Skip site navigation (1) Skip section navigation (2)

Re: Joel's Performance Issues WAS : Opteron vs Xeon

From: Christopher Browne <cbbrowne(at)acm(dot)org>
To: pgsql-performance(at)postgresql(dot)org
Subject: Re: Joel's Performance Issues WAS : Opteron vs Xeon
Date: 2005-04-25 22:03:52
Message-ID: m3acnmpl87.fsf@knuth.cbbrowne.com (view raw or flat)
Thread:
Lists: pgsql-performance
Martha Stewart called it a Good Thing when merlin(dot)moncure(at)rcsonline(dot)com ("Merlin Moncure") wrote:
>> In practice, we have watched Windows evolve in such a fashion with
>> respect to multiuser support, and, in effect, it has never really
>> gotten it.  Microsoft started by hacking something on top of MS-DOS,
>> and by the time enough applications had enough dependancies on the way
>> that worked, it has essentially become impossible for them to migrate
>> properly to a multiuser model since applications are normally designed
>> with the myopic "this is MY computer!" model of the world.
>
> Completely false.  NT was a complete rewrite (1993ish) and was
> inherently multi-user with even the GDI running as a user level
> process (no longer however).  The NT kernel was scalable and
> portable, running on the Alpha, MIPS, etc.

Completely irrelevant.  When Win32 was deployed, the notion that more
than a tiny fraction of the users would be running Win32 apps on
multiuser platforms was absolutely laughable.  It continued to be
laughable until well into this century, when Microsoft ceased to sell
systems based on MS-DOS.

> However, you do have a point with applications...many win32 developers
> have a very bad habit about expecting their apps to install and run as
> root.  However, this is generally not a problem with Microsoft stuff.
> In short, the problem is really people, not the technology.

Reality is that it is all about the applications.

Microsoft spent _years_ pushing people from MS-DOS to Windows 3.1 to
WfW to Windows 95, and had to do a lot of hard pushing.

The result of that was that a lot of vendors built Win32 applications
for Windows 95.

None of those systems supported multiple users, so the usage and
experience with Win32 pointed everyone to the construction of single
user applications.

At that point, whether Windows NT did or didn't support multiple users
became irrelevant.  Usage patterns had to be oriented towards single
user operation because that's all Win32 could be used to support for
the vast majority that _weren't_ running Windows NT.
-- 
let name="cbbrowne" and tld="gmail.com" in String.concat "@" [name;tld];;
http://linuxfinances.info/info/x.html
But  what can  you  do with  it?   -- ubiquitous  cry from  Linux-user
partner.  -- Andy Pearce, <ajp(at)hpopd(dot)pwd(dot)hp(dot)com>

In response to

pgsql-performance by date

Next:From: Josh BerkusDate: 2005-04-25 22:11:41
Subject: Re: half the query time in an unnecessary(?) sort?
Previous:From: Tom LaneDate: 2005-04-25 21:10:55
Subject: Re: [HACKERS] Bad n_distinct estimation; hacks suggested?

Privacy Policy | About PostgreSQL
Copyright © 1996-2014 The PostgreSQL Global Development Group