RE: Windows platform stability (was: Stability horrible in CODE)

["Phil" <sublime78ska@xxxxxxxxx>]

Well that's impressive.

I don't think I've ever, ever, ever met a PC person who understands how to
do what you described.

Phil

> -----Original Message-----
> From: code400-l-admin@midrange.com
> [mailto:code400-l-admin@midrange.com]On Behalf Of Buck Calabro
> Sent: Tuesday, October 30, 2001 11:25 AM
> To: code400-l@midrange.com
> Subject: Windows platform stability (was: Stability horrible in CODE)
>
>
> >Has there been anyone who's made the migration
> >from 98SE to 2000 who can testify to the stability?
>
> I'll make one broad post on Microsoft OS stability in general and
> then shut
> up.
>
> I've been playing in the PC space since the MITS Altair.  I've built S-100
> machines and written my own kernel (the hard way.)  My experience
> with MS is
> therefore a bit atypical.  When I got an IBM PC running DOS 2.1(?) I was
> amazed at how often the thing went casters-up.  Then I started doing some
> research on the boot process they used.  After some determined tinkering
> with AUTOEXEC.BAT and CONFIG.SYS, I was able to find the magic driver load
> sequence that kept the machine going all day long.
>
> Some of that early tinkering and experimentation still applies
> today.  High
> resolution video and sound cards came somewhat late to the IBM
> PC-compatible
> platform, and have been historically a sore spot.  The drivers just don't
> work that well in combination with other software, and failures can be
> spectacularly unpredictable.  That's why loading the drivers in the right
> order can be helpful, even today.  When I get a new box, one of the first
> things I do is hit the registry and make sure that I know how the drivers
> are getting loaded.  Nowadays that's not an easy task but my work
> PC (Win98
> 2E) stays running for weeks at a time.  I had the same experience with Win
> 3.1 (386sx-33) and Win 95 (Pentium 75).  I have never seen an off
> the shelf
> install of any MS OS run that well.
>
> The other major contributor to overall stability seems to be "minimise
> memory leaks."  When I browse an IBM manual in PDF form, I generally
> download it and view it offline.  When done with that manual, I close the
> document window but not Acrobat.  I minimise Acrobat, expecting
> that I'll be
> using it again soon.  Most definitely, I don't open/close IE all day long.
> No matter what version, IE has memory leaks.  I open my Notes client once
> and leave it open all day long.  When I get in tomorrow, I'll look at my
> free resources - if it drops below 50% I'll re-boot as a prophylactic
> measure.  That's a PII-450 with 128 megs of RAM.
>
> At home, I was the only user on the machine for a long while, so I kept my
> ancient Win 3.1 installation going.  When the kids started
> pounding on it, I
> was pleasantly surprised to see that it ran just fine with a once in the
> morning reboot.  I kept is for so long that they couldn't play the games
> they wanted, so I upgraded to NT4.0 with a boot menu for true DOS.  THAT
> took some tinkering with, but once done, I had separate logins
> for all 4 of
> us, and the stability was quite good.  I had a Blue Screen once a week or
> so.  I recently went to Win 2K - after rooting out the IM clients and Real
> Player from automatically starting it's remarkably stable.  Maybe once a
> month it'll go Blue Screen, even with the kids and Mom banging away on it.
> That's on a P-200 with 128megs of RAM.
>
> I do NOT think that Code in and of itself is unstable - I had a devil of a
> time with Notes and Access until I got the right sound card
> drivers and then
> my life was Much Better.  Yes, sound card drivers.  Some goofy
> conflict with
> the video in combination with THIS app using THAT memory while THAT app
> tries to use THIS memory.  A little bit busy with a .pdf download, an
> interrupt services a little late and crunch.  The thing that
> surprises me is
> not that people see crashes, but that the thing actually runs at
> all.  Fire
> up a debugger sometime and look at all the processes that are supposed to
> honour gentleman's agreements and play nice together.  To this
> day, I do not
> believe that the PC is ready for off the shelf use.  I think
> we're still in
> the experimenter's age, and if you want to run complex apps, you should do
> some serious reading and tinkering to tune your particular
> machine for your
> particular mix.  That's why some people (like me) report Code being stable
> and wonderful and others (like Mike) complain that it's terrible.
>  It's the
> combination of memory leaks, drivers and application mix all
> taken together
> that contribute to the situation.
>
> All done.  Sorry for the bandwidth.
> _______________________________________________
> This is the CODE/400 Discussion & Support (CODE400-L) mailing list
> To post a message email: CODE400-L@midrange.com
> To subscribe, unsubscribe, or change list options,
> visit: http://lists.midrange.com/cgi-bin/listinfo/code400-l
> or email: CODE400-L-request@midrange.com
> Before posting, please take a moment to review the archives
> at http://archive.midrange.com/code400-l.
>


_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com