development corrupts

Weird Browser Behavior

2010-05-21T06:18:49Z

(Or: when your focus listener interferes with your click listener) Demo.

Stumbled across it while I was working on the new update page, and it held me up a while. Eventually, I decided to use another approach.

Playing with the thought of submitting it as a bug, but it may be expected quirky behavior -- at least, it behaves the same way in both Firefox and Opera. Chrome, meanwhile, does not suffer from the problem, but that may be sheer accident: Chrome doesn't trigger the focus event when you use the mouse, only when you use the keyboard. So in Firefox/Opera, the click event doesn't work; in Chrome, it's the focus event that doesn't show up.

Browsers: they are strange animals.

PS. If you edit the URL to see previous versions of the fiddle, you'll see that I originally thought that the presence of images had something to do with it, ehehe.

comments

Issue while updating db

2010-04-04T18:19:54Z

I ran into this issue while updating my database on my dev environment:

Failed to read auto-increment value from storage engine

That happened when I was trying to insert something new into logproplist. When I checked logproplist, I saw that the propid had skipped from 31 to 255, for the latest property (community moderator thing). information_schema.AUTO_INCREMENT had 256 as its value. Now I have no idea what that means, or why it happened, but it was causing me trouble.

I deleted all references to the property in logprop2, deleted the property itself in logproplist, reset the auto increment for logproplist, and then reran the database update command. First two steps may have been unnecessary; I did them before I managed to google up the reset fix.

(In the process, I think I figured out that ALTER_TABLE modifies the information_schema somehow).

As far as I can tell, it's not something that everyone is going to go through (that is, it appears to be a bug in my MySQL server version, triggered by some funky other stuff), so I'm not posting in dw_dev to spread the word/fix. Just here in my journal so I don't forget :-)

comments

Snow Leopard network woes (problem connecting to another computer on the network)

2010-02-08T16:54:19Z

Okay, here's one thing that bothers me: occasionally, I stop being able to connect to my brother's computer on our home network. Let's call this computer "brother".

So I'll still see "brother" under the Shared section of the Finder sidebar, but double-clicking on it gives me a "Connection failed", and clicking "Connect as" on the upper right either has no visible effect, or gives a pop-up with the text "The server “brother” may not exist or it is unavailable at this time. Check the server name or IP address, check your network connection, and then try again."

In Console (Applications > Utilities > Terminal), there's this error message in the system log:
SharePointBrowser::handleOpenCallBack returned 64

Judicious googling tells me that indicates the remote server is down, which is not true! My laptop just thinks it is. No amount of flushing dnscache (whether by doing dscacheutil -flushcache, or killall -HUP mDNSResolver) could fix it. It's also not a network setup issue, because I've been able to connect to this computer before and if I restart, etc, I'll be able to connect to it again.

I just spent the evening digging into this, because it frustrates me so, and I think I've finally got some solutions that work well enough that I want to share (also note them down so I can find them when I next need them!):

( three things that worked for me under the cut )

And there! I haven't managed to google up *any* solution to this kind of issue (except for advice to restart, but again that's not always feasible), so I've tried to put down everything that I could, and maybe it can help someone else in the future.

Future self, if this happens to you again, please remember to look at this entry.

comments

Ghosts in the interests code

2010-01-09T13:17:19Z

Bah, the issues with interests are slowly driving me to distraction. Two requests about interests within hours of each other! I feel that those two... three? issues have been going on long enough, so I spent the afternoon digging into interests code.

No progress on any of the issues, unfortunately, and I'm left feeling that there are, indeed, ghosts in the interests code. Without a reliable method of replication, I can't even tell whether it's memcache or actual stored values that are the problem.

I have the feeling that when we figure out what's wrong, it will be a FLASH OF SUDDEN INSIGHT HOW COULD I HAVE MISSED IT. But today, I'm well and truly stumped past the point of unproductivity.

*shakes fist*

Will get you, yet.

comments

I have the specter of a commit and ditch hanging over my head

2009-11-02T19:22:28Z

But I cannot replicate on my 'hack. I am stumped!

comments