Thread: The Yellow Brick Road
Reassurances notwithstanding, recent developments in the PostgreSQL community still concern me. I understand Tom would like to postpone this thread. > We wanted to postpone the discussion until Great Bridge was out in > the open and could allow Rusty Friddell, their counsel, to answer > questions about his suggestions directly. (And just to defuse any > fears beforehand, there will be no license changes without full > discussion and consensus from the pghackers community. This > decision is not core's to make, but the community's.) - Tom Lane I'm certainly interested in what Rusty has to say. But please don't ask the PostgreSQL community to stop discussing this issue until Great Bridge speaks. Sorry to sound cynical and jaded, but it seems an ominous portent that we should be asked to keep our mouths shut until Daddy Warbucks has his say. --- > Yes. BSD-style licensing is clearly more acceptable to businesses > than GPL-style, as the Postgres community understood all along. I > think GB's choice of Postgres as the database they wanted to work > with is not unrelated to that. - Tom Lane This is not clear at all. As evidenced by what? A more appropriate question may be: what side of the business equation are you talking about, the buyer or the seller? Great Bridge has indicated their intent to keep _all_ source they develop completely open: > We have no interest in any kind of proprietary fork. As far as code > goes, everything we write will go straight back into the open source > stew, for proper review by the Committed. - Ned Lilly I mean no offense to Ned, but while this statement sounds very reassuring, and I'm sure he's sincere, as far as the law is concerned, it has no legally binding significance whatsoever. That is what licenses and copyrights are for. So my question is: if you really mean what you say, why don't you release PostgreSQL under the GPL? The situation at hand is exactly the type of situation the GPL is intended to address - namely, to provide assurance to the community at large that nobody obtains proprietary ownership of source code. Because a BSD-style license is more acceptable to business? The only way I can see that a BSD-style license is more acceptable to business, is if that business wants to reserve the right to obtain proprietary ownership by simply extending the code. What other advantage is there? --- I'm also concerned about how recent developments may affect the PostgreSQL team financially. Core developers especially, but other contributers as well. > One thing we have agreed to is that there must not be an unseemly > fraction of core members working for the same company. With six > people on core, probably about two working at the same company would > be a reasonable limit. - Tom Lane What happens if a small fraction of the PostgreSQL team become disproportionatly wealthy? When a couple of team members show up at the annual PostgreSQL barbeque in new Lexus SUV's, and the rest rattle in in rusty station wagons? When some team members kids go to good private schools, and others are stuck in underperforming public school districts? When the kids' college is paid for, vs. being indebted for the rest of your life? When health care isn't an issue, vs. becoming an omnipresent concern? When taking care of your elderly parents is easy, vs. not even being able to afford a visit? You get the point. I don't really expect anyone to speak openly about their financial situation. It's really a private matter. But I can't help supposing that should such a financial disparity arise withing the PostgreSQL team, that it would have (unpleasant) repercussions. --- I am extremely grateful to all those who have made PostgreSQL the wonderful program that it is. In thanks, I feel like I'm pouring cold water on your head. I also mean no disrespect to Landmark. The weather channel has lubricated many a conversation between me and my Grandmother. I wish only the best to the PostgreSQL team, and to Landmark and it's subsidiaries. But whatever you do, please don't upset the dynamic that made PostgreSQL what it is today. -Ron Peterson-
Ron Peterson wrote: >Reassurances notwithstanding, recent developments in the PostgreSQL >community still concern me. I understand Tom would like to postpone >this thread. > >> We wanted to postpone the discussion until Great Bridge was out in >> the open and could allow Rusty Friddell, their counsel, to answer >> questions about his suggestions directly. (And just to defuse any >> fears beforehand, there will be no license changes without full >> discussion and consensus from the pghackers community. This >> decision is not core's to make, but the community's.) > - Tom Lane > >I'm certainly interested in what Rusty has to say. But please don't ask >the PostgreSQL community to stop discussing this issue until Great >Bridge speaks. Sorry to sound cynical and jaded, but it seems an >ominous portent that we should be asked to keep our mouths shut until >Daddy Warbucks has his say. I didn't read Tom's words like that at all. He seems to be saying that discussion _had_ been delayed until the recent announcement; so that Great Bridge's counsel would be free to speak, if he had anything to say. Before the announcement, he would not have been free to say anything. There is no reason to hold back on discussion now the announcement has been made. ... >So my question is: if you really mean what you say, why don't you >release PostgreSQL under the GPL? The situation at hand is exactly the >type of situation the GPL is intended to address - namely, to provide >assurance to the community at large that nobody obtains proprietary >ownership of source code. The developers are not free to change the licence of the code inherited from UCB, though they could put their own additions under the GPL, which would have the effect you want. Any commercial body wanting to make a proprietary version would then have to start from the latest pure BSD release. I shouldn't worry, though. Even if the whole core team went renegade and sold out to the evil empire, a new team would take the latest free release and carry on from there. -- Oliver Elphick Oliver.Elphick@lfix.co.uk Isle of Wight http://www.lfix.co.uk/oliver PGP key from public servers; key ID 32B8FAA1 ======================================== "Search me, O God, and know my heart; try me, and know my thoughts. And see if there be any wicked way in me, and lead me in the way everlasting." Psalms 139:23,24
> Reassurances notwithstanding, recent developments in the PostgreSQL
> community still concern me. I understand Tom would like to postpone
> this thread.
