>>I always slightly cringed when phrase "vfp tables" was used instead of "dbfs". Because the SMB2 shit (IMO a completely intentional move by M$ to squeeze out any ISAM database which may rely on file sharing - bTrieve and many others - in order to sell SQL servers), they became unreliable, ...
>>
>I am definately not saying MS is smelling like roses, but I watched SMB with a sharp eye coming from NT4 to W2K.
>Under NT4 network speeds of NetBEUI in small networks was light years ahead of TCP/IP. Large networks were flooded, but on small networks easiest speedup for LAN based ISAM was NetBEUI.

In my experience, small network was 3 boxes max. Once we added the fourth, the shoutouts from each box have flooded the cables to the point of being completely unusable. So we had to create a boot floppy for dos7 (aka windows 98 without the windows part) with TCP driver, to make the network usable. We then just installed that on each machine... dos7 was the solution then. Don't remember if we had more than six machines on a network like that. Perhaps a couple of cases, but we installed Lantastic there. But this is period of 1998-99.

But it all worked perfectly, the speed was OK, the record locking functioned flawlessly as long as the network connection was stable. And it was, except in some weird environment, where some medical equipment (a Röntgen or some others) would cause interference, or the network cables were laid close to some big electric motor's power cables, or someone would be welding nearby.

>Changed with W2K, as SMB came to TCP/IP. But then surfaced more and more questions on app speed ttally disgusting with many people logged in... SMB always band aid plastered on the sore of network being 10-20 times slower than local disc.

I don't even remember what was the setup in those days, whether we used a loader or not, and I don't quite remember how many workstations were there in places where I remoted to users' networks - which wasn't much, I didn't do support. In-house network where we coded was top notch for the time. The slowness was usually caused by something stupid the programmers did, and with some clever trick I could sometimes deliver a factor of 20.

>>>Whenever the problems inherent to that approach surfaced, vfp and not the bad practice was blamed...
>>...but instead of blaming M$ and the SMB2 cockup, it was always set up to blame VFP.
>
>So yes there were "customer wishes" for better SMB (mostly from those unaware of technological reason).
>Making a total cesspool of SMB2: yupp, there may have been some intent against ISAM, but at that time the horse was dying already (Vista times)

Now this is a bit hard to follow, because the rest of this conversation is in the other half of our forum brain, aka UT, but what I got from it is that it suffered from attacks. And probably bad design in the first place, because I can't imagine M$ throwing its best brains on the problem of... writing it with the purpose of being compatible with Samba - not only that it had a huge Not Invented Here written all over it, but it also originated in the *nix world, perhaps even Linux, which in those times M$ wouldn't touch with welder's gloves. Which, I guess, may be the source of the problem.

Why couldn't they just reproduce the locking schema which worked perfectly in dos7?

>Time frame similar here: bought C64 as my finger skills always were better to pokink instead of gentle typing...

I come from the dark side: bought ZX... because that's what existed in 1983; c64 came later. And somehow the ZX looked sexier, with its black-white-rainbow box, compared to all drabs and grays of c64.

> And then wrote programs in Basic and Pascal as statistical analysis via calculator was... error prone, esp. with my typing temprament.

Of all the languages I tried on the ZX, I loved the best to play with Forth. That was an incredible experience.

>When I realized that most students could not fathom what just seemed logical for me, I increased my funds writing homework for programmers and similar students, while trying out grown up statistical software on university machine.

My then colleagues generally didn't have the money to offer, but a bottle of whiskey would find its way, so my roommate and I did have some fun with Fortran.

>>And I'm still using it for my pet projects :).
>
>lucky you to have time for such...

Isn't that the purpose of retirement, to finally have time for all those things?