Tablespaces

Compile error?

gmake[2]: Entering directory
`/space/1/home/chriskl/pgsql-server/src/timezone'
gcc -O2 -fno-strict-aliasing -g -Wall -Wmissing-prototypes
-Wmissing-declarations -I../../src/include -c -o pgtz.o pgtz.c -MMD
In file included from ../../src/include/storage/bufmgr.h:20,
from ../../src/include/storage/bufpage.h:18,
from ../../src/include/access/htup.h:17,
from ../../src/include/tcop/dest.h:64,
from ../../src/include/utils/guc.h:17,
from pgtz.c:26:
../../src/include/storage/relfilenode.h:29: syntax error before `.'
../../src/include/storage/relfilenode.h:29: syntax error before `.'

Chris

Gavin Sherry wrote:

Gavin Sherry wrote:

Attached is my latest patch implementing tablespaces. This has all the
functionality I was planning for 7.5.

A few minor points I happened to notice while reading through the patch:

+  * To simply initialisation and XLog activity, have create and 
maintain
+  * a symbolic link map in data/pg_tablespaces.

Grammar errors.

+ void
+ TblspcCreateDbspace(Oid tbloid)
+ {
+ #ifndef HAVE_SYMLINK
+ 	return;
+ #endif
+ 	struct stat st;
+ 	char		*dir;

If HAVE_SYMLINK is undefined, this is a syntax error (at least in
C89, which is what we ought to limit ourselves to). Similar problems
elsewhere in the same file (tablespc.c)

+ 	dir = (char *) palloc(strlen(DataDir) + 14 + 10 + 10 + 10 + 3 + 1);
+ 	sprintf(dir, "%s/pg_tablespaces/%u/%u", DataDir, tbloid,
+ 				MyDatabaseId);

Is the length of that buffer right? At the least the addition is a
little weird (why are you adding 10 three times for two numeric
variables?) I noticed another buffer allocation (linkloc) that
looked dubious at first glance as well.

+ char realnewpath[MAXPGPATH];

This is somewhat pedantic, but how do we know that MAXPGPATH >=
PATH_MAX (the minimum safe size of the second argument to
realpath(), at least on my local system)?

-Neil

Attached is an updated patch, fixing a compile error which my compiler
didn't seem to detect/suffer from and incorporating fixes to problems
raised by Neil.

Thanks,

Gavin

Just reminding someone to review this some time...it does all seem to
work very well :)

Chris

Gavin Sherry wrote:

Yes, I was wondering if it was ready for application. Tom, you want to
eyeball it?

---------------------------------------------------------------------------

Christopher Kings-Lynne wrote:

Just reminding someone to review this some time...it does all seem to
work very well :)

Chris

Gavin Sherry wrote:

Attached is an updated patch, fixing a compile error which my compiler
didn't seem to detect/suffer from and incorporating fixes to problems
raised by Neil.

Thanks,

Gavin

------------------------------------------------------------------------

---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match

---------------------------(end of broadcast)---------------------------
TIP 1: subscribe and unsubscribe commands go to majordomo@postgresql.org

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

Bruce Momjian <pgman@candle.pha.pa.us> writes:

Yes, I was wondering if it was ready for application. Tom, you want to
eyeball it?

I do. I'd like to get the composite-type-column stuff out of the way
first, and then I'll buckle down to reviewing patches (this one and the
other major patches in the queue...) Should be able to start on that
within a couple days.

regards, tom lane

Your patch has been added to the PostgreSQL unapplied patches list at:

http://momjian.postgresql.org/cgi-bin/pgpatches

I will try to apply it after review..

---------------------------------------------------------------------------

Gavin Sherry wrote:

Attached is an updated patch, fixing a compile error which my compiler
didn't seem to detect/suffer from and incorporating fixes to problems
raised by Neil.

