Friday, March 30, 2007 by Dave Winer.
Until a few days ago, I had two servers at a colo facility in Somerville; they were my original life-after-UserLand servers, started in 2003. One was used for weblogs.com and the other for everything else. Eventually weblogs.com migrated off the server onto a Linux machine at ServerMatrix in Dallas. Then when my business with Adam Curry was forming, in 2004, I rented two servers for our podcasting work -- on Sunday those servers are finally going away.
After this corner-turn I will go from seven servers to three. One of the remaining servers is running only SYO, which I must admit is suffering from lack of attention. But there is some small hope of reviving the project.
At the same time Manila sites are seriously under attack, I'm not up on the exact vector or what the company may be doing to deal with the threat. However, we are having a board meeting a week from today. Writing about the situation with Russo & Hale here apparently had a good effect. It's the old sunlight as disinfectant thing.
Meanwhile, the Manila server I set up at Harvard when I was there is back up, but only for a very short period of time. The fragility of this server is one of the things that's inspiring my interest in future-safe systems. Back in 2003, with the help of Jack Russo, representing UserLand, we transferred the RSS 2.0 spec to Harvard, for safe-keeping through the years, hopefully decades, maybe longer. I think now, in 2007, it needs one more little bit of help to turn the corner to be relatively safe for the next century or so (knock wood, praise Murphy).
I volunteered to convert the RSS 2.0 site from a dynamic Manila-hosted site to a static Apache-hosted site. I'm moving all the images and sample XML files into sub-folders of the static HTML pages. We will also offer the whole site as a zip archive so people can download it in total. I will also provide the Harvard webmaster with an htaccess file that redirects the old urls to the new ones. I'm trying to anticipate the nit-picking that such a change is likely to cause, but I'm hopeful that with the recent Kathy Sierra mess, people will be reluctant to get so personal about it. All I'm doing is helping my friends at Harvard deliver on their commitment to provide an unchanging home for a spec that a lot of people build on. I am being careful not to change one word of the spec. Yet I'm sure, as I write this, that some people will spin as if they were on a Sunday morning politics show, hoping to confuse good people into thinking that changes were made. That's just the context our work takes place in, there's nothing to be done about it, other than hope people understand.