oracle pl/sql -> sql server

[prunoki]

Hello,

I am an SQL server newbie. Our company has a massive application
written in PL/SQL. I need to port parts of it to SQL Server.

- Which SQL server version should I choose, to have a reasonable chance
of porting?

- Could you recommend any best practices, tools? I use Toad for SQL
development on Oracle, I am looking for something similar for SQL
server.

Regards,

Hegyvari Krisztian

[Simon Hayes]

The latest production version of MSSQL is 2000; 2005 will be released
at the start of November. Which one of those you choose depends on how
compelling you find the features in 2005, and how cautious you are
about porting to a version 1 product. However, given that MSSQL 2000 is
already 5 years or so old, it would certainly worth be worth
considering 2005.

You should probably spend some time reviewing the product information
here, which includes migration guides and tools for porting from
Oracle:

http://www.microsoft.com/sql/default.mspx

As with any significant migration, tools will only take you to a
certain point - porting DDL and data may be relatively straightforward,
but procedural code is another matter, so you should be prepared to do
a certain amount of work by hand. And of course there are all the
issues of working with a completely new architecture, being able to
run/monitor/backup etc a new platform, the security model etc.

If your company has no previous experience of working with MSSQL, it
would probably be worth looking for someone who has, otherwise there's
a good chance that the migration will not be successful (or it may be
functional but perform unacceptably, have no robust security etc).
Personally, I would not be confident about porting an application to a
platform which I don't know, and your management needs to understand
the risks involved in that approach.

Simon

[DA Morgan]

prunoki wrote:

Hello,

I am an SQL server newbie. Our company has a massive application
written in PL/SQL. I need to port parts of it to SQL Server.

- Which SQL server version should I choose, to have a reasonable chance
of porting?

- Could you recommend any best practices, tools? I use Toad for SQL
development on Oracle, I am looking for something similar for SQL
server.

Regards,

Hegyvari Krisztian

Depending on what you mean by the word "port" I would suggest you not
do it.

If your Oracle application uses sequences ... SQL Server doesn't have
them. If your Oracle application uses the full range of triggers ...
SQL Server doesn't have them. If your Oracle application uses many
different table and index types ... SQL Server doesn't have them. If
your application assumes the Oracle transaction model, locking model,
and read-write consistency ... SQL Server doesn't have them.

It would be far better to rewrite the application based on the Oracle
version to take advantage of optimizing SQL Server's capabilities.
People that try to port, either way, almost always make big expensive
messes.

An application built in an environment where reads don't block writes
and writes don't block reads will fare very poorly in one that doesn't
have the same architecture.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Greg D. Moore \(Strider\)]

"Simon Hayes" <sq*@hayes.ch> wrote in message
news:11**********************@o13g2000cwo.googlegr oups.com...

The latest production version of MSSQL is 2000; 2005 will be released
at the start of November. Which one of those you choose depends on how
compelling you find the features in 2005, and how cautious you are
about porting to a version 1 product. However, given that MSSQL 2000 is
already 5 years or so old, it would certainly worth be worth
considering 2005.

Also given that SQL 2005 has several design changes that specifically are
aimed at it "taking on Oracle" (like changes to transaction concurrency) I'd
go with SQL 2005.

Also, MS has a product on their website (I don't recall the exact URL) that
will take an Oracle app and port to SQL Server.

Personally I wouldn't trust any tool to do more than do a basic port of the
schema. I'd still review ALL your code to make sure it works in a similar
fashion on SQL Server as it does on Oracle (concurrency being a big one,
sequences another).

If your company has no previous experience of working with MSSQL, it
would probably be worth looking for someone who has, otherwise there's
a good chance that the migration will not be successful (or it may be
functional but perform unacceptably, have no robust security etc).
Personally, I would not be confident about porting an application to a
platform which I don't know, and your management needs to understand
the risks involved in that approach.
I would second this.


Simon

[DA Morgan]

Greg D. Moore (Strider) wrote:

Also given that SQL 2005 has several design changes that specifically are
aimed at it "taking on Oracle" (like changes to transaction concurrency) I'd
go with SQL 2005.

In short ... after complaining bitterly for years about how wrong
Oracle's architecture is ... they've copied it. Flattery perhaps but
where is the apology for all of the criticism's heaped on Oracle for
so many years?

