Since i was preparing a response, and now one of Ladislav is here, i TOFU.
- Ladislav is a guru here. If he says "we" we agree. Or he does not say it. :) - He is right again. Mostly. Debatable. Depends if guru Gabriele steps in. Or guru /me, sometimes. Or you. ;) - Spelling is important. (At least for rebol. Humans have to ignore my typo= s.) Decisive here is how rebol handles it dictionary. And rebol makes an entry for each different spelling. And that entries decide about alias or not. - Your 'aliases has a subtle bug. Try ctx: context [w1: none] probe aliases? 'w1 in ctx 'w1 =3D=3D true Words with the same spelling, but different contexts are not 'same? . (As expected Ladislav spots it immediate, demoed by using 'use) (Thats what beeing guru is all about! ;) - You miss one indirection. So your examples about multiple words pointing to strings are OT: The global context is only the words and a value-slot for each. And in that slots are the references to strings etc. Word "points" to slot points to string. (At least in the way "we" use "global context" ;) - Quick note to global context (in the way "we" - but i said that): Is special. It contains all words ever loaded. Usual contexts (objects) contain only the words they need, the global context contains everything (One instance for each spelling). Because it is used for such meta-things like loading sourcecode and 'alias. - Hohoho too! And i looked up "cordial", good addition to my dictionary. ;) - Good investigation. Looks like you spend a lot time examining rebol. Hope you like it here. - About Ladislavs Version spelling?: IMHO "mold to-word word" does the same. Shorter. aliases?: Technically different cases result in auto-alias. Should we compare with case? On 5/9/06, Ladislav Mecir <[EMAIL PROTECTED]> wrote: > > Pierre Johnson napsal(a): > > First, text messaging can be such a limiting medium. > > > Yes, it looks that there was a misunderstanding somewhere. I thought I > found where it was but I was probably wrong, so let's try again: > > Now, excuse me, but it's not me who needs "HTH" and a referral > > to the REBOL docs. > > > That is possible, let's have another look. > > Again, to review the flow of messaging [with notes added]: > > > > ---- from the first ----- > > A recent exchange on the REBOL email list made me revisit > > Ladislav Mecir's article "Rebol Words and Contexts, alias > > Bindology" ++ note: see that word "revisit"? it implies i've > > read it at least twice ++ > > > > [http://www.fm.vslib.cz/~ladislav/rebol/contexts.html] > > > > Mr.Mercir writes: > > > > "Two words are equal, if they have equal spelling, or if they > > are aliases ... This knowledge can be used to define a Rebol > > function which can tell us whether two words are aliases." > > > > His word and definitional: > > > > aliases?: func [ > > {find out, if word1 and word2 are aliases} > > word1 [any-word!] > > word2 [any-word!] > > ] [ > > found? all [ > > equal? :word1 :word2 > > not equal? spelling :word1 spelling :word2 > > ] > > ] > > > > which depends upon: > > > > spelling: func [ > > {return the spelling of a word} > > word [any-word!] > > ] [ > > if word? :word [return mold word] > > if set-word? :word [return head remove back tail mold :word] > > next mold :word > > ] > > > > But is this truth? > > > > > > --- end from the first --- > > > > This isn't truth. Spelling has nothing to do with it! > > > > > As it seem to me, you are saying, that the sentence: "Two words are > equal, if they have equal spelling, or if they > > are aliases." is incorrect. Correct me, please, if I didn't understand yo= u properly at this point. > > > > The correct test has its basis here and here alone: > > > > aliases?: debate [ > > word1 [any-word!] > > word2 [any-word!] > > ][ > > either (not same? word1 word2) and (equal? word1 word2) [ > > true > > ][ > > false > > ] > > ] > > > > OK, let's try: > > aliases-2?: func [ > word1 [any-word!] > word2 [any-word!] > ][ > either (not same? word1 word2) and (equal? word1 word2) [ > true > ][ > false > ] > ] > > I would say that the above looks like your definition of the meaning of > the word 'ALIASES?, which differs from my definition above. As always, > no definition can be wrong, they are just different. Here is an example: > > word1: 'a > word2: use [a] ['a] > aliases? word1 word2 ; =3D=3D false > aliases-2? word1 word2 ; =3D=3D true > > > 1. Two words are aliases if and only if their definitionals > > ("values") are equal AND they point to the exact same (one and > > only one) definitional ("value") existing within REBOL > > controlled memory. > > > > > As I pointed out above, a different definition exists. I don't say your > definition is wrong (I don't think a definition *can* be wrong), but I > have got a good reason to keep my original definition - it takes into > account the symbolic nature of aliasing as mentioned by > http://www.rebol.com/docs/words/walias.html > > Again, Mr. Mecir, even if you would have wrote correctly in your > > "bindology" article, your concept of what is an alias in REBOL > > and how to detect it (your functional definitional) are both not > > truth, i.e., they're false, incorrect. > > > This is a point of distinction. What I am saying is, that your > definition of ALIASES? differs from mine. I am quite sure, that both my > as well as your definition cannot be incorrect, although one of them may > be more convenient than the other. > > -- what you should have wrote correctly (but would still be > > false)---------------------------------------------------- > > What is necessary for two words to be *aliases*? Two words are > > *aliases* if they are equal *have the same symbols in the exact > > same sequence* and *don't have the same face-value spelling*. > > > This sheds some light into it for me. It looks, like you are saying, > that I didn't define the meaning of: > > 1) "two words are aliases" > 2) "two words are equal" > > in the article. You are right, I didn't define these, because I wanted > to use the same meaning that can be found in > http://www.rebol.com/docs/words/walias.html . I offered the ALIASES? > function as a formalization, but another interpretation is possible, as > you pointed out. > > The final point is I've remained cordial through this exchange > > and find your replies amusing in your complete missing what is > > clear as if defending a mistake is more important than the > > mistake itself. > > > > > Thanks for your feedback and sorry for any offense I might cause by > misunderstanding your proposal. > > -L > > -- > To unsubscribe from the list, just send an email to > lists at rebol.com with unsubscribe as the subject. > > --=20 -Volker "Any problem in computer science can be solved with another layer of indirection. But that usually will create another problem." David Wheeler -- To unsubscribe from the list, just send an email to lists at rebol.com with unsubscribe as the subject.
