GSoC 2009 Project Ideas Hints

This page is to prioritize projects from larger list of ideas collected for Tcl GSoC 2009 that are most likely to succeed and/or benefit the larger community. Students shouldn't afraid proposing ideas not listed bellow if they find them inspiring. But we will try to give students some hints about selected project so they can have better perspective on what is needed and how it could be accomplished.

The list is not comprehensive, but rather a short-list for projects that we want to highlight.


tkosiak It seems that this "Most WANTED / Also important" thing didn't work out for a time being - nobody wanted to contribute.

In no particular order my private picks:

  1. Tcl string handling
  • Better string sentence and title case handling
  • Extended Unicode
  • ropes
  1. HTTP handling
  • HTTP/1.1 client
  • tclhttpd
  1. Tcl Application Server see GSoC Application Server Project for details
  2. Regexp engine cleanup
  3. Megawidget development
  4. Tcl/Tk image handling and manipulation
  5. TkPrinting (continued from last year GSoC)
  6. XLS read/write
  7. GTK related things: GLADE, Gnocl, GTK Tile Theme
  8. TDBC Driver Development
  9. Jacl
  10. RoR clone for AOLserver
  11. Graphs
  12. Tcllib
  13. Tklib
  14. Pure Tcl network protocols implementation: SSH or/and Bittorrent

But hey! I've just saved this to edit them further tomorrow. Don't look at them yet ;)


Here is a place for discussion:

tkosiak I feel that this year it is good to focus on:

 1 string handling modernization (titlecase, unicode, ropes, etc)
 2 regexp engine
 3 http protocol handling code unification
 4 Tcl application server infrastructure
 5 TkPrinting w/Tile
 6 XLS handling in Tcl
 7 Bittorrent & SSH in pure Tcl
 8 native GTK, QT, Cocoa Title themes
 9 Tcl on bare metal (embedded Linux / Solaris / QNX) [tkosiak]: I will elaborate on this more after doing community work.

and some more I haven't noticed yet

JH Tk Demo Expansion would be better, but #9 above less so. #6 may be nice for a niche project, but not sure overall. #5 I think should be classified as hard ... unlikely to succeed. Jacl is a perfectly acceptable choice from the larger list also.

tkosiak I disagree about #5 TkPrinting if it will extend what was done last year --tkosiak)