Report: Computer hijacking on the rise

TalkBack 37 of 144:: Next »; « Previous

Thread View
Flat View

MS simply does NOT bother: to design a system according to its intented and expected purpose. They invest in 'convencing' users that the need what they (make them) want.

MS moved into the silly and flimsy workgroup level client/servers at the end of '80's/beginning '90's. And tey did after the made IBM give them the OS/2 v2.0 code. OS/2 was designed and built from scratch by a company with very long record and know-how in multi-user, transaction oriented systems. Very pertinently, IBM opted to pretty much follow a watered-down kernel model from UNIX.

At that time, you recall MS broke its relationships with IBM. Apparently the OS/2 v2.0 was intellectually a heavy load for them to assimilate. Recall that MS was supposed to come up with OS/2 v3.0. Instead of that they shamelessly used all the code (they could comprehend) and they came up with win 3.1x.

When they understood the code a little better (actually they asked the VAX/VMS lead architect to understand it for them) they started working on Win NT 3.X which finally after 100s of rounds of massive bug fixes it became Win NT 4.0 ( mid 1996? ) which was the FIRST stable pc os by MS.

They never understood the multi-user, time-sharing concerns. And they did NOT bother since a TRUE multi-user Win NT would mean that MANY users could log on and work on ONE PC. (That would had made their pals PC manufacturers upset, since they were forced to install ms windows on ALL PCs anyway anmd they wanted something in return.)

For MS each user needs at least ONE PC (h/w) and a complete copy of the s/w environment (MS windows) and APPLICATION SUITE (ms office, etc.) In this way EACH user has to spent money for ALL three (prices vary by volume and/or aggreements)

However, this FAT PC cient never created the need for MS to make a decent multi-user system, where people are reasonable protected from each other and that there are well-defined admin accounts which manage the system.

Win XP is maybe their first shot at it; Longhorn their 2nd one.

Recall that even LONG before UNIX, in the early ro mid 60's the then mainframes had to come up with the sharing of the extremely expensive h/w among users. I recall CTSS, MULTICS (whch became the force for which UNIX was the counter-force), and many others from DEC and other mini-computer manufacturers. In the late 60's, IBM came out with the various VM and TSO which started offering 'time-sharing options' on top of mamouth batch or mixed systems.

Multi-user/Time-sharing technology is now 30+ good years old. There is no magic into it and it is well understood. MS did NOT bother: they could get away with it with their FAT pc client model.

MS is STILL offering a workgoup platform. They are still trying to figure out the multi-user model.

They are also getting punished for their own underhanded mistakes: the kernel-ification of apps has created a huge mess in terms of safeguarding the stability and security of misbehaving apps and users. Now they realize that they need a well defined user app which runs in the kernel. This cannot happen CLEANLY until the Longhorn which is a REWRITE. Also, they keep making implementation more complex with the addition of middle-ware like their .NET.

It is a mess which they created intentionally or they simply let grow with light mind. MS will have VERY hard time converting its platform to a truely multi-user and truely secure system. They will follow the same path as Intel, which realized that the x86 is a dead-end and developed a completely incompatible new ISA (IA64). They then realized that the switch-over cost is HUGE and reverted back to playing catch-up with AMD.

Fast growth on FLIMSY platforms means a tall and shaky structure which some time is bound to collapse under its OWN weight. I have no doubt that this will happen.

-m; Posted by: michael_t Posted on: 07/11/05 You are currently: a Guest | Members login | Terms of Use