TBoS 3rd edition

[Mark Tarver]

I've prepared the final edition of TBoS(3rd edition).  I revised over 50% of the first edition to make the second, but for the second I'm only changing about 10% - indicative that this book may have (probably has) reached the end-state.   Note I don't want to keep revising this, but I do want the book to as good as I canmake it.

I mention it here because fuzzywozzy's question re concurrency is answered in this text.   Here are the changes.

* the errata have been fixed

* the chapter on model checking has been replaced by a chapter on abstract and semi-abstract datatypes. In particular I show how to implement an ML-style 

  (algebraic datatypes) type discipline on top of Shen.   I thought this would be more interesting.

* the final chapter has been substantially changed, improved and and brought forward into the text.   I spend this chapter explaining how to control the proof process in type checking and tracing errors; hence the use of (spy +) is explained early (following Aditya's suggestion)

* there is a new appendix C on concurrency.   In it I propose a very portable and (IMO) elegant concurrency model based on a 2 instruction extension of KL.

* there is a new appendix on native calls; based on the understanding reached last month about standardising native calls.

I may do a couple of videos on the new stuff. 

Mark

[Kean Lau]

Mark, any news on when the TBoS 3rd edition will be ready for publication?

[Racket Noob]

Dr. Tarver,

Please, change the font for Shen programs in the new edition to some of the non-proportional fonts (Courier new or something similar).

I have both editions of TBoS and i think the programs would look much nicer and easier to read in Courier (or some similar non-proportional font) then in Arial which is completely unsuitable for programming purpose.

RacketNoob  

[Mark Tarver]

This party depends on a grant application I have filed.  At the moment the project has no money and w.o. support and the appropriate hardware (i.e. a multicore machine), I cannot develop concurrent KL which is based on a very compact 2 instruction extension.   In the event that the grant is not successful and the appendix on concurrency is cut, it would be possible to produce the next edition by this summer.    If there is funding, then the autumn.    In either event I think that this will be the last edition of TBoS for a considerable time, since the specification is by now very stable and the presentation is streamlined to be as clear as possible.

Mark