AltME groups: search

it would be nice if make-dir/deep worked with ftp! I use this instead:
make-dir-deep: func [ f /local t u p h][

 h: copy f until [exists? h: copy/part h next find/reverse back tail 
 h "/"]
	t: parse find/tail f h "/"
	u: copy h
	foreach p t [if not exists? append u join p "/" [make-dir u]]
	]

(again, just thinking out loud.. I am deep in my head resolving a 
get_username OS call, to get proper logon user on windows) hehe

were there so deep changes to kernel, that those already working 
things don't work?

petr, the deep changes are, i guess, to make Carl's life easier. 
(i.e. compiler and string)

I hope this is useful for someone

REBOL []

rest: func [s [series!]][skip s 1]

define-object: func [
	spec [block!] 
	/local 

  arg-spec ctx-spec object-name constructor-name predicate-name attributes
		spec-rule type-spec continue? w
][
	arg-names: copy []

	continue?: [none] ;used to stop parsing
	name-rule: [set w word! (insert tail arg-names w)]

 type-rule: [set w word! (unless datatype? attempt [get w] [continue?: 
 [end skip]])]

 spec-rule: [name-rule some [name-rule opt [into [some [type-rule 
 continue?]]]]]

	if any [
		not parse spec spec-rule
		arg-names <> unique arg-names
	][
		make error! "invalid spec"
	]

    object-name: to-string first arg-names
	constructor-name: to-word join 'make- object-name
	predicate-name: to-word join first arg-names '?
	attributes: rest arg-names

	arg-spec: copy []
	foreach itm attributes [
		insert tail arg-spec reduce [
			to-word join itm '-value
			either block? w: select spec itm [w][[any-type!]]
		]
	]

	ctx-spec: copy []
	arg-names: extract arg-spec 2 1
	repeat i length? attributes [

  insert tail ctx-spec reduce [to-set-word attributes/:i to-get-word 
  arg-names/:i]
	]

	;create constructor function
	set constructor-name make function! 

  compose [(reform ["Makes a new" uppercase object-name "object with 
  attributes" mold attributes]) (arg-spec)]
		compose/only [make object! (ctx-spec)] ;body

	;create predicate function
	set predicate-name make function! 

  compose [(reform ["Determines if value is a" uppercase object-name 
  "object"]) value [object!] /local types]
		compose/deep/only [
			either (attributes) = rest first value [
				foreach itm (attributes) [
					unless any [

      [any-type!] = types: select (arg-spec) to-word join itm '-value
						find types type?/word value/:itm
					][return false]
				]
				true
			][
				false
			]
		] 
]

a find/deep would be very useful... recursive searching of blocks 
and objects. it could allow you to find a value in large objects 
and blocks quickly

find/deep is long requested feature, but it was said as not being 
as trivial, as blocks can contain self-references etc ...

>> b: next next [1 2 3]
== [3]
>> head b
== [1 2 3]
>> head copy b
== [3]
>> c: [9]
== [9]
>> insert/only c b
== [9]
>> c
== [[3] 9]
>> d: copy/deep c
== [[3] 9]
>> head d/1
== [1 2 3]

So why does 'copy/deep don't ?

; Q1: copy/deep will give you new strings:
block: [none none none]
new-block: copy/deep replace/all block 'none {}

i guess you need copy/deep and poke ... copy pick ...

if the arrays have multiple dimensions, you need copy/deep. And if 
there are objects inside, those objects are not copied, then you 
need explicit work (copying all objects in a loop using 'make)

so, copy/deep, and, as volker suggests, you're likely to need to 
clone the objects too.

when you need a reduce/deep and there isn't one, what do you use 
instead?

reduce [ 'blah reduce [  to-word "desired-literal-word" reduce [to-word 
"deep-literal
-word"]] ]   ;ss this works...just talking to myself...nevermind

hmmn, reduce/deep or reduce/nested would be more elegant nonetheless.

is compose/deep/only an option? Also a reduce/deep would be short, 
if you need it.

trying to reduce a set of nested blocks ... compose/deep/only would 
not reduce the inner blocks, but leave them as they are...

compose/deep [ (a) [ (b) ] ] ; would work

I guess reduce/deep would not be very hard to implement...was just 
surprised there isn't one already...;-)

depends if ou code some themplate, then maybe compose/deep. if its 
data, maybe better reduce/deep.

sorry,
  compose/deep [ blah [  (to-word "desired-literal-word") ] ]

and to be defensive, compose/deep/only.