Only regarding details of a proposed license change. You aren't asking
about that specifically, so here goes... ;)
Oh, to just touch on the issue: the GB-sponsored proposal is intended
to clarify the intent of the BSD license. It does not fundamentally
change it, and should be evaluated on whether it reinforces your
understanding of the existing license. If it doesn't, then the wording
will need to be changed. It will be short and sweet, and will include
all of the words of the current license.
But please, let's not get into specifics beyond that until we all see
it and have something detailed to discuss.
> I'm certainly interested in what Rusty has to say. But please don't ask
> the PostgreSQL community to stop discussing this issue until Great
> Bridge speaks. Sorry to sound cynical and jaded, but it seems an
> ominous portent that we should be asked to keep our mouths shut until
> Daddy Warbucks has his say.
That wasn't the intent, but I (and Tom, and everyone else) realize
that is how it might sound. Let's start talking...
> > Yes. BSD-style licensing is clearly more acceptable to businesses
> > than GPL-style, as the Postgres community understood all along. I
> > think GB's choice of Postgres as the database they wanted to work
> > with is not unrelated to that.
> This is not clear at all. As evidenced by what? A more appropriate
> question may be: what side of the business equation are you talking
> about, the buyer or the seller? Great Bridge has indicated their intent
> to keep _all_ source they develop completely open:
> > We have no interest in any kind of proprietary fork. As far as code
> > goes, everything we write will go straight back into the open source
> > stew, for proper review by the Committed.
> I mean no offense to Ned, but while this statement sounds very
> reassuring, and I'm sure he's sincere, as far as the law is concerned,
> it has no legally binding significance whatsoever. That is what
> licenses and copyrights are for.
Right.
> So my question is: if you really mean what you say, why don't you
> release PostgreSQL under the GPL? The situation at hand is exactly the
> type of situation the GPL is intended to address - namely, to provide
> assurance to the community at large that nobody obtains proprietary
> ownership of source code.
We have had extensive discussions of BSD vs GPL in the context of Open
Source software on the -hackers mailing list in the past (and I'm sure
on other lists too; I tend to live on -hackers so remember those). And
the consensus is that we could go around in circles forever (been
there, done that ;) on BSD vs GPL, since they both have strong points.
The tie-breaker is that Postgres was written at Berkeley, came with a
fully-formed BSD license, and so we will leave it as that in honor of
the UCB achievements and contributions (OK, I made up that last part).
> Because a BSD-style license is more acceptable to business? The only
> way I can see that a BSD-style license is more acceptable to business,
> is if that business wants to reserve the right to obtain proprietary
> ownership by simply extending the code. What other advantage is there?
An "advantage" mentioned by others is that the BSD-style license makes
it difficult to *inadvertently* violate the license, which means a
business can keep its nose clean even if an employee makes a mistake.
> I'm also concerned about how recent developments may affect the
> PostgreSQL team financially. Core developers especially, but other
> contributers as well.
Yes, that is a concern. As you might guess, the approach by Great
Bridge caused *lots* of discussion and opened new areas of possible
conflict and disagreement among the steering committee. And this will
continue to be a possible problem.
