RISC OS - so long and thanks for all the fish.
----------------------------------------------
Well, it's a sad day indeed. Far from being a toy OS
as many will claim (never actually having used it,
of course) it should be alive and well, with two new
machines being launched in the last couple of tears or
so, and  both forks of the OS (yes, that's right, TWO
forks) being updated, and one them in the process of
being open-sourced. The future should be bright, the
future should be green.

So why drop it then?
--------------------
Well, part of  the reason is the clue above. Two forks
of the OS for a minority platform? It's difficult to
believe that such silliness could even start, never
mind carry on. The simmering "dispute" (for want of a
better word) between the two camps has been going on
for several years and was just a bit of a side-show
until I bought an A9Home for the business. Ok, so I
was aware the OS was "a bit beta", but I wasn't expecting
it to be unable to rememebr half of it's configuration
settings, or be lacking pre-installed printer drivers,
or crashing losing data to often for comfort. Though
those points are irritating in the extreme, well, never
mind, they can be over-come - the real killers are these
two issues.
In the eight months or so that I've had the A9, and
despite all the known bugs and problems there have been
precisely NO updates at all. That may not matter to your
average hobbyist but it is a disaster for someone wanting
to use the machine "for real".
The other issue, not un-realted, came about when I bought
the C/C++ complier, for use on the A9 and the Risc PC to
32-bit the apps I have written that use !Prophet to help
run the business, to release them, and to develop them
further. I happily installed it and..... it didn't work.
One is supposed to get personalised copy of the actual
compiler from the Iyonix site, which I did. Following
the insturctions to ready it for use it failed miserably -
the protection was still in place (set it "absolute" and
then run it do this). Never mind, at only 50ukp the money
wasted this time was not a patch on the 650ukp wasted on
the A9. I'd cpould simply go back to the old version,
couldn't I?. Except the files had been deleted from the
hard drive by the installer - including the copies in the
back-up !boot  directory. Now panicking as I had no way
of keeping my software up-to-date, there followed an
exchange of e-mails the Iyonix people, which still resulted
in no working complier. I did get it to run once, but the
hard drive then died and guess which file DisckNight couldn't
recover! Not that it mattered really, in less time than it
had taken to try to sort this one issue out I had got Ubuntu
Linux, with ROX, up and running on one of the PC's - with
a working complier and Quasar accounts. More on this later.
The galling thing about this was really the fact that when
I mentioned that I intended to run the compiler on the A9,
the chap at Iyonix though it was amusing - clearly not
a problem. I wish I could treat my customers like that!
This was, of  course, on top of the petty bickering and
short sightedness of thosewho have appointed themselves
as the Keepers Of The Knowledge. I know exactly what they
will say as well, the same tired arguments will get trotted
out again, and doubtless it will ever more be so until the
only audience is themselves. The problem is they are wrong -
demonstrably. Is the ROS user base increasing or decreasing?
OK, so FireFox is wonderful (though NetSurf is far sllicker),
but whether the Keepers like it or not, the vast majority of
users consider Flash, video, etc to be part of the browser -
donning teflon shoulder pads and saying "not my problem" is
simply burying you head in the sand. Trumpeting FireFox,
omittingto mentiom the lack of video etc will have a negative
ettect rather than a positive one. Can it actually do
anything that NetSurf or Fresco can't? In effect no, apart
from getting the layout nicer sometimes. Is that going to impress
someone used to another platform? Like hell. They'll try to go
to YouTube or whateve and just return to wherever they came
from. So that's the "home" market scuppered.
What ROS needs is the dull, boring, stuff to get into markets
where such things are not required, even deprecated. Accounts,
drivers for devices like the Dymo label printers, Point Of Sale
software (we had that, and very good it was too, but users have
to be able to integrate into their accounts and stock control
for it to be of real use - no one these days is going to bother
printing out end of day stuff and then re-enter it into their
accounts software - and not everyone uses bar codes in a shop).
This is where the ROS could score well, it's UI is far easier
to use than others, and the lack of video streaming etc could
be viewed as a plus in the "work" market.
The real danger to ROS from not having these things is not
their abscence per se though. As mentioned above, the compiler
and A9 debacles pushed me into actually installing Linux. I
didn't want to, but I had to - I've needed a modern multi-user
accounts system for a long time now, but I've desperetley
battled on with ROS and !Prophet for far longer than is
sensible (bear in mind we are talking *serious* use here,
use that can make or break a company, not mucking about with
something in the evening out of a sense of loyalty or seeing
if it can do this or that). The complier fiasco was the last
straw, and so the deed was done.  Like, I strongly suspect,
so many other ex-ROS users, the experience was an eye opener.
It's a different  world out there - just one example will
suffice to show why ROS is not now an option. My web site
generator program used to take 9 hours to generate the site.
The same code, ported to Linux, does it in roughly 50 SECONDS.
It's not just speed either. There are those who bang on about
Zap etc (I used to too) - well, I have three editors installed,
two of which do most of what  Zap does, and one (part of the
Anjuta project suite) that does far more. Draw? I though there
was nothing like it, but Xara Xtreme (apart from the duff name)
does it all and more (except import Draw files, which is a bit
of a bummer). It (the Linux version) is even directly supported
by Computer Conepts now. The only application I'm missing to
date is Impression, but then I haven't really looked yet.
Even IglooFTP is very nearly as easy top use as FTPc.