Thanks,

Gavin

Content-Description:

[ Attachment, skipping... ]

---------------------------(end of broadcast)---------------------------
TIP 9: the planner will ignore your desire to choose an index scan if your
joining column's datatypes do not match

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

-----Original Message-----
From: pgsql-hackers-owner@postgresql.org
[mailto:pgsql-hackers-owner@postgresql.org] On Behalf Of
Zeugswetter Andreas SB SD
Sent: Friday, March 05, 2004 1:20 AM
To: jearl@bullysports.com; tswan@idigx.com
Cc: Bruce Momjian; Tom Lane; Greg Stark;
pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: Re: [pgsql-hackers-win32] [HACKERS] Tablespaces

First of all, symlinks are a pretty popular "feature."

Even Windows

supports what would be needed. Second of all, PostgreSQL

will still

run on OSes without symlinks, tablespaces won't be available, but
PostgreSQL will still run. Since we are all using

PostgreSQL without

My idea for platforms that don't support symlinks would be to
simply create a tblspaceoid directory inplace instead of the
symlink (maybe throw a warning). My feeling is, that using
the same syntax on such platforms is important,
but actual distribution is not (since they will most likely
be small systems).

I know of bot SQL*Server and Oracle database systems on Win32 with
hundreds of millions of rows and many hundreds of gigabytes of space.
These are production systems, run by fortune 500 companies.

I expect that PostgreSQL systems on Win32 will have multiple 64-bit CPU
systems, with 16 gigs or so of ram, and a terabyte of disk, not long
after 7.5 is released (unless problems with PostgreSQL on that platform
turn up).

Is that what you have in mind when you say "small systems"?

I expect that one year after release, there will be ten times as many
PostgreSQL systems on Win32 as all combined versions now on UNIX flavors
(of course, that is a SWAG, but I think a sound one)

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten times as many
PostgreSQL systems on Win32 as all combined versions now on UNIX flavors

I surely hope not. Especially not multi-gig databases. The folks
running those should know better than to use Windows, and if they
do not, I'll be happy to tell them so.

regards, tom lane

Tom Lane wrote:

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten times as many
PostgreSQL systems on Win32 as all combined versions now on UNIX flavors

I surely hope not. Especially not multi-gig databases. The folks
running those should know better than to use Windows, and if they
do not, I'll be happy to tell them so.

Admins often don't have a choice, but a company strategy to use win
only. Deciding on the platform before examining the app's requirements
is always a bad idea, but that's what happens. Respecting this,
suggesting "don't use win32 for high performance pgsql databasing" is
equivalent to "don't use pgsql".

Regards,
Andreas

With the rule system and two underlying tables one could make it work by
hand I think.

The rule system could be used to do this, but there was some discussion of
using inherited tables to handle it. However neither handles the really hard
part of detecting queries that use only a part of the table and taking that
into account in generating the plan.

I think the consensus should be to add smarts to the planner to include
static constraint information to reduce table access.

e.g if you have a constraint "acol integer, check acol < 5"
and you have a query with a "where acol = 10" you could reduce that
to "where false". This would help in all sorts of situations not only
partitioned/inherited tables. I am not sure what the runtime cost of
such an inclusion would be, so maybe it needs smarts to only try in certain
cases ?

Andreas

"Zeugswetter Andreas SB SD" <ZeugswetterA@spardat.at> writes:

e.g if you have a constraint "acol integer, check acol < 5"
and you have a query with a "where acol = 10" you could reduce that
to "where false".

I think part of the question is how much work do you put into checking this.
Checking constant known values like above is probably not too expensive.
Checking for ranges like "where acol between 5 and 10" is probably doable. And
that might be enough for partitioned tables. I think that's about all Oracle
bothers to check, for example. More complex where clauses and check
expressions might be hard to prove are true or false.

But then the work's still not done, you still have to add an optimization that
prunes members of a UNION ALL (or equivalent if it's done using inherited
tables or some other infrastructure) if they are known to provably produce
zero rows.

And then there are more subtle cases. Like if the query is "where acol = ?".
Then you know it only has to read one partition, but you don't know which one
at compile time. And it's important to handle that case because that might be
the only clause. So knowing that you only need one partition might be the
difference between a sequential scan of one partition, or an index scan of
many thousands of records because they're only a small percentage of the
entire table.

