16
edits
(→Other) |
No edit summary |
||
Line 22: | Line 22: | ||
== F: Feeling fine == | == F: Feeling fine == | ||
* | *IRC/mailing lists: familiar with using IRC. have also used (subscribed and posted)the octave-maintainers list | ||
*Mercurial: new to using mercurial but I have the hang of it(thanks to Jordi) | |||
*Mediawiki: quite familiar with mediawiki including (slightly) advanced features such as templates | |||
*Average working knowledge of make and gcc. Haven't used gdb much. | |||
* | *I work much better in a group (of 2/3 than individually) and prefer to have regular (constructive) feedback. Since both of those are inherently a part of OSS I should be quite happy to continue contributing to Octave after GSoC during most of my free time. | ||
* | |||
== O: Only out of interest == | == O: Only out of interest == | ||
I guess I have covered most of the topics above. My two main gripes with octave are: | |||
*Not all of the image processing built-in functions are not supported including some of the more commonly used ones like the weiner filter | |||
* | *Octave is much slower than Matlab especially for non-vectorized code (which have many loops) | ||
* | |||
== P: Prerequisites == | == P: Prerequisites == | ||
My primary OS is linux (Arch linux distro if that matters). I also have access to Windows 7 professional. | |||
I should have access to the internet 24x7 unless I am on the road/travelling (which is a remote possibility). I have two computers (laptop + Desktop) and have complete access to them 24x7. | |||
== S: Self-assessment == | == S: Self-assessment == | ||
* | *I am quite capable of working alone for extended periods but I prefer working in small groups and periodic review/criticism. Also working alone for a long time tends to get boring. | ||
*I prefer to have a sort of holistic view of the work ahead but prefer to write code iteratively/incrementally with small improvements and testing between each step. I usually jot down my algorithm and functional blocks on paper before I start writing the code. Also usually I tend to prefer simple working algorithms(even if the have poor performance) initially and to later replace them with more complicated but efficient ones. | |||
* | *I don't mind throwing out chunks of my work if it doesn't fit with the overall project. I have worked in a student run lab for 5 years and we did on occasion have to chuck out weeks worth of work but what I learnt from it was that the effort is never wasted and that dumped code tends to find applications elsewhere later. | ||
* | |||
== Y: Your task == | == Y: Your task == | ||
I am interested in two tasks: | |||
* | *Expand the built-in image processing functions to support more/most of those already there in matlab. | ||
* | *Work on JIT compilation and expand it to work with matrices/other data types than scalars. | ||
I am quite sure that I will have enough time for only one of the two and am leaning more towards JIT since that will be more useful in general. | |||
<!--[[Category:Summer of Code 2012]]--> | <!--[[Category:Summer of Code 2012]]--> |
edits