So much for innovation in Redmond. Why invent when you can copy.
But then why would anyone want these features in a version 1.0 product
when they could have them fully tested in a version 10? Go figure.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Martijn Tonies]

I am an SQL server newbie. Our company has a massive application
written in PL/SQL. I need to port parts of it to SQL Server.

- Which SQL server version should I choose, to have a reasonable chance
of porting?
I'd say, try the 2005 beta.
- Could you recommend any best practices, tools? I use Toad for SQL
development on Oracle, I am looking for something similar for SQL
server.

Database Workbench ( www.upscene.com ) can do both Oracle and
MS SQL Server. SQL Server 2005 support is in the making.

As for "simply porting" your application -- that will be hard, believe
me. Both are quite different. For example, MS SQL Server only has
"after action" triggers.

--
With regards,

Martijn Tonies
Database Workbench - tool for InterBase, Firebird, MySQL, Oracle & MS SQL
Server
Upscene Productions
http://www.upscene.com
Database development questions? Check the forum!
http://www.databasedevelopmentforum.com

[Alexander Kuznetsov]

>So much for innovation in Redmond. Why invent when you can copy.

I feel your pain, but on the other hand I think Oracle did not invent
SQL. IBM did, correct? Why invent the wheel anyway?

[John Bell]

> As for "simply porting" your application -- that will be hard, believe

me. Both are quite different. For example, MS SQL Server only has
"after action" triggers.

That is not correct there are instead of triggers as well.

John

[Martijn Tonies]

> > As for "simply porting" your application -- that will be hard, believe

me. Both are quite different. For example, MS SQL Server only has
"after action" triggers.

That is not correct there are instead of triggers as well.

Right John, thanks for correcting that.

Anyway, no "before" triggers ;-)

--
With regards,

Martijn Tonies
Database Workbench - tool for InterBase, Firebird, MySQL, Oracle & MS SQL
Server
Upscene Productions
http://www.upscene.com
Database development questions? Check the forum!
http://www.databasedevelopmentforum.com

[DA Morgan]

Alexander Kuznetsov wrote:

So much for innovation in Redmond. Why invent when you can copy.

I feel your pain, but on the other hand I think Oracle did not invent
SQL. IBM did, correct? Why invent the wheel anyway?

Actually IBM didn't invent it either. You need to look further back
into the history books for that.

But it is true that Ellison worked on System/R, IBM decided it
wasn't interested. And Ellison, legally, walked out with a couple
of partners and his knowledge and built it himself.

The difference is that Ellison didn't spend years criticizing
SQL and relational theory before adopting it. Microsoft has been
criticizing, for many years, the very features it will now tout
as innovations. That is quite a different situation methinks.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[DA Morgan]

John Bell wrote:

As for "simply porting" your application -- that will be hard, believe
me. Both are quite different. For example, MS SQL Server only has
"after action" triggers.

That is not correct there are instead of triggers as well.

John

What SQL Server lacks is BEFORE triggers, DDL triggers, and SYSTEM
triggers.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[prunoki]

Yep, I actually mean rewriting. I would like to reuse the logic behind
the code and stay as close to the original as possible.

K

[Craig Kelly]

"DA Morgan" wrote:

Actually IBM didn't invent it either. You need to look further back
into the history books for that.

Could you clarify the history part? I thought Chamberlin and Boyce
developed SEQUEL for System/R using (in part) Codd's relational work... and
all three of those guys were at IBM at the time. What am I missing?

Craig

[Erland Sommarskog]

DA Morgan (da******@psoug.org) writes:

What SQL Server lacks is BEFORE triggers, DDL triggers, and SYSTEM
triggers.

Right on first. The second and third are added in SQL 2005.

--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp

[DA Morgan]

Erland Sommarskog wrote:

DA Morgan (da******@psoug.org) writes:

What SQL Server lacks is BEFORE triggers, DDL triggers, and SYSTEM
triggers.

Right on first. The second and third are added in SQL 2005.

I don't comment on products that are, as yet, unreleased.

Until version 1.0 is officially released it does not exist.
Any feature you see in Beta may or may not make it to the final.

The one thing that fascinates me about the SQL Server community
is that as a Beta tester for Oracle I am prevented by NDA from
saying a word about what I see. You folks, on the other hand
run around talking about it like you were signed on by the public
relations department.