A first version:
reduce-deep: func[blk /local][
 blk: reduce blk
 forall blk[
  if block? blk/1[
   blk/1: reduce-deep blk/1
  ]
 ]
 blk
]

probe reduce-deep [ 1 + 2 [ 3 + 4 ] ]
probe reduce-deep [ a: [1 + 2] a/1] ; limitation, does not work

reduce-deep: func[blk /local][
 forall blk[
  if block? blk/1[
   blk/1: reduce-deep blk/1
  ]
 ]
 reduce blk
]
probe reduce-deep [ 1 + 2 [ 3 + 4 ] ]
probe reduce-deep [ a: [1 + 2] a/1] ; limitation, does not work
[3 [7]]
[[3] 3]

this breaks then, because inner blocks are evaluated first:
  a: 2 probe reduce-deep [ a: 1 [a]]

reduce-deep: func [
	block [block!]
	/local queue result file eval-state sub-result
][

	; Initialise
	queue: make list! 10
	result: reduce []


 ; Loop until we exhaust all unfinished blocks and we exhaust current 
 block.
	until [

		either empty? :block [
			; If finished this block, but more to process - add our
			; result to higher block result and set that block as current.
			if not empty? queue [
				set/any 'sub-result get/any 'result
				set [result block] queue/1
				insert/only tail result get/any 'sub-result
				queue: remove queue
			]
		][

			; Process current block item.

			either block? block/1 [
				; Save current level to process later,
				; set this new block as current.
				queue: head insert/only queue reduce [result next block]
				result: reduce []
				block: block/1
			][
				; Evaluate item.
				eval-state: do/next block
				insert/only tail result eval-state/1
				block: eval-state/2
			]
		]

		all [tail? block tail? queue]
	]

	; Return final result
	result
]

Or doesn't deep copy.

Chris was ahead of me there -- a working solution is
  block: copy/deep [do [a]]

You would need to copy/deep.

Should programmers concern about copy/deep when creating new objects 
from a spec? It seems strange that they do.

The easiest rule is to write make object! copy/deep spec instead 
of make object! spec

But is this appropriate, why not just have copy/deep by default on 
make object! ?

Perhaps it could come under 'Gotchas' -- it's not a bug so much as 
a 'feature' of Rebol values and contexts?  Just as much as 'copy 
does not automatically 'copy/deep...

regarding the behaviour: I think, that it might be optimal to have 
a function like CONTEXT deep copying by default with eventual /no-copy 
refinement?

My view is that Carl made this a conscious choice, knowing that advanced 
users could do their own copy/deep when they need to, and it won't 
come up most of the time anyway.

context: func [
    "Defines a unique (underived) object."
    blk [block!] "Object variables and values."
][
    make object! copy/deep blk
]

a similar issue exists for USE (use copy/deep spec) , and FUNC (solved 
by my CLOSURE)

closure, associative arrays, construct with deep copy, load on values 
that don't exist in system object, others. Some of the bugs that 
have been fixed are due to our need for more power, stable and predictable 
interpreter.

BrianH: do you think, that you could give a simple example using 
MAKE OBJECT! that would break using the deep copying variant of CONTEXT?

I find myself changing the mindset with REBOL every few years, because 
for a long time I was afraid of for example, using PARSE. PARSE is 
so central and important that it can change the way you work with 
REBOL, if you have stayed away from it. I had the same experience 
when starting to use the SDK and when starting to do networking stuff 
in REBOL to let scripts communicate with eachother.


It's not just a new set of ideas that turn up that lets me add to 
existing scripts, but doing the same scripts in entirely different 
ways. I feel I know about 30-40% of REBOL. :-) It's so damn deep.

Yes, I can only swim so deep before I run out of air :-)


I read a snippet of code related to the system object and think, 
Wow! I didn't even know that existed.

As you say, Parse itself is such a powerful thing. What I find inspirational 
is to sit down in front of the console and just explore ideas.

REBOL is like a little, magic and very deep lake high up in the mountains. 
It doesn't look much on the surface, but you'll be surprised, again 
and again.


