Thomas Lockhart <lockhart@alumni.caltech.edu> writes:
> The original sources, and the modified sources *that we know about*
> (someone could have and probably has taken the source code, modified
> it, and not contributed back the changes) are always fair game to be
> taken out of open source status.
The above could easily be misinterpreted. I believe the point Thomas
meant to make is that anyone is free to make a derivative version that
they choose not to release as open source. It does *not* mean that
someone can take away your right to use existing code that was already
released with a Berkeley-style license.
> The original copyrights are still valid and travel with the code.
> However, afaict they are designed to release UC Berkeley from
> liability and to preserve some credit for the original work, not to
> allow Berkeley to assert ownership control over derivative sources
> (into which category I think the current PostgreSQL tree falls, so to
> speak).
Right, and I think that it's past time that the Postgres group (in
the person of PG Inc, or some other entity if that's what a majority
want) explicitly make the same statements that UC Berkeley has made.
Anyone here want to be on the hook for liability when some big
company's database crashes? Not me...
> I think Marc is concerned that there be someone or something able to
> represent the current code tree, and to prevent hijacking of the
> PostgreSQL (and perhaps Postgres) names from this open source group.
That's actually quite a separate issue. Trademarking the name
"PostgreSQL" might be a good idea to prevent some random bozo from
claiming ownership of it. (In reality, I think any attempt by
someone else to register that name as a database trademark at this
point could easily be shot down, but it would be far cheaper to
register the mark pre-emptively than to file suit against someone's
predatory registration.) Yet ... on the other hand, if it were a
trademark then someone *could* buy it off PG Inc. If that happened,
we (the community) would still have the rights to use the Postgres code,
but we'd have to find another name for it ;-).
regards, tom lane