--
greg

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten times as many
PostgreSQL systems on Win32 as all combined versions now on UNIX flavors

I surely hope not. Especially not multi-gig databases. The folks
running those should know better than to use Windows, and if they
do not, I'll be happy to tell them so.

This is a prejudice that we should try to avoid. Yes, Windows is lacking
on so many levels, but that really isn't the point.

A good box running Win2K or XP Server, with no internet connectivity, and
no user applications, can really perform and be reliable. Would I choose
this? Hell no, but there are HUGE amount of people who either don't know
any better or have no real choice.

The REAL bonus here is getting PostgreSQL in their hands. Right now, for
the small to medium business running Windows, Microsoft has a virtual lock
with SQL Server. SQL Server is expensive and a real PAIN.

Giving Windows users PostgreSQL with a good set of .NET, ODBC, and JDBC
drivers loosens the Microsoft stranglehold, just a little bit. If they
develop their application with MSSQL, there is a good chance it will never
use any open source software and always run on Windows. If they develop
their application using PostgreSQL, there is a better likelyhood that
other open source projects will be used, AND that should the requirement
be to upgrade the system, a wider range of OS and hardware options will
present themselves.

-----Original Message-----
From: pgsql@mohawksoft.com [mailto:pgsql@mohawksoft.com]
Sent: Friday, June 11, 2004 9:39 AM
To: Tom Lane
Cc: Dann Corbit; Zeugswetter Andreas SB SD;
jearl@bullysports.com; tswan@idigx.com; Bruce Momjian; Greg
Stark; pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: Re: [pgsql-hackers-win32] [HACKERS] Tablespaces

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten

times as many

PostgreSQL systems on Win32 as all combined versions now on UNIX
flavors

I surely hope not. Especially not multi-gig databases. The folks
running those should know better than to use Windows, and

if they do

not, I'll be happy to tell them so.

I know better than to tell people to change their operating system.
Linux is a great OS, and people familiar with it will do exceedingly
well. But there are 40 million computers sold in a year, most of which
have some flavor of Windows installed. People know how to use and
administer them, and they have all their applications in Windows. They
are not going to change for ideological reasons. Also, it isn't just
DBAs that need to implement database systems. Suppose, for instance,
that I want to write an accounting package. I can use PostgreSQL as a
base and save my customers thousands of dollars. If I tell them, "Now,
you need to reformat your machine and install Linux" that would not be
very popular. But they don't even need to know about the database. And
they should not have to care about the OS. A database and an operating
system are both things to help get work done. Believe it or not, lots
of large companies depend on Windows OS.

Personally, I am technology neutral. My position is "use whatever you
like."

This is a prejudice that we should try to avoid. Yes, Windows
is lacking on so many levels, but that really isn't the point.

Every OS has advantages and disadvantages. The applications for Windows
are many and mature. The tool sets available for Linux are extensive
and usually free. If you want real 24x7x365.25 then MVS cannot be beat.
The file versioning and protections of OpenVMS are something that all
operating systems should have modeled.

A good box running Win2K or XP Server, with no internet
connectivity, and no user applications, can really perform
and be reliable. Would I choose this? Hell no, but there are
HUGE amount of people who either don't know any better or
have no real choice.

And there are knowledgeable people who understand Windows, Linux and
many other operating systems who choose Windows because it is the best
choice for their company.

The REAL bonus here is getting PostgreSQL in their hands.
Right now, for the small to medium business running Windows,
Microsoft has a virtual lock with SQL Server. SQL Server is
expensive and a real PAIN.

It is expensive and a multi-user system ramps the cost. But it is
easier to administer than PostgreSQL. Hopefully, autovacuum will remove
most of this discrepancy.

Giving Windows users PostgreSQL with a good set of .NET,
ODBC, and JDBC drivers loosens the Microsoft stranglehold,
just a little bit. If they develop their application with
MSSQL, there is a good chance it will never use any open
source software and always run on Windows. If they develop
their application using PostgreSQL, there is a better
likelyhood that other open source projects will be used, AND
that should the requirement be to upgrade the system, a wider
range of OS and hardware options will present themselves.