Scan comp.databases.ms-sqlserver and you find post after post
about the Beta. Scan comp.databases.oracle.server and you will
find hardly a peep about 10gR2 (now released) but not a single
word about 11. Your is the only technology community I can think
of that publishes books about vapourware: Quite amusing.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Erland Sommarskog]

DA Morgan (da******@psoug.org) writes:

I don't comment on products that are, as yet, unreleased.

Until version 1.0 is officially released it does not exist.
Any feature you see in Beta may or may not make it to the final.
One should keep in mind that the SQL 2005 beta is very late in the cycle.
The September CTP that was released this week is most likely the last
public drop before the whole thing launches Nov 7th.

It is highly unlikely that any features would be dropped at this point.
Keep in mind that there are quite a few installations out there, which
already are running SQL 2005 in production, both inside and outside
Microsoft. (What has happened though is that one feature, Database
mirroring, will only be for "evaluation" initially.)
The one thing that fascinates me about the SQL Server community
is that as a Beta tester for Oracle I am prevented by NDA from
saying a word about what I see. You folks, on the other hand
run around talking about it like you were signed on by the public
relations department.
Guess what? Microsoft has NDAs as well, and when I first got access to
beta 1 two years ago I was under NDA. However, Microsoft did release
beta 1 publicly at PASS already in November 2003. Beta 2 was distributed
with MSDN. If you want to try it (OK, just joking :-), the latest CTP
is available on http://www.microsoft.com/sql/2005/productinfo/ctp.mspx.
Scan comp.databases.ms-sqlserver and you find post after post
about the Beta. Scan comp.databases.oracle.server and you will
find hardly a peep about 10gR2 (now released) but not a single
word about 11. Your is the only technology community I can think
of that publishes books about vapourware: Quite amusing.

I would not call software that actually runs and is available to the
public "vaporware". Anyway, from a commercial point of view, it appears
to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Since I have never used any Oracle product, I have no experience of they
run their betas. I would be suprised though, if they never make any betas
publically available, as most software vendors appear to do that at some
point in the cycle.

--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp

[DA Morgan]

Erland Sommarskog wrote:

DA Morgan (da******@psoug.org) writes:

I don't comment on products that are, as yet, unreleased.

Until version 1.0 is officially released it does not exist.
Any feature you see in Beta may or may not make it to the final.
One should keep in mind that the SQL 2005 beta is very late in the cycle.
The September CTP that was released this week is most likely the last
public drop before the whole thing launches Nov 7th.

That may well be true. But is still no guarantee that what is in the
Beta is in the production release or that it has been implemented in
the same way.
It is highly unlikely that any features would be dropped at this point.
I agree. But unlikely is far different from "impossible."
Keep in mind that there are quite a few installations out there, which
already are running SQL 2005 in production,
Then they deserve what they get. Anyone that would put a Beta into
production is a mashochist or a fool and demonstates a lack of
professionalism.

To use in for testing ... makes sense. To use it for limited development
might make sense. To move it to production is insanity.

The one thing that fascinates me about the SQL Server community
is that as a Beta tester for Oracle I am prevented by NDA from
saying a word about what I see. You folks, on the other hand
run around talking about it like you were signed on by the public
relations department.

Guess what? Microsoft has NDAs as well, and when I first got access to
beta 1 two years ago I was under NDA. However, Microsoft did release
beta 1 publicly at PASS already in November 2003. Beta 2 was distributed
with MSDN. If you want to try it (OK, just joking :-), the latest CTP
is available on http://www.microsoft.com/sql/2005/productinfo/ctp.mspx.

So it really isn't Beta at all. It is pre-release copies: Basically
marketing and public relations.

Scan comp.databases.ms-sqlserver and you find post after post
about the Beta. Scan comp.databases.oracle.server and you will
find hardly a peep about 10gR2 (now released) but not a single
word about 11. Your is the only technology community I can think
of that publishes books about vapourware: Quite amusing.

I would not call software that actually runs and is available to the
public "vaporware".

The point is that the features it contains do not necessarily reflect
the production release. Capabilities may be removed. Capabilities may
be modified. Capabilities may be added. My point is that it isn't
"real."

Anyway, from a commercial point of view, it appears to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.
Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.
Since I have never used any Oracle product, I have no experience of they
run their betas. I would be suprised though, if they never make any betas
publically available, as most software vendors appear to do that at some
point in the cycle.

