b0rked page on pgsql.com

Given that I'm considering a move from mysql to [postgres|interbase], this b0rkedness is not very encouraging!



christo

Well thats mainly cause some troll hasnt programmed the page right or setup postgres right. He is also using pg_exec, that command was depricated a while back

troll? Would't that be one of the postgres team? And it's been like that for ages and nobody has noticed... They want us to use their database, but they don't know how to use it themselves, let alone check their pages and fix errors.

Bit of a pointless rant, I know, but I'm feeling tetchy


christo

I know this was just a light-hearted jab but, AFAIK, while some of the people at pgsql.com contribute to Postgres development and the community, I don't believe it's the main postgres site; but is the main contact for commercial support. Try going to postgresql.org instead.



Did you already send them an email notifying them of the errors on their site?

-b

I'll have to admit, I have found the pgsql.com site disappointing, too. It has always seemed slow and undependable. A few weeks ago, they were down for literally several days, because of a hard drive crash, and they had some goofy looking "out of service" cartoon on the site, with a message about having to ship a hard drive to a server in... Panama!

Notice that if you nslookup the site, it is part of the Hub.org network. In fact, it is the same IP address as hub.org. Hub.org's site doesn't exactly inspire confidence either. It loads slowly, and looks like it comes from 1996. Notice if you click on About Us, there is a blurb at the top of the page saying

Referral Program
Off Shore Web Hosting in Panama
Sign for 12 Months, Pay for 10

Somehow (no offense intended, please), Panama just doesn't seem like the place for a dependable worldwide internet presence. I'm willing to bet that the whole bandwidth available to the country is the equivalent of one T3, or maybe an OC-3. Maybe I'm wrong, but still, you can't convince me that you can get the same internet reliability there as in some of the larger American or European countries.

While the main postgresql.org site does look a little more presentable, an nslookup will show that it is also being served by hub.org, in the same IP block. Meaning, I suppose, Panama, too. I suspect this is one of the primary reasons for the many complaints about unreliability and slow response of both postgresql.org and pgsql.com.

Now, I have been a PostgreSQL user for a couple years, and absolutely no doubts about the quality of PostgreSQL, and of the people involved. I'm sure the developers are too busy to worry about websites, and the website people are too busy to do much about making a more impressive web presence. But... this does send the wrong message to international corporate management people who might be considering PostgreSQL. I would say this is one area where the PostgreSQL project needs a lot of help.

If any PostgreSQL insiders are reading this, please understand that I am not pointing fingers, and that we here are very appreciative of the work done in PostgreSQL. So, I guess the only real question to ask is: how can we Devshed-ers help? Hmm....

(hope I haven't ruffled too many feathers...)

I'm sure sending a link to this thread to postgresql's advocacy email addresses would get this to insiders, or sending it to the appropriate pg mailing list, but as far as their hosting goes I, personally, can't do anything to help out as I don't have access to that type of bandwidth/hardware. Although if I did, the PG people would be the first I'd share that with as both the community and product are outstanding.

-b

Thought this might interest some.

http://www.varlena.com/GeneralBits/
Updates for the week of July 14.

PG, was chosen as the best database by Linux Journal, which is nice but the real reason I'm posting this is because in the post titled, " Opinions from O'Reilly's Open Source Conference" there are a few points regarding which areas some of the PG team will be working on to increase their advocacy efforts.

Specifically:


Hopefully these will address some of christo's misgivings regarding the pg site.

-b

nice feedback, bcyde - well found and well remembered. I was hoping it was more of an organisational thing rather than a representation of $competence within the pgsql camp.

I notice that the b0rked page I linkd to has since been fixed and reading those notes from the o'reilly conference is encouraging to say the least..


christo

Yeah I emailed one of the webmasters about that page when you brought it up. Forgot to let you know when he replied a little while ago.

Still I'm hoping the advocacy efforts go forward with a strong push as it's always nice to see a good product thrive. One of the other major things that people have problems with is not realizing that the default PG settings need to be tuned (its default config is setup for maximum compatibility instead of speed), which is why there are sometimes horrible discrepancies when running benchmarks. Modifying settings like shared memory have huge effects on performance, yet up until recently there were very few documents about how to tweak these settings in the postgres.conf. If you read further on that link I posted, there's a couple recently released articles about how to performance tune PG. I'm hoping they get even more in depth and more friendly to new users, but it's definitely a step in the right direction.

-b