Microsoft dominates because they offer real value (the world is not
completely full of idiot CEOs -- they make decisions based on profit).
The open source community is closing the gap, but it has a long way to
go. I don't see Microsoft as the dark side of the force or anything.
Actually, the approach of PostgreSQL and ACE is (too me) the most
superior. The GPL approach is far too confining, and getting a black
box that will be a terrible mystery if it breaks are not nearly so
pleasant.

Instead of telling people how to do their jobs, I suggest the approach
of providing the best possible tools and letting them decide how to use
them.

-----Original Message-----
From: pgsql@mohawksoft.com [mailto:pgsql@mohawksoft.com]
Sent: Friday, June 11, 2004 9:39 AM
To: Tom Lane
Cc: Dann Corbit; Zeugswetter Andreas SB SD;
jearl@bullysports.com; tswan@idigx.com; Bruce Momjian; Greg
Stark; pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: Re: [pgsql-hackers-win32] [HACKERS] Tablespaces

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten

times as many

PostgreSQL systems on Win32 as all combined versions now on UNIX
flavors

I surely hope not. Especially not multi-gig databases. The folks
running those should know better than to use Windows, and

if they do

not, I'll be happy to tell them so.

I know better than to tell people to change their operating system.
Linux is a great OS, and people familiar with it will do exceedingly
well. But there are 40 million computers sold in a year, most of which
have some flavor of Windows installed.

How many billions of cigarettes are sold? How many Big Macs? Popularity
does not imply quality or safety.

People know how to use and
administer them, and they have all their applications in Windows. They
are not going to change for ideological reasons.

This is interesting, since when is ideology *not* the american way? Have
you looked at politics lately?

Also, it isn't just
DBAs that need to implement database systems. Suppose, for instance,
that I want to write an accounting package. I can use PostgreSQL as a
base and save my customers thousands of dollars. If I tell them, "Now,
you need to reformat your machine and install Linux" that would not be
very popular. But they don't even need to know about the database. And
they should not have to care about the OS. A database and an operating
system are both things to help get work done. Believe it or not, lots
of large companies depend on Windows OS.

I've been in the trenches for a while now, and I haven't met a single CIO
that is comfortable with Windows. They hate the cost, they hate the
viruses, they hate the instability. The only thing they hate more is being
isolated on an island. Fortunately Linux is becoming less obscure.

Personally, I am technology neutral. My position is "use whatever you
like."

I would call myself "neutral" to a point, but when I have to give advice,
I have to tell the truth. A little Linux goes a long way.

This is a prejudice that we should try to avoid. Yes, Windows
is lacking on so many levels, but that really isn't the point.

Every OS has advantages and disadvantages.

Some more than other.

The applications for Windows
are many and mature. The tool sets available for Linux are extensive
and usually free. If you want real 24x7x365.25 then MVS cannot be beat.
The file versioning and protections of OpenVMS are something that all
operating systems should have modeled.

A good box running Win2K or XP Server, with no internet
connectivity, and no user applications, can really perform
and be reliable. Would I choose this? Hell no, but there are
HUGE amount of people who either don't know any better or
have no real choice.

And there are knowledgeable people who understand Windows, Linux and
many other operating systems who choose Windows because it is the best
choice for their company.

I seriously do not know anyone, including myself, that would choose
Windows on technical merrits alone. I know some need to choose it for
"killer" application requirements, but not on merrit.

As for best choice for their company, I can't even say that with a
straight face.

The REAL bonus here is getting PostgreSQL in their hands.
Right now, for the small to medium business running Windows,
Microsoft has a virtual lock with SQL Server. SQL Server is
expensive and a real PAIN.

It is expensive and a multi-user system ramps the cost. But it is
easier to administer than PostgreSQL. Hopefully, autovacuum will remove
most of this discrepancy.

Having dealt with both, as well as MySQL, DB2, and Oracle, I not sure I
agree with that statement. As long as MSSQL is installed correctly the
first time, it may be OK.