We are trying very hard to "Do the Right Thing" here, and that
includes balancing personal decisions and choices against continuing
involvement with Postgres as an avocation (used to be a hobby, but I
hope you'll agree that when one spends several hours a day for > 3
years it has gone beyond a hobby ;)
> > One thing we have agreed to is that there must not be an unseemly
> > fraction of core members working for the same company. With six
> > people on core, probably about two working at the same company would
> > be a reasonable limit.
> What happens if ...
All valid concerns. These will be continuing issues for each of us.
otoh we each have been free to do what we could with Postgres until
now, and it has been pointed out by others that there are *already*
people using Postgres and even building businesses around this tool.
So GB just brings that into focus by having a "big gun" jump into the
fray, and we'll have to see if they can follow through on their plans
to build a successful business.
> I am extremely grateful to all those who have made PostgreSQL the
> wonderful program that it is. In thanks, I feel like I'm pouring cold
> water on your head.
Thanks. And we need a shower every once in a while ;)
> I also mean no disrespect to Landmark. The weather channel has
> lubricated many a conversation between me and my Grandmother.
> I wish only the best to the PostgreSQL team, and to Landmark and it's
> subsidiaries.
> But whatever you do, please don't upset the dynamic that made PostgreSQL
> what it is today.
That last point is in fact the dominant topic in discussions among the
steering committee regarding the GB announcement. We *think* that
Postgres can withstand stresses that will come up, but the "dynamic"
is something we realize we have (slowly realized, to be sure) and are
not certain how we got it (I think a lucky coincidence of people,
timing, and Postgres code). We do not want it to be lost, and each of
us is committed to doing whatever it takes, perhaps -- but not
certainly -- short of hari-kari, to keep it.
Regards.
- Thomas
--
Thomas Lockhart lockhart@alumni.caltech.edu
South Pasadena, California
> I am extremely grateful to all those who have made PostgreSQL the
> wonderful program that it is. In thanks, I feel like I'm pouring cold
> water on your head.
>
> I also mean no disrespect to Landmark. The weather channel has
> lubricated many a conversation between me and my Grandmother.
>
> I wish only the best to the PostgreSQL team, and to Landmark and it's
> subsidiaries.
>
> But whatever you do, please don't upset the dynamic that made PostgreSQL
> what it is today.
About the only major contribution I can make here is to say that the
core team is very concerned about keeping things exactly the same. Our
San Francisco meeting was primarily to explore those issues with
Landmark. We want only the best for PostgreSQL. We are doing our best
to make sure that the addition of Landmark only _helps_ PostgreSQL.
Landmark has the same goal, so we think this is attainable.
As far as GPL vs. BSD, we have already gone over that in the past. As
an example, the nature of our code is that if a company develops a data
type and wants to distribute it to some customers, they are required to
freely distribute that data type to everyone, even if it took them
months to write it. We generally feel that is too much of a burden, and
hampers use of PostgreSQL by companies that want to distribute versions
of PostgreSQL with such enhancements. Now, we don't think Great Bridge
or any one else wants to do that, but we honor their right to do it.
As far as getting rich, well, that is no one's plan. We do feel that
Great Bridge will allow some developers to spend more time on
PostgreSQL, and properly supervised, that is a good thing for the
project as a whole.
Someone already mentioned that if I get on Letterman with my PostgreSQL
book, he is going to be upset, so I guess I am going to have to mention
the major PostgreSQL developers if I ever get on the show. Doesn't
leave much time to talk to Dave. :-)
There is one other issue that seems obvious to me. PostgreSQL really
doesn't have many options with any company wanting to commercially
support PostgreSQL. We can be hostile, but that doesn't seem good.
I franky was afraid some IPO-spin company would come in, do a terrible
job supporting PostgreSQL, give us a bad name, and leave. I am pleased
Landmark is the one wanting to do this, because I see them in for the
long haul (25M is a lot for a startup). I know they will do they best
job they can. I have a book here that is their corporate guidebook,
"Built to Last", by Collins and Porras, and impressed by its contents.
Even the Landmark website has an interesting "Core Values" section:
http://www.landmarkcom.com/culture/core.html
So, basically, yes, this is a challenge, and the best we can do is to
try to make it positive for PostgreSQL. I believe this can be done.
