Original Lisp Eliza added to Eliza Gen collection

[jshr...@gmail.com]

With the permission and assistance of the author himself, Bernie Cosell, I have added the original Lisp Eliza to the Eliza Generations collection. Cosell wrote this Lisp version of Eliza at BBN in the mid-late 1960s. (Weizenbaum’s original was written in about 1966 in SLIP, a long-dead Fortran-based symbol processing package.)

See: http://elizagen.org/index.html

Thanks to Peter Seibel for connecting me with Bernie Cosell.

!!YOU CAN HELP!!

One way that you can help is by writing to Deborah Cotton (cot...@hq.acm.org) at the ACM permissions office and encourage them to open source Weizenbaum’s paper, which is still inaccessible under copyright protections.

Second, I’m hoping to create a “perfect” OCR of this code and then macrify it to run on CL with as little modification as possible. You can help create the codebase for this by choosing a single page at random from among the 48 TIFFs, manually entering the code as precisely as possible (including indentation) into a text file, and then emailing it to me: jshr...@stanford.edu. If you decided to do this, here are a few details worth attending to: So that we get good coverage, please really choose a TIFF file at random, e.g., via the moral equivalent of (1+ (random 48)). Please don’t just OCR the TIFF file; I’ve already tried this with very high end OCR tools, and they make terrible encodings of this sort of thing! It would help me do the reconstruction is you put the name of the TIFF file in a leading comment. Finally, if you would like to be explicitly acknowledged for your efforts, please include a comment line for yourself as well. The codebase will be released on github or some such public location. Then you’ll be able to help actually hack it!

Finally, if you know of open source Elizas, in any language, roaming around the net, please send me links to them so that I can update the “more recent” section of the page.

Thanks!

Cheers,
‘Jeff

[Pascal J. Bourguignon]

A very commendable effort!

--
__Pascal Bourguignon__
http://www.informatimago.com/

[tar...@google.com]

On Friday, November 15, 2013 10:45:29 AM UTC-8, jshr...@gmail.com wrote:

> Second, I’m hoping to create a “perfect” OCR of this code and then macrify it to run on CL with as little modification as possible. You can help create the codebase for this by choosing a single page at random from among the 48 TIFFs, manually entering the code as precisely as possible (including indentation)

I will take on
Eliza-19720424-DOCTORp1-00of06.TIF
Eliza-19720424-DOCTORp1-01of06.TIF
Eliza-19720424-DOCTORp1-02of06.TIF
Eliza-19720424-DOCTORp1-03of06.TIF
Eliza-19720424-DOCTORp1-04of06.TIF
Eliza-19720424-DOCTORp1-05of06.TIF
Eliza-19720424-DOCTORp1-06of06.TIF

[jshr...@gmail.com]

Wow! Well, that was much more effective than I had expected! I've got well over half of the thing already xscribed by numerous volunteers. If you still want to contribute, here's what I'm missing (I think -- apologies if someone actually did any of these and I haven't correctly accounted your contribution -- I do have it, I'm just eye-balling what's missing at the moment):

All the source code is complete! (I know, that's the fun part... :-)

Still missing are pages 7-13 and 17-20 of the 1969 script,
and 1-3 and 5-8 of the 1972 script.

Thanks again!

'Jeff

[jshr...@gmail.com]

Thanks for everyone's help on this. All the code is xscribed. All I need to do now is to stitch it together. Stay tuned!

[Carlos]

[jshr...@gmail.com, 2013-11-18 08:03]

> Thanks for everyone's help on this. All the code is xscribed. All I
> need to do now is to stitch it together. Stay tuned!

Hi,

are you going to make the original code (not your adaptation) available?

--

[jshr...@gmail.com]

It is already in github (posted elsewhere in cll and linked to the page):

https://github.com/jeffshrager/elizagen

And it is under discussion elsewhere in cll. However, you have reminded me that I neglected to post the very first, unadterated version in the repo. I'll do that as soon as I get to a computer.

[jshr...@gmail.com]

Okay, to Carlos's excellent (if indirect) point, I've added a version of the raw transcripts to the repo.