New Work

Jim Younger reported a user complaint that the Two Gyro script didn't handle decimal seconds. I modified the script to handle this properly. He also said that a new set of plots was being delivered and wanted to know how my script would hanndle them. I told him and advised him to come up with a consistent versioning scheme.

Then I started looking at Mike Wiggs' problems. I found out why the astroweb pages were not being excluded from the bad link check. (I had modified the output file insted of the curators script.) I then started looking at his problems with the metrics upload script. So far I haven't found anything.

Finished Two Gyro

Jim tested my revised Two Gyro script and pointed out a few problems. I fixed them and a few I had found on my own. I told him to test it again and if he finds no problems, we can put it in the pipeline. After finishing that I started looking at Mike Wigg's problems with the metrics script.

Two Gyro Script

I fixed the problem that Edwin reported with the new search scripton Friday. It turned out to be an easy fix. According to Mike, we're going to be able to move it to production this week. I then started debugging the modifications to the two gyro script. Most everything works, except I can't get it to retain values entered in the form when results are returned.

Between Two Moons

I found and fixed the problem Mike had located in the web search. It was happened because I was handling repeated xml tags incorrectly. Edwin discovered a problem that only occurs on a Mac. I will have to test that next week. In the afternoon I coded an upgrade to the Available Science Time and Orientation Tool that allows the plots to be displayed in a specified order and allows plots to be stored outside the Zope server.

The Final Agony

I mad a final fix to the web search code, tested it, and decided everything was done. I placed the Python files in CVS and created a list of files to move for Mike. Mike moved the files to testing. We had some back and forth, but no major problems until Mike used a test search string that caused a Unicode conversion error. I've started trying to track that down, but so far have had no success.

Bug Squashing

Mike reported a problem with selecting an alternate search catalog that required that I rework the entire logic of catalog selection. It mostly works now, but I'll have to handle the case of a missing catalog value correctly. Jim Younger reported a problem with the image conversion tool I had written. The tiff library had gone missing. I solved the problem by creating a hard link to the library under the name it was expecting.

Clean Up

I continue to work on the problems that Mike identified with the Web Search. I modified the text format and eliminated the supporting column. The latter introduced problems that I haven't been able to solve yet. I also started looking at the problem with the procedure catalog. In the afternoon I tested Leigh's conference registration code and reported the results.

Permissions and Formatting

Julia asked m to configure TWiki so that only specified users could access it. That's done. Mike asked for some formatting changes to the Web Search results. I'm in the middle of ding that.

Search Done

I've finished working on integrating the search engine into Zope. The last issue was allowing the code to set alternate databases. I did this by repurposing the "scope" variable already in the code. Adding sp_scope to a searchpoint now redirects all local searches to the database it points to.

Home Stretch

I solved the problem with local searches and turned to the last major problem: specialized search databases. modified the code and started debugging. The first problem I've encountered is my query returns a search form instead of an xml file. I'll have to figure that out tomorrow.