minimal functionalities (with 
> Jabbercom).
> Any hint wold be appreciated.

One thing to consider is whether it wouldn't be more productive to spend 
your time improving an existing open-source client (many are listed on 
the Projects pages at DevZone) instead of creating a new one. Think 
about what it is that really interests you -- if it's a particular 
feature that doesn't exist, you could work on adding it to a client that 
already works instead of spending 90% of your time re-implementing the 
basic stuff. In the grand scheme of things it's better to have a small 
number of really good clients rather than a large number of less 
polished ones.

Of course there are still valid reasons for creating new clients -- if 
(1) you want to support a platform that there's no client for yet, or 
(2) you want a fundamentally different user interface, or (3) you think 
the existing clients aren't good enough to work on, or (4) business 
reasons make it infeasible to use GPL'd software. (I'm building my own 
client due to reasons 2 and 4, the first based on my own opinions, the 
latter on the opinions of the customer I'm developing for.)

--Jens