Never. The vast majority of software companies never make Betas
available to any other than qualified testers that will actuall use
their software for purposes of testing.

Anyone thinking they can get a copy of Oracle 11 should be prepared to
buy Larry a new boat. I doubt anything less will put it into their
hands today or tomorrow.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Hugo Kornelis]

On Fri, 16 Sep 2005 08:39:56 -0700, DA Morgan wrote:

(snip)

Anyway, from a commercial point of view, it appears

to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.

Since I have never used any Oracle product, I have no experience of they
run their betas. I would be suprised though, if they never make any betas
publically available, as most software vendors appear to do that at some
point in the cycle.

Never. The vast majority of software companies never make Betas
available to any other than qualified testers that will actuall use
their software for purposes of testing.

Anyone thinking they can get a copy of Oracle 11 should be prepared to
buy Larry a new boat. I doubt anything less will put it into their
hands today or tomorrow.

Hi DA,

Did you google for "public beta" before posting this? I got over 2
million hits. Just the first two pages had links to public beta programs
of (among others) Norton AntiVirus 2006, Novell OpenEnterprise Server,
Macromedia Flash Player, IBM Lotus Notes/Domino 7, and many others.

These companies apparently take quality serious, and take their
customers serious. They use the possibilities Internet offers to ensure
that their products are tested by a number of testers, and on a number
of configurations, that would never be achieved in a closed beta
program.

It's sad to learn that Oracle, apparently, still values secrecy over
quality.

Best, Hugo
--

(Remove _NO_ and _SPAM_ to get my e-mail address)

[Alexander Kuznetsov]

>> I feel your pain, but on the other hand I think Oracle did not invent

SQL. IBM did, correct? Why invent the wheel anyway?

Actually IBM didn't invent it either.

I have big problems trying to believe that, but I'm no good in history
and willing to listen and learn

[Erland Sommarskog]

DA Morgan (da******@psoug.org) writes:

That may well be true. But is still no guarantee that what is in the
Beta is in the production release or that it has been implemented in
the same way.
Only in the sense that there is no guarantee that what is in the RTM
version is also in SP1. OK, so Microsoft is still free to make breaking
changes, but the impact on the code base has to be small, not to
endanger the quality of the product.

Keep in mind that there are quite a few installations out there, which
already are running SQL 2005 in production,

Then they deserve what they get. Anyone that would put a Beta into
production is a mashochist or a fool and demonstates a lack of
professionalism.

Well, if you downloaded the beta from microsoft.com, and put a system
on a production from that, you would not only be unprofessional - you
would also be violating the license. There is a some sort of general
go-live license for SQL 2005, but it applies the Express Edition only.

Those who run other editions of SQL 2005 in production do that as part
of a Microsoft programme, where they have close interaction with Microsoft.
Certainly, you have to be a bit brave, but it's not insanity.

Guess what? Microsoft has NDAs as well, and when I first got access to
beta 1 two years ago I was under NDA. However, Microsoft did release
beta 1 publicly at PASS already in November 2003. Beta 2 was distributed
with MSDN. If you want to try it (OK, just joking :-), the latest CTP
is available on http://www.microsoft.com/sql/2005/productinfo/ctp.mspx.

So it really isn't Beta at all. It is pre-release copies: Basically
marketing and public relations.

Yes, that is certainly part of it. But not only. Microsoft very clearly
are interested in the feedback from the user community. Both with regards
to bugs, as well as opinions on features.
Anyway, from a commercial point of view, it appears

to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.

I can't escape the feeling that your attitude is that becuase Microsoft
does it, it's bad. That's not a very professional attitude.

I have never had reason to try software from SAP and IBM, but I have
also been involved in beta programmes for PowerDesigner, a data-modelling
tool from Sybase. And I have never signed an NDA about that.

If IBM, SAP or Oracle prefer to keep their betas closed until the product
is released, they may feel that that business model fits them well. I
don't believe for a second that think public betas as "unethical".
--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp

[DA Morgan]

Hugo Kornelis wrote:

On Fri, 16 Sep 2005 08:39:56 -0700, DA Morgan wrote:

(snip)

Anyway, from a commercial point of view, it appears

to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.

Since I have never used any Oracle product, I have no experience of they
run their betas. I would be suprised though, if they never make any betas
publically available, as most software vendors appear to do that at some
point in the cycle.

