[GEM Development] An opportunity to ask some questions :)

[Michael Bernstein]

Hi Shane,

> I'm glad there is some activity on the list at the moment, it gives me
> the chance to take a quick poll :)

Ok, i also want to give you a little bit feedback.

> Section 1 - General Questions
> 
> 1.1 - Are you developing software to run under GEM (any version)? If
> No go to 1.2

Yes.
 
> 1.1.1 - What version of GEM are you developing for? (eg: OpenGEM, DR
> GEM/3, Atari GEM)

Atari GEM
 
> 1.1.2 - What tools are you using?

Turbo C for Atari (includes a IDE), RCS (from DR and RsourceMaster)

> 1.1.3 - Are they free? (as in 'really' free - released under a
> free-ware or open source license)

No

> 1.1.4 - Where do you get your documentation? From the tools you use,
> from other developers, from google?

TOS.HYP (ST Hypertext, also online available)
Atari Profibuch ST STE TT

> 1.2 - Do you *need* to use GEM? If No go to 1.3

No

> 1.2.1 - Why?

Because this os the "standard OS" for my Atari computers.
 
> 1.2.2 - What is stopping you from changing to another OS and UI ?

Seee 1.3.1

> 1.3 - Do you enjoy using GEM? If No go to 1.4

Yes

> 1.3.1 - Why?

The GEM on my Atari Computers have a "easy to use" feeling which i miss
on other systems. I can also write my own software and nothing is hidden
inside of huge librarys. For my small programs good enough.

Windows is more colorfull, Mac hides internals, Linux are great for my
servers but with KDE goes the windows way.

Its like the first love. All others miss something to feel again like
the days i started with my first Atari. Maybe computers are to common
and my knowledge to big to give again the feeling of this magic of
explore the computer.

> 1.3.2 - What would you change about current GEM versions if you could?

For Atari, MiNT with Posix functions N.VDI and new AES was a good way.
For PC GEM i would like to see the same way for GEM to let both systems
be compatible. This will allow to compile software with little effort
for both systmes. But i dont know, how it is possible with limited
resources. Maybe only with DOS extender?

For both systems VDI drivers. On Atari f.vdi was a free replacement.
But it lacks some features of N.VDI

> 1.4 - Please give me a free form comment about what you think of GEM.

A nice small gui which was easy to use. And easy to programm at low
level. This means i can learn the basic on GUI programing without
frameworks wich hides the system.

But i see not much need for GEM as a small GUI. This is because a
small system of today is much more powerfull than a old DOS PC. As an
example: Tanenbaum is working on Minix 3. Minix was a "leightweight"
mini unix which can published in a book. The version 3 was made to
get a small UNIX like OS for small systems. But they also use X11 as
a GUI and they dont see the need of a smaller GUI.

It seems for me, it is difficult to find ai new place for GEM.

> Section 2 - Specific Questions
> 
> 2.1 - Are there standard function names for the VDI/AES operations available?

Yes. The names are defined by DR and Atari as a standard for portable
GEM development. You can look inside my PC GEM library to get the
standard names.

> 2.1.1 - Would you be upset if the names changed (ie: for VDI function
> #1 from v_opnwk to vdiOpenWorkstation)?

No

> 2.1.2 - Would it be better to make the older names available even if
> different names were defined?

For my needs, the old names are good enough.

> 2.2 - Are there any OO frameworks for GEM development available?

I dont know about such a framework. But maybe there is one available
for Atari computers.
 
> 2.2.1 - If so, what languages (and compilers) are they available for?

If yes, i expect a framwork for the GC.

> 2.2.2 - Do you think they are important for developing GEM applications?

I think no. A good framework can make development more easy and faster.
But it eats more memory. And this is a limited resource for DOS software.

> 2.2.3 - What languages (and compilers) would you like to see them for?

I am satisfied with the compilers a have.

> 2.3 - Imagine this scenario: You can write applications for Linux that
> use the standard GEM API and the output can be viewed on any external
> device with the appropriate client program. Would this be useful for
> you?

This is the idea behind X11! I have some UNIX machines and a X server
on every machine which should show the output. And this works with less
overhead than VNC. I would like more the idea of map the GEM API with
a library to the X11 systems. This will allow to compile GEM software
on a UNIX system. But i dont have to set up a different graphical gui
and can use all benefits of X11.

Regards
Michael Bernstein