🚀 go-pugleaf

Peter Ceresole <peter@cara.demon.co.uk> wrote:


> I still use sys 7.5.2 on a 750/100 and in general it's beautifully stable.
> However as I try more stuff I can tell it needs a few new hooks... So I
> shall be getting sys 8 and the update to Norton. But...

Is there a particular reason you've stayed with 7.5.2 so long? I found
just upgrading through to 7.5.5 made my 7500 hugely more pleasant
to use, and didn't cost anything either.
> 
> I get maybe one type 11 error a week. Often less. But in the past two
> weeks, while experimenting with RealPlayer 5 and other stuff I have had my
> problems with whatever it was that corrupted my system and stopped me
> running Norton when booted from the HD.
> 
> Each time these type 11 errors have shown up, they have tended to persist
> when I reboot. Type 11. Reset (Ctrl-Cmd-start) and some problem or other
> crops up again- usually the crash on running Norton from the HD. Sometimes
> something else- however, there is never any disk or file damage and if I
> shut down and wait a few minutes, then restart, the machine (including
> Norton) will run happily for days until I do something odd again and have
> another short but irritating run of type 11 errors.
> 
> It behaves as though the memory weren't resetting on startup. But I thought
> that this was done automagically. Am I wrong? If I restart via the
> keyboard, can a problematic memory state persist?

I've noticed the very same thing in the past on my 7200/90. It would
seem to get in a temper every so often and I might spend a day
performing restarts every 20 minutes or so. I don't see how memory
can remain unreset across a restart, at least memory that matters, as
that will have fresh, correct information written to as the system
software reloads. All I could think was that I was using some
particularly crash prone combination of apps which was causing the
problems
-- 
'ric

In article <1d1n2cm.1s88e38c796kcN@rm-dynf1-138.rmplc.co.uk>,
btgsch@rmplc.co.uk (ric) wrote:

>Is there a particular reason you've stayed with 7.5.2 so long?

Yes. It just worked so very well. Stable, and with Speed Doubler running
(made a huge difference- without it I'd have lost patience and upgraded
ages ago) the finder fairly steams along. The machine is lovely and slick
and I never had any reason to upgrade. What's happening now is that I am
beginning to need (read "want") extra bits which say they need 7.5.3 or
above, so I may as well go to sys 8. Especially as my experience of other
Macs running 7.5.5 is that they work no better than 7.5.2/Speed Doubler. 

>I've noticed the very same thing in the past on my 7200/90. It would
>seem to get in a temper every so often and I might spend a day
>performing restarts every 20 minutes or so.

This isn't nearly as bad as that- but it can't be the applications I am
using because they are mostly the same all the time.

I shall live with it and upgrade soon.


-- 
Peter

Peter Ceresole <peter@cara.demon.co.uk> wrote:

[snip]

> It behaves as though the memory weren't resetting on startup. But I thought
> that this was done automagically. Am I wrong? If I restart via the
> keyboard, can a problematic memory state persist?

My Mac Plus (which uses system RAM for the screen) will often show the
last screen picture if you've got to turn off and turn on (rebooting
without a programmer's switch).  Yes, RAM is quite able to maintain its
state for a while after turning it off, and I'm not sure that Macs *do*
clear it; my Mac Plus and 512Ke certainly don't.

Rowland.

-- 
Remove the animal for my email address:   rebecca@astrid.dog.u-net.com
Sorry - the spam got to me.          
PGP pub key A680B89D; fingerprint: F5605EAE F56F2C0B 818F3319 5227767F
                                   What?  Don't ask me - I don't know.

In article <B0C34504966824854@cara.demon.co.uk>,
peter@cara.demon.co.uk (Peter Ceresole) wrote:

> Each time these type 11 errors have shown up, they have tended to persist
> when I reboot. Type 11. Reset (Ctrl-Cmd-start) and some problem or other
> crops up again-

Type 11 errors are usually down to (pause for to create tension ...)

                            loose RAM chips !

Yes, folks.  If you have untrackable type 11 errors (and I mean
untrackable ... you should have tried INITs, new systems, etc.)
you may well have loose RAM.

Things to try: turn your system off, pick it up, shake it.  Any
change ?  Try giving it a sharp thump it while it's on (preferably
with the hard disk in powered down).  Do you get an instant Type 11 ?

If you're brave, open it up, take out the memory clean the contacts
with a rubber, and put the chips back in all swapped around.  Any
change ?  If you're not brave, pay me and I'll do it for you (and
help to diagnose the problem if it's not due to loose RAM).

