[softwarelist] Re: File problems

On 19 Nov at 22:17 David Pilling <flist@xxxxxxxxxxxxxxxxxxx> wrote in
message <Ya4eEkAefNYFFwdB@xxxxxxxxxxxxxxxxxxx>

> In message <gemini.j8zqpq04fu8s003s2.tony@xxxxxxxxxxxxxx>, Tony van der
> Hoff <tony@xxxxxxxxxxxxxx> writes
> > David has indicated that he may put in some work to resolve the
> > incompatibilities, but I believe he got sidetracked from this intention.
> > David, would you be interested in discussing a commercial tie-up to port
> > OP to QT/Linux?
> 
> I've got a PC here which I rebuilt and then installed Linux on. I
> installed it, and that was it, never turned it on since, due to
> non-computing events.
> 
> I am open to offers. I suspect I would be happier with Linux than I am
> with Windows. However as an outsider (to Linux) I don't understand the
> market. I don't understand the Windows market come to that.
> 

'Understanding' the various markets is a tall order for anyone. I believe
the Linux market to be much more web orientated than, say, Windows, which
tends to rely on popular mags to get the message across. Whilst Linux is
becoming more established on the Desktop, the market is much smaller than
Windows (But, of course, very significantly larger than Risc OS, and unlike
the latter, is growing exponentially). There tends to be a bias in favour of
FOSS, although there are some very powerful commercial applications around. 

The only DTP of any significance is Scribus, and whilst free, is highly
ideosyncratic, and offers only a tiny fraction of what OP can do. I believe
there is room for a 'killer' DTP app on Linux, and that OP could well be
that package. But <shrug>, what do I know?

> As to QT... the problem is that DTP depends very intimately on the
> operating system. Although OP was written for RISC OS with no thoughts of
> portability, moving it to Windows would not have been much work - except
> for how close one has to get to Windows.
> 
Now there I have no immediate answer. My feeling is that, given OP renders
quite nicely under wine, a native implementation can only do better. I also
rather suspect that the work you've put in to the Windows port (Presumably
Visual Studio) will be very helpful in a Linux port. Linux, (and, for that
matter, QT) has a very great advantage in being open source, and proximity
to the OS is very much more transparent than Windows, and there is also very
much more opportunity to examine the source of other graphics packages,
notably the Gimp, to see how others have tackled specific problems. 

Realistically, I'm sure there is a lot of work involved, which may come to
nought, but you have proved it can be done for Windows. From my experience,
whilst the GNU/Linux tools are not as polished as the Windows ones, the
underlying OS is much easier to work with, and if anything, must be easier
than what you have successfully achieved. However, in the end it will be a
matter of 'suck it and see', I'm afraid.

> For example you can't just paint text on screen, if you do it looks awful,
> you've got to start fiddling with how many pixels appear between
> characters - that is just one example it goes on and on.
> 
> So does QT offer the sort of control that is needed?
> 
No, QT is just a set of GUI libraries. It's benefit is providing the
necessary infrastructure of windows, system calls, etc for the desktop
system. There are other libraries out there to facilitate graphics
programming, which may help provide the necessary control over screen
rendering. 

You have the advantage of intimate knowledge of the application and its
requirements, and my inclination would be to let you get on with the job you
have magnificantly performed over the years. However, with all respect, you
appear to have become, I hope only temporarily, stalled for various reasons.
I also believe that having built a machine, and installed Linux, is a far
step from porting a major application to it, even for a man of your
undoubted ability. I believe it would take at least a year of quite intense
use of the system to become sufficiently familiar with it to contemplate
such a project. 

So, with no knowledge of the internals of OP, but with some years experience
of developing for GNU/Linux, I'm offering you a helping hand. I really don't
know how successful it'll be, but I'm prepared to give it a try.

I must own up to not being entirely altruistic. For a start, I have been 
running a systems consultancy which has used OP for all of its technical
documentation, for many years, and having built up en extensive library of
templates and style sheets, I would like to continue to do so. Since my
in-house systems are all Linux based, having to fire up my trusty Risc PC,
or my bastard Windows laptop is a disincentive to getting the documentation
written. Secondly, I would expect some quid-pro-quo, over which I'll go to
private email. Thirdly, This would be an ideal project for my next step into
the depths of Linux application programming.


-- 
Tony van der Hoff        | mailto:tony@xxxxxxxxxxxxxx
Buckinghamshire, England

Other related posts: