Here's one for Google. If you use Camino to browse, and more relevantly if you store important bookmarks in Camino, you might want to know about a nasty little bug that could pop up sooner or later.
Camino appears to have a problem with its bookmarks file format. It saves out your entire list of bookmarks as an XML file,
bookmarks.plist, and there's a bug that causes this file to become corrupt, which looks like this:
After you hit the OK button, all your bookmarks — whether in the main menu or the bookmark bar — disappear from the browser.
All's not lost though. Camino attempts to save the file as
bookmarks-corrupted.plist, which is located in
~/Library/Application Support/Camino/ (where
~ is your user directory). This file should contain what you lost, just in a format that Camino can't read anymore.
It appears that Camino saves out this file every time you close the browser. Whatever the cause of the bug is, the result is that it saves this file as non-well-formed XML. Your bookmark data is intact, but the XML isn't valid, so Camino chokes on it and throws up an error.
There are two options at this point. One, you can debug the XML. Two, you can use the URLs within this file to manually reconstruct your bookmark list (though presumably you'll be smart enough to do this in another browser or on delicious, if you go this route).
I ended up debugging the XML. Through trial and error (deleting large swaths of bookmarks at a time, loading Camino to see if I got the error or not, and a lot of restoring the corrupted file) I was able to pinpoint two or three locations where the XML was invalid. In one case, the angle brackets were reversed. Instead of seeing this:
I noticed the first right-facing bracket was reversed:
In the second case I wasn't patient enough to track down the specific error, so I just deleted the 10 or so bookmarks that were causing the problem and reconstructed them manually.
The main take-away from this experience is that Camino users ought to back up their bookmarks on a regular basis, or store them in another browser until this bug is fixed.
Somehow, and I'm just guessing here, but somehow my email address ended up on a centralized mailing list for public relations firms representing the tech industry. All of a sudden I'm getting in the neighbourhood of 2 or 3 press releases directly emailed per day, all from different people at various PR agencies. This started around a week ago.
From what I can tell, these are legit emails from real humans; they're mostly of the product announcement/request for review variety. They're usually promoting software or web sites. They're personalized, and they don't really strike me as overly spammy. But they also don't appear to have any obvious opt-out mechanism.
I'm more curious than annoyed... for now. I'm sure this is just business-as-usual for PR types, though it's going to get old real quick. I've emailed back a few of them to find out if there's a central source. I'll report back here if I get an answer.
I received confirmation that my email address is indeed in a PR database called Cision. I'll try to get my name removed over the next few days, and update again if I manage to accomplish that. Meanwhile, the press releases keep on coming.
a thoughtful reaction from someone in the PR industry who clearly gets it.
Bits are flying out across the planet as you read this, updating name servers and propagating the happy news that no, in fact, csszengarden.com has not fallen victim to squatters, hackers, or the general underbelly of the web.
This morning I received initial contact from my registrar that the domain was up for renewal... in the form of an expiry notice. Then there were the dozens and dozens of other emails from all of you out there wherever you are. Thanks for your notes, questions, conspiracy theories, and general goodwill, but rest assured the domain has been renewed for another chunk of years, and soon whatever you're seeing will be replaced by whatever you're supposed to see, if that hasn't happened already.
Which leads me to my question — who's everyone using for domain registration these days? After years of increasingly questionable service I'm ready to bite the bullet and move everything to a serious business concern, instead of the current discount shop who appears ready to switch on the revenue generator search engine spam the minute I screw up.
I've made some minor updates to this site you might be interested in knowing about.
Gravatars appear to be back up again and coping, so I've switched over to using them instead of the kludgy manual solution I developed a few months back. Given the service's track record I haven't exactly raced to scrap the previous method just yet, but we'll see how this holds up over the next little while.
Since my redesign last year I've felt like the logo was missing something, an extra touch that would throw it over the top. I could never put my finger on it, until I came up with the little beauty you see now gracing the header. I know it was Picasso who said great artists steal, but I can't help but think this is a wholly unique idea.
Long time readers will remember a few years back I ran a links sidebar I called The Dailies. Due to the manual nature of updating it at the time, and the increasingly inaccurate name as my publishing schedule dwindled, I eventually killed it.
Now thanks to the power of del.icio.us I'm giving it another shot. Think of the legacy name as being a stab at continuity rather than accuracy, and we'll be just fine.
Two of these things are true. One is not. I'm sure you'll figure out which is which. (Update: if you said #2, give yourself a hand. See this Flickr group for context.)