Subject | RE: [ib-support] Re: New Firebird Release 1.0.0 Beta2 Available |
---|---|
Author | Lee Brown |
Post date | 2001-08-23T20:49:02Z |
First off, thanks for the link. I didn't know this effort was going on.
Good news.
I think your point regarding the hands-on availability is a good one and yes
it makes
it easier to fix things yourself -- provided you have the skills or have the
skills in-house.
Consider the case where I'm a small business, my IT dept knows how to
install OS's, but doesn't
have the expertese to modify them. If I'm running a windows platform, I can
easily find somebody
conversant with whatever version I have. Now admittedly the
componentization makes it more difficult,
but even that is usually pretty fixed too. With a Linux system that has
been modified, my outsourced
personel could have a much worse time of it trying to figure their way
through the customizations. OK,
it's a bit of a self-defeating argument and personally I think it's actually
pretty equal on both sides.
Yeah, sorry for the "Linux variants" I should really be referring to all of
them
as GNU-Linux (the kernel is different, but most of the other pieces are
typically
always GNU implementations.)
Returning to the old trend (yes I do remember those days!) I'm looking
forward to
the days when the OS can microcode the CPU so it doesn't matter who's
CPU you want to run on or what OS you want to run. But in the end it always
seems to come down to (usability * performance) which unfortunately tends
to drive the market to writing an OS for a CPU. The only example I can
think
of actually is Visual Age (software, but same principal) -- great idea, but
when
I first tried it back in 96 it was totally unusable unless you had the
latest hardware
because it was windowing on top of windowing; OK I could write one
application
and deploy it on many platforms (sound familiar), but that flexibility
(usability
in the above formula) was simply outweighed by the fact it was too slow
(performance).
Linux seems to moving from the back end toward the desktop from what I've
observed.
McDonalds in the US has just signed a whole bunch of linux machines for some
restraunts
to connect to a central site and Ford Europe has done something similar.
Dell has
decided not to put GNU-Linux on the desktop. But it is coming, that's for
sure.
Regards,
Lee Brown
-----Original Message-----
From: Paul Schmidt [mailto:paul@...]
Sent: Thursday, August 23, 2001 14:52
To: ib-support@yahoogroups.com
Subject: RE: [ib-support] Re: New Firebird Release 1.0.0 Beta2 Available
Lee:
Good news.
I think your point regarding the hands-on availability is a good one and yes
it makes
it easier to fix things yourself -- provided you have the skills or have the
skills in-house.
Consider the case where I'm a small business, my IT dept knows how to
install OS's, but doesn't
have the expertese to modify them. If I'm running a windows platform, I can
easily find somebody
conversant with whatever version I have. Now admittedly the
componentization makes it more difficult,
but even that is usually pretty fixed too. With a Linux system that has
been modified, my outsourced
personel could have a much worse time of it trying to figure their way
through the customizations. OK,
it's a bit of a self-defeating argument and personally I think it's actually
pretty equal on both sides.
Yeah, sorry for the "Linux variants" I should really be referring to all of
them
as GNU-Linux (the kernel is different, but most of the other pieces are
typically
always GNU implementations.)
Returning to the old trend (yes I do remember those days!) I'm looking
forward to
the days when the OS can microcode the CPU so it doesn't matter who's
CPU you want to run on or what OS you want to run. But in the end it always
seems to come down to (usability * performance) which unfortunately tends
to drive the market to writing an OS for a CPU. The only example I can
think
of actually is Visual Age (software, but same principal) -- great idea, but
when
I first tried it back in 96 it was totally unusable unless you had the
latest hardware
because it was windowing on top of windowing; OK I could write one
application
and deploy it on many platforms (sound familiar), but that flexibility
(usability
in the above formula) was simply outweighed by the fact it was too slow
(performance).
Linux seems to moving from the back end toward the desktop from what I've
observed.
McDonalds in the US has just signed a whole bunch of linux machines for some
restraunts
to connect to a central site and Ford Europe has done something similar.
Dell has
decided not to put GNU-Linux on the desktop. But it is coming, that's for
sure.
Regards,
Lee Brown
-----Original Message-----
From: Paul Schmidt [mailto:paul@...]
Sent: Thursday, August 23, 2001 14:52
To: ib-support@yahoogroups.com
Subject: RE: [ib-support] Re: New Firebird Release 1.0.0 Beta2 Available
Lee:
On 23 Aug 2001, at 13:06, Lee Brown wrote:
> This is exactly the reason why Linux is going to be slow taking off.
> Remember the article about Intel Itanium and how Linux got there
> first, etc. Well as this post demonstrates, most people aren't going
> to bother with the hassle required to get it there until the OS
> vendors make it happen. Of most particular interest to me was the
> comment about losing all the customizations by installing a new
> version of an OS. That seems like another trap that'll make people
> wary of using Linux for some things. "Yes I want a free OS, but I
> have to add stuff to the kernel for some things, what an upgrade
> headache." Packages are making it easier, but are still not
> consistent across different Linux implementations which I would image
> will make the vendors loath to write half a dozen install scripts that
> all have to be kept in sync not only with each other, but also with
> the different versions of the different linux variants. One
> application could easily have 10 install scripts. Yet again, what a
> headache.
I will admit that Linux is more like a 57 Chevy then a Japanese SUV,
in that it's much more hands on then Windows is, but that is also
it's biggest strength, when something fails to work as it should, the
whole system is open to examination. Unlike Windows where often the
only recourse is to reinstall the operating system.
I prefer to refer to "linux variants" as Linux based operating
systems, because realistically Redhat Linux and OpenLinux and Suse
Linux and the others are different operating systems with different
features, but running the same engine. Sort of like a Chevy, Pontiac
and Buick are different cars, in that the badges, are different, the
body has a slightly different shape, and the interior looks
different, but everything else is the same. The reason they are
different is different markets, the Pontiac is for the single person,
sporty and fun, the Chevy is more practical, common to see a car seat
or a couple of ankle-biters in the back seat. The Buick, well your
dad buys the Buick (actually that's moms doing, he would rather buy a
Porsche).
It is returning to an old trend, where the O/S engine is driven by
the CPU vendor, rather then by an O/S vendor. Itanium and AMD's
version are just the beginning. The computer vender will soon be
able to take Intel's Kernel and compiler, combine these with some
publicly available bits and build a custom O/S intended for that
particular model. I expect that companies like Redhat will move from
commercial O/S sales to doing this kind of customization service for
computer vendors. I expect that soon the Linux Standardization
Project will have at least a working document, see
http://paradigm.uor.edu/linux/standard/ for what they are doing so
far, so this may become less of an issue in the future.
Paul
Paul Schmidt,
Tricat Technologies
Email: paul@...
Website: www.tricattechnologies.com
Yahoo! Groups Sponsor
To unsubscribe from this group, send an email to:
ib-support-unsubscribe@egroups.com
Your use of Yahoo! Groups is subject to the Yahoo! Terms of Service.