Giving Windows users PostgreSQL with a good set of .NET,
ODBC, and JDBC drivers loosens the Microsoft stranglehold,
just a little bit. If they develop their application with
MSSQL, there is a good chance it will never use any open
source software and always run on Windows. If they develop
their application using PostgreSQL, there is a better
likelyhood that other open source projects will be used, AND
that should the requirement be to upgrade the system, a wider
range of OS and hardware options will present themselves.

Microsoft dominates because they offer real value (the world is not
completely full of idiot CEOs -- they make decisions based on profit).

FACT: Microsoft dominates because they break the law.

The open source community is closing the gap, but it has a long way to
go. I don't see Microsoft as the dark side of the force or anything.

Then you have not had your company stomped on by them. You have not worked
on technologies like "Go Computing."

Actually, the approach of PostgreSQL and ACE is (too me) the most
superior. The GPL approach is far too confining, and getting a black
box that will be a terrible mystery if it breaks are not nearly so
pleasant.

GPL vs BSD is a long debate.

Instead of telling people how to do their jobs, I suggest the approach
of providing the best possible tools and letting them decide how to use
them.

We should provide people with the right tools, true, but we are bound by
our conscience to inform them about Windows' failures.

-----Original Message-----
From: pgsql@mohawksoft.com [mailto:pgsql@mohawksoft.com]
Sent: Friday, June 11, 2004 1:37 PM
To: Dann Corbit
Cc: Tom Lane; Zeugswetter Andreas SB SD;
jearl@bullysports.com; tswan@idigx.com; Bruce Momjian; Greg
Stark; pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: RE: [pgsql-hackers-win32] [HACKERS] Tablespaces

-----Original Message-----
From: pgsql@mohawksoft.com [mailto:pgsql@mohawksoft.com]
Sent: Friday, June 11, 2004 9:39 AM
To: Tom Lane
Cc: Dann Corbit; Zeugswetter Andreas SB SD; jearl@bullysports.com;
tswan@idigx.com; Bruce Momjian; Greg Stark;
pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: Re: [pgsql-hackers-win32] [HACKERS] Tablespaces

"Dann Corbit" <DCorbit@connx.com> writes:

I expect that one year after release, there will be ten

times as many

PostgreSQL systems on Win32 as all combined versions

now on UNIX

flavors

I surely hope not. Especially not multi-gig databases.

The folks

running those should know better than to use Windows, and

if they do

not, I'll be happy to tell them so.

I know better than to tell people to change their operating system.
Linux is a great OS, and people familiar with it will do

exceedingly

well. But there are 40 million computers sold in a year, most of
which have some flavor of Windows installed.

How many billions of cigarettes are sold? How many Big Macs?
Popularity does not imply quality or safety.

Right. It implies volume. That was the only point I was making. If
everyone is wearing suede shoes, you will have trouble selling shoe
polish.

People know how to use and
administer them, and they have all their applications in Windows.
They are not going to change for ideological reasons.

This is interesting, since when is ideology *not* the
american way? Have you looked at politics lately?

I am also politically neutral and have not voted since I was 18 as a
matter of conscience.

Also, it isn't just
DBAs that need to implement database systems. Suppose, for

instance,

that I want to write an accounting package. I can use

PostgreSQL as a

base and save my customers thousands of dollars. If I tell them,
"Now, you need to reformat your machine and install Linux"

that would

not be very popular. But they don't even need to know about the
database. And they should not have to care about the OS.

A database

and an operating system are both things to help get work done.
Believe it or not, lots of large companies depend on Windows OS.

I've been in the trenches for a while now, and I haven't met
a single CIO that is comfortable with Windows. They hate the
cost, they hate the viruses, they hate the instability. The
only thing they hate more is being isolated on an island.
Fortunately Linux is becoming less obscure.

Personally, I am technology neutral. My position is "use

whatever you

like."

I would call myself "neutral" to a point, but when I have to
give advice, I have to tell the truth. A little Linux goes a long way.

