Re: How proprietary is the "COBOL file system"

On Wed, 3 Oct 2007 17:31:25 +1300, "Pete Dashwood" <dashwood@xxxxxxxxxxxxxxxxxxxxxxxxx>
wrote:

Did you have any comment on the video? I'd be interested in your opinion.

I was disappointed. I expected better from a Microsoft visionary.

I wasn't expecting a theoretician who produces beautiful design with crappy
implementation. Universities have a surplus of that type. But I wasn't expecting a code
monkey either, the type who tries to overcome weak design with skillful implementation.
That's Anders. His test for design quality is totally pragmatic, throw it against the code
wall and see what sticks.

I was also put off by his solipsism. If he didn't think of an idea, it's trivial. There he
was 'inventing' parallel processing, as though no one had thought of it before. He was
'inventing' query processing. He doesn't acknowledge that databases have been doing both
for twenty years, and he dismisses their query processing as something anyone could write
in a short time.

Anders and the interviewer agree that Query Expressions, Functional Programming and Lamda
are ideas from older unsuccessful languages (such as Anders' own Prolog), but only
Microsoft has the skill to dust them off and make them useful. What hubris. If they want
to implement others' ideas, which is fine, they should stick to models that are fully
developed.


.

Relevant Pages

  • Re: How proprietary is the "COBOL file system"
    ... Ah, I was expecting nothing, never having heard of him before I saw the ... and he dismisses their query processing as something ... Anders and the interviewer agree that Query Expressions, ... Microsoft has the skill to dust them off and make them useful. ...
    (comp.lang.cobol)
  • Re: WriteFile()
    ... > The point is that while expecting a blocking IO call to WriteFile to ... expectations of the design. ... > WriteFile to return an error if not all of the bytes are written under all ... NON-BLOCKING is not a BLOCKING CALL ...
    (microsoft.public.win32.programmer.kernel)
  • Disappointment
    ... So I come back from work expecting to read Daniele's ... "Alas the fate of the one-button mouse in today's multibutton world. ... "Click on the left side to use Mighty Mouse in its simplest, ... seems to me to be a bad day for UI design. ...
    (uk.comp.sys.mac)
  • Re: Falcon first stage finished
    ... untrained peanut-gallery eyes. ... design. ... I've seen others heated to a yellow glow. ... OK - I guess I was just expecting it to be a bit more uniform. ...
    (sci.space.policy)
  • Re: Remote View to a table with more than 255 fields
    ... Good point Anders. ... it is due to bad design. ... Sietse ...
    (microsoft.public.fox.programmer.exchange)