On 15 Apr 2022 as I do recall,
Derek Haslam wrote:

> Powerbase v.10.42 (11-04-2022) is now available for download from my
> website www.boulsworth.co.uk and is, thanks to a lengthy exchange with
> Martin Avison, a thoroughly cleaned up version which is expected to
> perform well on all recent versions of RISC OS including those with
> the latest version of the Wimp.
>
I'm curious as to what on earth the problem could have been, since so
far as I'm aware the database front-end is a basic WIMP icon-based
display that one wouldn't expect to have redraw issues....

--
Harriet Bazley == Loyaulte me lie ==

I like work; it fascinates me; I can sit and look at it for hours.

In article <c0448bd959.harriet@bazleyfamily.co.uk>,
Harriet Bazley <harriet@bazleyfamily.co.uk> wrote:
> On 15 Apr 2022 as I do recall, Derek Haslam wrote:

> > Powerbase v.10.42 (11-04-2022) is now available for download from
> > my website www.boulsworth.co.uk and is, thanks to a lengthy
> > exchange with Martin Avison, a thoroughly cleaned up version
> > which is expected to perform well on all recent versions of RISC
> > OS including those with the latest version of the Wimp.

> I'm curious as to what on earth the problem could have been, since
> so far as I'm aware the database front-end is a basic WIMP
> icon-based display that one wouldn't expect to have redraw
> issues....

I am not sure Derek reads this newsgroup, so I will say...

The redraw problem appeared to be caused because although the PRM
says that Wimp_RedrawWindow must be the first Wimp call after a
WimpPoll, there were many calls to it in the program that did not
abide by this rule.

It appears that this caused no problems up to about Wimp v5.76 (29
Jun 2020), but v5.80 showed redraw problems. Exactly which change
provoked this I never managed to work out, but changing PB to abide
by the rules fixed the redraw problem. And that enabled me to help
with the original weird Report problem that Chris posted.

Martin

--
Martin Avison
Note that unfortunately this email address will become invalid
without notice if (when) any spam is received.

In article <59d99aab61News03@avisoft.f9.co.uk>,
Martin <News03@avisoft.f9.co.uk> wrote:

> [...] although the PRM says that Wimp_RedrawWindow must be the first
> Wimp call after a WimpPoll [...]

Even before the 0 reason?
The redraw in my program has glitches too. Could it be a matter of swapping
Wimp calls? Or is it just my usual lack of understanding, after all?

Paul

--
http://riscos.sprie.nl

In article <59d99aab61News03@avisoft.f9.co.uk>, Martin
<News03@avisoft.f9.co.uk> wrote:
> In article <c0448bd959.harriet@bazleyfamily.co.uk>, Harriet Bazley
> <harriet@bazleyfamily.co.uk> wrote:
> > On 15 Apr 2022 as I do recall, Derek Haslam wrote:

> > > Powerbase v.10.42 (11-04-2022) is now available for download from
> > > my website www.boulsworth.co.uk and is, thanks to a lengthy
> > > exchange with Martin Avison, a thoroughly cleaned up version which
> > > is expected to perform well on all recent versions of RISC OS
> > > including those with the latest version of the Wimp.

> > I'm curious as to what on earth the problem could have been, since so
> > far as I'm aware the database front-end is a basic WIMP icon-based
> > display that one wouldn't expect to have redraw issues....

> I am not sure Derek reads this newsgroup, so I will say...

He can't get access via his ISP - BT. Have any BT users out there got
round this problem? I'm sure Derek would like to know.

--
Chris Newman

On 15 Apr, Chris Newman <cvjazz@waitrose.com> wrote:

> He can't get access via his ISP - BT. Have any BT users out there got
> round this problem? I'm sure Derek would like to know.

Free: http://news.aioe.org/
EUR10 per annum: https://news.individual.net/overview.php

There are probably others.
--
Brian Howlett - Email to From: address deleted unseen
-----------------------------------------------------
"I'm Brian, and so's my wife..."

In article <59d9b66e56Paul@sprie.nl>,
Paul Sprangers <Paul@sprie.nl> wrote:
> In article <59d99aab61News03@avisoft.f9.co.uk>,
> Martin <News03@avisoft.f9.co.uk> wrote:

> > [...] although the PRM says that Wimp_RedrawWindow must be the
> > first Wimp call after a WimpPoll [...]