For me, if I was going to start a company, Linux is a technically
superior solution for a server in my view. This is especially true due
to license reasons. If I want a thousand users on a machine, the cost
for a Windows solution dwarfs any reasons I can think of not to switch
to Linux. However, if a company does not have personnel trained to
administrate Linux machines and applications, then something else might
be a better choice for them. ("Fire all your workers and hire new
ones." does not work)

This is a prejudice that we should try to avoid. Yes, Windows is
lacking on so many levels, but that really isn't the point.

Every OS has advantages and disadvantages.

Some more than other.

And yet each choice can have different weights depending upon who is
using it, for what reasons, and other business factors.

The applications for Windows
are many and mature. The tool sets available for Linux are

extensive

and usually free. If you want real 24x7x365.25 then MVS cannot be
beat. The file versioning and protections of OpenVMS are something
that all operating systems should have modeled.

A good box running Win2K or XP Server, with no internet

connectivity,

and no user applications, can really perform and be

reliable. Would I

choose this? Hell no, but there are HUGE amount of people

who either

don't know any better or have no real choice.

And there are knowledgeable people who understand Windows,

Linux and

many other operating systems who choose Windows because it

is the best

choice for their company.

I seriously do not know anyone, including myself, that would
choose Windows on technical merrits alone. I know some need
to choose it for "killer" application requirements, but not on merrit.

Religious arguments are hard to fight when one person is unable to
listen.

As for best choice for their company, I can't even say that
with a straight face.

Then for those customers for which it is the best choice, you would give
them bad advice.

The REAL bonus here is getting PostgreSQL in their hands.

Right now,

for the small to medium business running Windows, Microsoft has a
virtual lock with SQL Server. SQL Server is expensive and a real
PAIN.

It is expensive and a multi-user system ramps the cost. But it is
easier to administer than PostgreSQL. Hopefully, autovacuum will
remove most of this discrepancy.

Having dealt with both, as well as MySQL, DB2, and Oracle, I
not sure I agree with that statement. As long as MSSQL is
installed correctly the first time, it may be OK.

It is the easiest of those database systems you named to install and
administrate.

Giving Windows users PostgreSQL with a good set of .NET, ODBC, and
JDBC drivers loosens the Microsoft stranglehold, just a

little bit.

If they develop their application with MSSQL, there is a

good chance

it will never use any open source software and always run

on Windows.

If they develop their application using PostgreSQL, there

is a better

likelyhood that other open source projects will be used, AND
that should the requirement be to upgrade the system, a wider
range of OS and hardware options will present themselves.

Microsoft dominates because they offer real value (the world is not
completely full of idiot CEOs -- they make decisions based

on profit).

FACT: Microsoft dominates because they break the law.

FACT: Every blanket statement is an over-generalization.
;-)

Microsoft has done some very seedy and evil things. The destruction of
STAK comes to mind. As far as domination by giving things away, that is
ludicrous. You can download Linux and most Linux applications for
nothing. In essense, they give everything away (really, you are buying
service when you purchase a Linux CD) and MS has to compete with 'free'
-- not just for the tack-on stuff but for the entire system and all the
applications. It shows you how well done some of their stuff is, that
it can compete in that arena.

The open source community is closing the gap, but it has a

long way to

go. I don't see Microsoft as the dark side of the force or

anything.

Then you have not had your company stomped on by them.

I have had to deal with many things that were unpleasant and caused
directly by MS doing something strange or bad. I have lost days of work
because of it.

You
have not worked on technologies like "Go Computing."

I don't even know what 'Go Computing' is, so you are right about that
one.

Actually, the approach of PostgreSQL and ACE is (too me) the most
superior. The GPL approach is far too confining, and

getting a black

box that will be a terrible mystery if it breaks are not nearly so
pleasant.

GPL vs BSD is a long debate.

Clearly won by BSD for me (no contest whatsoever), and others like GPL
better.

Instead of telling people how to do their jobs, I suggest

the approach

of providing the best possible tools and letting them decide how to
use them.

We should provide people with the right tools, true, but we
are bound by our conscience to inform them about Windows' failures.

It must be nice to be young and still see everything as black and white
with no shades of gray. For those who think that Windows should be
canned, Gates should be burned at the stake, and Linux should rule the
world, I have no problem with their opinions. We all get to choose what
we like and dislike. I think that the typical Linux fan is WAY over the
top both in seeing the advantages with rose colored glasses and turing
opposition molehills into mountains. But passion is good, and I like to
see it. If it were not for the passion of the Linux crowd, there would
be a far less interesting competitor for MS and a far less interesting
toolset to use with it.

Maybe the thread should go to some advocacy channel at this point.

My reason for jumping in was to show that:
1. PostgreSQL will have a exponential leap in possible sites when it
opens up to Win32 systems
2. There will be huge installations on Win32 systems, like it or not.

Some other things to keep in mind:
1. The average Windows user is far, far less computer saavy than a
Linux (or other flavor of UNIX user) and hence, there will be a big load
of "deer in the headlights" users coming on board.
2. On the plus side, there are millions of good developers familiar
with Windows. Some of these may become involved with the PostgreSQL
project and give added value.

-----Original Message-----
From: pgsql@mohawksoft.com [mailto:pgsql@mohawksoft.com]
Sent: Friday, June 11, 2004 2:41 PM
To: Dann Corbit
Cc: pgsql-hackers@postgresql.org; PostgreSQL Win32 port list
Subject: RE: [pgsql-hackers-win32] [HACKERS] Tablespaces

[snip]

Microsoft has harmed the computing industry more than any
single factor that I can remember. I've seen a lot of it,
from DEC to Wang, and microsoft has single handedly wiped out
more computing innovation in 20 years than any 10 other companies.

Stac, Go Computing, Netscape, BeOS, and the list grows for as
long as you think about it.

Netscape and BeOS are still around. I think Linux is a bigger blow to
BeOS than MS.
Stak was unbelievable and MS got a tap on the wrist compared to the harm
caused.
I will have to look up Go Computing to see what all that is about.

It isn't wide eyes passion, I've programed computers since jr
high school, on a PDP-8/e. It is what I love to do, and it is
what I make my living doing.

Some things are important in life. Spending a few extra
dollars *NOT* going to Walmart is one small thing you can do
to improve the world. Taking advantage of every LEGITIMATE
opportunity to move a person or project off Windows is one
small step one can do to improve our industry.

If they are moved to another platform for their benefit or for the right
reasons there is nothing wrong with it. If it is because of your own
ideology and not for the benefit of the client then it is harm to them
and immoral. IMO-YMMV

Maybe the thread should go to some advocacy channel at this point.

Yes.

My reason for jumping in was to show that:
1. PostgreSQL will have a exponential leap in possible

sites when it

opens up to Win32 systems 2. There will be huge installations on
Win32 systems, like it or not.

Some other things to keep in mind:
1. The average Windows user is far, far less computer saavy than a
Linux (or other flavor of UNIX user) and hence, there will be a big
load of "deer in the headlights" users coming on board.

Total koolaid induced delusion. Dumb users are dumb users,
"deer in the headlights" looks come from flashing VCR clocks.
Competent professionals can handle a few twists. The switch
from DOS Windows (3x,9x,ME) to XP was just as traumatic.

You are totally wrong about that. 'Dumb users' are people who don't
care to become computer saavy. Often because they don't need to.
Someone who can't program their VCR may be able to do brain surgery on
you. Like Will Rogers said, "Everyone is ignorant, only in different
areas." If people don't want to become computer experts, we should not
try to force them to become so. You and I enjoy computers but other
people just want what the computer can deliver and don't care to learn
how it got there.

2. On the plus side, there are millions of good developers

familiar

with Windows. Some of these may become involved with the

PostgreSQL

project and give added value.

Having been a Windows developer since version 1.03, with DOS
and CP/M before that, I can say with complete authority that
most Windows developers are not "good." The worst I've seen
is Charles Petzold, and he sets the bar.

