Re: Who's a "Corporate Sponsor"?

Josh Berkus wrote:

>Folks,
>
>We've been laying out a new "Corporate Sponsors" page on WWW for the new web
>site. However, some debatable issues have arisen on who "qualifies" to be on
>the page. We've agreed, there, on listing:
>
>1) Any company that sponsors a PostgreSQL major contributor's time;
>2) Any company that has contributed a feature or significant add-in since
>2000;
>3) Any company which pays for or donates infrastructure resources for the
>project.
>
>
>
Fair enough. Lets expand these to the questions below.

>The issues that aren't clear are:
>1) do all mirrors get listed?
>
>
I think so. Why not? They are contributing.

>2) does documentation "count" as much as code?
>
>
Absolutely.

>3) do add-ins count if they are completely externally hosted?
>
>
Marc and others have made compelling arguments against this. However, I
think that we can address these arguments without preventing add-ins
from being listed. See below.

>4) If yes to (3), do they still count if they are not OSS?
>
>
Definitely not.

It seems to me that we need clear guidelines for the inclusion of
add-ins. I would suggest:

1) The add-in must provide database access, development, or
administration functionality, or otherwise extend the backend (for
example, by offering a replication solution).

This would allow for inclusion companies sponsoring PL's, replication
systems, admin tools, database drivers, object types, etc.

2) It must be open source or for documentation be open content. We can
now debate whether allowing commercial distribution should be a
requirement of listing ;-) We may want to list a number of approved
licenses. (For example, I don't think that if something is licensed
under terms similar to Qmail it should count either, but that is another
story.)

3) The project must list their core developers.

4) Only companies sponsoring the core developers can be listed.

The idea here is to require a substantial contribution, not just a "One
of our employees submitted a patch to PgAdmin III last year and it was
accepted."

5) The project may be required to list a minimum number of user
testimonials with verifiable contact information. This is designed to
limit the "We contributed a set of 5-dimensional geometric objects that
nobody uses, but we should still be listed."

>Since scrolling space on our web page is not exactly a scarce resource, I'm
>inclined to say "yes, yes, yes and no". It benefits *us* to list as many
>companies as possible, because it shows how widely used and supported
>PostgreSQL is to potential new users. So we have little incentive to be
>"stingy" with listings; for the same reason, I wouldn't suggest "expiring"
>companies unless they go out of business.
>
>
>
Might be a good idea to require companies to reapply from time to time,
though. It may help to give additional incentive to contribute. The
advocacy group could then from time to time nominate a "top 20"
contributors list. It might give some companies an economic boost if
they can use that in their marketing.

>The reason for the last "no" is that vendors of commercial software, no matter
>how closely tied to PostgreSQL, are not "contributing"; they are at best
>complimenting Postgres for mutual benefit. It also removes some "incentive"
>for companies in the "PostgreSQL space" to OSS their software, which of
>course we want them to do.
>
>
>However, before you give an opinion, you should be aware that under that set
>of rules, Elein's company Varlena LLC would not be listed with "corporate
>sponsors", despite providing the quite valuable "General Bits". While very
>useful, 100% PG-oriented, and free, GB is "all rights reserved" and hosted
>entirely at Varlena.com.
>
>I'm bringing this up not to pick on Elein -- especially as GB could change its
>status at any time per Elein's post: http://www.varlena.com/GeneralBits/ --
>but because this "borderline" situation is liable to arise again. We
>already had something exactly analogous with CommandPrompt's online
>publishing of Practical PostgreSQL; again, useful to the community and free,
>but externally hosted and not OSS. This issue would apply equally if, for
>example, EMS HiTech offered pgExporter under a free shareware license. It
>would be nice, useful, popular, but still not a "contribution".
>
>So, opinions?
>
>
>
Again, I think that this example is exactly the reason why we need a set
of clear policies, such as the development of a list of accepted
licenses for both software and documentation contributions. This avoids
disappointment and hard feelings :-)

Best Wishes,
Chris Travers
Metatron Technology Consulting