[Ilugc] Internet service provider info
- From: prabhu@xxxxxxxxxxxxxxxxxx (Prabhu Ramachandran)
- Date: Thu, 8 Jan 2004 17:29:57 +0530
"RS" == Sridhar R <Sridhar> writes:
>> 2. Figure out what you really want to do with a
>> well defined set of goals before doing it.
RS> I know some Soft. Engg. I have already done the
RS> requirements analysis stuffs. Everything is feasible.
RS> (except problems with embedding glade within IDE)
The question is who are you targetting and what "holes" do you
perceive in the existing set of tools? What are you going to with
this new tool? If one knows Python its easy enough firing up an
editor and coding away or using the interactive interpreter. There
are too many editors, and enough highly interactive interpreters for
Python. If you are looking to generate a GUI builder, then again
there are lots of tools for that.
>> 3. Its better if you join an existing project
>> rather than start on your own. Did you look at
>> http://vb2py.sf.net or http://boa-constructor.sourceforge.net/
>> ? Of course, if you can
RS> Already saw those two projects. vb2py converts VB
RS> code to wxpython code. But the aim of my project is not just
RS> alowing VB programmers to use Python but also .....
The devil is in the detail.
RS> Boa! Again wxpython. You know ... wxwindows is so
RS> _buggy_ and thus wxpython. Too many wrappers, too many
Huh? From all my experience with wxPython, it is certainly not buggy.
It is a moving target but thats a different issue. GTK has also been
a moving target.
As regards bindings, pyGTK is a binding too so there is no big
difference in that regard between pyGTK and wxPython. Where are there
"too many bindings" I cannot understand. Trouble with GTK is that its
still not as cross platform as wxPython is (atleast when I checked
BTW, have you looked at PythonCard?
RS> bindings. But the real intention of VP anyway differs from
RS> Boa, which is ..
RS> VP is aims to be an _ease_of_use_ IDE and
RS> programming language (that's Python) like VB and much flexible
RS> not as VB.
There is no dearth of Python based IDE projects (IDLE, Boa, Eric
etc.). Have you read this:
Perhaps what you want to do is generate another IDE thats GTK
specific. Which brings me back to question 2. Where is the clear
analysis of what you want to do, and what current projects don't
already do or intend to do? Why is your solution better than say
>> Unless you are either a totally incompetent programmer or a
>> god, you typically will commit infrequently. So I'd think
>> dialup access is quite ok for this sort of thing. Continuous
>> access does help though.
RS> But each time I dialup (even for uploading a 1 byte
RS> file), it makes up equal to one telephone call!
So, you do the obvious, queue up commits and do it in one go when you
have a substantial amount to commit. You usually have to connect to
get email, so this should not be a big deal.
Other related posts: