To: tt@bu.edu In-reply-to: <199809040318.XAA04057@pm0.bu.edu> (message from Tom Toffoli on Thu, 3 Sep 1998 23:18:36 -0400 (EDT)) Subject: answers to questions Reply-to: nhm@mit.edu References: <199809040318.XAA04057@pm0.bu.edu> --text follows this line-- Hi Tom, Here are some answers to your questions: > After one has specified > lut-src address > why does one have to say > sa-bit reg! ? > Wouldn't it be simpler and clearer to say something like > SA-bit ? > > Could you define the word SA-bit for me? "sa-bit" is a register, and it gets set like every other register, by first constructing the default step-list entry, and then specializing it as necessary -- thus in the CAM-8 machine language, the specializing words always occur *after* the register name. The word you're asking for breaks this convention. It also has a bad property, which is that it differs only in capitalization from an existing word, but works differently. Your word could be defined, for example, as : SA-bit (s address-bit -- ) layer-mask @ sa-bit layer-mask ! reg! ; Note that "layer-mask" is a global variable that is set by the word "field" and reset (to "all layers") by all register-name words. > I'm trying to help Mentrasti & Co in Rome to program CAM-8. > Who should they contact to have a copy of Adorni's > CAM documentation in Italian? Why don't they contact Adorni directly? You should also take a look at "http://luisamiller.ce.unipr.it/cam8/cam8page.html" > Can you tell me what other documentation files there are on CAM-forth, > besides what is in the listings themselves? "ftp://im.lcs.mit.edu/cam8/doc.tar.gz" > Is there a way to send the identity table to a cam lut without > explicitly constructing the identity rule, naming it, and compiling it? The identity table is predefined in the system as "identity". When data bits must be permuted as part of the "assemble-fields" processing, this "identity" table is downloaded in a permuted order. These permuting operations are defined in "step-perm.fth". Incidentally, whenever a new table is defined using "create-table", it is initially filled with the identity table. Thus you wouldn't have to make an identity rule, name it, and compile it in any case. > How come file>field seems to have the same effect as file>cam, > in the sense that the entire cell is overwritten? This was a bug in the most recent full release of the system. Get the latest system updates from "ftp://im.lcs.mit.edu/cam8/changes_7_31_98" > How do I tell the machine to use only 1 module? > (1 is #modules no longer works as it used to) The word you want is "single-module". You should look in the "demos" directory for examples of the use of CAM Forth. The word "single-module" is used in "/cam8/demos/shear/shear.exp", for example. You might also look at "smooth/diffuse.exp" for an "assemble-fields" example ("life/fast-life.exp" is a more complicated example). There are many good examples that haven't been moved into the demos directory. It's time we updated this -- I plan to do that this week and also put the changes into the ftp directory. Let me know what other questions you have. Take care, Norm