Thread: Version Numbering

As we know, most hindrances to OSS adoption are psychological. For that
reason, do you think that maybe 7.4, if it contains native windows support,
should be called 8.0?

Normally, I'm against version inflation, I think Linux 2.6 should be 2.6 and
not 3.0, MSN 8 should be MSN 3, and IE 6 should never have been created, or
at most, version 5.

But I do know it will be no small deal for me when it is available, and
don't I think I'm alone. I can see the headlines now: "PostgreSQL 8.0 - The
best ORDBMS now available on Windows!"

Then after that, all I need is a way to sync the row differences (insert,
delete, updates only) since last sync. If I could 'play' a transaction log
to one or more remote databases, then archive it, I'd be happy. Of course
I'd need Win<->Unix compatibility, because remote machines would most likely
be windows, and they'd be syncing to our unix master server. This is because
most of our clients don't have a persistent link. They sync with us, we
audit, fix, tally, and send it back. At the point of sending it down, all
sites have consistency.
Does something like this exist now? Right now we get all the data, for a few
kilobytes of changes!

Thanks,
J

PS. We're not using PostgreSQL right now, but I want to be real soon!

Jason,

> But I do know it will be no small deal for me when it is available, and
> don't I think I'm alone. I can see the headlines now: "PostgreSQL 8.0 - The
> best ORDBMS now available on Windows!"

This is *not* a discussion I want to have on Hackers.   There are already
programmers on Hackers who aren't keen on the whole idea of an advocacy
group; asking them to change version numbers for marketing reasons would be
pouring gasoline on the fire.

> Then after that, all I need is a way to sync the row differences (insert,
> delete, updates only) since last sync. If I could 'play' a transaction log
> to one or more remote databases, then archive it, I'd be happy. Of course
> I'd need Win<->Unix compatibility, because remote machines would most likely
> be windows, and they'd be syncing to our unix master server. This is because
> most of our clients don't have a persistent link. They sync with us, we
> audit, fix, tally, and send it back. At the point of sending it down, all
> sites have consistency.
> Does something like this exist now? Right now we get all the data, for a few
> kilobytes of changes!

Point-in-time restore is a goal for 7.4.   Whether this will extend to
log-based replication is beyond me.

--
-Josh Berkus
 Aglio Database Solutions
 San Francisco

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> As we know, most hindrances to OSS adoption are psychological. For that
> reason, do you think that maybe 7.4, if it contains native windows support,
> should be called 8.0?

No. This would never fly. Postgres has a reputation of sensible version
numbering. I think it would have to be something more major to the
underlying codebase than native Windows support before we jumped up
to 8.0.

Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200212130903

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE9+ejevJuQZxSWSsgRAvEaAKDv0OuZsBU8JTftibU0RO11UAIN1ACfcdwV
z3I5ESCM/8ZX94RyugrVDk0=
=f1RM
-----END PGP SIGNATURE-----

On Fri, 2002-12-13 at 08:55, greg@turnstep.com wrote:
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> > As we know, most hindrances to OSS adoption are psychological. For that
> > reason, do you think that maybe 7.4, if it contains native windows support,
> > should be called 8.0?
>
> No. This would never fly. Postgres has a reputation of sensible version
> numbering. I think it would have to be something more major to the
> underlying codebase than native Windows support before we jumped up
> to 8.0.
>

I think you meant to say "stubborn" version numbering, not sensible.
There were rumblings that 7.1 should have been 8.0 back in the day, and
I think a very good argument could be made that 7.3 should have been 8.0
(considering it broke all the client apps by adding schema support).

That said, it doesn't look like 7.4 is likely to cause that much
upheaval based on the big items on the list, so your right, a 7.4 seems
sensible if it stays at that.  OTOH, if they rewrite the communications
protocol I might be willing to revisit the discussion.

Robert Treat

> No. This would never fly. Postgres has a reputation of sensible
version
> numbering. I think it would have to be something more major to the
> underlying codebase than native Windows support before we jumped up
> to 8.0.

I disagree.


Most likely windows support will mean major rewrites of the code base,
for example there is discussion in hackers about switching from process
model to thread based model.

That being said, I think there is a lesson to be learned from Microsoft
and other successful software vendors.  Marketing concerns outweigh all
other concerns when trying to make decisions about things which are
visible to the outside world.  There is a very natural reasoning that
8.0 is the first version which was built to take over the database
world.

Do not underestimate how important windows compatibility is to the small
/medium business world.

Because postgres grew out of highly technical roots, technical concerns
stood behind most decisions like version #s and such things.  By this
point, the technical stuff is pretty much in line and its time to start
thinking like the big boys.  This probably will seem very distasteful to
the hacker types, but this is natural and even expected :)

For many commercial products there are probably two or even three sets
of version #s maintained for the product, one for the marketing folks,
one for the technical people (kept secret) and code revisions.



Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200212130903

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE9+ejevJuQZxSWSsgRAvEaAKDv0OuZsBU8JTftibU0RO11UAIN1ACfcdwV
z3I5ESCM/8ZX94RyugrVDk0=
=f1RM
-----END PGP SIGNATURE-----


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

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> Most likely windows support will mean major rewrites of the code base,
> for example there is discussion in hackers about switching from process
> model to thread based model.

A 'major rewrite' of the codebase to support Windows is just not going
to happen in 7.4. The major change in 7.4 is going to be replication.
As Robert pointed out, adding schemas *may* have been a reason to
bump up the number, but the changes from 7.3 to 7.4 do not warrant it.
I will also revisit if the communications protocol is changed,
but arguing that the version number should be changed for marketing
reasons is never going to get you anywhere in the Postgres community.

