>you still didn't give me the actual trap message you're seeing.

 

That's a bit difficult, given that I currently escaped to my Windoze partition.
But to make you happy I'll set myself up for yet another crash (again).
From memory I do know it should be the classical SOCKET.SYS trap as described in the official IBM bug report.

>This is *my* list, and bad behavior is simply not something to be tolerated.

 

And this is *my* inbox and BS gets BS in return.
I also inhereted a lack of tact, albeit from my Grandfather who was known for his short fuse when it came to BS.

>If you're driving a patched together VW Beetle which constantly stalled,

 

That >Beetle< would be a $5000+ system with +10k programs and +2 million files on it.
Where I come from you call that a compu HUMMER, not a Beetle.
And this thing is not >patched together<, I happen to do computer consulting for a living and the boxes I maintain are worth a lot more than your whole setup.
So I think I can claim that I know what I'm doing.

 

>Maybe after you ran something else...?

 

Analyzed that myself. But it strictly happens after heavy NW traffic.
Even when avoiding FireFox and using e/g/ only Opera, Mozilla or FTP does it (sometimes) happen.

 

>You did not mention anything about a VPN connection before.
>Could we just try to troubleshoot the problem with a bare TCP/IP config for a single adapter and a stripped down CONFIG.SYS?

 

That just increases NW traffic, the problem used to occur even before I used VPN (albeit less frequent).
I got a MAHOR problem with stripping down my Config.sys, as it is very big to begin with and I really couldn't say which parts I could "live w/o" with.
I repeat, I *need* that box for work, BIG PROJECT work, with VERY impatient clients - the kind of people who accept an "my box just crashed" excuse only ONCE.

 

>"People" as in...who? I didn't. Neil didn't. Andy didn't. Steve didn't.

 

The first two people who replied simply told me how silly I was to blame OS/2 for crashes that must have been caused by those FireFox XPIs.

>I try to approach my system as I approach my clients' systems, by remaining calm and going through things methodically,
>using a very simple setup at first.

 

I'm not sure about your systems, but when I get to a client's working configuration in production mode I can't just strip it all down and then use a "try & error" approach to put it back together again.
THOSE ARE HOT SYSTEMS, that are not allowed more than 1 hour downtime during any one month of the year.
And my box is, too. I simulate those client boxes on my system and when I get a call, I better be ready or else...

>Blaming the software for what may very well be the fault of something else - as the rest of us are not seeing
>what you are seeing - is simply unprofessional.

 

What *is* >unprofessional< is to release a TCP/IP stack that stinks to the public.
And I doubt its the fault of "something else<, as I already wrote a DOZEN TIMES that I checked both time stamps and sizes of all relevant TCP/IP files, that need to be patched.
According to IBM's own bug report there should no longer be a SOCKET.SYS error happening on my box PERIOD.

 

>You did not mention that you even had Windows installed on the system before, so again, without the whole story, how can you expect better feedback for your problem?

 

Heeelloo, I wrote seperate partition(s). So what does one have to do with the other?!!

 

>Again, I would suggest a spare drive with a clean install to use as a baseline comparison.

 

With 60GB of program and 160GB of data files I'd be out of business after the 3rd system move with your approach.
Having week long downtimes every time I switch over is not acceptable to me or my clients.
The Windoze issues with HAL is one of the reason why I don't like to use Windoze as my main OpSys.
Because I have to re-install all my Windows programs EVERY TIMY I MOVE ONTO A NEW LAPTOP!!

Not to mention the +10k WPS objects I got.

 

> "Online encryption?" You mean like Cryptstream? Or some other 3DES (or similar) VPN encryption which rides over TCP?

 

I mean like all of the above together and then some that not even you know about.
I work on highly sensitive data, both government and corporate, and utilize both commercial and military grade encryption technology.

 

>Okay, so again, I say that you need to troubleshoot on a clean, plain vanilla setup

.

OK, I see we won't come together on this one.
I NEED MY BOX FOR WORK! I have neither the time, leisure nor patience to set up a box just for playing around and then do a pesky week or month long "try & error"approach to figure out what's happening.
Do you have ANY idea how long it would take me to get to the bottom of this if I'd do that with my 10+ programs?
It'll be easier for me to just install VPC for Windoze or Linux (which I got) and start moving my OS/2 stuff over to a more stable platform.

 

>Fixing it may very well take more than a day, as well.

 

One more day like this and I might well be out of business!

 

>Told...by whom? And what - aside from time - is wrong with a fresh install of some apps?

 

Listen, I really(!) DISLIKE it when eople repeat their point over and over and simply REFUSE to accept my reply.
Not sure if you want to play games here, but right now I just won't repeat MY POINTS about re-installing a 320+GB system from scratch.
And I got that info directly from Serenity (I think it might have even been BOb himself).

 

>There comes a point, however, when a fresh install is good for the soul.

 

In my line of work you'd last for about 5 MINUTES.
I couldn't even recall when anyone ever told a client running a +500GB HOT production system >a fresh install is good for the soul< and lived to tell the tale.
The only time when I'll do a fresh install will be when I'm moving OFF OS/2!

 

>I really don't want to continue this rapidly downhill sliding OT discussion.

 

Me neither as I can see that you just refuse to accept my point and keep coming back to that notion of me just re-installing my whole system from scratch and then trying to find the problem via "try & error" of +10k software components.
And that while I got clients with hot production systems worth millions of $ on the phone.

I could just as well fire myself right here & now.

I'll e-mail Roderick and see what happens.