--
Bruce Momjian | http://www.op.net/~candle
pgman@candle.pha.pa.us | (610) 853-3000
+ If your life is a hard drive, | 830 Blythe Avenue
+ Christ can be your backup. | Drexel Hill, Pennsylvania 19026
People may be wondering what happened to the licensing discussion. The answer is that we are either going to transfer the license to a PostgreSQL foundation, if one is ever created, or update our BSD license to match the more modern BSD license used by FreeBSD. > Reassurances notwithstanding, recent developments in the PostgreSQL > community still concern me. I understand Tom would like to postpone > this thread. > > > We wanted to postpone the discussion until Great Bridge was out in > > the open and could allow Rusty Friddell, their counsel, to answer > > questions about his suggestions directly. (And just to defuse any > > fears beforehand, there will be no license changes without full > > discussion and consensus from the pghackers community. This > > decision is not core's to make, but the community's.) > - Tom Lane > > I'm certainly interested in what Rusty has to say. But please don't ask > the PostgreSQL community to stop discussing this issue until Great > Bridge speaks. Sorry to sound cynical and jaded, but it seems an > ominous portent that we should be asked to keep our mouths shut until > Daddy Warbucks has his say. > > --- > > > Yes. BSD-style licensing is clearly more acceptable to businesses > > than GPL-style, as the Postgres community understood all along. I > > think GB's choice of Postgres as the database they wanted to work > > with is not unrelated to that. > - Tom Lane > > This is not clear at all. As evidenced by what? A more appropriate > question may be: what side of the business equation are you talking > about, the buyer or the seller? Great Bridge has indicated their intent > to keep _all_ source they develop completely open: > > > We have no interest in any kind of proprietary fork. As far as code > > goes, everything we write will go straight back into the open source > > stew, for proper review by the Committed. > - Ned Lilly > > I mean no offense to Ned, but while this statement sounds very > reassuring, and I'm sure he's sincere, as far as the law is concerned, > it has no legally binding significance whatsoever. That is what > licenses and copyrights are for. > > So my question is: if you really mean what you say, why don't you > release PostgreSQL under the GPL? The situation at hand is exactly the > type of situation the GPL is intended to address - namely, to provide > assurance to the community at large that nobody obtains proprietary > ownership of source code. > > Because a BSD-style license is more acceptable to business? The only > way I can see that a BSD-style license is more acceptable to business, > is if that business wants to reserve the right to obtain proprietary > ownership by simply extending the code. What other advantage is there? > > --- > > I'm also concerned about how recent developments may affect the > PostgreSQL team financially. Core developers especially, but other > contributers as well. > > > One thing we have agreed to is that there must not be an unseemly > > fraction of core members working for the same company. With six > > people on core, probably about two working at the same company would > > be a reasonable limit. > - Tom Lane > > What happens if a small fraction of the PostgreSQL team become > disproportionatly wealthy? When a couple of team members show up at the > annual PostgreSQL barbeque in new Lexus SUV's, and the rest rattle in in > rusty station wagons? When some team members kids go to good private > schools, and others are stuck in underperforming public school > districts? When the kids' college is paid for, vs. being indebted for > the rest of your life? When health care isn't an issue, vs. becoming an > omnipresent concern? When taking care of your elderly parents is easy, > vs. not even being able to afford a visit? You get the point. > > I don't really expect anyone to speak openly about their financial > situation. It's really a private matter. But I can't help supposing > that should such a financial disparity arise withing the PostgreSQL > team, that it would have (unpleasant) repercussions. > > --- > > I am extremely grateful to all those who have made PostgreSQL the > wonderful program that it is. In thanks, I feel like I'm pouring cold > water on your head. > > I also mean no disrespect to Landmark. The weather channel has > lubricated many a conversation between me and my Grandmother. > > I wish only the best to the PostgreSQL team, and to Landmark and it's > subsidiaries. > > But whatever you do, please don't upset the dynamic that made PostgreSQL > what it is today. > > -Ron Peterson- > -- Bruce Momjian | http://candle.pha.pa.us pgman@candle.pha.pa.us | (610) 853-3000 + If your life is a hard drive, | 830 Blythe Avenue + Christ can be your backup. | Drexel Hill, Pennsylvania 19026