> For many commercial products there are probably two or even three sets
> of version #s maintained for the product, one for the marketing folks,
> one for the technical people (kept secret) and code revisions.

I don't think "secret" version numbers will work well on an open-source
project. :) You'll also need much better arguments than "that's the way
Microsoft does it" to convince people to adopt arbitrary version
numbering.

Version numbering is a red herring. We have much bigger fish to fry,
such as the lousy website, the bad documentation, and the lack of
built-in replication.

Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200212131008

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE9+fhlvJuQZxSWSsgRAjvXAJ9xxXt7mfsHWPSbcQ17nuBgggiVjQCgk4F6
E/SBhx6HwigCWxbnAjhHyPI=
=2DZH
-----END PGP SIGNATURE-----

On Fri, 2002-12-13 at 09:56, Merlin Moncure wrote:
> Most likely windows support will mean major rewrites of the code base,
> for example there is discussion in hackers about switching from process
> model to thread based model.

No, it won't. My understanding is that the Windows port will have a
minimal effect on how PostgreSQL runs on Windows (and it certainly won't
introduce threading on Unix, which is widely viewed as a bad idea).

Cheers,

Neil

Le Vendredi 13 Décembre 2002 15:56, Merlin Moncure a écrit :
> Do not underestimate how important windows compatibility is to the small
> /medium business world.

IMHO, this is not so important.

End-users are looking for are good performance, extensive features,
reliability and real information (not like MySQL which markets would-be
features before implementation. SAP-DB says they have 100 developpers
working, but in the real world only ... three of them commit code to CVS). DO
NOT BELIEVE users believe such crap. In the end, MySQL and SAP-DB will suffer
from these lies.

As for PostgreSQL :
- 7.4 means : look, we are reliable, we provide a real product, we will still
be here in 20 years... That's why you can choose us.
- 8.0 means : Hey man, look at my product, it's new... (but what will you do
when you reach numbre 15?)...

I vote for 7.4

On the converse, I am concerned by companies which rename PostgreSQL into some
stu... names. Example : RedHat dabase. Why not Nasa database, Corn Flakes
database or Hutu database? This destroys the image of PostgreSQL because
users are confused.

This is why pgAdmin II & III have a special open-source activist licence.
RedHat will never get pgAdmin and rename it to some stu... name.

OK, this was just my 2 cents.
Jean-Michel POURE

On Fri, 2002-12-13 at 10:03, greg@turnstep.com wrote:
> Version numbering is a red herring. We have much bigger fish to fry,
> such as the lousy website, the bad documentation, and the lack of
> built-in replication.

Exactly! Beside the fact that a lot of people (including much of our
current user base) find marketing-oriented versioning to be annoying,
the key point is that it's really not a good way to spend marketing
time: there are LOTS of more productive things for the advocacy group to
be working on.

Regarding documentation: can you suggest some specific areas for
improvement?

Cheers,

Neil

Well said...
how can I help?

> Version numbering is a red herring. We have much bigger fish to fry,
> such as the lousy website, the bad documentation, and the lack of
> built-in replication.

Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200212131008

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE9+fhlvJuQZxSWSsgRAjvXAJ9xxXt7mfsHWPSbcQ17nuBgggiVjQCgk4F6
E/SBhx6HwigCWxbnAjhHyPI=
=2DZH
-----END PGP SIGNATURE-----



---------------------------(end of broadcast)---------------------------
TIP 2: you can get off all lists at once with the unregister command
    (send "unregister YourEmailAddressHere" to majordomo@postgresql.org)

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1


> Regarding documentation: can you suggest some specific areas for
> improvement?

I am working on a rewrite of the idocs, which should help
immensely. The documentation is generally good, it is
just terrible to navigate and find anything. I also think
some more examples throughout the docs would be useful.

In my ideal world, you type in www.postgresql.org and see a search
box on the main page. You type in "pg_hba" and get instantly
redirected to a page about the pg_hba.conf file. You type in
"How do I change the value of a sequence?" and get directed
to a page with a quick summary of sequence functions, and links
to more details.

Greg Sabino Mullane greg@turnstep.com
PGP Key: 0x14964AC8 200212131039

-----BEGIN PGP SIGNATURE-----
Comment: http://www.turnstep.com/pgp.html

iD8DBQE9+f/JvJuQZxSWSsgRAmUeAKDqRMxuWWbf1bKKE6hStHEXa6he3QCfdNwC
NOeTF/uOfzh7Ny8pzsw4LUk=
=5eNe
-----END PGP SIGNATURE-----

How does the license stop anyone from marketing a renamed product?

I looked at the license for pgAdmin II
( http://pgadmin.postgresql.org/pgadmin2.php?ContentID=2 )
and it don't see that explicitly stated.

Is it a consequence of (a) anyone must reprint the pgAdmin Public License
plus (b) condition 5, that "The names and trademarks of the authors and
copyright holders must not be used in advertising or otherwise to promote
the sale, use or other dealing in this Software without specific, written
prior permission."?  So when you put those two together you get the
no-rename feature?

    douglas

Jean-Michel POURE wrote:

> [...]
>
>
>This is why pgAdmin II & III have a special open-source activist licence.
>RedHat will never get pgAdmin and rename it to some stu... name.
>
>OK, this was just my 2 cents.
>Jean-Michel POURE
>

http://pgadmin.postgresql.org/pgadmin2.php?ContentID=2

"8. The pgAdmin Development Team may end the licence agreement for any reason,
without notice and without prejudice. In this case the licencee only retains
the right to use the software for personal or internal use. Within 10 days of
notification, the licencee agrees to end distribution of the software or to
pay $10,000 (ten thousand American dollars) for each copy sold."