It's a good exercise (maybe not for the totally newbie) to read some 
of the scripts, Carl has produced. They can be found e.g. in the 
Library. You find things like this one, that I trampled over in his 
color-code.r script:

set [value new] load/next str

load/next ... !!?? cute! :-)

And as you are pointing out now...the issue is goes deep enough to 
warrant discussion among high level REBOL programmers, although perhaps 
not at the 'tieing shoelaces' level.  :)

Yeah, I explained to Robert that the group here may have to suffer 
through some 'Tier B' advice  :)  There is a knack to explaining 
the deeper issues that can only come with enlightenment.   I have 
been trying to explain that the visible none is not the value none, 
but lack the vocabulary, based on the lack of deep understanding 
and experience.  It will come.

;; this is neat

 mold make object! ["test"]   ;; warning: to preserve a spec block 
 be sure to use    obj: make object! copy/deep specBlk

;; copy/deep issues are rampant in Smalltalk ( if you get the impression 
that I think Smalltalkers neglect Rebol, yer rite )

More and more I think that was is not obvious is no longer obvious 
once it is obvious

There is an 'active' LISP tutorial that would be a good model for 
a 'Rebol for newbies'

I would like to use the approach taken in the 2.3 "Official Guide" 
book to introduce unit testing in Rebol for TDD "from-the-get-go"

In Smaltlalk we used to count on newbies exploring in a workspace: 
we reaped a culture where people thoght the point of O-O was to write 
subclasses and create deep hierarchies like, say, Collection.  What 
was obvious was just wrong.  Messages were the point, not classes, 
let alone sub-classing.  Am I wrong to suggest to anyone new: "buy 
as used copy of "The Official Guide" " ?  For Oz, which is so much 
like Rebol, I do not hesitate to recommend Peter Van Roy's CTM from 
MIT Press.  Scheme has 'Little Schemer' and 'Simply Scheme'   The 
latter would be my model for an interactive tutorial in which you 
LEARN.  Smalltalk was supposed to be about how we model things ( 
how we learn how things interact )

I think it fair to say that it failed.  Classes were not the point. 
 Objects were not the point.  Things went wrong early on in abandoning 
the Actor Model in early 70's     I am hoping Rebol3  is getting 
it right ;-)   ( Io, the language, is quite inspiring ( www.iolanguage.com 
) but I still think Oz is a great intro to Rebol (they, too, lack 
an effective learning tool to "think in Oz " )

But is any valid deep path only the forst word has to have  a value 
associated with it.
  e.g. 
     block/tag1/tag2/tag3

Here is my code:

append top text
pdf: copy []
page: compose/deep [
    textbox [  
        font Courier 4.5 as-is (:top) 
    ] 
] 
append/only pdf page
write/binary %deposit-slip.pdf layout-pdf pdf
browse %deposit-slip.pdf

split-text: func [ txt n 
	/local frag result bl
][
    bl: copy []
    result: copy ""

    frag-rule: compose/deep [ copy frag (n) skip (print frag append result 
    frag) copy frag to #" " (if not none? frag [ print frag append result 
    frag ]
append bl copy result clear result) ]
    parse/all txt [ some frag-rule ]
    bl
]

compose/deep [ copy frag (n) skip (print frag append result frag) 
copy frag to #" " (if not none? frag [ print frag append result frag 
]
append bl copy result clear result) ]


looks suspicious to me - don't forget that your parens are there 
for two different purposes. My guess is, that you don't need compose?

Trying to do some macro expansion in text ...

This is not working :(

expand-macros: func [tmp [string!] macros [block!]
	/local white-rule rule len lexp
] [
	white-rule: charset [#" " #"^/"]
	foreach [macro expansion] macros [
        len: length? macro
		lexp: length? expansion
		rule:  compose/deep copy [

            [ to here: white-rule (macro) white-rule ( change/part here expansion 
            len  ?? macro) lexp skip ]
            to end
        ]
		parse/all tmp [some [rule]]
	]
	tmp
]

just a quick idea:

FORALL is implemented as mezzanine function. It calls FORSKIP which 
is mezzanine also. As you can see, it's probably not the fastest 
method. So here's the idea. Cannnot be FORALL rewritten to make it 
faster and is it possible to use PARSE to do this?
So I tried and came up with this simple function:

parall: func [
	'word body
	/local data
][
	data: get :word
	parse data compose/deep [

  some [(to set-word! word) any-type! (to paren! [do bind body :word])]
	]
]

(parall is just a shortcut for parse version of forall).


this is very simple function written in five minutes and not very 
well checked, it needs some additional work (eg. it does not return 
same value as 'forall etc).