Simon.
-- 
Simon Slavin -- Computer Contractor. | Excessive quoting from Shakespeare:
   http://www.hearsay.demon.co.uk    |                  gag me with a folio.
Check email address for spam-guard.  | -- sclerc@ernie.bgsu.edu (Susan Clerc)
Junk email not welcome at this site. | I receive about 2.5 junk emails a day.

Simon Slavin <slavins@hearsay.demon.co.uk.NOJUNK> wrote:

> In article <B0C34504966824854@cara.demon.co.uk>,
> peter@cara.demon.co.uk (Peter Ceresole) wrote:
> 
> > Each time these type 11 errors have shown up, they have tended to persist
> > when I reboot. Type 11. Reset (Ctrl-Cmd-start) and some problem or other
> > crops up again-
> 
> Type 11 errors are usually down to (pause for to create tension ...)
> 
>                             loose RAM chips !
[snip]

So *that's* what `miscellaneous hardware exception error' means...

> If you're brave, open it up, take out the memory clean the contacts
> with a rubber, and put the chips back in all swapped around.

Someone like me would suggest using a squirt of pukka switch cleaner
(you can get it in aerosol cans from Maplin and possibly Tandy) on the
RAM SIMM contacts, (rubbed with a lint-free wipe for preference, cotton
bud if not), and something similar for the SIMM socket.  Make sure you
leave a thin film of lubricant when you've finished - I tend to squirt,
rub, and squirt again.  Just a squirt of switch cleaner without any
rubbing is worthwhile - it dissolves crud, which is then shifted out the
way when you re-fit the SIMM.  I use this stuff when I'm fitting *new*
ICs, SIMMS, and the like.  

The advantages of switch cleaner over using a rubber are many: it
doesn't remove any of the *very* thin gold plating you often get on SIMM
contacts; it doesn't leave any traces of rubber on any of the contacts;
it cleans the socket as well as the SIMM; and best of all, it leaves a
thin layer of oil over the contacts.  This film moves out of the way of
the electrical contacts, and forms a nice barrier against air-born
contamination - it keeps the air away from the metal, so it's difficult
for any gases that might want to oxidise or tarnish anything to actually
get in contact with anything that might want to be oxidised or
tarnished.  All in all, very good stuff.  It even works wonders on
variable resistors - I've turned an old audio amplifier, a 150MHz
dual-beam Textronix oscilloscope, and an old Roberts radio from
`unusable' to `lovely' just by squirting loads of the stuff into the
cases of the various variable resistors and working them back and forth.
All the excess noise you got moving the knobs vanished and they're both
great now.  (A fiver for the radio; the amplifier and 'scope were free.
Replaced the amp's 741s with TL071s and it's really rather excellent - a
Texan II for those who are into that sort of thing).

It can be worthwhile shoving bits of loo roll in corners that you think
might get hit with excess switch cleaner before you squirt away -
leaving patches of any oil over the PCB and circuitry of an oscillscope,
for example, is rarely a good idea.  Try and keep it on and around the
contacts.

[snip]

Rowland.

-- 
Remove the animal for my email address:   rebecca@astrid.dog.u-net.com
Sorry - the spam got to me.          
PGP pub key A680B89D; fingerprint: F5605EAE F56F2C0B 818F3319 5227767F
                                   What?  Don't ask me - I don't know.

Simon Slavin <slavins@hearsay.demon.co.uk.NOJUNK> wrote:
[]
> Type 11 errors are usually down to (pause for to create tension ...)
> 
>                             loose RAM chips !
[]
> Things to try: turn your system off, pick it up, shake it.  Any
> change ?  Try giving it a sharp thump it while it's on (preferably
                            ^^^^^ ^^^^^
> with the hard disk in powered down).  Do you get an instant Type 11 ?

AKA "Engineer-induced physical actuation" :-)

rgds, Alan

ric <btgsch@rmplc.co.uk> wrote:

> Peter Ceresole <peter@cara.demon.co.uk> wrote:
> > I still use sys 7.5.2 on a 750/100 and in general it's beautifully stable.
[]
> Is there a particular reason you've stayed with 7.5.2 so long? I found
> just upgrading through to 7.5.5 made my 7500 hugely more pleasant
> to use, and didn't cost anything either.
> > 
> > I get maybe one type 11 error a week. Often less.
[]
> > Each time these type 11 errors have shown up, they have tended to persist
> > when I reboot. Type 11.
[]
> I've noticed the very same thing in the past on my 7200/90. It would
> seem to get in a temper every so often and I might spend a day
> performing restarts every 20 minutes or so. I don't see how memory
> can remain unreset across a restart, at least memory that matters, as
> that will have fresh, correct information written to as the system
> software reloads. All I could think was that I was using some
> particularly crash prone combination of apps which was causing the
> problems

