Re: I have time, what needs to be done?

[Jorge Stolfi <stolfi@xxxxxxxxxxxxx>]

    > I am very much interested in making a substantial contribution
    > to the project timewise. 
    
I think the most urgent thing to study is the gallows bit correlation.
(Map each word to the number of gallows letters it contains, then
study the statistics of the resulting digit string.)

As you must all be aware of by now, the result consists of 0's and 1's
almost exclusively, and there are surprisingly long runs of 1's and
0's. This correlation between successive words severely constrains the
VMs theories, crypto or otherwise.

Some questions that occur to me:

  What is the distribution of the lengths of the runs?
  
  Are the statistics similar in both "languages"? In all sections?
  
  Do the line/paragraph/page breaks affect the run-length statistics?
  
I should have been doing these statistics myself, but I have been
busy with useless work (meaning the stuff I am paid to do).
Perhaps in a few months...

    > One big limitation I have right now, is that I no longer have
    > access to a Sun, and I haven't taken the time to get really
    > familiar with my PC clone, it's mainly a toy to write e-mail,
    > pirate music and play on Chinese web sites.  Is there anyway to
    > run all of those neat little grep and cat routines on this
    > machine without having to shell out some big bucks to buy a
    > language?

It is possible to install both Windows and Linux on the same PC, so
that you can switch between them at boot time. I haven't done it
myself (I use a Sun workstation), but the PCs in our student labs are
configured that way, and som are many of our faculty machines.

All the best,

--stolfi