So let's do some test (using Ladislav's %timblk.r):
>> n: copy [] repeat i 100 [append n i]

== [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 4...
>> time-block [forall n [b: n/1 + 1]] 0.05
== 3.623046875E-4
>> time-block [parall n [b: n/1 + 1]] 0.05
== 3.814697265625E-6
>> 3.62e-4 / 3.81e-6
== 95.0131233595801

95x faster? whooo....

and what about bigger block?

>> n: copy [] repeat i 10000 [append n i]

== [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 
25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 4...
>> time-block [forall n [b: n/1 + 1]] 0.05
== 3.540625E-2
>> time-block [parall n [b: n/1 + 1]] 0.05
== 3.7994384765625E-6
>> 3.54e-2 / 3.8e-6
== 9315.78947368421

9000x ? omg...

comments?

I tried to enclose parse in loop 1 [] and it seems to handle break. 
I guess you'll probably prove me wrong, Brian :)

parall: func [
	'word body
	/loc data
][
	loop 1 [
		data: get :word
		parse data compose/deep [
			some [(to set-word! word) skip (to paren! [do body])]
		]
	]
]

re: continue - this is not r3 ;)


>> n: [1 2 3 4 5] parall n [if n/1 = 4 [break/return "break"] if 
n/1 > 4 [print "bad"]]
== "break"

this is possible, but my guess, that Steeve would call it ugly too:


>> once-only: func [rule] [use [rule'] copy/deep [rule': :rule [rule' 
(rule': [end skip])]]
]

>> rule: [(a': once-only a b': once-only b c': once-only c) 3 [a' 
| b' | c']]

== [(a': once-only a b': once-only b c': once-only c) 3 [a' | b' 
| c']]

To do block parsing, I suggest:

>> once-only: func [:rule] [use [rule'] copy/deep [rule': :rule [rule' 
(rule': [end])]]]

>> rule: [(a': once-only 'a b': once-only 'b c': once-only 'c) 3 
[a' | b' | c']]
>> parse [c b a] rule
== true

I see. That's also why my block parsing give false result with less 
input. So for block-parsing, it should be:

>> once-only: func [:rule] [use [rule'] copy/deep [rule': :rule [rule' 
(rule': [end skip])]]]

>> once-only: func [rule] [use [rule'] copy/deep [rule': :rule [rule' 
(rule': [end skip])]]]

>> rule: [(a': once-only 'a b': once-only 'b c': once-only 'c) 3 
[a' | b' | c']]
>> parse [a b c] rule
** Script Error: Invalid argument: a

How do you build parse rules?

rule: copy []
word: "abc"

;; Want to compose the block to look like this  ["abc" (print ["found 
word len" 3])] 

insert tail rule compose [(word) (print "found word len (length? 
word))]  no go - obvious

I've tried  compose/deep [(word) (to paren! [print ["found word len" 
(length? word)])]  but length? word doesn't get composed, it gets 
included in the to paren! expression  compose/only/deep same thing


I guess the question is what is the technique to compose something 
that is to include paren! expressions with data inside that needs 
to be composed?  **With binding of course**  :)

reduce [word to paren! compose/deep [print ["Found Word of Length" 
(length? word)]]]

compose [(word) (to paren! compose/deep [print ["found word len" 
(length? word)]])]

compose/deep before the to paren! block, makes so much sense...ummm, 
now.  Now that I've seen it.  :)

It now passes all my non-regression tests, but I've done several 
deep changes, so I'll consider it beta until the end of the year 
before declaring it the new official version.

I can't say I've been 'using' Rebol for long, but I've been playing 
for quite a while now.  I discover something new every time I open 
up the system.  It's too cool how RT has put something as wide and 
deep as the ocean into a cup, a cup with blinking lights no less.

Brian, right, so I have to workaround all the time, write slow code 
with deep parse recursions, and all those funny and nice things. 
Or, give up and pretend REBOL was PHP.

>> unmush/deep [rSfeCs[Nse[i1v5x10l50c100d500m1000]twCi~J[JnCN]Kk+elJn[alN-j 
N0]'jJn]pK+j]

== [r s fe c s [n: se [i 1 v 5 x 10 l 50 c 100 d 500 m 1000] tw c 
i ~ j [j: n cn] k: k + el j n [al n - j n: 0] 'j j: n] p k + j]

read-thru: func [

    {Read a net file from thru the disk cache. Returns binary, else none 
    on error.}
    url [url! file!]

    /progress callback {Call func [total bytes] during transfer.  Return 
    true.}
    /update "Force update from source site"
    /expand "Auto-decompress after transfer."

