Week 12 (26 Nov.)

(Note: this post is 1 day late. Sorry for the inconvenience.)

Last week I was in ATHENS week, so I didn't work on the paper most of the time.

I finished re-writing the first half of the paper. I added all the new changes and the implementation details. But as of now, everything from subsection "libraries compared to GASAL2" to more or less the end of the paper is completely copy-pasted from the previous GASAL paper. This was what I saw, it's not my doing, but Nauman's. What I re-wrote was also completely copy-pasted from the previous paper, so at least this part is new and actually brings information on the work I did on GASAL2. But the rest is, to put it bluntly, auto-plagiarism. Hence, it's not publishable.

At the moment, there is no improvement in terms of performance for GASAL2. My work on GASAL2 consisted in making the whole library actually readable, extensible, and humanly accessible to work on. (It wasn't the case before: keeping 3 files of 1000 lines each is not what anyone would call "humanly accessible"). For the performance part of the paper, not only do I not have the authorization to run any measurements on the servers, but also they may not even make sense (why would you compare results from a 2017 high-end server with a 2013 GPU ?).

My changes are on a private copy on Overleaf, the link has been only disclosed to concerned people. I am using GitHub synchronization, but the GitHub repository I am using is private. Invitations can be dispatched on demand.

For the rest of the week, I will try to see what I can do or if I can start something else.