So will I miss ROS?
-------------------
Yes. Big Time. It's very hard to give up something you've
supported come hell or high water for so long, but with ROX
providing (with a few very irritating exceptions) a ROS like
desktop, I'll survive.
What I won't miss is being told my future by the Keepers, who
do a very good impression of completely failing to uderstand
the difference between those who use a computer to do various
tasks because the *have* to, and those who do so because they
either want to or just see it as a hobby. Fiddling while
Rome burns and all that....

The reposses will an amalgum of:

"Developers know best....."
"You don't understand....."
"Zap can do......."
"I run my buisiness (with only half a dozen transactions a
day rather than hundreds)......."
etc etc.

which all just ignore three simple facts - you've lost
*another* user and failed to ask yourselves why, you've also
lost your most visible user to the outside world, and I
*do* understand (you might not like it, but it's true), that's
why I've been forced to move.
Not one will ask the only questions that really need to be
asked - what can we do to retain, or even gain, users? Which
market(s) can we address bearing in mind the technical
limitations of the hardware?

But that's ok, hey, who needs users when making clever
points about the semantics of someones news post is so
much mor fun!

Richard

In article <1174296326.494271.37...@o5g2000hsb.googlegroups.com>,
   beamends <sa...@beamends-lrspares.co.uk> wrote:

[Snip]

Terribly sad to read your post.

Thank you for taking the time to explain what happened and how you feel.  

Your comments encapsulate the whole problem - the ROS world is becoming the
MG Owners Club*, a bunch of geriatrics and youthful enthusiasts desperately
trying to keep the old machines on the road whilst insisting that the
machines are just so nuch better than this modern stuff.  Like the MG
owners, most of them seem to use something else for serious stuff.

I still harbour a hope that somehow those vital bits of missing software
will appear or that someone will find a way of cloning MartinW, David P and
others so that a real avalanche of new software will appear.

Good luck and I hope that you keep some connection with the RO world.

Cheers

Alan

--
Alan Calder, Milton Keynes, UK.

In message <1174296326.494271.37...@o5g2000hsb.googlegroups.com>
          "beamends" <sa...@beamends-lrspares.co.uk> wrote:

<snip>

Sorry to lose you, Richard. As you know, you and I have had similar
views many times about RO, and, sadly, I agree with all you say.