Never. The vast majority of software companies never make Betas
available to any other than qualified testers that will actuall use
their software for purposes of testing.

Anyone thinking they can get a copy of Oracle 11 should be prepared to
buy Larry a new boat. I doubt anything less will put it into their
hands today or tomorrow.

Hi DA,

Did you google for "public beta" before posting this? I got over 2
million hits. Just the first two pages had links to public beta programs
of (among others) Norton AntiVirus 2006, Novell OpenEnterprise Server,
Macromedia Flash Player, IBM Lotus Notes/Domino 7, and many others.

These companies apparently take quality serious, and take their
customers serious. They use the possibilities Internet offers to ensure
that their products are tested by a number of testers, and on a number
of configurations, that would never be achieved in a closed beta
program.

It's sad to learn that Oracle, apparently, still values secrecy over
quality.

Best, Hugo

Flash Player ... now there's a serious piece of software.

Just kidding. I guess I my views are somewhat outmoded given that I
deal in large line-of-business commercial applications. Did you find
any for Oracle? DB2? Informix? Sybase? ... Didn't think so.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[DA Morgan]

Alexander Kuznetsov wrote:

I feel your pain, but on the other hand I think Oracle did not invent
SQL. IBM did, correct? Why invent the wheel anyway?

Actually IBM didn't invent it either.

I have big problems trying to believe that, but I'm no good in history
and willing to listen and learn

Familiarize yourself with google. ;-)
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[DA Morgan]

Erland Sommarskog wrote:

DA Morgan (da******@psoug.org) writes:

That may well be true. But is still no guarantee that what is in the
Beta is in the production release or that it has been implemented in
the same way.

Only in the sense that there is no guarantee that what is in the RTM
version is also in SP1. OK, so Microsoft is still free to make breaking
changes, but the impact on the code base has to be small, not to
endanger the quality of the product.

That would be a good thing. But SQL Server 2005 is a version 1.0
product: The changes are massive.

Keep in mind that there are quite a few installations out there, which
already are running SQL 2005 in production,

Then they deserve what they get. Anyone that would put a Beta into
production is a mashochist or a fool and demonstates a lack of
professionalism.

Well, if you downloaded the beta from microsoft.com, and put a system
on a production from that, you would not only be unprofessional - you
would also be violating the license. There is a some sort of general
go-live license for SQL 2005, but it applies the Express Edition only.

I agree. But just a day ago in this thread the statement was made that
it was being done as you will recall.
Those who run other editions of SQL 2005 in production do that as part
of a Microsoft programme, where they have close interaction with Microsoft.
Certainly, you have to be a bit brave, but it's not insanity.
My point was development leading to production with a Beta. That is, as
you said above unprofessional.

Guess what? Microsoft has NDAs as well, and when I first got access to
beta 1 two years ago I was under NDA. However, Microsoft did release
beta 1 publicly at PASS already in November 2003. Beta 2 was distributed
with MSDN. If you want to try it (OK, just joking :-), the latest CTP
is available on http://www.microsoft.com/sql/2005/productinfo/ctp.mspx.

So it really isn't Beta at all. It is pre-release copies: Basically
marketing and public relations.

Yes, that is certainly part of it. But not only. Microsoft very clearly
are interested in the feedback from the user community. Both with regards
to bugs, as well as opinions on features.

Feedback to Microsoft would be both ethical and professional. But we
have feedback on usenet. I think that is an entirely different thing.

Anyway, from a commercial point of view, it appears

to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.

I can't escape the feeling that your attitude is that becuase Microsoft
does it, it's bad. That's not a very professional attitude.

Not at all. I teach at a university. I am an equal opportunity critic.
I would criticize any company that did it.
I have never had reason to try software from SAP and IBM, but I have
also been involved in beta programmes for PowerDesigner, a data-modelling
tool from Sybase. And I have never signed an NDA about that.
Can't comment because I have never heard of the product.
If IBM, SAP or Oracle prefer to keep their betas closed until the product
is released, they may feel that that business model fits them well. I
don't believe for a second that think public betas as "unethical".

