Rich Mellor writes:
> I have just been having a look through the list of subscribers to this
list
<>
> As for those others, what a software house, such as mine, needs to know is
> why you are not customers. I appreciate that not everyone on this list
> purchases commercial software, and certainly not the type of software I
sell.
> However, maybe there are some suggestions for other software which I
should
> be providing which would appeal to these people.
>
> ****** Let us know - your software houses and authors all need feedback
if
> we are to survive in this market ********
What we need, I think, are new applications - and/or improvements to
existing ones - and programs that address the fact that the world has
changed considerably since the final collapse of the QL software market
(approx 8-5 years ago, depending on how you reckon). There are lots of,
mainly free, utilities out there: File finders, viewers, simple text
editors, calculators, etc, etc, but very few real applications that justify
so many utilities. Applications that people would like to use, and need to
use, would lead to all kinds of spin-off effects as we would then need more
drivers, tools and yes, utilities, which us wannabe programmers seem to
enjoy producing. Applications, however, are not easy to produce: You
need good programmers, or even teams of programmers, good tools, and
last but not least, a market for them! A catch 22 situation.
I dont think we can ever expect to see anything like Corel Draw or Word or a
video editing suit on the QL. But I do think there is a need for simpler
applications that do a good job with a minimum of fuss, at a good turn of
speed, and in a no-nonsense, stable environment, provided: They are of a
high quality, well supported, non-mercenary, ingenious, superior. If we try
to emulate Windoze or Linux or Mac we are going to end up being always
dissatisfied, always feeling inferior and in a continuous state of terminal
decline. My own feeling is that I dont really need, dont even want! 65k
colored sprites, vector fonts, fade-out menus or scented icons. I dont mind
24 million colours in a drawing program or route-finder, or postscript fonts
in (some) wordprocessors, but I feel I can live a full and wholesome life
without all that where I dont need it.
To achieve that, though, we need good tools. The standard of programming
tools is totally inadequate, because Im not advocating that we remain in the
dark age of floppies and monochrome and obscene commandline incantations. To
me at least, even comparative to Windoze and Linux, SMSQ/E on QPC or Q60 is
about sufficient for what I have spelt out above. Yes, there is work to do
in those areas too, but the fact that there is no simple, standard way to
knock up a main window with some buttons and a few menus is a major
handicap. The lack of advanced structures and types in S*Basic also doesnt
help. Sure, C is no doubt great, but it isnt "native" to the QL
environment, at least not as it stands (though there are some good examples
of programs that, probably through considerable effort, have managed to hide
that fact. Ported programs invariably dont achieve an adequate degree of
"nativeness" or integration, but may be acceptable where no other
alternative exists).
Good tools will make possible that even people who arent ace programmers,
but that have good concepts or solutions and a certain amount of technical
aptitude and discipline, can produce adequate applications. Now please dont
think that this whole rant is merely to advertise my miserable little
utility, FastFind, in the vain hope that I then can bask in the glory of the
admiration of my peers! It is true that I did want to make a statement with
it. It is a toy, vastly over-specified in comparison to any utilitarian
merit it may have. Only if I were planning to sell thousands of copies would
it justify the effort that went into producing it (and it will never be
finished according to spec!) What I wanted was to demonstrate - to sketch,
as it were - the sort of thing I should like to see in a "modern" QL
program: It has icons, cascading menus, (some) colour, interactive window
adjustment; employs advanced structures and non-Sbasic data types. It is
run-time configurable and has multi-session memory functionality in some
areas, re-usable dialogues, plug-in facilities for external programs (only
partially implemented) an auto-installation program, etc, etc. Most of this
is, unfortunately, written as one-offs whereas in fact a lot of it could
have been be made up of pick'n mix, standard components that should reduce
development time by 50-90%!. But such components are just not available;
each programmer will have to write, and re-write, their own. What a waste of
effort and resources! However, writing such components in a re-usable format
is no trivial task; I know, Ive been wandering in that wilderness for years!
It would take even an ace programmer such as our TT half a morning to write
a halfway decent menuing system.
So if anyone wants to know what to develop for, or next to, or into SMSQ/E,
Id say that we need some of the things or sorts of things Ive tried to
outline above. If there are any ace programmers left in our camp and theyre
still with me, I hope theyre listening, and unless they have a better plan
(lets hear it!) know what needs doing.
Sorry Rich, for hijacking your thread, but I do believe I have tried to
respond to part of your question, at least, in a very general way. I can
assure you all that if it had been possible for me to say all this in a few
snappy sentences, I would have!
Per
