[GRASS-dev] GRASS-TNG

[Radek Bartoň]

On Wednesday 21 of February 2007 18:05:07 tlaronde at polynum.com wrote:
>
> And where exactly do you think it is?

Correct me if I'm wrong since I don't know KerGIS as well as you, but compared 
to GRASS-TNG KerGIS is:

It is not object oriented nor component oriented.
It is independend in mean that its program itself.
It is not a proove of concept or some kind of research how to improve GRASS 
itself.
It is not only a framework for module development.

> Unix dates back 1970. It's more and more the reference.

It's reference from outside point of view and from concepts but i. e. Linux is 
inside different.

> UML on the other side is a contemporary subject of laughs (see acmqueue,
> they have some great papers about "dead by UML fever").

UML is above all what it's acronym says:

Modelling - Tool to make a model of a system before it exists in practically 
minimal time according to time of implementation.
Language - Kind of formalism to describe this models.
Unified -   Providing unambiguous semantics even that this is at least 
discutable. 

> If you want to improve GRASS, you must learn first what GRASS is and how
> it was supposed to work as a whole when there was a core _developers_
> team. Beauty is homogeneity. Distinct pieces that fit perfectly one with
> each other in an unity.
>
> If you want to learn GRASS, you must go back to its core.
>
> That's what KerGIS has done.
>
> I already changed a lot of things. I already reverted some changes I
> made because I found that I did not correctly understand how things
> work. KerGIS will end being significantly different from GRASS from the
> engineering point of view. And despite of that, it will still be GRASS
> because it will be one step nearer GRASS' very nature. It will be an
> evolution of GRASS.
>
> ESRI presented its products as a 3 stages line, from cheap to expensive:
>
>
> 1) Visualizing and updating SIG informations;
>
> 2) Being able to produce maps;
>
> 3) Being able to analyze data.
>
> GRASS/KerGIS is the more powerful, since it's naturally stage 3.
> Improving tools for 1 and 2 can be done (these tools already exist even
> if they are not sexy).
>
> And more than once, people with all the bells and whistles of shiny
> commercial GUI were unable to do the work while KerGIS/GRASS was able to do
> it.
>
> I find it rather---say---disgusting to read that whenever people find
> something not obvious with GRASS, they think this is GRASS fault. And
> the last thread about throwing away v.digit is typical.
>
> Despite what people may think, I'm less conceited than the majority of
> the people I read here. I still think that CERL GRASS is a golden gift.
> And when I found a mistake, I do not blame the developers, but I'm just
> proud to have the chance to become one of them making KerGIS/GRASS
> even better.
>
> But I'm not born tired, that's probably why I do not feel comfortable
> in the contemporary "europe". You are probably right then : since I do
> not belong to this low epoq, I do not feel compelled to produce the
> "software of the XXI century". I prefer to work on an eternal one...
>
> Even if I'm the only one. People are not "impressed" by GRASS? But who
> is impressed by them?

Maybe I feel it same way? I like GRASS as a user even or just without GUI. 
It's modules, possibility to make a command line script for some piece of 
work and for more other reasons which are not important in this discussion.

The only I don't like about GRASS is it's code itself. Any modification in 
modules code is a nightmare for me.  Maybe it's just because I'm used of 
confortance of object oriented and dynamic languages but maybe it's just that 
there are no or at least poor programmers' documentation and there is nobody 
who wants to document that vast of code nor who completely undertands to it.

I just want a GIS which is beauty from outside as GRASS is and even from 
inside as GRASS is not according to my and few other people is not. Beauty is 
very subjective word so I see beauty of code in it's easiness, cleanness, 
power and documentation.

-- 
Bc. Radek Bartoň

Faculty of Information Technology
Brno University of Technology

E-mail: xbarto33 at stud.fit.vutbr.cz
Web: http://blackhex.no-ip.org
Jabber: blackhex at jabber.cz