IMHO, here is how the license can be understood :

pgAdmin II & III can be downloaded freely, source + binary, on pgAdmin web
site. It cannot be included on a distribution (like RedHat) without prior
written permission. We did not want to go too far in the license definition,
hence we (pgAdmin Team = mostly Dave and other developpers - i.e. people who
write code and work on pgAdmin everyday) can decide on a case by case basis.

This license was driven mostly by RedHat renaming PostgreSQL into RedHat
database, a few days after the development of pgAdmin II started.

Recently, RedHat did a lot of harm to KDE3 in their 8.0 edition. Did you find
any Kpackage in RedHat 8.0? In a few months, people may have to pay to
downlad the software they wrote on their own!

What happened to PostgreSQL and KDE3 will never happen to pgAdmin. pgAdmin is
owned by a community of free developpers. Like other free software
initiatives, they are gifts to Humanity.

Cheers,
Jean-Michel POURE

Jean-Michel POURE wrote:
<snip>
> What happened to PostgreSQL and KDE3 will never happen to pgAdmin. pgAdmin is
> owned by a community of free developpers. Like other free software
> initiatives, they are gifts to Humanity.

Hear, Hear.


> Cheers,
> Jean-Michel POURE


--
"My grandfather once told me that there are two kinds of people: those
who work and those who take the credit. He told me to try to be in the
first group; there was less competition there."
- Indira Gandhi

On Sat, 14 Dec 2002, Justin Clift wrote:

> Jean-Michel POURE wrote:
> <snip>
> > What happened to PostgreSQL and KDE3 will never happen to pgAdmin. pgAdmin is
> > owned by a community of free developpers. Like other free software
> > initiatives, they are gifts to Humanity.
>
> Hear, Hear.

Just a comment ... there is a "benefit" to RedHat calling it RedHat
Database ... if/when they decide to switch to another DB, it won't have
near as much impact as GB dying had, since few outside of the community
with associate the two :)

Folks,

> This license was driven mostly by RedHat renaming PostgreSQL into
> RedHat
> database, a few days after the development of pgAdmin II started.
>
> Recently, RedHat did a lot of harm to KDE3 in their 8.0 edition. Did
> you find
> any Kpackage in RedHat 8.0? In a few months, people may have to pay
> to
> downlad the software they wrote on their own!

I don't agree with the Red-Hat-bashing sentiment expressed in this and
elsewhere.  Sure, RH did rename their PostgreSQL version *for branding
and marketing reasons*.  However, they are employing at least one
full-time PostgreSQL hacker, and their "value-add" administration tools
have, at this point, all been contributed to the community.  Red Hat,
regardless of their *marketing* decisions on what versions and
alterations of OSS software they distribute, have *always* made the
source available to their Linux products.

The same cannot be said of SuSE, Caldera, or TurboLinux, all of whom
have *proprietary* software and tools they use to support their
services business.  Yet because Red Hat is more successful re-branding
OSS than other distributors are with selling proprietary extentions,
Red Hat gets all the flack.

Personally, I use SuSE because I don't like BlueCurve and I feel that
SuSE does a better server distribution.  But I feel strongly that Red
Hat deserves recognition for maintaining their dedication to OSS, and
not a bunch of grumbling about "the Microsoft of Linux".

-Josh Berkus

On Sat, 14 Dec 2002, Josh Berkus wrote:

> Folks,
>
> > This license was driven mostly by RedHat renaming PostgreSQL into
> > RedHat
> > database, a few days after the development of pgAdmin II started.
> >
> > Recently, RedHat did a lot of harm to KDE3 in their 8.0 edition. Did
> > you find
> > any Kpackage in RedHat 8.0? In a few months, people may have to pay
> > to
> > downlad the software they wrote on their own!
>
> I don't agree with the Red-Hat-bashing sentiment expressed in this and
> elsewhere.  Sure, RH did rename their PostgreSQL version *for branding
> and marketing reasons*.  However, they are employing at least one
> full-time PostgreSQL hacker, and their "value-add" administration tools
> have, at this point, all been contributed to the community.  Red Hat,
> regardless of their *marketing* decisions on what versions and
> alterations of OSS software they distribute, have *always* made the
> source available to their Linux products.

I hope you weren't including me in that 'RH-bashing', as it wasn't my
intent ... I personally think that GB did alot more harm to PgSQL then
good, since they worked hard for a short period of time to promote
themselves as "PostgreSQL", and then flop'd ... RH, with their renaming,
is indirectly protecting us from 'shifts in the marketing wind', so that
if something does happen in the future that makes it politically
unfeasible to back PgSQL (the MySQL guys buy them out), ppl don't see "yet
another company go under" in relation to PgSQL ...

Le Samedi 14 Décembre 2002 21:21, Josh Berkus a écrit :
> I don't agree with the Red-Hat-bashing sentiment expressed in this and
> elsewhere.  Sure, RH did rename their PostgreSQL version *for branding
> and marketing reasons*.  However, they are employing at least one
> full-time PostgreSQL hacker, and their "value-add" administration tools
> have, at this point, all been contributed to the community.  Red Hat,
> regardless of their *marketing* decisions on what versions and
> alterations of OSS software they distribute, have *always* made the
> source available to their Linux products.

IMHO, RedHat is missing basic economical concepts :

1) The question asked is the notion of brand. What is brand, what is it for
and what does it represent?

* RedHat answer is "nothing", a brand is a "sticker on a box". This is far
away from PostgreSQL notion of excellence.

* People are looking for original brands, not copies. RedHat distribution is
an "original concept" because they invented rpm and worked together thousands
of software packages. On the converse, RedHat database is not an original
concept, but a pure copy of existing software.

* RedHat do not seem to understand that PostgreSQL is here to stay and
probably for a long time. Like the Rolling Stones, PostgreSQL developpers
will probably be here in 20 years. How long can RedHat stick to a silly name
like "RedHat database" ? IMHO, their invesment in the "RedHat database" brand
is a whaste of time and money, because it is deemed to disappear on the long
run.

2) Also, the question asked is the notion of "fairness", "written-unwritten"
rules and ultimately "violence" :

* Marketing a product which took XXXX man-year, only paying for one developper
salary, is a gift from the PostgreSQL community to RedHat. Employing a
PostgreSQL developper does not prevent them from being fair.

* The notion of fair/unfair is an unwritten rule. When you meet a Greatbridge
employee, who comes a long way by car, and annouce the same day that RedHat
will release a "RedHat database", without telling this Human Being that you
decided to do "Eye in the Eye", you also break the rule.

* Their attitude shows that RedHat management is probably violent. For those
who worked in severeal companies, you know what I mean : some companies are
violent, others are not.

* The problem with violence is that you never know when it is going to stop.
RedHat breaking the rules will probably end in a disaster for them if new
contrib software is released under the pgAdmin license. Candidates?

3) Ultimately, there is a chance that Redhat is not a violent company, but
simply have stupid marketing staff. There are some signs that RedHat database
marketing team is stupid :

* RedHat database is marketed like a food product, probably by
non-technicians, who employs "marketing recipies" without understanding the
underlying economical rules.

* There is a real problem with price. You cannot market an empty concept like
RedHat database at that price. When users understand they have been robbed,
this destroys the image of RedHat.

* RedHat does not have a real service offer (like database migration, software
engineering, etc...) and only concentrates on pushing (empty) boxes.

Cheers,
Jean-Michel

On Sun, 15 Dec 2002, Jean-Michel POURE wrote:

> Le Samedi 14 Décembre 2002 21:21, Josh Berkus a écrit :
> > I don't agree with the Red-Hat-bashing sentiment expressed in this and
> > elsewhere.  Sure, RH did rename their PostgreSQL version *for branding
> > and marketing reasons*.  However, they are employing at least one
> > full-time PostgreSQL hacker, and their "value-add" administration tools
> > have, at this point, all been contributed to the community.  Red Hat,
> > regardless of their *marketing* decisions on what versions and
> > alterations of OSS software they distribute, have *always* made the
> > source available to their Linux products.
>
> IMHO, RedHat is missing basic economical concepts :

[snip]

Red Hat employ. Tom and some other programmers who work on Postgres. They
don't appear to drag Tom around the world like a show pony. In fact, Red
Hat do not do that to any of the programmers they employ. They have to
make money from that and if that means selling software and services, so
be it. Anything Red Hat does in marketing the Red Hat Database will help
PostgreSQL -- it increases the market for open source databases and it
increases the people's exposure to PostgreSQL.

Lets talk about more important issues. Please.

Gavin

Hi all,

Has anyone been in contact with the people at Wild Open Source
(http://www.wildopensource.com) about the Enterprise Solutions Centre I
mentioned? If not, I will contact, get details and post to list.

Gavin

I"m starting to put together some materials for a media kit, and I realized
that I don't have most of the materials in my possession.

I need the following things the most:
 - Printable logos, or the URL of where to find them
 - Any advantages/features above and beyond the list on the advocacy site
 - A few more case studies, if there are any
 - A list of awards PostgreSQL has won
 - An overview of the PostgreSQL "organization" (kind of like the About
Company XYZ section you'd find in a corporate media kit)

If anybody can help me with any or all of these items, I'd appreciate it.

Thanks,
Lucas Alexander

On 14 Dec 2002 at 21:52, Justin Clift wrote:

> Jean-Michel POURE wrote:
> <snip>
> > What happened to PostgreSQL and KDE3 will never happen to pgAdmin. pgAdmin is
> > owned by a community of free developpers. Like other free software
> > initiatives, they are gifts to Humanity.
>
> Hear, Hear.

I differ in opinion here. KDE is a community project as well. Nothing stops
anyone to ship a modified version of PGAdmin as long as contributions are masde
open.

Of course, RedHat have good reason to customize KDE given the exposure that
project has. I don't think PGAdmin have exposure of that scale. For this
reason, it might continue to ship uncustomized.

Bye
 Shridhar

--
Cheit's Lament:    If you help a friend in need, he is sure to remember you--    the
next time he's in need.

Le Dimanche 15 Décembre 2002 23:14, Gavin Sherry a écrit :
> Red Hat employ. Tom and some other programmers who work on Postgres. They
> don't appear to drag Tom around the world like a show pony. In fact, Red
> Hat do not do that to any of the programmers they employ. They have to
> make money from that and if that means selling software and services, so
> be it. Anything Red Hat does in marketing the Red Hat Database will help
> PostgreSQL -- it increases the market for open source databases and it
> increases the people's exposure to PostgreSQL.