    /check {Update only if version, checksum/secure, or date/size do 
    not match.} info

    /to "Specify a file target, not cache." local-file [file! none!]
    /local file data purl loc-path
][
    vbug ['read-thru url info]
    if none? file: path-thru url [return none]
    if local-file [file: local-file]
    if all [not update exists-thru?/check file info] [
        if error? try [data: read/binary file] [return none]
        return data
    ]
    if file? url [
        if error? try [data: read/binary url] [return none]
        return data
    ]
    loc-path: first split-path file
    if data: read-net/progress url :callback [
        if not exists? loc-path [make-dir/deep loc-path]
        if all [expand find/match data "rebpress"] [

            if error? try [data: decompress skip data 8] [return none]
        ]
        write/binary file data
        if all [check block? info info/2] [

            error? try [set-modes file [modification-date: info/2]]
        ]
    ]
    vbug ['read-thru-ok? found? data]
    data
]

Joe you are asking a question that finds its answer in a completely 
different model.  It reminds of the joke "What I meant to say, was, 
Mother, would you please pass the salt,' (look it up).


The answer is to throw away the brochure (page) model of the web, 
and move to web 2.0, where there is a cohesive (continuous) model.


The UI is complete separated from the backend, and the UI is a single 
entity, that is persistent during the session.  Everything else is 
simply a pain.


Most sites are horizontal (shallow) as opposed to vertical (deep). 
 And most are still modeling on the brochure (page) as opposed to 
the space (like a desktop).

prebol: func [code "changes code" /local p f] [
	parse code rule: [
		any [
			p: #include set f file! (
				p: change/part p load f 2
			) :p
			| into rule
			| skip
		]
	]
	code
]
; changes code, use copy/deep if needed

t1: now/precise
save %test.r [The embedded stuff]
p: prebol [Hello [World #include %test.r here] we are ]
print[difference now/precise t1]

prebol: func [code /local p f rule] [
	if file? code [code: load code]
	parse code rule: [
		any [
			p: #include set f file! (
				p: change/part p prebol load f 2
			) :p
			| into rule
			| skip
		]
	]
	code
]
; changes code, use copy/deep if needed

t1: now/precise
save %test.r [The embedded and #include %test2.r stuff]
save %test2.r [Subembedded]
p: prebol probe [Hello [World #include %test.r here] we are]
print [difference now/precise t1 newline mold p]

wind: compose/deep [
	button "Ana" [do %anamonitor.r]
	obj: object
	return 
	button "UP" [titlebar/size: titlebar/size + 10x0 show titlebar]
	text (join "REBOL/View :  " system/version)
]
display "Test du style de l'objet LadyBird"  wind
do-events

do %rebgui.r
display "RebGuiTest" compose/deep/only [
        label "Query" t: text 40 "" return
		a: area 100x50 "In the beginning, "
        button "Insert" [
				insert a/text "Title{}"
				; edit-text a enter

    insert tail a/text "And from a barely documented VID, a need for 
    something bare but easy to document: RebGui.  And the user saw it 
    was good."
				; edit-text a enter
				show a
        ]

do [at face/pane/3/text 10 insert tail face/pane/3/text "there was 
View 1.3.  "]
] ; End Display
do-events

How to move the cursor to inside the brackets of Title{->Here<-}.

How to enter a newline, and more generally use the function defined 
in rebgui-edit.r into the area?
** Script Error: edit-text has no value
** Where: action
** Near: edit-text a enter

More generally, how to enter a string and move the cursor after the 
insertion of it.  Thanks for an answer to this 'II am New' question 
pertaining to RebGui.

Graham, are you sure the disappearing text-list slider problem wasn't 
just the fact that all rows fit? Try the following code:

b: copy []
repeat i 1000 [insert tail b i]

display "Test" compose/deep [
	text-list #H data [1 2 3 4 5 6 7 8 9 10]
	text-list #H data [(b)]
]


When you maximize the window you'll see that the sliders only appear 
if needed.

To everyone:

What characterize a good learning book? Do you prefer thick books 
with deep explanation and many examples, or do you prefer the thin 
book with the essentials? Look at your collection of technical book; 
about computer languages, OSs, databases or what you have. Which 
ones do you like, and which ones is no-good?

