Charles L. wrote:
> Michael Letterle wrote:
>> Charles,
>>
>> I like how you laid out your project and whatnot.  My "dream" is to make 
>> zliby more then just a zlib compression library, but a more general 
>> compression library and command line tool.  Your compression code seems 
>> to be more in the same style as my decompression implementation, any 
>> qualms against me incorporating some of your stuff?
>>
>> I also added Park to the zliby ruby-forge project, if you'd like I can 
>> add you as well, or add you as a collaborator on the zliby github 
>> project.
> 
> Hi Michael,
> 
> Sure feel free to add me on the rubyforge/github, or just copy the 
> relevant files back in if you prefer. There's a few small fixes needed 
> to the deflate code, but after that it might be a good candidate for 
> inclusion.
> 
> The idea of a general ruby compression library is nice. With rubyzip, 
> and a pure ruby zlib, all thats probably missing is bzip2, and bwt is 
> fairly straight forward (I believe I saw a python implementation in 
> around 100 lines).
> 
> I'm really hoping some of that javascript vm goodness will find its way 
> to ruby though (or psyco equivalent or whatever), because without it, 
> pure ruby compression (or any algorithmic code) remains unsuitable for 
> general purpose use.

I didn't notice much of a slowdown with Park's version when installing 
gems or building gems, though I didn't do any sort of formal benchmarking.

It would be interesting to see some real benchmarks with MRI, JRuby and 1.9.

Regards,

Dan