Public Betas are not unethical. Using them as the basis for books and
production systems is I hope you would agree.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Erland Sommarskog]

DA Morgan (da******@psoug.org) writes:

Just kidding. I guess I my views are somewhat outmoded given that I
deal in large line-of-business commercial applications. Did you find
any for Oracle? DB2? Informix? Sybase? ... Didn't think so.
You may have a point there so far, that it depends a little on which market
you aim at. I don't know about Oracle, but SQL Server may be different
since Microsoft aimes a wider market spectrum. Basically, anything from
the local sports club to large enterprise systems, whereas Oracle does
not provide anything similar to MSDE/SQL Express, that is a free engine
for small-scale applicaiton. (But I since I don't follow Oracle at all,
I may be totally wrong.)

If the market mainly is companies of some size, public betas are maybe
not that meaningful. There are other channels that may be better.

A competitor to SQL Server in the low-end market, and also in the
middle market is MySQL. And sure enough, if you go to http://www.mysql.com
you can download their 5.0. Which is a beta version.
Public Betas are not unethical. Using them as the basis for books and
production systems is I hope you would agree.

No, I don't agree. If there is an interest in a beta product, I see
nothing wrong with writing a book about it. (And while encouraged by
Microsoft, the books are independent publications.) As for production
systems, I would expect Oracale as well put quite a few installations
live on their new versions, both inside and outside Oracle before they
make an official release. If they do not, I would be very worried as a
potential customer. There simply is no such thing is no such test as a
production environment...

--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp

[DA Morgan]

Erland Sommarskog wrote:

No, I don't agree. If there is an interest in a beta product, I see
nothing wrong with writing a book about it. (And while encouraged by
Microsoft, the books are independent publications.) As for production
systems, I would expect Oracale as well put quite a few installations
live on their new versions, both inside and outside Oracle before they
make an official release. If they do not, I would be very worried as a
potential customer. There simply is no such thing is no such test as a
production environment...

Oracle does huge Betas. But they are "real" beta programs not open to
the general public and requiring that the testers sign NDAs and report
back results: In other words they are real Betas. I'd be interested to
know, over the last 3.5 years, what percentage of the people with SQL
Server 2005 Beta actually provided any real feedback on the product.

But I am surprised you see no problem with a book about a Beta. A Beta
is guaranteed to not be production ready. Guaranteed to not be the
final release code. A book about a Beta is guaranteed to be incorrect
unless wholly superficial. It just doesn't compute to me.
--
Daniel A. Morgan
http://www.psoug.org
da******@x.washington.edu
(replace x with u to respond)

[Serge Rielau]

DA Morgan wrote:

Erland Sommarskog wrote:

No, I don't agree. If there is an interest in a beta product, I see
nothing wrong with writing a book about it. (And while encouraged by
Microsoft, the books are independent publications.) As for production
systems, I would expect Oracale as well put quite a few installations
live on their new versions, both inside and outside Oracle before they
make an official release. If they do not, I would be very worried as a
potential customer. There simply is no such thing is no such test as a
production environment...

Oracle does huge Betas. But they are "real" beta programs not open to
the general public and requiring that the testers sign NDAs and report
back results: In other words they are real Betas. I'd be interested to
know, over the last 3.5 years, what percentage of the people with SQL
Server 2005 Beta actually provided any real feedback on the product.

But I am surprised you see no problem with a book about a Beta. A Beta
is guaranteed to not be production ready. Guaranteed to not be the
final release code. A book about a Beta is guaranteed to be incorrect
unless wholly superficial. It just doesn't compute to me.

Isn't that for the purchaser of the book to decide? Presumably most
readers are consenting adults. On the DB2 side we have both closed and
open betas, also with select customers going into production earlier.
What makes a closed beta more "real" than a closed beta that turns into
an open beta (as Erland describes).
I too am surprised that people buy these books, but that tells me more
about demand than wrongness. Keep in mind that while the installations,
in average may be smaller than Oracle's and DB2's I wouldn't be
surprised if the number of developers is actually higher given the
SMBness of the install base. These folks want to get on the beta train
to learn. Nothing wrong with that.

Cheers
Serge
--
Serge Rielau
DB2 SQL Compiler Development
IBM Toronto Lab

[Hugo Kornelis]

On Fri, 16 Sep 2005 21:58:23 -0700, DA Morgan wrote:

Hugo Kornelis wrote:

On Fri, 16 Sep 2005 08:39:56 -0700, DA Morgan wrote:

(snip)

Anyway, from a commercial point of view, it appears

to me that Microsoft is doing the right thing. By making betas of the
new product available early on, more people get to play it, learn it
and know it, and will thus be more inclined to deploy it early on.