Dear Gavin,

I agree with you as regards employment and RedHat marketing efforts. My only
concerns are that :
1) Changing PostgreSQL name to RedHat database is unfair and useless.
2) RedHat may close RedHat database sources in a near future.

Kpackage dissapeared from RedHat 8.0. People will soon have to pay to download
RPM packages.

As regards the PostgreSQL license, do you think it is ***technicaly***
possible to close RedHat database sources? I don't know the PostgreSQL
license enough to answer the question. Maybe you know the answer, can it be
done by law?

Cheers,
Jean-Michel

On Mon, 16 Dec 2002, Jean-Michel POURE wrote:

> Le Dimanche 15 Décembre 2002 23:14, Gavin Sherry a écrit :
> > Red Hat employ. Tom and some other programmers who work on Postgres. They
> > don't appear to drag Tom around the world like a show pony. In fact, Red
> > Hat do not do that to any of the programmers they employ. They have to
> > make money from that and if that means selling software and services, so
> > be it. Anything Red Hat does in marketing the Red Hat Database will help
> > PostgreSQL -- it increases the market for open source databases and it
> > increases the people's exposure to PostgreSQL.
>
> Dear Gavin,
>
> I agree with you as regards employment and RedHat marketing efforts. My only
> concerns are that :
> 1) Changing PostgreSQL name to RedHat database is unfair and useless.

I disagree. Under the terms of the license they can change the
name. Presumably, they don't think it is a useless idea. Its their
problem. It does not have a detrimental effect on PostgreSQL.

> 2) RedHat may close RedHat database sources in a near future.

So? They're allowed to. The license PostgreSQL was *given* (I want to make
that very clear: 'given', 'bestowed upon', 'without which it would not
exist as you know it') allows Red Hat to do so.

More importantly, you have no proof that they will. Red Hat have a
commitment to open source. None (!) of their revenue streams / business
models use closed source code. When you speculate, without evidence, that
someone or something is going to do something, what you are saying is
slanderous. That is against the law.

Besides, their are several products which already ship PostgreSQL closed
source.


> As regards the PostgreSQL license, do you think it is ***technicaly***
> possible to close RedHat database sources? I don't know the PostgreSQL
> license enough to answer the question. Maybe you know the answer, can it be
> done by law?

As above.

Can we *please* focus on the issue at hand: PostgreSQL advocacy. So far,
lots of people seem to prefer to attack others -- whether they be MySQL AB
or Red Hat. This is a sign of immaturity and lack of experience.

While we're sitting around discussing political matters, people like Tom,
Bruce, Peter, Tatsuo, Neil, Joe, Lamar, Christopher (along with all the
other developers, documentors and translators) are putting in serious man
hours, doing work which *directly* benefits PostgreSQL and its users. And
they get a lot of work done. You only see these people talking about
other databases or products when planning how to implement features better
than they have, or when they're preparing migration documentation.

LETS GET ON WITH THE JOB.

Gavin

Le Lundi 16 Décembre 2002 10:44, Gavin Sherry a écrit :
> LETS GET ON WITH THE JOB.

Your opinion is that RedHat will not close RedHat database souces. On the
converse, I think they will.

Please consider RedHat 8.0 software installer, KDE 3.0 fork, etc... These are
obvious moves towards reducing access to source code. RedHat wants to please
the financial community and is driving away from its roots.

Sorry Gavin, I have some business experience founding several companies. My
companies are still alive. I have a huge respect for PostgreSQL hackers work
and skills.

Here is the opinion of a 32 year-old guy who got screewed several times before
he actually understood the power of greed in business :

MAKE SURE TO SETTLE PROBLEMS BEFORE THEY ACTUALLY ARRIVE.

To avoid PostgreSQL image being broken in the future, some PostgreSQL new
developements should be released under ***activist*** licenses comparable to
pgAdmin license.

It would be a win-win situation :
- RedHat benefits from PostgreSQL software whever they keep PostgreSQL name,
- PostgreSQL group benefits from RedHat support as regards employment.

I see no conflict. Otherwize, a conflict may appear whenever PostgreSQL
becomes really successfull. In business, this is usually the case :
everything is OK when you don't earn money, people fight when money is at
stake.

Cheers,
Jean-Michel POURE

Jean-Michel, you seem to have some GPL vs. BSD issues you need to
resolve *within yourself*.  PostgreSQL is BSD licensed, and AFAIK
everyone on the core group is happy with that decision. If you don't
like the side effects of what a BSD licensed project entails, you need
to either find a different project, or start lobbying the core group to
change the license; either way the discussion is moot here on
pgsql-advocacy.

