🚀 go-pugleaf

In article <memo.895245@cix.compulink.co.uk>, amclintock@cix.compulink.co.uk ("Cray Communications, Alex McL") writes:
|> In-Reply-To: <727580969snz@cryton.demon.co.uk> altman@cryton.demon.co.uk (Hugo Fiennes)
|> 
|> Hard Disk Whinge
|> I don't like my IDE drive/controller
|> which I bought from ICS.
|> 
|> 1) 25second boot up time (with RiscOs2)
So optimise your boot sequence...  What are you doing during bootup?  My
system takes far longer than that to spring into life.  Or are you talking
about spin up time?  In which case 25 seconds is probably about par for the
course...


|> 2) window bugs in filer, paths longer
|>    than 40 chars (or so) cause windows larger
|>    than they need to be. (big gaps to the right)
|>    Or perhaps this is a riscos bug.
Well this is a feature of the window manager.  The minimum width of any
window is the width it's title bar had when it was created.  Subsequent
changes to the title text make no difference.   This feature is present in paint
too, and every program I've written.  It's not HCS's fault.  Don't know how
to get around it apart from creating the window with a very small title, and
then changing the title afterwards.
Come to think of it, I'm pretty sure the filer DOES do this - you can shrink
the windows, it's just their default size which is large enough to see the
full filename.


|> 3) The occasional bad drive error.
|> 
Can't help you on that.  Might be your own programs or others with hard
wired drive numbers in.  (Illegal but it happens.)  Particularly easy to
get this happening with expansion of system variables - you might have
something hardwired to a floppy drive number causing this error.  Again don't
leap down HCS's throat.

|> I still haven't got a satisfactory answer to these
|> questions. I suppose I can't really complain
|> though since I have had it for three months+
Well who have you asked before now?


|> Does anyone know why the "*map" command
|> fills my screen with bits of free space.
|> Shouldn't it all be in one continuous block?
|> *compact reduces it a bit...
Yes it should do.  The E-type format maintains a pattern of lots of smaller
blocks near the middle of the disk (logically speaking, so I mean the tracks
half way along as opposed to the hub) and larger blocks out towards the edge,
so small files like Obey ones which need to be accessed fast for slick
directory opening can be found with the minimum seek distance, and large ones
which are going to take a while to load anyway get stuck in the less
desirable areas of hard disk real estate.
So don't complain, this is a Filecore feature anyway...


Ian Griffiths

In article <1993Jan26.130538.5927@infodev.cam.ac.uk> idg10@phx.cam.ac.uk writes:
>Well this is a feature of the window manager.  The minimum width of any
>window is the width it's title bar had when it was created.  Subsequent
>changes to the title text make no difference.   This feature is present in paint
>too, and every program I've written.  It's not HCS's fault.  Don't know how
>to get around it apart from creating the window with a very small title, and
>then changing the title afterwards.
>Come to think of it, I'm pretty sure the filer DOES do this - you can shrink
>the windows, it's just their default size which is large enough to see the
>full filename.

There is a minimum size field in the window definition. Zero means
that RiscOS should calculate the value from the shape of the
scrollbars, the size of title bar needed to display the entire title,
and the number of other icons around the window edge. If you set it to
some other sensible value ("Work area" menu option under !FormEd) then
the title bar will not be used to limit the window's minimum size.

This may be in the "little known fact" category, but it's in the
RiscOS2 PRMs!

--Clive.

In article <memo.895245@cix.compulink.co.uk>
amclintock@cix.compulink.co.uk writes:

>Hard Disk Whinge
>I don't like my IDE drive/controller
>which I bought from ICS.
>
>1) 25second boot up time (with RiscOs2)
>2) window bugs in filer, paths longer
>   than 40 chars (or so) cause windows larger
>   than they need to be. (big gaps to the right)
>   Or perhaps this is a riscos bug.
>3) The occasional bad drive error.
>
>
>I still haven't got a satisfactory answer to these
>questions. I suppose I can't really complain
>though since I have had it for three months+
>
>Does anyone know why the "*map" command
>fills my screen with bits of free space.
>Shouldn't it all be in one continuous block?
>*compact reduces it a bit...

I think perhaps you will be happier with your HD interface when you have
spent a little more time with it. No offense meant, but it does look as if
you have not been using the Archimedes for very long, so perhaps some
things are worrying you when they should not. I have had an ICS IDE drive
and controller for well over a year now and have been very happy with it -
it works reliably and fast. To answer your questions, though:

1) Boot-up will take longer than on a floppy-only system, as the computer
must wait for the hard drive to spin up. Until this has happened, the IDE
Filer cannot read the disc name to display underneath the disc icon.

2) This is a feature of the main filer application (which is in the main
OS ROM); filer windows belonging to ADFS and other filing systems also
behave in this way.

3) Bad drive errors are usually caused by badly-written software, not the
filing system itself. You presumably only have a singe IDE drive attached;
this will have a device number of 4. Some badly-written software
(particularly games) assume the selected filing system is ADFS and try to
read files explicitly from device 0 (the floppy drive).

Finally, the *map command displays the free space as a list of "bits", as
the free space on the disc is fragmented over the disc. This is normal,
and RISC OS does a fairly decent job keeping fragmentation to a minimum.

Graham

In article <memo.895245@cix.compulink.co.uk> amclintock@cix.compulink.co.uk writes:

>In-Reply-To: <727580969snz@cryton.demon.co.uk> altman@cryton.demon.co.uk (Hugo
> Fiennes)
>
>Hard Disk Whinge
>I don't like my IDE drive/controller
>which I bought from ICS.
>
>1) 25second boot up time (with RiscOs2)
some other IDE cards do this (DT software?)... I can't see a reason myself,
most IDE drives become ready from powerup within 5-10 seconds max.

>2) window bugs in filer, paths longer
>   than 40 chars (or so) cause windows larger
>   than they need to be. (big gaps to the right)
>   Or perhaps this is a riscos bug.
That's RISC OS...

>3) The occasional bad drive error.
That's the drive?
>
>Does anyone know why the "*map" command
>fills my screen with bits of free space.
>Shouldn't it all be in one continuous block?
>*compact reduces it a bit...
There's a maximum free chunk size, but you don't see it often (empty, big
disks.. it's something like 16mb chunks?), but there are always 'bits'
of free space. *compact doesn't compact any more when some oscure formula
works out that it's not worth the hassle (chunks big enough already? 
not too many chunks?)

Hugo
-- 
-----------------------------------------------------------------------------
                                     "Did I say that? My lips didn't move..."
Hugo Fiennes - The Serial Port                      altman@cryton.demon.co.uk
                                                or altman@cix.compulink.co.uk
-----------------------------------------------------------------------------