Fortunately, the July backup which I had thought was unusable turned out to be usable after all, and HTYP (along with a few other items) has been restored from that backup.
The hard drive was declared hopeless-for-now by Data Recovery Systems, whom I think I can recommend (despite not being able to recover the data) simply because they were very easy to deal with and have apparently done all necessary physical repairs to the drive at essentially no cost (other than parts), making it possible that I might be able to recover the data later on. The drive is supposedly on its way back here, where I'll take another look at it and see if it's recovered to a stage where the tools I have available might be able to do something with it.
We had a major hard drive failure on the server last week, which had apparently been preceded by a massive failure to back things up properly for quite some time.
I should know by 11/28 or so whether the contents of the hard drive can be recovered; I've sent it off to a professional drive recovery place.
Yes, we have proper backups now. I need to set up a test server to make sure that the backups are sufficient to reconstruct everything if this should happen again -- but I have verified that all databases are being properly dumped to a separate drive every night.
In the meantime, I'm going to proceed on the worst-case assumption and begin rebuilding the site from cached copies of pages and anything else I can find. All of the cached pages I have saved (but not yet added back to the site) are in http://htyp.org/rescued -- if you know of any content that should be rescued and hasn't been, please email me at email@example.com. Feel free to search Google's cache for the article:
- site:htyp.org keyword
To retrieve the cached copy, hover over the search result you want, and click on the grey right double-angle-bracket that appears off to the right side. You should see an image of the page appear -- possibly in its present form -- along with a "cached" link. If you click on that, you will get Google's cached copy.
More on this story as it develops.
- 2007-05-05 Finished and installed SpamFerret two days ago. It has already intercepted over 150 spam attempts across all open wikis, and only one or two new patterns have gotten through. I expect this ratio to improve over time. There is no noticeable lag in the saving of edits, even though it is checking 28 patterns before allowing each save.
- 2007-04-30 Currently fighting an onslaught of spam (about a dozen vandalized pages per day); the SpamBlacklist extension seems to have stopped working. I've decided it's time to write and deploy my own solution, SpamFerret.
- 2006-06-30 Finally got the shortened URLs working properly (so this page is htyp.org/Current_events instead of htyp.org/index.php/Current_events); Drupal's auto-configuration provided the clue (it comes with, or generates, its own .htaccess files, which showed the correct form for using mod_rewrite)