Robert Treat


On Mon, 2002-12-16 at 06:17, Jean-Michel POURE wrote:
> Le Lundi 16 Décembre 2002 10:44, Gavin Sherry a écrit :
> > LETS GET ON WITH THE JOB.
>
> Your opinion is that RedHat will not close RedHat database souces. On the
> converse, I think they will.
>
> Please consider RedHat 8.0 software installer, KDE 3.0 fork, etc... These are
> obvious moves towards reducing access to source code. RedHat wants to please
> the financial community and is driving away from its roots.
>
> Sorry Gavin, I have some business experience founding several companies. My
> companies are still alive. I have a huge respect for PostgreSQL hackers work
> and skills.
>
> Here is the opinion of a 32 year-old guy who got screewed several times before
> he actually understood the power of greed in business :
>
> MAKE SURE TO SETTLE PROBLEMS BEFORE THEY ACTUALLY ARRIVE.
>
> To avoid PostgreSQL image being broken in the future, some PostgreSQL new
> developements should be released under ***activist*** licenses comparable to
> pgAdmin license.
>
> It would be a win-win situation :
> - RedHat benefits from PostgreSQL software whever they keep PostgreSQL name,
> - PostgreSQL group benefits from RedHat support as regards employment.
>
> I see no conflict. Otherwize, a conflict may appear whenever PostgreSQL
> becomes really successfull. In business, this is usually the case :
> everything is OK when you don't earn money, people fight when money is at
> stake.
>
> Cheers,
> Jean-Michel POURE
>
> ---------------------------(end of broadcast)---------------------------
> TIP 5: Have you checked our extensive FAQ?
>
> http://www.postgresql.org/users-lounge/docs/faq.html

On Mon, 16 Dec 2002, Jean-Michel POURE wrote:

> As regards the PostgreSQL license, do you think it is ***technicaly***
> possible to close RedHat database sources? I don't know the PostgreSQL
> license enough to answer the question. Maybe you know the answer, can it
> be done by law?

More to the point ... who cares?  As much as I respect Tom Lane, he is but
one person amongst many developers ... so even if Redhat were to close
*its* database sources, and take Tom away from us, albeit a sad day, there
are others that would fill the gap ...

Personally, I think the bad karma resulting from that would cause them to
think twice very quickly, but hey, its been known to happen ...

On Mon, 16 Dec 2002, Jean-Michel POURE wrote:

> MAKE SURE TO SETTLE PROBLEMS BEFORE THEY ACTUALLY ARRIVE.

'K, what problem?  That RedHat closed RedHat Database, something that
those they are marketing to most likely don't even know is PostgreSQL, so
won't even put two and two together, or really care?

Have to agree with Gavin ... don't we have better things to do with our
time t hen worry about a problem that has a small chance of ever being an
issue?

Le Lundi 16 Décembre 2002 14:38, Robert Treat a écrit :
> Jean-Michel, you seem to have some GPL vs. BSD issues you need to
> resolve *within yourself*.  PostgreSQL is BSD licensed, and AFAIK
> everyone on the core group is happy with that decision. If you don't
> like the side effects of what a BSD licensed project entails, you need
> to either find a different project, or start lobbying the core group to
> change the license; either way the discussion is moot here on
> pgsql-advocacy.

OK, I understand there are more important issues, so I stop here : new
developments are welcome under a license similar to pgAdmin.

The BSD license cannot be changed because it requires the agreement of all
contributors (at least in European law), which is impossible.

/End

On Mon, 2002-12-16 at 10:06, Jean-Michel POURE wrote:
>
> OK, I understand there are more important issues, so I stop here : new
> developments are welcome under a license similar to pgAdmin.
>

As far as advocacy is concerned, new developments are welcome under any
license.

Robert Treat

Jean-Michel POURE wrote:
> Your opinion is that RedHat will not close RedHat database souces. On the
> converse, I think they will.
>
> Please consider RedHat 8.0 software installer, KDE 3.0 fork, etc... These are
> obvious moves towards reducing access to source code.

Every piece of software on my RedHat 8.0 CD comes with source code. No
exceptions. The RedHat 8.0 software installer is GPLed. Its home page
lives here: <http://rhlinux.redhat.com/anaconda/>. The source code is a
link away.

The KDE 3.0 fork...maybe you don't understand one of the critical points
of Free Software, so I'll help you out. Anyone can change anything about
the code that they don't like. That goes for distributors and end users.
RedHat has chosen to exercise that right. If you don't like it, you can
avoid using RedHat or exercise your own right to change it back. You can
argue about whether or not it's a change for the better, but they
unquestionably have the right to do it; the license is quite clear on
the subject. They distribute the patches they used. They are in full
compliance with the GPL.

There is no basis for saying that any of these is a move toward reducing
access to source code, much less an "obvious" one.

I agree with Gavin Sherry; what you are saying is slander. Please take
it elsewhere.

I'm grateful to RedHat for the time they've put into PostgreSQL, and
they can call it whatever they damn well please.

Scott

Scott Lamb wrote:
> I'm grateful to RedHat for the time they've put into PostgreSQL, and
> they can call it whatever they damn well please.

I agree that they can, and it is their decision.

But by doing so their efforts at promoting their database do nothing
to help spread the adoption of PostgreSQL.  That is, they do nothing
to improve the mindshare of PostgreSQL.  In fact, their efforts
*hinder* PostgreSQL's mindshare.

The probability that something will be chosen for new projects is
almost directly related to that something's popularity, so by
marketing PostgreSQL under a different name RedHat is making it more
difficult for PostgreSQL to gain popularity: they're essentially
causing PostgreSQL to compete against itself for customers.

It would be exactly like RedHat naming their distribution "RedHat OS"
instead of "Redhat Linux".  Many people who purchased "RedHat OS"
would have no idea that it's Linux under the hood, and as a result
they would not be able to recommend Linux as a solution to others even
if they were ecstatic about "RedHat OS" -- because they wouldn't know
that "RedHat OS" *is* Linux.

I would be much happier if they called RHDB something like "RedHat
PostgreSQL Server", just like they call their distribution "RedHat
Linux".


