Pages: 1 
|
 |
|
Author
|
Topic: Rx Handbook, Draft of Revision 2 (Read 3345 times)
|
Charles Childers
Administrator
Sr. Member
Karma: +2/-0
 Offline
Posts: 745
|
I uploaded a draft of my work in progress on updating the Rx Core documentation. You can read it (about 400k, in PDF format) at http://retroforth.org/doc/rx-handbook.pdf
This covers the current changes since the original Rx Core used in 9.0.x, and also adds some expanded content and new sections to the handbook.
I'd appreciate feedback on this! I personally think it's an improvement, but more input is needed as always.
|
|
|
|
|
Logged
|
|
|
|
helmar@spirit
Newbie
Karma: +0/-0
Offline
Posts: 2
|
crc,
good work 
Well I still dont understand why the classes:
.DATA - as a class that compiles "literals"
.FORTH - as a class that describes the normal colon-defintions
.SELF - as a class that implements something very similar to "IMMEDIATE-words" from ANS
.INLINE - as the class for words that inline their code
are not enough? Really - I dont understand it. I dont get the point. Help me a litttle...
-Helmar
PS: The mail for forgotten passwords seems to take a long time... I did not get it...
|
|
|
|
|
Logged
|
|
|
|
Charles Childers
Administrator
Sr. Member
Karma: +2/-0
Offline
Posts: 745
|
I have the .class as a bit of additional support for the classes themselves. With some of my work on a debugger, it makes it easy to identify the classes that are defined.
As to .macro, I chose to keep this separate from .self, since I only want macros to have a compile-time action. Rather than doing state @ if .... ;then around the definitions, I decided to have a separate class for this.
I will check and see if I can find out why the server is not sending out the password emails; nothing is stuck in the queue at this point, so I'm not sure where the problem is 
|
|
|
|
|
Logged
|
|
|
|
Charles Childers
Administrator
Sr. Member
Karma: +2/-0
Offline
Posts: 745
|
Sending passwords should now work again; I still had the server set to the old server I used to use. 
|
|
|
|
|
Logged
|
|
|
|
helmar@spirit
Newbie
Karma: +0/-0
Offline
Posts: 2
|
How about a class that ensures compilation? This would be really great as extension to the four I''ve enumerated. Eg. IF or FOR or DO that cause compilation in every case - until a word that turns off compilation - something like THEN - occurs.
OBSOLETED  : What's on with the notification mails, I dont know - I still did not get it. Could it be my mail-service provider did not accept the mail? Well, it would be very unlikely but it could be.
-Helmar
|
|
|
|
|
Logged
|
|
|
|
Helmar
Library Contributor
Full Member
Karma: +1/-0
Offline
Posts: 129
TUCK what??? SWAP OVER!!!
|
Hahhhh... Nice to be me again 
-Helmar
|
|
|
|
|
Logged
|
|
|
|
|
|
|
|
Chuck Adams
Contributor
Jr. Member
Karma: +0/-0
Offline
Posts: 24
|
Hear hear -- I've always found it lame that most forths don't let you use loops and conditionals in the interpreter. Of course switching these on requires some kind of block behavior, which forth provides in word definitions, but the interpreter has no concept of.
I'm not sure it calls for yet another class to enable that behavior, as we already have a word that does the job of switching the compiler for a "block" on, namely ::. It does unfortunately leak the temporary definition, but that can be fixed by wrapping it in another word. No classes involved in that approach.
|
|
|
|
|
Logged
|
|
|
|
Helmar
Library Contributor
Full Member
Karma: +1/-0
Offline
Posts: 129
TUCK what??? SWAP OVER!!!
|
Chuck,
the implementation as class would have the big advantage that new words that require compilation could be easy to implement by simply assigning the class to them. One other big advantage in solving this problem with classes is that the user can reflect that behaviour easier if he is able to see the class of the word. My HelFORTH has a coloured "words" for that purpose - I do think a nice "blue" would be still available for such a class
-Helmar
|
|
|
|
|
Logged
|
|
|
|
Charles Childers
Administrator
Sr. Member
Karma: +2/-0
Offline
Posts: 745
|
I fixed a minor mistake in Appendix 1 today, (had the wrong class on ^), so an updated copy is now online.
In addition, the word information is now accessible via help in the 9.1 development build. (http://retroforth.org/projects/901.tar.gz). At present it only covers the words in the Rx Handbook, but this will be expanded to cover all words in Retro soon.
|
|
|
|
|
Logged
|
|
|
|
Charles Childers
Administrator
Sr. Member
Karma: +2/-0
Offline
Posts: 745
|
Yet another updated copy is now online. This time the contents have been reorganized, and I added a new section on the threading model used. It's getting closer to ready I think...
|
|
|
|
|
Logged
|
|
|
|
Pages: 1 
|
|
|
|
|