Sunanda;  I'm starting to take a deep interest in RitC.  But I have 
doubts.  Doubts that need to be squashed.  Sadly, Ontario (a fairly 
vast province) has standardised curriculum now.  It's all schools 
or no schools here.  I'm not a fan, the excuse was that some kids 
in some boards were getting sub-par educations, ignoring the fact 
that some boards were providing above-par educations and instead 
picking a middle-of-the-road bland education for all.

REBOL []


use [fixup-rule label-rule label-fixup-rule label-error-rule here] 
[
    ; Initialize the intermediate rules
    label-rule: make block! 0
    label-fixup-rule: make block! 0
    label-error-rule: make block! 0
    ; Build the fixup-rule based on the opcode-rule
    fixup-rule: copy/deep rebcode*/opcode-rule
    parse fixup-rule [
        some [
            here: 
            lit-word! block! '| (

                unless find ['bra 'brat 'braf] here/1 [insert here/2 [label-error-rule 
                |]]
            ) |
            lit-word! 'word! '| (

                unless 'label = here/1 [here/2: [label-error-rule | word!]]
            ) |
            lit-word! | '| | 'block! | 'series! |
            'word! (here/1: [label-error-rule | word!]) |
            'any-type! (here/1: [label-fixup-rule | any-type!]) |
            into ['integer! '| 'word! | 'word! '| 'integer!] (
                insert here/1 [label-fixup-rule |]
            ) |
            block! (insert here/1 [label-error-rule |])
        ]
    ]
    ; Replace the fix-bl function

    rebcode*/fix-bl: func [block /local labels here there label rule] 
    bind [
        labels: make block! 16
        block-action: :fix-bl
        if debug? [print "=== Fixing binding and labels... ==="]
        parse block [
            some [
                here:
                subblock-rule (here/1: bind here/1 words)
                |

                'label word! (here/1: bind here/1 words insert insert tail labels 
                here/2 index? here)
                |
                'offset word! integer! (
                    here/1: bind 'set words
                    here/3: 3 + here/3 + index? here

                    if (here/3 < 1) or (here/3 > 1 + length? block) [
                        error/with here "Offset out of bounds:"
                    ]
                )
                |
                opcode-rule (here/1: bind here/1 words)
                |
                skip (error here)
            ]
        ]
        either 0 < length? labels [
            label-rule: make block! length? labels

            foreach [key val] labels [insert insert tail label-rule to-lit-word 
            key '|]
            clear back tail label-rule

            label-fixup-rule: [there: label-rule (there/1: 2 + select labels 
            there/1)]

            label-error-rule: [label-rule (error/with here "Cannot use label 
            here:")]
            rule: fixup-rule
        ] [
            rule: opcode-rule
        ]
        parse block [
            some [
                here:
                ['bra word! | 'brat word! | 'braf word!] (

                    if not label: select labels here/2 [error/with here "Missing label:"]
                    here/2: label - index? here
                )
                |
                rule
                |
                skip (error here)
            ]
        ]
    ] rebcode*
]

You can also reference the block of offsets through a word. Labels 
are not converted then - you must use numeric offsets that you count 
by hand (or in a compiler). Since these offsets are relative to the 
end of the branch statement, this block is only useful in one location. 
Also, when the rewrite phase comes back and they start using rewrite 
rules again, those hand-calculated offsets will likely be wrong. 
In theory, this could be used to implement a multi-state machine, 
but that kind of thing is deep magic that you should be doing with 
the parse engine anyways. It is theoretically possible to fill the 
block at runtime, which would technically be a computed branch, but 
this is so slow, awkward and unnecessary as to be ridiculous, especially 
for a branch block that can only be used from one location.

just that if you have some bugs or wandering about specifics, he 
has deep knowledge of it, he might even be able to give you hints 
on how to go around a few missing things.

I haven't done deep research, but mainly I view them as something 
like a "helper" for meditation, self-hypnosis, etc.

Not deep enough toin Erlang to judge about problems, but if you can 
explain :)

maybe there is some setting for that, dunno .... Windows denerves 
me sometimes with so called - rought czech translation - delayed 
write was not successfull. Not sure how it happens, but somewhere 
deep in your profile there is a dir for such a feature, and if there 
is some file, you can see annoying messages each time Windows starts.

jaime, how deep can you go with it? I don't feel too inclined to 
dive into objC programming right now

I watched a program recently about crossing the street there.  The 
trick being to begin to cross, and keep your speed exactly the same 
at all times.  Which they demonstrated.  Two women holding hands 
(one was a westerner) cross the very wide street with motorcycles 
and mopeds and tiny cars whizzing by.  Scared me to simply watch 
it.