--
Kevin Brown                          kevin@sysexperts.com

On Mon, 16 Dec 2002, Kevin Brown wrote:

> Scott Lamb wrote:
> > I'm grateful to RedHat for the time they've put into PostgreSQL, and
> > they can call it whatever they damn well please.
>
> I agree that they can, and it is their decision.
>
> But by doing so their efforts at promoting their database do nothing
> to help spread the adoption of PostgreSQL.  That is, they do nothing
> to improve the mindshare of PostgreSQL.  In fact, their efforts
> *hinder* PostgreSQL's mindshare.

How?  Do they have a team of non-paid developesr working on their code?

> I would be much happier if they called RHDB something like "RedHat
> PostgreSQL Server", just like they call their distribution "RedHat
> Linux".

And if, at some point in the future, they drop PostgreSQL and take on
MySQL?  Or Oracle?  I'd rather have that kinda hidden behind the name
myself ... that kinda thing would do more damage then not naming it RedHat
PostgreSQL in the first place, no? :)

BTW, can someone explain to me how this whole threads helps us in any way?
Just cause you don't like the name, they aren't going to change it ... so
wouldn't that energy be better spent in other aresa/

Marc G. Fournier wrote:
> On Mon, 16 Dec 2002, Kevin Brown wrote:
> > But by doing so their efforts at promoting their database do nothing
> > to help spread the adoption of PostgreSQL.  That is, they do nothing
> > to improve the mindshare of PostgreSQL.  In fact, their efforts
> > *hinder* PostgreSQL's mindshare.
>
> How?  Do they have a team of non-paid developesr working on their
> code?

That has nothing to do with it.

You *have* to think of this from the perspective of the customer,
because in marketing it's the *only* perspective that counts.

If the customer acquires RHDB, they'll think they've got RHDB.  Only
if they bother to do a bit of digging will they find that they
actually have PostgreSQL.

Now suppose that a friend of theirs comes to them for a database
engine recommendation.  The customer, being very happy with RHDB, will
recommend it to their friend without reservations.  And so their
friend acquires RHDB itself.  The userbase of RHDB grows over time
through this.

But relatively few of these people know that RHDB is actually
PostgreSQL, so the mindshare of PostgreSQL does not benefit from the
mindshare of RHDB.  Similarly, the mindshare of RHDB does not benefit
from the mindshare of PostgreSQL.  Because they are named differently,
uninformed people who are evaluating database engines *will treat them
as separate, competing products*.

And mindshare is one of the factors (perhaps the biggest one) many
people use to decide which product to use.  And they're not entirely
wrong for doing so, either, because the mindshare a product has is
directly related to how well supported (in terms of availability of
tools, knowledge, etc.) that product is.

> > I would be much happier if they called RHDB something like "RedHat
> > PostgreSQL Server", just like they call their distribution "RedHat
> > Linux".
>
> And if, at some point in the future, they drop PostgreSQL and take on
> MySQL?  Or Oracle?  I'd rather have that kinda hidden behind the name
> myself ... that kinda thing would do more damage then not naming it RedHat
> PostgreSQL in the first place, no? :)

Oh, I'll not argue that naming it RHDB has certain benefits for
RedHat!  But that's not exactly the point.  I'm looking at this from
the standpoint of what's good for PostgreSQL, not necessarily what's
good for RedHat.  They're not necessarily the same thing, as this
thread illustrates.

And besides, "RedHat Oracle" has a nice ring to it, don't you think?
:-)

> BTW, can someone explain to me how this whole threads helps us in
> any way?  Just cause you don't like the name, they aren't going to
> change it ... so wouldn't that energy be better spent in other
> aresa/

It helps us because, by making this issue clear and the consequences
of product naming clear, we illustrate to those who advocate
PostgreSQL why they should discourage distributors from labelling
their PostgreSQL-derived database engine something that doesn't
include "PostgreSQL" in the name.

This is an advocacy list, after all, so marketing strategies and
information are what it's all about, yes?  :-)



--
Kevin Brown                          kevin@sysexperts.com

On Sun, 2002-12-15 at 18:27, Lucas Alexander wrote:
> I"m starting to put together some materials for a media kit, and I realized
> that I don't have most of the materials in my possession.
>
> I need the following things the most:
>  - Printable logos, or the URL of where to find them

http://www.pgsql.com/propaganda/

>  - Any advantages/features above and beyond the list on the advocacy site

without getting too technical how about:

* available on more than a dozen platforms.

>  - A few more case studies, if there are any

I've got one in the works, check it out at
http://advocacy.postgresql.org/casestudies/monsterlabs/

I'd like some feedback before we put up links to it.

>  - A list of awards PostgreSQL has won

Marc should have a list of these. We did win the 2002 Linux New Media
Award.

>  - An overview of the PostgreSQL "organization" (kind of like the About
> Company XYZ section you'd find in a corporate media kit)
>

Hmm... this is a tricky one. I'd say that the main leadership is whats
commonly referred to as the "core group" which includes Marc Fournier,
Tom Lane, Vadim Mikheev, Bruce Momjian, and Jan Wieck.  After that you
have highly active folks in different areas like Neil Conway, Peter
Eisentraut, and Joe Conway on the development side, and Justin Clift and
Josh Berkus on the advocacy side, though I don't think there is a strict
organization at that point, since a lot of people are involved in
multiple areas.

> If anybody can help me with any or all of these items, I'd appreciate it.
>

Hopefully this can get you started. Sorry for the slow reply.