Charles Petzold is a decent programmer. I have read his books and he
knows what he's talking about. He no W. Richard Stevens or Donald
Knuth, but I would hire him to do a job.

pgsql@mohawksoft.com wrote:

We should provide people with the right tools, true, but we
are bound by our conscience to inform them about Windows' failures.

It must be nice to be young and still see everything as black and white
with no shades of gray.

I wouldn't call 41 very young.

For those who think that Windows should be
canned, Gates should be burned at the stake, and Linux should rule the
world, I have no problem with their opinions. We all get to choose what
we like and dislike. I think that the typical Linux fan is WAY over the
top both in seeing the advantages with rose colored glasses and turing
opposition molehills into mountains. But passion is good, and I like to
see it. If it were not for the passion of the Linux crowd, there would
be a far less interesting competitor for MS and a far less interesting
toolset to use with it.

Actually, I am not a wide eyed passionate Linux zealot. Like my support
for John Kerry, I gladly choose the better side of mediocrity over extream
evil, it is nothing more than pure practicality.

Well, call me extreme evil too. Then I guess PostgreSQL is partly pure
evil, or partly extreme evil, or something like that.

Of course, if you meet me, I don't appear so. We are taught to hide our
evil so effectively.

-- 
  Bruce Momjian                        |  http://candle.pha.pa.us
  pgman@candle.pha.pa.us               |  (610) 359-1001
  +  If your life is a hard drive,     |  13 Roberts Road
  +  Christ can be your backup.        |  Newtown Square, Pennsylvania 19073

We should provide people with the right tools, true, but we
are bound by our conscience to inform them about Windows' failures.

It must be nice to be young and still see everything as black and white
with no shades of gray.

I wouldn't call 41 very young.

For those who think that Windows should be
canned, Gates should be burned at the stake, and Linux should rule the
world, I have no problem with their opinions. We all get to choose what
we like and dislike. I think that the typical Linux fan is WAY over the
top both in seeing the advantages with rose colored glasses and turing
opposition molehills into mountains. But passion is good, and I like to
see it. If it were not for the passion of the Linux crowd, there would
be a far less interesting competitor for MS and a far less interesting
toolset to use with it.

Actually, I am not a wide eyed passionate Linux zealot. Like my support
for John Kerry, I gladly choose the better side of mediocrity over extream
evil, it is nothing more than pure practicality.

A diversity of platforms in the market place creates jobs, increased
security (any particular exploit does not wipe out a vast majority of
targets.), and feeds innovation and competition.

Microsoft has harmed the computing industry more than any single factor
that I can remember. I've seen a lot of it, from DEC to Wang, and
microsoft has single handedly wiped out more computing innovation in 20
years than any 10 other companies.

Stac, Go Computing, Netscape, BeOS, and the list grows for as long as you
think about it.

It isn't wide eyes passion, I've programed computers since jr high school,
on a PDP-8/e. It is what I love to do, and it is what I make my living
doing.

Some things are important in life. Spending a few extra dollars *NOT*
going to Walmart is one small thing you can do to improve the world.
Taking advantage of every LEGITIMATE opportunity to move a person or
project off Windows is one small step one can do to improve our industry.

Maybe the thread should go to some advocacy channel at this point.

Yes.

My reason for jumping in was to show that:
1. PostgreSQL will have a exponential leap in possible sites when it
opens up to Win32 systems
2. There will be huge installations on Win32 systems, like it or not.

Some other things to keep in mind:
1. The average Windows user is far, far less computer saavy than a
Linux (or other flavor of UNIX user) and hence, there will be a big load
of "deer in the headlights" users coming on board.

Total koolaid induced delusion. Dumb users are dumb users, "deer in the
headlights" looks come from flashing VCR clocks. Competent professionals
can handle a few twists. The switch from DOS Windows (3x,9x,ME) to XP was
just as traumatic.

2. On the plus side, there are millions of good developers familiar
with Windows. Some of these may become involved with the PostgreSQL
project and give added value.

Having been a Windows developer since version 1.03, with DOS and CP/M
before that, I can say with complete authority that most Windows
developers are not "good." The worst I've seen is Charles Petzold, and he
sets the bar.