ISTR, one of the "fixes" in 7.5.5 & 7.6 was that they're would
be fewer Type 11 errors on PPC boxes. The cynical have suggested
that's just because they are correctly identified as other specific
errors and not just reported as miscellaneous hardware exception 
type 11 errors...

rgds, Alan 

In article <19971229182345168015@newtechd.demon.co.uk>,
alan.frame@acm.org (Alan W. Frame) wrote:

> ISTR, one of the "fixes" in 7.5.5 & 7.6 was that they're would
> be fewer Type 11 errors on PPC boxes. The cynical have suggested
> that's just because they are correctly identified as other specific
> errors and not just reported as miscellaneous hardware exception 
> type 11 errors...

No, that's actually what Apple described it as doing.  If the 680x0
emulator crashed in the earlier systems, it was always reported as
a Type 11 error.  The fix meant that the PPC would analyse how the
emulator had crashed and report the appropriate error message.  I
don't think there was any suggestion that there'd be fewer crashes.

... except that, given the correct error message, the user would be
more likely to be able to work out why the crash had happened and
figure out how to prevent it from happening again.

Simon.
-- 
Simon Slavin -- Computer Contractor. | Excessive quoting from Shakespeare:
   http://www.hearsay.demon.co.uk    |                  gag me with a folio.
Check email address for spam-guard.  | -- sclerc@ernie.bgsu.edu (Susan Clerc)
Junk email not welcome at this site. | I receive about 2.5 junk emails a day.

Simon Slavin <slavins@hearsay.demon.co.uk.NOJUNK> wrote:

> In article <19971229182345168015@newtechd.demon.co.uk>,
> alan.frame@acm.org (Alan W. Frame) wrote:
> 
> > ISTR, one of the "fixes" in 7.5.5 & 7.6 was that they're would
> > be fewer Type 11 errors on PPC boxes. The cynical have suggested
> > that's just because they are correctly identified as other specific
> > errors and not just reported as miscellaneous hardware exception 
> > type 11 errors...
> 
> No, that's actually what Apple described it as doing.  If the 680x0
> emulator crashed in the earlier systems, it was always reported as
> a Type 11 error.  The fix meant that the PPC would analyse how the
> emulator had crashed and report the appropriate error message.  I
> don't think there was any suggestion that there'd be fewer crashes.
> 
> ... except that, given the correct error message, the user would be
> more likely to be able to work out why the crash had happened and
> figure out how to prevent it from happening again.

Also, a type 11 always forced a restart, while a type 1 or 2 frex gives
you at least the chance to save your work in other open apps etc. If
you're adequtely brave, you can often carry on OK.
-- 
'ric

Simon Slavin <slavins@hearsay.demon.co.uk.NOJUNK> wrote:
> alan.frame@acm.org (Alan W. Frame) wrote:
> 
> > ISTR, one of the "fixes" in 7.5.5 & 7.6 was that they're would
> > be fewer Type 11 errors on PPC boxes. The cynical have suggested
> > that's just because they are correctly identified as other specific
> > errors and not just reported as miscellaneous hardware exception 
> > type 11 errors...
> 
> No, that's actually what Apple described it as doing.

Ah! - I wasn't sure if that was the official story or not.
- I was NDA'd at the time.

> If the 680x0
> emulator crashed in the earlier systems, it was always reported as
> a Type 11 error.  The fix meant that the PPC would analyse how the
> emulator had crashed and report the appropriate error message.  I
> don't think there was any suggestion that there'd be fewer crashes.

The observation was "fewer Type 11 errors", and that much
is true...:-/

> ... except that, given the correct error message, the user would be
> more likely to be able to work out why the crash had happened and
> figure out how to prevent it from happening again.

But you must have heard the comparison between Mac error codes
(as far as "users" are concerned) and going to the doctor?

If you got to the doctor with an ailment, then he tells you
what it's called in Latin and sends you off none the wiser as to 
what caused it... :-)

I don't know how much use error codes are to punters[1] - they
can't look up IM and at least determine which manager/package
is to blame...

rgds, Alan
[1] Unless we can get it to say "Get rid of all that crap
you installed from some dodgy cover CD, and get on with
your work"