Other software companies seem to do well keeping their Betas as Betas.
I don't see SAP or IBM or anyone else thinking what you describe is
ethical.
Since I have never used any Oracle product, I have no experience of they
run their betas. I would be suprised though, if they never make any betas
publically available, as most software vendors appear to do that at some
point in the cycle.

Never. The vast majority of software companies never make Betas
available to any other than qualified testers that will actuall use
their software for purposes of testing.

Anyone thinking they can get a copy of Oracle 11 should be prepared to
buy Larry a new boat. I doubt anything less will put it into their
hands today or tomorrow.

Hi DA,

Did you google for "public beta" before posting this? I got over 2
million hits. Just the first two pages had links to public beta programs
of (among others) Norton AntiVirus 2006, Novell OpenEnterprise Server,
Macromedia Flash Player, IBM Lotus Notes/Domino 7, and many others.

These companies apparently take quality serious, and take their
customers serious. They use the possibilities Internet offers to ensure
that their products are tested by a number of testers, and on a number
of configurations, that would never be achieved in a closed beta
program.

It's sad to learn that Oracle, apparently, still values secrecy over
quality.

Best, Hugo

Flash Player ... now there's a serious piece of software.

Just kidding. I guess I my views are somewhat outmoded given that I
deal in large line-of-business commercial applications. Did you find
any for Oracle? DB2? Informix? Sybase? ... Didn't think so.

Hi DA,

A very short search on Google resulted in a link to an article about the
DB2 public beta program:
http://esj.com/enterprise/article.aspx?EditorialsID=954

I didn't find any for the others, but I only spent about half a minute
trying.

My point is, regardless of which companies do and which don't, that
public beta programs:

a) will attract a much higher number of beta testers than closed beta
programs, even if a high percentage of the beta testers won't really
push the product and/or won't report bugs. Suppose one million people
download the product; 98% of them (attempt to) install and only 25% of
the rest will actively test the product and report bugs. That leaves you
with a total of 245,000 active beta testers. How does that comapre to a
typical closed beta program?

b) will provide prospective customers with a way to comment on new
features before it's too late to change. Closed beta's are typically
only done with the existing user base; getting early feedback from
prospective customers could be extremely valuable in a competitive
market.

c) will also serve as a good way to get potential customers to try the
new features and maybe become interested in the product. In this regard,
a public beta is no different fro a time-limited evaluation edition. Are
they unethical too?

Of course, products that have minimum hardware requirements that costs
thousands of dollars and products that are so complicated that they
require a staff of trained experts to setup and run are not suited for
public beta's. But for a product with the relatively low hardware
requirements and the ease of use of SQL Server, a public beta test is a
great way toa ccomplish the three aforementioned benefits.

Best, Hugo
--

(Remove _NO_ and _SPAM_ to get my e-mail address)

[Erland Sommarskog]

Serge Rielau (sr*****@ca.ibm.com) writes:

Isn't that for the purchaser of the book to decide? Presumably most
readers are consenting adults. On the DB2 side we have both closed and
open betas, also with select customers going into production earlier.
What makes a closed beta more "real" than a closed beta that turns into
an open beta (as Erland describes).
Yeah. I don't know why DA comes with this rubbish about "real" betas. Had
he worked at the Oracle marketing department, it would have made sense.
Now it only serves to give an unprofessional impression of himself.
I too am surprised that people buy these books, but that tells me more
about demand than wrongness. Keep in mind that while the installations,
in average may be smaller than Oracle's and DB2's I wouldn't be
surprised if the number of developers is actually higher given the
SMBness of the install base. These folks want to get on the beta train
to learn. Nothing wrong with that.

Yes, I think you nailed it there.
--
Erland Sommarskog, SQL Server MVP, es****@sommarskog.se

Books Online for SQL Server SP3 at
http://www.microsoft.com/sql/techinf...2000/books.asp

[Alexander Kuznetsov]

> Familiarize yourself with google. ;-)

1. IMO the remark is neither polite nor professional

2. Don Chamberlin on his site provides a link to:

"... System R is a database system built as a research project at IBM
San Jose Research (now IBM Almaden Research Center) in the 1970's.
System R introduced the SQL language and also demonstrated that a
relational system could provide good transaction processing
performance..."

3. Do you have any better arguments?