One Last Thing
I added a hidden field to my plain text exception, which stores the length of the text, so it can't be lost by a user updating the value of NAXIS1. I returned to looking at the mapping code in drizzle. Have plans to write a python function to create a mapping image once I can figure out the ... More
Build Process
I spent most of the day trying yo get drizzlepac to build, with a lot of help with Megan. It builds now, but still fails most of its tests. I'm also trying to identify the python calls it makes, so I can factor them out. More
Slight Progress
With Megan's help I got the drizzlepac tests to run. I also started writing the calc_pixmap python method, but ran into my lack of knowledge of numpy, so I started reading up on that. More
More Tests
I worked with Megan to get the regression tests gpr drizzlepac straightened out. Otherwise, stared at the wcs code. More
Testing Away
I finished writing calc_pixmap, which creates an image mapping the pixels of one image into another and then wrote the first test case, which I am trying to get to run. More
A Learning Opportunity
I spent the day learning the ins and outs of WCS by debugging my test case. More
First Test
I got the first test case to run. The problem was an ussue with which version of WCS to link to. More
Last Test
I finished writing and passing the tests for calc_pixmap and went onto figuring out how to interface it to the existing code. More
Rewriting
After some more study of the code, I began editing drizzlepac to use calc_pixmap instead of the current callback. So far the top level / interface methods are converted, still need to convert the code that does the actual mapping. More
Finished Rewrite
I finished modifying the drizzlepac code to use the new pixmap data. Am going to write tests, but first need to understand the C framework. More
This and That
I reviewed the code I had written yesterday and tried to build it. I ran into a few problems that are similar to the ones I had two weeks ago and will wait for Christine's help to resolve them, as she did two weeks ago. In the mean time I reviewed Mark's single sign on authentication code and sent ... More
Onward
I fixed the compile errors in the drizzlepac code and started writing unit tests. More
Building
I have finished the first set of unit tests and am trying to get them to build. I am having some issues with pandokia, since I seem to be the first person to use it to test C code. More
Linking
I got the library and its tests to build successfully, but found out that to run them I will need to write a Python wrapper that invokes them. More
Wrap that Rascal
I wrote the C and Python code to wrap the unit tests and the Python test case to invoke it. While it all compiles and links properly (I think) there are probems when running, it says that code was not installed properly. So I will need some help identifying those issues. More
Moar Testing
With Mark's help I identified the compiling and linking problems and was able to move forward. I am currently trying to identify problems and fix them to get the unit tests to pass. More
Tests Run
All the tests run successfully, but the test suite crashes with a segmentation violation when exiting, which is not good. I'm trying to figure out how to use gdb to solve the problem. Complicated by the fact I'm running the tests through python. More
gdb vs lldb
I spent the day trying to figure out the cause of my segmentation problem. It says it's a free of an uninitialized pointer. Wanted to debug the problem with gdb, but it seems like gdb is no longer supported on the Mac and I will need to use lldb. More
Not Quite Yet
I renamed the stsci_sphere package to spherical_astronomy pending its release. There are still some small problems that will need to be addressed before I can say it's done. I briefly looked at the segmentation violation problem in drizzlepac based on an early morning hunch. No success from the ... More
Found Bugs
In the morning I did some more sleuthing of the problems with spherical_geometry. In the afternoon I found the cause of my segmentation violation and moved in to write my first test case for check_over. More
Moving On
We tagged and released the spherical_geometry package to the wider world. I had to rewrite my tests for check_over when I found that you can only have one source file with C tests linked into the C library, because the test package uses global variables. More