I can't quite work out why Iyonix Ltd don't take much more of a lead
in software development - it's in their interests as they'd sell more
machines, and they own the O.S. I also can't see why Ad6 went to RO
Ltd instead of Castle for their version of the OS. They must be very
frustrated at RO Ltd's inability to complete the OS for them in
anything like a sensible time.

Is it really a bad business plan for Castle/Iyonix Ltd, Ad6, and maybe
even CJE to get more involved in serious professional software
development to sell more of their products and open up new markets?

Jeremy.

--
Jeremy Brayshaw     <><

In article <1174296326.494271.37...@o5g2000hsb.googlegroups.com>, beamends

There is much that you have said that is correct - and much that is wrong,
both general and specific. What is correct should be taken to heart and put
right and you should be fairly, but strongly, tackled on the bits that may
mislead. Sadly that will not happen. It won't happen because anyone trying to
discuss such problems honestly will be publicly villified. In fact that's the
only real problem out of the whole lot; your quibbles *could* have been put
right had honest public criticism been allowable - and publicly countered if
you included unreasonable comments. As it is most developers work shut away
from the public because working in public just isn't worth the aggravation;
and some of the few that do risk it make sure they get their retaliations in
first - and that just escalates the problems.

The answer is simple. Use Usenet properly or don't use it at all - ie scrap
all csa.groups and portals and lists. If we are going to use it then answer
all questions politely; don't make attacks on individuals; don't criticise
products because you own a rival product; appreciate that in the RISC OS
market all devlopers and suppliers can best flourish through the success of
the others.

But that only works if everyone agrees and everyone follows those rules.

I'll risk two criticisms:

ONE: I offered to help sort out one of your problems (and that may have led to
the whole being sorted). I needed a copy of your bug report but apparently
that had been limited to telephone conversations. If anyone has a critical
problem then it really is essential that they are properly documented. 1.
There is a far better chance that your problem will be sorted out and your
expense *not* wasted, and 2. It would benefit the rest of us rather than
having to wait longer for someone else to encounter the problem, report it
properly, have it identified, and get the bug sorted.

TWO: You relied on Usenet rumour to inform you of development. This isn't your
fault, but the misinformation does mean that decisions are made on bad
premises.

--
        John

In article <1174296326.494271.37...@o5g2000hsb.googlegroups.com>, "beamends"

  I didn't cry! And the A9 does everything I must have - including the Castle Compiler.
I would like better printing facilities - but that can wait.

--
    _                     _________________________________________
   / \._._ |_  _     _  /'       Orpheus Internet Services
   \_/| |_)| |(/_|_|_> /          'Internet for Everyone'
_______ | ___________./       http://www.orpheusinternet.co.uk

On Mar 19, 9:25 am, "beamends" <s...@beamends-lrspares.co.uk> wrote:

<snip>

I am sorry to see you go Richard.

That actually, with respect, is not really the problem (look at how
many variants of Linux there are and they seem to work, also the bulk
of RISC OS users are *still* using 3.7 or earlier, then you have 4.xx
and 5.xx, within Select you'll have people who are up to date and some
who have lapsed out at earlier stages). This is something that would
have existed even if Castle had *never* released RO5.

I have a simple rule that I apply to *every* machine I've ever bought,
- never buy the earliest ones - wait a year. I did that with Iyonix
and on PCs. Me a guinea pig - no thank you.

Beta means that - it's capable of release to a limited set of users
before main release. RO4.XX as on A9 was a work in progress - it's a
bit unfair (no matter how miffed you feel) to assume any particular
level of completion. Beta meant it wasn't complete some things will
and somethings wont work. As to the delay in updates that is a matter
for ROL and Ad6 - its not in their interests to delay on this (so I'd
expect things to progress overtime - albeit perhaps not rapidly enough
to keep you onboard).

Thing is I would *not* buy a machine that I knew was incomplete, so
why did you ?

