title: Sun's Latest Brag link: https://onemoretech.wordpress.com/2007/11/14/suns-latest-brag/ author: lembobro description: post_id: 611 created: 2007/11/14 19:00:07 created_gmt: 2007/11/14 19:00:07 comment_status: open post_name: suns-latest-brag status: publish post_type: post

Sun's Latest Brag

I couldn’t help myself. After reading a recent entry on Symas’s blog, I forwarded the link to my bosses with the comment “Howard Chu chews up latest Sun DS brag ad”.

Yeah, yeah, I know their blog is shamelessly self-promoting, but having read Howard’s posts on various mail lists over the years (particularly over on the OpenLDAP Software List), I’ve found he’s usually dead-on in his assessments of this kind of stuff.

Of course they save the best for last, and my favorite part of the post is it’s closing paragraph:

Sun plugs away without refactoring and reengineering DSEE. We expect their progress to continue to be impeded by long-standing structural and/or coding complexity barriers mentioned elsewhere by Sun itself. This brag tape sounds good on the surface but the truth is something else entirely, just like their claims of being Open in this brave new world of open source. If Sun would ever come out from hiding behind their “do-not-publish-benchmark-results” EULA clause all of these empty claims would evaporate in the harsh light of reality. Fortunately the truth isn’t that hard to find. Increasing numbers of sophisticated directory-using enterprises are converting to OpenLDAP and we expect this trend to continue to accelerate.

Not much left for the imagination here. But then no one should expect engineers to be diplomats (unless they’re bucking for management). Apart from the pyrotechnics (and probably aided by them), the rest of the article makes some important points not only on Sun’s product in particular, but LDAP server design in general.

There’s a lot that Sun could do to blunt, or even eliminate, these criticisms. The first would be to get serious about refactoring their product instead of running off trying to “make history” with the first commercial pure Java LDAP server. The Netscape line of directory servers, despite my frequent criticisms of them on this blog, are worth saving. Although I think Red Hat has done an admirable job of refactoring the Netscape product they purchased (through their open source Fedora Directory Project), this is a huge market that Sun appears poised to neglect even further. The second would be to remove those ridiculous EULA restrictions on publishing benchmark results. If they don’t then it’s probable that someone else is going to eat their lunch.

Copyright 2004-2019 Phil Lembo