> Even before the 0 reason?
> The redraw in my program has glitches too. Could it be a matter of
> swapping Wimp calls? Or is it just my usual lack of understanding,
> after all?

Not quite sure what you mean.

When a Wimp_Poll returns reason code 1 (RedrawWindow), then
Wimp_RedrawWindow must be the first Wimp call made, and it should not
be called at any other time.

Does that help?

--
Martin Avison
Note that unfortunately this email address will become invalid
without notice if (when) any spam is received.

In article <59d9b67c56cvjazz@waitrose.com>,
Chris Newman <cvjazz@waitrose.com> wrote:
> In article <59d99aab61News03@avisoft.f9.co.uk>, Martin
> <News03@avisoft.f9.co.uk> wrote:

[Snip]

> > I am not sure Derek reads this newsgroup, so I will say...

> He can't get access via his ISP - BT. Have any BT users out there
> got round this problem? I'm sure Derek would like to know.

I suggested www.eternal-september.org to him, but I do not know if he
followed that up. I use it to fill in missing posts from giganews!

--
Martin Avison
Note that unfortunately this email address will become invalid
without notice if (when) any spam is received.

In article <59d9b91b68News03@avisoft.f9.co.uk>,
Martin <News03@avisoft.f9.co.uk> wrote:

> When a Wimp_Poll returns reason code 1 (RedrawWindow), then
> Wimp_RedrawWindow must be the first Wimp call made, and it should not
> be called at any other time.

Ah, now I understand what you mean - and I sin against that rule indeed.
Before calling Wimp_RedrawWindow, I do a Wimp_GetWindowInfo call. Should I
really move that one after the redraw call? If I do, things go horribly
wrong. So, I still may not understand you.
(Apologise for drifting off the topic.)

Paul

--
http://riscos.sprie.nl

On 16 Apr, Paul Sprangers wrote in message
<59d9f56d3fPaul@sprie.nl>:

> In article <59d9b91b68News03@avisoft.f9.co.uk>,
> Martin <News03@avisoft.f9.co.uk> wrote:
>
> > When a Wimp_Poll returns reason code 1 (RedrawWindow), then
> > Wimp_RedrawWindow must be the first Wimp call made, and it should not be
> > called at any other time.
>
> Ah, now I understand what you mean - and I sin against that rule indeed.
> Before calling Wimp_RedrawWindow, I do a Wimp_GetWindowInfo call. Should I
> really move that one after the redraw call?

What details do you need to find from Wimp_GetWindowInfo to perform the
redraw, that you don't get in the block returned by Wimp_RedrawWindow and
Wimp_GetRectangle?

> If I do, things go horribly wrong.

Define "horribly wrong"... which might help us help you. :-)

--
Steve Fryatt - Leeds, England

http://www.stevefryatt.org.uk/

In article <59d9f56d3fPaul@sprie.nl>,
Paul Sprangers <Paul@sprie.nl> wrote:
> In article <59d9b91b68News03@avisoft.f9.co.uk>,
> Martin <News03@avisoft.f9.co.uk> wrote:

> > When a Wimp_Poll returns reason code 1 (RedrawWindow), then
> > Wimp_RedrawWindow must be the first Wimp call made, and it should
> > not be called at any other time.

> Ah, now I understand what you mean - and I sin against that rule
> indeed. Before calling Wimp_RedrawWindow, I do a Wimp_GetWindowInfo
> call. Should I really move that one after the redraw call? If I do,
> things go horribly wrong. So, I still may not understand you.
> (Apologise for drifting off the topic.)

I am only quoting what is in the PRM and the StrongHelp manual!
What do you need from WindowInfo before you enter the redraw loop?

--
Martin Avison
Note that unfortunately this email address will become invalid
without notice if (when) any spam is received.

In article <mpro.rafg4d01cbkc603j7.news@stevefryatt.org.uk>,
Steve Fryatt <news@stevefryatt.org.uk> wrote:

> > If I do, things go horribly wrong.

