Complete agreement, Steve.
I'd like to add that there wouldn't be a real, important change in OS design until someone goes full circle and implements what IBM has had on mainframes for decades. I'm talking about batch processing; a customer controlled database of installed OS and application component information; performance controls and a monitor; a real job-entry-subsystem; near 100% RAS (reliability/availability/serviceability), and thorough documentation on everything that happens inside the OS. And I'm sure there's a few other things I'm not mindful of at the moment.
Windows architecture made desktop applications real and ubitquous , no question about that, but beyond the desktop is another story altogether.
Bill
> > So the question is: do you start working with an OS with a future
> > now, or do you milk your current OS for a few more years, knowing full
> > well that it will be obsolete soon.
>
> I have a personal observation on this, if I may. I visit and spend
> time with several Fox development shops every year, year after year.
> I see a definite pattern.
>
> Without exception, all the shops I've seen that are runaway successes,
> measured either by the number of their customers, the number of their
> users, or the wealth of their owners, are milking it. To me the
> correlation is clearly evident: the older and more fundamental the
> system, platform, or development philosophy, the more profitable the
> shop tends to be.
>
> In other words, virtually nobody among the multi-millionaires to near
> millionaires I know is remotely concerned with LongHorn. They are
> concerned about their customers. None of them, to the extent that I
> know them, ever peddled technology. They leveraged their insight into
> applications and, in the process, generated superlative value for
> their customers, their people and, of course, themselves.
>
> Conversely most (if not all) the shops I've visited in the past decade
> that are either struggling, or not able to pull it over the top, as it
> were, are invariably the rainbow chasers. The common thread here is
> disappointing results given the pots of money they invest, the time
> they forsake, and the risk they voluntarily assume to be slogging on
> the bubble. Put another way, if anybody knows a truly successful
> rainbow chasing development or consulting shop, one that consistently
> delivers superlative value to its customers and its owners, I'd like
> to see it.
>
> And I think Bill Anderson is right: the impetus to upgrade windows
> gets smaller with each new version, and moreover the installed base of
> "Windows.Prior" gets ever larger. When the Economist Magazine looked at
> Windows XP on August 9th 2001, the article headline was "Another year,
> another Windows", which pretty much sums up the situation.
>
> In the quote I pasted at the top of this post strikes me: "...knowing
> full well that it will be obsolete soon". What complete
> disillusionment! If LongHorn gets two digit percentage penetration of
> all the boxes used in business, I'll be very, very surprised. See,
> two years after that, another new OS will ship, and it will take its
> increasingly smaller slice of the installed-base pie, and so on. The
> reality in the field is this: We've long since passed the point where
> an OS upgrade offered anything really notable, nevermind sustainably
> valuable.
>
> The same can be said of the boxes those OS' run on: We've long since
> passed the point where upgrading a server or a workstation or a laptop
> offered anything really notable, nevermind sustainably valuable.
>
> Spot the pattern here?
>
> If the past two years have taught us anything, it's this: Business
> isn't on the technology-for-technology's-sake boat ride anymore.
> Those days are over. It's all about delivering real value and real
> benefits. This is, of course, as it should be.
>
>
> **--** Steve (Still on Win2K, as are all the customers I've worked
> with so far this year and all of last year, and I know I won't be
> upgrading anything OS-wise anytime soon. There's no need to. No
> sliver of a need. Zero. XP? 2003? Longhorn? Whatever. Wonderbread
> is Wonderbread.)
>
>
--- StripMime Report -- processed MIME parts ---
multipart/alternative
text/plain (text body -- kept)
text/html
---
©2004 Bill Arnold |
|
