Two weeks after WPX CW and I haven’t been in the shack. I tried to upload the log to LotW but there was a glitch somewhere. The log didn’t make it to the LotW database. I sent a mail to the ARRL and it just got fixed now. Speaking of logs #1: still on top of things in the UBA Prefix Hunt. I didn’t expect that since apart from the contests I haven’t done much DX’ing so far this year. Let’s hope WX and propagation will allow for some WARC fun this summer when I am at home. That prefix hunt game is fun and makes me work stations I’d otherwise ignore. Like a German on 40m. But if it’s a prefix I’ll call him now.
Speaking of logs #2: I finally got in the right state of mind to debug the UBA log checking code. It turned out to be a simple problem with a simple solution. The problem was an anomaly between CTY.DAT and Clublog’s CTY.XML. Some KV4 prefix results in “U.S. Virgin Islands” in one file and “VIRGIN IS.” In the other. Since I use both files and cross-feed information (long story), the strings did not match which in turn resulted in a function returning a zero length string which in turn screwed up the scoring algorithm.
I got it nailed down and fixed in half an hour so I could finish calculating and compiling the scores. I made some more changes to the code after that to make it stable and then imported all the CW logs. This went slow. My laptop had been processing Cabrillo files for almost seven hours and it had only parsed 650 out of 1100 logs. I took a look at the code and found a checking routine that took for ages and was more or less obsolete. I get smarter with each version HI. I rewrote this part. It was mostly removing some code and it worked fast as lightning after that. I’ve said it before and I’ll say it again: the code would run MUCH faster if everyone used a decent program to make Cabrillo and not mess with it afterwards. Then I could break down the Cabrillo by using fixed position of the elements of the QSO. Now I need to run through the QSO information one character at a time which of course takes forever.
After three years of digging through piles of UBA contest logs I have achieved pretty bullet proof code to get everything out of the Cabrillo when it’s in there. Indeed some logs lack important information. This year another new phenomenon showed up. A log where (only) four contacts out of the whole log began with “QSO;” in stead of “QSO:” per Cabrillo specs. A semicolon versus a regular colon. It always boils down to bad software (or ill configured) or people messing up the log by hand.
Last week I received a nice certificate from WRTC 2010 in Russia. Aah, WRTC – worth a dedicated rant here! Later…
And my special WPX software project has been put on hold seriously. In fact it has nothing to do with WPX but that contest was a test case. More on that later.