Robert Treat

    Sorry for adding a new message to this thread that leads to nowhere...

> If the customer acquires RHDB, they'll think they've got RHDB.  Only
> if they bother to do a bit of digging will they find that they
> actually have PostgreSQL.
>

    Just having a look at RedHat's site, I found the RHDB product page [1],
where the word "PostgreSQL" appears 4 times.

    I followed the link that took me to the project's site [2] and there I
found "PostgreSQL" several more times and a banner [3] where under "Red
Hat Database 2.1" they have placed "Powered by PostgreSQL", and it seems
that it is also present on the product's box. They seem to present their
product always followed by sentences as "powered by PostgreSQL" and the
like.

    Reading RHDB manuals, the screenshots in [4] make me think that actualy
any user installing it is clearly adviced of the PostgreSQL presence.

    Anyway, PostgreSQL's license allows Red Hat to rename it to whatever
they want and they are not obliged do so clear mentions of PostgreSQL.
The fact that they mention it once and again for marketing purposes
should make us happy.


[1] http://www.redhat.com/software/database/
[2] http://sources.redhat.com/rhdb/
[3] http://sources.redhat.com/rhdb/images/rhdb2.png
[4]
http://www.redhat.com/docs/manuals/database/RHDB-2.1-Manual/install/guiinst.html


    Best regards,

--
Alberto Caso Palomino
Adaptia Soluciones Integrales
http://www.adaptia.net
alberto.caso@adaptia.net

Alberto Caso wrote:
>     Reading RHDB manuals, the screenshots in [4] make me think
> that actualy any user installing it is clearly adviced of the
> PostgreSQL presence.

Then I stand corrected, and don't have a problem with RedHat's
methods.


--
Kevin Brown                          kevin@sysexperts.com

> >  Reading RHDB manuals, the screenshots in [4] make me think
> > that actualy any user installing it is clearly adviced of the
> > PostgreSQL presence.

Hi guys,

Do you realize RedHat is a company? I own two companies myself and Business is
Business. Therefore, IMHO, RedHat attitude is not difficult to understand:

Take RHDB for example : the source code is GPL, but it receives only RedHat
employee contributions. I would be very surprised if non-RedHat employees had
access to their CVS. Of course, you can always send patches, but these
patches become RedHat propriety and are then incorporated into RHDB.

RedHat owns RHDB completely. At any moment, it deserves the right to change
RHDB license completely. RedHat has the right to do so.

Do you really think RedHat are a philantropic company? These guys have and
need to earn money to continue operation. On the converse, PostgreSQL group
are a philantropic organization, even if its members are making business.
That's a big difference.

Therefore, I would like to make things clear : RedHat attitude might evolve
tomorrow because of financial objectives. There are clear signs that RedHat
is preparing to close its distribution sofware (=RedHat own applications),
ask money to download packages, shrink existing KDE, etc... And this might
well be the case of RedHatDB.

And what if RedHat are bought by a large vendor during a hostile stock-market
offer? Do you think investors will say "OK, RHDB deserves to stay
open-source". They wron't bother with this kind of things...

The situation is : the gun is in pointed to the back of your head, but you
cannot see it because you are looking forward. "RedHat never killed me,
nothing serious man!". Of course, they never killed you necause "you are
still alive". Too simple? You are warned and I won't say more to avoid
bothering you all.

Cheers, Jean-Michel POURE

Jean-Michel POURE wrote:
>>> Reading RHDB manuals, the screenshots in [4] make me think
>>>that actualy any user installing it is clearly adviced of the
>>>PostgreSQL presence.
>>
>
> Hi guys,
>
> Do you realize RedHat is a company? I own two companies myself and Business is
> Business. Therefore, IMHO, RedHat attitude is not difficult to understand:
>
> Take RHDB for example : the source code is GPL, but it receives only RedHat
> employee contributions. I would be very surprised if non-RedHat employees had
> access to their CVS.

Jean-Michel,

You'd made it *really* clear that you don't like Red Hat's way of doing things,
but it *really* looks like you're arguing to make a point without checking the
truth of the things you're saying.

We've probably all done that accidentally now and again, but you're *really*
making a bad impression in this particular thread by doing it over and over.

The stuff you say in other topics of conversation is really knowledgable and
on track, so please don't reduce your noise-to-information ratio overall by
continuing trying to argue something, with a completely bad/incorrect set of
examples.  :-(

Anyway, your point about CVS has further information here:

http://sources.redhat.com/rhdb/cvs.html

Red Hat appear to have an internal development CVS and also one that external
people can access.


 > Of course, you can always send patches, but these
> patches become RedHat propriety and are then incorporated into RHDB.

Correct.  And Red Hat makes the source code for RHDB available to everyone,
and will *very* likely continue to do so.  They're trying to prove their
model of Open Source business, and closing the sources would almost
guarantee a rather large amount of end-user backlash.

> I won't say more to avoid bothering you all.

Regardless of whether time proves you correct or not, thank you.

Everyone understands what you're saying, and you don't need to push it further.

:-)

Regards and best wishes,

Justin Clift


> Cheers, Jean-Michel POURE
>
>
> ---------------------------(end of broadcast)---------------------------
> TIP 6: Have you searched our list archives?
>
> http://archives.postgresql.org


--
"My grandfather once told me that there are two kinds of people: those
who work and those who take the credit. He told me to try to be in the
first group; there was less competition there."
- Indira Gandhi