|Apr. 23rd, 2009 02:12 pm Not a good week so far|
This hasn't been a very good week so far. The fact that I have a head cold, sinus infection or really bad allergies is a part of it. But most of the problem is that my main task for the week, getting a build out the door by Wednesday, has been nothing but trouble.
On Monday, I got a late start with the build because I never received the notification that our part of the code was ready to go. Once I got started, it went fairly smoothly until I actually started compiling the code. Then I got a failure that seemed, at the time, inexpiable. After looking at a bunch of possible causes, I concluded that the code hadn't synced correctly.
On Tuesday, I started syncing the code again. But the network was acting slow. About the time the code got mostly synced, the network suffered a complete outage and left me with a code base of unknown quality. After the network restored, I tried to find a way to fix the sync without syncing everything over again and finally started compiling. Different failures. Write a quick batch file to delete everything, sync everything again and kick off the build.
On Wednesday, I discover that my batch file had died because the "clean everything" step had deleted the batch file. I started syncing the files again, but had a couple of missteps related to some code I thought I needed to pull in. Once I was finally pulling in the right code, the network -- or at least Perforce -- slowed down so the syncs took the rest of the day. I kicked off the compile and left for home.
Wednesday night, working from home, I got the same build failure from Monday. Finally, I tracked down a change that seems to only break my product. I patched that and kicked off a new build. By this time, it was almost 11:30, so I went to bed (robot_grrl is working the midnight shift this week, so she'd already left for work), taking the work laptop with me. About 12:30, one of the compiles needed assistance. My pager woke me up (as planned) and I got it going again. Then about 1:30 or 2:00 the other build had a new failure. I tried to sleep on it, but I couldn't. So about 2:30, I woke up and started working on the problem. By 4:20 or so, I gave up and sent an e-mail for help to my team and tried to get some sleep before I had to be up between 5:30 and 5:45 to take the kid to school.
This morning, I finally dragged myself out of bed about 5:45 or 5:50 and quickly checked my e-mail. One of the people in India had resolved my problem. (Actually there were a couple of e-mail exchanges before I took the kid to school to get a solution). I kicked off new builds and went to bed. When the builds were done, I tried loading one of them remotely and thinking it was OK, I started the automated testing. By the time I got up about 11:00 and checked, the tests had all failed. I did some debugging from home over lunch and a bit after noon, I headed into the office.
Once here, I determined that the build wasn't starting correctly. With some assistance, I identified that the problem was that I'd missed a patch that was needed -- due to incomplete patching instructions for this build. I've kicked off the updated builds, and am hoping that I'll somehow be able to get the builds out before I have to leave for the day.
Between all of this, I've missed my Tuesday karate class (too sick), will probably miss my class tonight (working late, still sick). I also missed my Weight Watchers meeting (I was at home when it started) and who knows what else. At least on Friday night my sister will visit, and on Saturday we have the Anime LA picnic to look forward to.Leave a comment