> Define "horribly wrong"... which might help us help you. :-)

[I transported this discussion to a new thread in
comp.sys.acorn.programmer, with the subject 'Wimp_RedrawWindow protocols']

Paul

--
http://riscos.sprie.nl

In article <c0448bd959.harriet@bazleyfamily.co.uk>,
Harriet Bazley <harriet@bazleyfamily.co.uk> wrote:
> On 15 Apr 2022 as I do recall,
> Derek Haslam wrote:

> > Powerbase v.10.42 (11-04-2022) is now available for download from my
> > website www.boulsworth.co.uk and is, thanks to a lengthy exchange with
> > Martin Avison, a thoroughly cleaned up version which is expected to
> > perform well on all recent versions of RISC OS including those with
> > the latest version of the Wimp.
> >
> I'm curious as to what on earth the problem could have been, since so
> far as I'm aware the database front-end is a basic WIMP icon-based
> display that one wouldn't expect to have redraw issues....

Derek Haslam has replied via me & made some remarks re this....

"No, I don't understand it and it puzzled Martin too. You're quite
right about the database record window having nothing but icons and
should only need user-redraw for the grid when in design mode. The
problem occurred with Wimp 5.79, which I think you and others such as
Harriet Bazley have, but *not* with Wimp 5.76 which I have on my
ARMbook with RO 5.27. Nor did it happen on my WinPC/Virtual Acorn
system with Wimp 5.61. I'd be interested to know if the same problem
has surfaced with any other software.

I tried issuing a Wimp_OpenWindow command, even though it oughn't to
have been necessary. That didn't work either unless a Wimp_CloseWindow
was issued first. Very strange indeed! The eventual solution was to
use Wimp_ForceRedraw.

The other problem which you originally communicated to Martin
(corruption of a report window with multiple occurrences of the same
field) was an interesting one. Martin pinned it down to the routine to
remove white space and created a single-record database which
illustrated the problem clearly. It would only occur if the "Total"
line was longer than the maximum *configured* length (not the maximum
*actual* length) of the first field. An unusual situation. I
implmented a workaround which prevented the corruption but wasn't
satisfied because it could truncate the footer under certain
circumstances. Then I tracked down the bit of the program which
excludes certain types of line (especially header lines) from the
white space removal. I found that a footer line was *not* so excluded.
Added that to the list and all was well.

Powerbase is a large program and I converted to to RISC OS (from a BBC
micro version) around 1991 so I sometimes find myself puzzled
regarding how things work!

v.10.42 is the newest version. I did a silly thing when I uploaded it
to the website, failing to remove my "reg" file and a Config file
which refers to my HD pathname. A warning from another user led me to
rectify that PDQ.

Regards,

Derek"

--
Chris Newman

In article <59da38ecaecvjazz@waitrose.com>,
Chris Newman <cvjazz@waitrose.com> wrote:
> Derek Haslam has replied via me & made some remarks re this....

I think you had better get in touch with Derek again about his website as
it says payment can be made through the now defunct APDL website.

--
Stuart Winsor

Tools With A Mission
sending tools across the world
http://www.twam.co.uk/

In article <59da77f9a5Spambin@argonet.co.uk>, Stuart
<Spambin@argonet.co.uk> wrote:
> In article <59da38ecaecvjazz@waitrose.com>, Chris Newman
> <cvjazz@waitrose.com> wrote:
> > Derek Haslam has replied via me & made some remarks re this....

> I think you had better get in touch with Derek again about his website
> as it says payment can be made through the now defunct APDL website.

Well spotted. I've sent a mail.

--
Chris Newman

In article <59da9cb850cvjazz@waitrose.com>,
Chris Newman <cvjazz@waitrose.com> wrote:
> In article <59da77f9a5Spambin@argonet.co.uk>, Stuart
> <Spambin@argonet.co.uk> wrote:
> > In article <59da38ecaecvjazz@waitrose.com>, Chris Newman
> > <cvjazz@waitrose.com> wrote:
> > > Derek Haslam has replied via me & made some remarks re this....

> > I think you had better get in touch with Derek again about his website
> > as it says payment can be made through the now defunct APDL website.

> Well spotted. I've sent a mail.

Derek has acknowledged & will fix it.

--
Chris Newman