[reportlab-users] Voice in the wilderness (was : Widget Libraries, anywhere?)

[Rasjid Wilcox]

On Friday 10 October 2003 00:40, Ian Sparks wrote:
> Dinu's original complaint was that there don't seem to be a lot of (any?)
> 3rd party widgets available for ReportLab despite RL being quite widely
> used.
>
> If the fundamental question is : Why arn't there more 3rd party widgets?
> "Lets build a website" is not an answer, its just changing the subject so
> we can talk about things we feel more comfortable with, like the relative
> merits of SF vs Wiki's.
>
> I'm not trying to rain on the parade here but I think that there isn't a
> market for 3rd party Widgets because there isn't a sufficient *demand* for
> 3rd party widgets - ReportLab is too complete and makes it too easy to
> "just do it" and move on.
>
> I think a website would help if the RL newsgroup was drowning in
> contributions and we needed a way to categorize and organize them but it
> doesn't look that way to me.
>
> If contributions were water we'd be dying of thirst. We intend to solve
> this by making a bucket?
>
> I can't decide if I'm a dry, old, cynical, git or if I'm the voice of
> reason. Maybe both.

Ian,

I think you should read http://www.benkler.org/CoasesPenguin.html. (Abstract 
at http://www.benkler.org/CoasesPenguin.html)

One of the points made there is that certain infrastructure needs to be in 
place for the open-source process to work.  Without that structure it just 
does not happen.  I commend the people at ReportLab for seeing this and 
taking steps to change the situation.

Secondly, I think there is quite a lot more that could be done with reportlab.  
As I see it, reportlab is very much like TeX, only much easier (and it 
doesn't do equations).  But it still could do a lot more, and it could be a 
lot more accessible.  It needs more "LaTeX" level tools and widgets to make 
things even easier and more accessible for the less technically inclined.

My guess is that part of the issue here is that most people (including many 
programmers) do not generalise a problem first.  They go: 'here is a problem' 
-> 'here is a solution'.  However, to generalise well, you have to work the 
other way.  You have to be able to think: 'here is a generalised solution' 
... and it solves this problem, and this problem, and this problem, and a 
whole lot of other problems that have not even been thought of yet.

I don't really want to argue this, since what I *really* want to be doing 
(after I get some more sleep) is writing some reportlab libraries and 
widgets, and then finding a good place to host them.

Good night,

Rasjid.

-- 
Rasjid Wilcox
Canberra, Australia (UTC +10 hrs)
http://www.openminddev.net