Re: win95 perscription: 50mg of reboot daily (was Re: DESgui 1.2?)

Colin L. Hildinger (
Wed, 21 May 97 13:02:00 -0500

On Wed, 21 May 1997 13:47:07 -0400, Rick Hornsby wrote:

>Sorry if this is a bit off topic, but its the only place I've been able to
>find a group of people who actually know what they're talking about when it
>comes to operating systems and such ...
>>At 23.19 05.20.97 -0500, Colin L. Hildinger wrote:
>>>On Tue, 20 May 1997 23:05:56 -0400, Rick Hornsby wrote:
>>Gee, I'd been up for 1.5 weeks until my modem locked up today (I
>>picked up the phone while it was dialing and it never let go of the
>>line). Just giving you a hard time. I have a simple solution that I
>>used to use when I ran Windows. Reboot every day, whether you think
>>you need it or not, it'll save you that lockup/reboot at the inopertune
>>time (like in the middle of a time save of your term paper -
>>trust me that sucks, because it corrupts your file. <MS Word has cause
>>a GPF in module SUX2BU.EXE>).
>hahah, sux2bu.exe :P
>Rebooting is a simple solution, but that really doesn't answer my question.
> Why on earth should I have to reboot every day? That seems like a waste.
>What is going wrong that I should have to do that? I've lost a point in
>each resource category just since last night.
>To answer David's question about the kernel update, yes I've had that
>installed for quite some time now, as well as a slew of other updates that
>make my OS look like a raggedy patchwork quilt.
>of 13 updates/patches ones that might be relevant include:
>kernel (, shell common dll, common dlg library, OLE 16/32
>Interoperablity library,network virtual file server, network redirector,
>NWSERVER virtual device.
>*sigh* I realize that in terms of the core OS, UNIX is far far superior.
>But what might be wrong with windows (my computer?) that would force me to
>reboot it so often, when a linux box could be left running and only
>rebooted for a few things, like after recompiling the kernel? If nothing
>else, is there a way to slow down the system degredation? Is NT any better
>at releasing terminated/killed processes' resources back to the machine?
>p.s. recognizing that UNIX is a better core OS, why don't I run it? 'cause
>I don't have the time to learn every little thing to make it work right,
>and I've not found anyone here that will teach me, or that I could call and
>say "help! whats wrong with this damn thing?!" If any OSU folks want to
>volunteer to be a mentor, it'd be appreciated ... perhaps even with
>occasional pizza ;)
Well, you'll notice from my .sig that I'm not a UNIX guru. I have
learned quite a bit about it from the DESCHALL list. Here's my order
of stability list (rated 1-10):

UNIX (in general) 9 (reboot very seldom)
OS/2 8 (reboot every week or two)
NT 6 (reboot every few days)
Win95 4 (reboot daily)
Win3.1 2 (reboot 3-4 times a day)

Note, that some software and poorly written hardware drivers can reduce
any of the systems to unstability. Consistent, well written programs
on well supported hardware can take OS/2 and NT a long way. I know of
NT servers that average 30 days uptime and I know of OS/2 servers that
average 6 months+ uptime. The numbers I gave above are based on my
experience for what I consider normal-heavy use with a wide variety of
software. For instance, I had a buggy OS/2 program with a memory leak.
After using it 10-20 times, my swap file would have grown to 30+ megs,
so I would shut down and reboot. I didn't have to, but as the swap
file grew I noticed a little bit of a performance hit. The latest rev
of this program fixed the memory leak and my machine is pretty darn
solid now.

OTOH - I've not seen any Windows 3.1 systems that could go a whole day
with out a Ga-PooF under a reasonable load, even when only using MS

Colin L. Hildinger

Games Editor - OS/2 e-Zine!

The Ultimate OS/2 Gaming Page

The Official Unofficial AWE32 and OS/2 Warp Page