I suspect he paused, which would be something most of us that are 
not used to that might do.


It is a deep shame, as an absent minded professor, one of my big 
fears is being hit by a car.  I used to walk home at about 3:00 a.m 
almost every night.  My office and home were 3k apart.  The streets 
were completely empty, and so I felt somewhat safe from cars.  One 
night in deep thought I crossed the street (totally ignoring the 
state of the light) and was almost hit by a racing car.  Often, when 
I would arrive at home I simply could not remember walking at all, 
I was so deep in thought.

The current library interface is servicable, but could be improved.

For example, char arrays in C structs are a real pain to deal with,

there is redundancy if you're importing a number of routines, from 

the same library, and extra work is required to deal with pointers
to values.

The biggest issue for me seems to be that I have to use COMPOSE 

heavily to get the results I want, or there's a lot of duplication
in struct and routine defs.

--- Easier Routine Declarations

The only thing I've addressed in my lib interface dialect is
making it easier to declare routines. I posted it to REBOl.org
for discussion:


http://www.rebol.org/cgi-bin/cgiwrap/rebol/view-script.r?script=lib-dialect.r

--- Pointers


This is maybe a bit of an extreme example, but I had to do it, so 
it's not purely theoretical.

    LPINT-def: [value [integer!]] none

    LPINT: make struct! LPINT-def none

This struct shows where a nested struct is needed.

    _FAX_JOB_PARAM-def: compose/deep/only [

        SizeOfStruct    [integer!]      ; DWORD structure size, in bytes

        RecipientNumber [string!]       ; LPCTSTR   pointer to recipient's 
        fax number

        RecipientName   [string!]       ; LPCTSTR   pointer to recipient's 
        name

        Tsid            [string!]       ; LPCTSTR   pointer to transmitting 
        station identifier

        SenderName      [string!]       ; LPCTSTR   pointer to sender's name

        ;SenderName      [struct! [value [string!]]]       ; LPCTSTR   pointer 
        to sender's name

        SenderCompany   [string!]       ; LPCTSTR   pointer to sender's company

        SenderDept      [string!]       ; LPCTSTR   pointer to sender's department

        BillingCode     [string!]       ; LPCTSTR   pointer to billing code

        ScheduleAction  [integer!]      ; DWORD job scheduling action code

        ;ScheduleTime    [struct! (SYSTEMTIME-def)]        ; SYSTEMTIME  
          time to send fax
            wYear       [short]
            wMonth      [short]
            wDayOfWeek  [short]
            wDay        [short]
            wHour       [short]
            wMinute     [short]
            wSecond     [short]
            wMilliseconds [short]

        DeliveryReportType  [integer!]      ; DWORD e-mail delivery report 
        type

        DeliveryReportAddress [string!]     ; LPCTSTR   pointer to e-mail 
        address

        DocumentName        [string!]       ; LPCTSTR   pointer to document 
        name to display
        CallHandle          [integer!]      ; HCALL reserved
        ;_PTR   Reserved[3]  [integer!]      ; DWORD must be zero
        _PTR-0  [integer!]      ; DWORD must be zero
        _PTR-1  [integer!]      ; DWORD must be zero
        _PTR-2  [integer!]      ; DWORD must be zero
        _PTR-3  [integer!]      ; DWORD must be zero
    ]
    _FAX_JOB_PARAM: make struct! _FAX_JOB_PARAM-def none
    _FAX_JOB_PARAM/SizeOfStruct: length? third _FAX_JOB_PARAM

    fax-complete-job-params: make routine! compose/deep/only [

        JobParams     [struct! (LPINT-def)] ; ptr to job information struct

        CoverPageInfo [struct! (LPINT-def)] ; ptr to cover page struct
        return:       [integer!]
    ] winfax.dll "FaxCompleteJobParamsA"

So, the API call returns pointers to structs containing the 
data we want; to get it we need to dereference the pointers
after the call.

    complete-job-params: func [
        /local
            params-ptr cover-ptr    ; API return pointers

            params cover            ; REBOL structs with data from API
    ][
        ; allocate return pointer structs for API call
        params-ptr: make-LPINT
        cover-ptr: make-LPINT

        ; make the API call

        reduce either 0 <> fax-complete-job-params params-ptr cover-ptr [

            ; get data from pointers returned by the API

            params: get-dereferenced-data params-ptr _FAX_JOB_PARAM-def

            cover:  get-dereferenced-data cover-ptr  _FAX_COVERPAGE_INFO-def
        ...



