On mySimon: Say You're One of Them
BNET Business Network:
BNET
TechRepublic
ZDNet
TalkBack 15 of 23:
Next »
« Previous
Really?
Does it take a $$ 6 billion + budget to realize that arbitrary and convoluted intercommunication between objects is bound to cost dearly in communication times and other overheads?

One needs to keep the communications cost under control. This is a basic constraint and not something inherent in the distrivuted object models. It is the convoluted implementation of MS objects that seems to be prohibiting the control of the cost. Then they apparently realized that different versions of their 'os' has problems with the installed DLLs and they had to push the DLLs out to the object clients and this made the cost prohibitive.

Now they come out and say: 'forget all the $$$ and resources you wasted developing on our super-dooper dickcom architecture; we are giving you NEW stuff; better stuff; we will take care of everything for you; we will... we will....' etc.

So folks, scrap all your infrastructure to develop on the robust MS object model. Because it is not robust and it is hardly a model of a system.

This is what happens when people do NOT design scalable s/w, nor a s/w that is capable of running across the Internet. Who could expect that s/w which runs with acceptable times over 100Mb LANs can scale equally well when ported (dumped really) to run on the general Internet? There should had designed a light-weight version for low BW/high leatency networks. But this would require some time for the MS 'engineers' to use their brain and figure out the resource cost for their system. And it would also call upon some basic common sense and engineering.

But who has time for this kind of crap? Just use ay of the wizards to dump GUI code out and ship it to customers; we will figure it out as the bugs and other problem reports are comming in....

No surprise here. This s/w 'design/development' model has worked so nicely for MS. They have been raking in huge profits out of what amounts to 10 year olds coding techniques.

Ofcourse, the ultimate blame is with the people who knowingly chose to utlilze the MS 'platforms' to develop code professionally.

Even after longhorn comes out (in the remote future) I am sure that NEW design flaws and malformations will come out and then the NEXT CYCLE of scrapping the crap will start all over again. That's OK for MS since there are allways going to be some poor people falling for the Barbie code 'professional development paradigm'.


-m

-m
Posted by: michael-t   Posted on: 01/27/04 You are currently: a Guest | Members login | Terms of Use
Reply to Story Reply to Message

Alert moderator to an offensive message

Subscribe to this discussion via Email or RSS

Progress is a good thing.  No_Ax_to_Grind | 01/27/04
You're right..the 2.6 kernel will have linux dominating servers  SloooeShflu | 01/27/04
Well, many fools believe so anyhow.  No_Ax_to_Grind | 01/27/04
I see you are busy sucking from that bulls hind...  B.O.F.H. | 01/27/04
2.6 is nice  Suicida| | 01/28/04
My MCSD's are on it...  Mike Cox | 01/27/04
you are on top of it Mr. Cox!  dg mh | 01/27/04
It's forward thinking like this...  Cardinal_Bill | 01/27/04
your certification junkies are on it!  nograin | 01/27/04
What a bunch of idiots.  DonnieBoy | 01/27/04
Back to the future?  techboy_z | 01/27/04
Not Quite  JimSatterfieldW | 01/27/04
Either way, it's dead in the water..  SloooeShflu | 01/27/04
Which Markets?  ShadeTree | 01/27/04
Really?  michael-t | 01/27/04
This says it all..  Patrick Jones | 01/27/04
I don't think Rotibi got it right  Larry Maccherone | 01/28/04
Why the long artical?  nite_w0lf | 01/27/04
Fiction from Fantasy  ParadigmOdyssey | 01/27/04
Who Cares  Tom Jones_z | 01/27/04
Better Shine More Light on It  Tom Jones_z | 01/27/04
Seems you care.  No_Ax_to_Grind | 01/27/04
What a moron  skunkarific | 05/01/04

What do you think?

SponsoredWhite Papers, Webcasts, and Downloads

advertisement
advertisement

SmartPlanet

Click Here