There were alternatives either (i). Upgrading an existing RPC to use
Select OR (ii). Choose an Iyonix - which pricewise is not a world
removed from the A9 and has some track record behind it.

But thing is the problem *might* be related, The compiler has to
*work* on the OS you use. The OS is the foundation upon which all
things run. If the OS isn't working (in your opinion) then there is
little reason to assume *any* particular application (or compiler)
would run properly on it either.

<snip>

Thing is if you mentioned you couldn't get Microsoft Visual C# to run
on an Apple Macintosh would you expect MS to fix it?

The guy hasn't developed the variant of RISC OS on the A9 he *can't*
possibly know what will/won't work on it. Sure you have an A9 for 8
months and you couldn't get it to work - so why should he.

With respect you knew A9 was beta and you still bought it. That's not
Ad6's fault, or ROL's fault. You had alternatives you ignored them.
That's your choice - but ultimately it has consequences such as the
ones you listed. Yes maybe some of the issues could have been
highlighted - perhaps some were but you perhaps ascribed them to
people with axes to grind - but ultimately you made YOUR choice and
now face the result.

If you really wanted to fix the issue i'd have (i) Not complained
publically here about the A9 (ii) Sold it and recouped some money
(iii). Bought an Iyonix and asked Castle to resend you your Compiler
keyed to that machines IP address. Then you'd have been up and running
and still on the RISC OS platform.

<snip>

For the love of Pete, some of those formats are *closed* you can
either not get them or if you attempt to use them you can get sued. If
you're prepared to pay for the development and license fees yes it's
doable - but if we can't we can't. You won't find HD on Linux either
(and if it appears it'll take some time, just like what happened with
DVD).

And firefox is what - a BROWSER - it displays WEBPAGES (like this
one). It's the plug in's that do the work. They're either proprietary
or rely on GPL code which can't simply be incorporated into the core
RISC OS.

Again how will *your* departure help, really. What's more if going why
make such a racket. Go quietly the rest of us are trying to keep the
thing going. The sad reality is ultimately if you want content you'll
have one choice - Microsoft. And that's because the Media/Content
providers know MS will try to keep their content *safe*. Linux they're
wary off and Apple don't really have a larege enough market presence.

<snip>

You bought a machine whose OS was not complete and expect compilers
and everything to work fine on it. Is that realistic?

As to Linux, it's a fine OS and I wish you well with it.

Good again I wish you well.

Not always

Yes, you don't.

We lost you because either (i) you had unrealistic expectations or
(ii) were misinformed (or a combination of the two). It's still no
less a pity

<snip>

Regards and best of luck.

Annraoi

This is a gross simplification of the the problem:  In the Linux
distribution world, almost all of the software packages, the kernel, and
OS infrastructure is built from identical or near-identical source code.
And where the source isn't identical, other Linux distributions can
cherry-pick from it.  What you end up with is pretty much identical
software delivered arranged slightly differently.  ie: completely
different to the ROL vs CTL versions of RISC OS.

<snip>

Mac OS is a completely different OS.  This also seems to go against what
you say above in that having two branches of the OS "is not really the
problem".

B.

On 19 Mar 2007 John Cartmell <j...@cartmell.demon.co.uk> wrote:

So we can expect the immediate cessation of all your off topic posts on
csa.* ? No I thought not.

---druck

--
The ARM Club Free Software - http://www.armclub.org.uk/free/
The 32bit Conversions Page - http://www.quantumsoft.co.uk/druck/

In article <848522c64e.dr...@druck.freeuk.net>, druck <n...@druck.freeuk.com>
wrote:

It's disappointing (but not surprising) that you on your own could ensure that
cooperation won't work. This is the bit you snipped:

"If we are going to use [Usenet] then answer all questions politely; don't
make attacks on individuals; don't criticise products because you own a rival
product; appreciate that in the RISC OS market all developers and suppliers
can best flourish through the success of the others.

But that only works if everyone agrees and everyone follows those rules."

--
        John