Getting the de-ref'd data is the real pain, and seems like it might
be unsafe in the way I did it, though it worked.

    get-dereferenced-data: func [

        {Given a pointer to memory, copy the target data into a REBOL struct.}

        pointer [struct!]   "LPINT structure whose /value is the data pointer"

        struct-def [block!] "The struct you want returned with data"
        /local struct data orig-pointer result
    ] [

        struct: make struct! compose/deep/only [ ; make wrapper struct
            sub [struct! (struct-def)]
        ] none

        orig-pointer: third struct              ; store original inner pointer

        change third struct third pointer       ; change inner pointer to 
        ref'd data

        data: copy third struct/sub             ; copy data from the inner 
        struct

        change third struct orig-pointer        ; restore inner pointer

        result: make struct! struct-def none    ; make result struct

        change third result data                ; change data in result struct
        struct: data: orig-pointer: none
        result
    ]


--- char arrays in structs, or as routine parameters

You can't just declare a fixed size block or string to do this, 
you have to (AFAIK), have individual elements for each item.
That's a huge pain if you have a 128 element array, so I end
up generating them dynamically. I think that was Cyphre's idea
originally, but I don't have notes on it.

    make-elements: func [name count type /local result][
        if not word? type [type: type?/word type]
        result: copy "^/"
        repeat i count [
            append result join name [i " [" type "]" newline]
        ]
        to block! result
    ]

    GUID: make struct! GUID-def: compose [
        Data1   [integer!]  ; unsigned long
        Data2   [short]     ; unsigned short
        Data3   [short]     ; unsigned short
        (make-elements 'Data4 8 #"@")  ; unsigned char
    ] none


--- MAKE-ing structs

How do other people make structs from prototypes? 

    make-struct: func [prototype /copy /with data] [
        make struct! prototype either copy
            [second prototype]
            [either with [reduce [data]][none]]
    ]


--- BSTR type

I've only needed it for one project, but it might be worth 
finding out if it would be worth adding BSTR support for
Windows, as a routine datatype.

I think with Swig you still need to be prepared to dive in pretty 
deep and swim around tying things together for a while.

Maxim: no, it does not solve the copy problem; however, closures 
need to copy deep the function body at each call; so, it somewhat 
solves the copy problem as a side effect (only for blocks and parens 
though, probably, as there's no need to copy anything else). (however, 
i would not rely on it, since for example it would be a useful optimization 
not to copy empty blocks or blocks that contain no words etc)

well I didn't say "solve"  ;-)  but by what I understand of closures 
(and I have been reading a lot lately) if 'closure does a deep copy 
without empty series optimisation, then we could rely on it.


actually, NOT copying empty series would make closures bugged, since 
all generated closures would referrence the same empty block from 
the context of the closure's definition.  which is in direct contradiction 
to the concept no?

Does the build dialect help with that situation where you have paraens 
inside an action block of the vid dialect, and you are using compose/deep 
on the whole layout?

I think it's easier to debug it in ghostscript. The printers I've 
tried this on will not print stack information. Perhaps only if I 
dive really deep into the capabilities of the printer, which I have 
no time for.

I haven't delved deep enough to understand if this'd be better written 
as a Cheyenne handler, though that would require forking the cgi 
script, which for the time being, also has to run under Apache.

Remember that I've done some deep changes in this version, so test 
it carefully before putting it online.

what was the mention of deep copying of function?

R3 now deep copies by default... but its at mezz so you can revert 
if you need.

What does it deep copies? Function arguments?

make function! spec copy/deep body

Doc;  Cheyenne is proving itself pretty worthy so far here at peoplecards.ca 
  Thanks.
Not delved too deep, but runs nice and smooth.

Henrik, most of the beaches are only 1-2 meers deep.

It's a deep tool.

I like to see more of the sheets from Ron... goes pritty deep..nice...

I plan to do something in my free time in a 3d engine .. I need that 
it works in a browser so unity seems the most obvious choice, adobe 
director is expensive and probably outdated, stonetrip / shiva seems 
even more IDE/GUI heavy than unity giving me even less controll (although 
I am not 100% sure bcause I haven't digged really deep in any of 
those.) Then there is Java (which has Ardor3d which is very early 
in the making , and jme which is somewhat "out") but starting with 
it would be at a lot more low level than unity and similar game engines, 
with no real docs and tools as far as I can see.