Evolution becoming unusable.

[Peter Damoc]

On Thu, 17 Mar 2005 15:05:35 +0200, Myles Green <rmg57 at telus.net> wrote:

> On Thu, 17 Mar 2005 13:29:23 +0100
> Matthew Thomas <mpt at myrealbox.com> wrote:
>
> <snip>
>
>> If anyone wants to develop an e-mail client designed for maximum
>> Gnomeness (the e-mail equivalent of Epiphany), I'd be interested in
>> helping design it (as I am with Epiphany).
>
> i've been using Sylpheed built with GTK2 for several months now - i got
> it from universe, i think - and it blends in very nicely with gnome and
> xfce4.2. with a little more work it should be up-to-task of being a
> light-weight and usable gnomish e-mail client. the only trouble is, i'm
> no code writer by any stretch of the imagination :-(

you might not be a code writer BUT you could become one... ;)
People think that writing code requires some amazing programming skills or  
some genetic predisposition...
Let me tell you... the only thing that "writing code" needs is "will" of  
course a good teacher and inclination toward exact sciences are a bonus.

If you think "is not possible for me to write an email app" let me tell  
you "You're wrong!"  everybody can write one of those... that's everybody  
with enough will power.
"How can I do that?" you might ask... well the fastest way "from Zero to  
Hero" is like this.

Learn GvR
Learn Python
Learn GUI programming with python. (pyGTK or wxPython)
Learn programing patterns (at least the ones involved in MVC)
Learn good programming habits (XP practices are a nice set)

How about this Mr. Shuttleworth? Create a developer factory!... instead of  
paying some developers to create software pay some teachers to draw such a  
path to lead eager people from Zero to Hero!

Maybe all this could be organised on levels...

Most of the current discontent with open source applications is related to  
GUIs... some are bloated some are hard to use some are plain ugly... but  
what most people don't know is the fact that what they see, the visible  
part or the View part from the MVC is in fact the easiest thing to do from  
that triad...  if the application is programmed using MVC...

I remember working in a firm where the person responsible for the View  
part could give us (the MC guys) work for a week in only few hours.  
Another thing I've learned is the fact that protocol boosts productivity  
and ultra-specialisation could create amazing speed in implementation.

Another problem that users face are bugs... omnipresent bugs... bugs that  
simply don't get fixed for years (just as OSNews's Eugenia complained in  
her editorials) and the only reason these bugs live so long is due to code  
bloat. Few people are capable of touching those bugs... and even fewer  
have the will to fix them.

In my naive view of the system such situations could be greatly improved  
if applications could use sound architecture and be implemented in an  
accessible language as python. Using a well documented MVC architecture  
for application combined with component specialisation (developers for  
views, developers for controllers, developers for models) could create a  
situation where if a bug appears... it could conceivable be traced to one  
of the layers (most probably it would be the Controller or the Model) and  
an a bunch of specialised programmers could be sicked on the problem to  
solve it...

Those that think "where would you find such a large mass of programmers?"  
should think about social interaction, social status... things could be  
set in motion with very little money... a 2 lines patch accepted in the  
main CVS tree is nothing to a seasoned programmer but to a highschool kid  
 from some obscure town in some country could become a hacker badge that  
would boost his social status skyward.... and IMHO a motivated enough kid  
would spend hours after hours hunting a bug out of shear passion.
Create a system where to bugs could be attribuited either reward money or  
some kind of bugpoints ... and have the first 10 bug sqashers receive some  
cool awards every year... iPods, iBooks, trips to Cannonical... whatever.

There is so much young energy wasted on games... Why? I think is because  
there aren't enough good teachers, because there is no such thing as a  
organised "developer factory" kids don't see a goal, or because they see a  
goal but cannot see the path toward that goal... or because the path looks  
so daunting that they feel they will never make it.

What if the path were to be clear? what if the begginer would know he/she  
will find a good mentor at each step?

GvR is simple enough to relax every begginer, Python gives them enough  
power to realise their own inner strength, a properly taught MVC  
architecture might enlightent them and make them realise that things are  
simpler than they appear or that they don't need to learn complex things  
in order to use them.

Joel Spolsky writes in "The Guerrilla Guide to Interviewing" that the  
description of perfect candidate for some job is:
"Smart, and Gets Things Done."
looking at this simple yet quintessential description I realise that I'm  
no candidate simply because altho I might be smart... I simply don't get  
things done... I've had some ideas and I've started implementing some of  
them only to take them to 90% and then abandon them completely... or  
starting with enthusiasm and getting stuck in design phase because of  
insecurity on my knowledge...
So... what can I do?... I've discovered that there are some people that  
know less than me on some subjects... I discovered that I really enjoy  
helping those people... I think its because some of their problems are  
small enough for me to fix in the timespan my attention could stay focused  
on some subject....

Sorry for such a long OT email.

-- 
Peter Damoc
Warrior Of The Light
http://www.sigmacore.net/