Jerry Spicklemire wrote:

As for the link below, Dr. G points out how effectively the ACS team leveraged the ready made tools built into AOL Server, and contrasts this with incorrect references to the Zope team's decisions to "build it all themselves". PG says they built their own Web Server, but it's actually Sam Rushing's "Medusa". Similarly, PG points out that the Zope team built their own RDBMS, but it's really Aaron Watters' "Gadfly".

Actually, what he said is that 'Zope built their own database management system', which probably refers to the ZODB, and could be considered a valid point. Personally, I think that Mr. Greenspun's greatest mischaracterization of Zope is a failure to even mention Zope's greatest assets, such as the integrated through-the-web management interface, and the direct URL-to-Object translation. It's interesting to note that at the end of the previous 'Strategy' Chapter: http://www.photo.net/building-community/strategy.adp , Mr Greenspun says (at the bottom of the page):

Final Words of Wisdom: Aim High

Set the strategy goals as high as possible. Here's an example minimum set:

- the site we're building should be more functional than Yahoo - more amazing than scorecard.org - have more nerd-appeal than Slashdot - have better customer service than Amazon.com - be more fun to work on than Transmeta

Why aim high? Projects only attract people who think that the goals are challenging. To have any chance of hiring people capable of achieving something good you must aim for something great. Transmeta, the Silicon Valley startup mentioned above, is actually a good example. The stated aim was to build a computer vastly faster than Intel's Pentium chip. This goal attracted many top-tier programmers, including Linus Torvalds, author of the Linux kernel. Did they achieve? No. Transmeta's chip was way late and way slow, a fraction of the speed of Intel's best. But it worked! If they had set the original goal lower, they'd never have attracted a team capable of building a working chip.

Think about the Space Shuttle. The goal was to get up into space reliably and then back down. The project attracted people capable of building a machine that normally does its job but sometimes explodes. If the Air Force had conceived the project, the goal would have been to build a machine capable of going into space and, once up there, killing people. That project would have attracted engineers capable of making a system that went up and down reliably but wasn't an effective weapon once in space.

Very few projects achieve all of their goals. So make sure that your partial success will be good enough by starting out with a broad canvas.

I think that Mr. Greenspun makes a very valid point here. I also think that regardless of ArsDigita's other goals as a company, that their goals for the ACS are set lower than Zope's. Thoughtfully, Michael Bernstein.