AltME groups: search

the numbers for the  FIND can be supplied from a copy of the block 
with UNIQE

yep, will give it a try, thx henrik...

I usually use Joel's tally function -- it saves me the time needed 
to write a faster function:
http://www.rebol.org/ml-display-message.r?m=rmlKDDS

I think the fastest would be a variation of radix sort, if you know 
the max/min..

Setup for example:
>> Data: array/initial 5000 does [random 200]

== [36 119 148 112 87 59 176 21 95 138 183 28 1 119 14 74 39 78 141 
99 56 71 47 174 92 81 157 182 122 123 98 73 35 170 150 11 183 8...

>> MaxVal: fold Data 0 :max ;Just use a simple loop, I defined fold 
long ago for simplicity
== 200

What I mean:
>> Vals: array/initial MaxVal 0

== [0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 
0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0...
>> foreach Val Data [Vals/:Val: Vals/:Val + 1]
== 27

>> repeat I MaxVal [if 0 <> Vals/:I [print rejoin [I {, } Vals/:I 
{ times}]]]
1, 24 times
2, 20 times
..........
199, 20 times
200, 16 times

hum, there is a bug, the last value (7) is not precessed, but u got 
the idea...

Don't know if it's faster, but it's a bug free and consumes less 
memory.

fast-tally: func [b [block!] /local c u i] [
	c: sort copy b
	u: unique c
	i: 1
	forall u [
		c: any [find c u/2 tail c]
		u/1: as-pair u/1 negate i - i: index? c
	]
	c: none
	head u
]

Should be a little better...

fast-tally: func [b [block!] /local c u i] [
	c: sort copy b
	u: unique c
	i: 1
	until [
		c: any [find c u/2 tail c]
		u/1: as-pair u/1 negate i - i: index? c
		tail? u: next u
	]
	c: none
	head u
]

izkata, in R2 maximum-of is a native function.
with that, your solution should be defintivly the fastest

this is great, very fast code. thanks a lot for this.
the code is used for generating dynamic parsing rules 
for financial market data. soon I will release a free
software based on this data.

Seems that Tally is faster with R3.
2 reasons:
SORT is faster
MAXIMUM-OF is not anymore native 

(btw some of us claimed that it was a bad idea to have maximum-of 
 not anymore native in R3, But will claimed in the desert, so now 
we can see the drawback)

I would definitely prefer a native fold-type function over maximum-of 
and minimum-of

We tried, but that has definitely been declared an R3/Plus function, 
not native. Making ACCUMLATE? native wouldn't help much - the mezzanine 
version is really fast already. You won't be able to get functional 
language speed even from a native because that requires compiler 
support, and REBOL doesn't have a compiler at all.

R3/Plus has a low barrier to entry, because it will be modular. People 
can just not include the modules they don't need.


Conversely, there will be a high standard for inclusion as mezzanine 
or built-in native code because you won't as easily be able to remove 
it if you don't need it. Many R2 mezzanines, and even some natives 
won't make the cut.

page: read/custom [ scheme: 'http host: "twitter.com" target: "direct_messages/new.xml" 
user: "my-twitter-id" pass: "mypassword" ]  [ POST "text=This was 
also sent from a Rebol&user=synapse_emr" ]


This sends a private tweet to a user ... not clear from the API docs 
what the call is to just tweet ... anyone know?

this is easy to figure out using firebug  ;-)


redirect the html page to a server you have cheyenne running and 
save out the whole http request, you will have url and post data 
:-)

But there is a  twitter restful api ... just can't find it there. 
  Must be blind.

speed is a priority especially with Rebol which can be really slow 
if you don't take care

functional:

reduce [ "hi " (either is-male? gender [ "boy" ] [ "girl" ]) , "how 
are you?" ]

vs imperativeish:
a: "hi "
either is-male? gender [ append a "boy" ] [ append a "girl" ]
append a " How are you?"
a

so although I got impression from Carl's blogposts that he thinks 
rebol should be practical vs. functional , I think a lot of it's 
power and elegance and practicality come from functional aspects 
of it (just expressing my opinion)

This works ... posting a message is actually called updating your 
status!


 page: read/custom [ scheme: 'http host: "twitter.com" target: "statuses/update.xml" 
 user: "my-twitter-id" pass: "mypassword" ]  [ POST "status=Playing 
 with REBOL and the Twitter API" ]

is there a quick way to tell if an integer overflows?

I'm missing an overflow? function:

overflow? 632479375863785946758934675892
== true

unfortunately this might not work... to-integer sometimes returns 
a decimal... unlesss its been fixed for 2.7.6

It should work - always return integer! or error! - if the input 
is a number! (as specified in the type test of the argument. TO-INTEGER 
would  only have that problem with string input.

any good reason apart from the fact that data is a block that this 
is not allowed?

forskip data: [ 1 2 3 4 ] 2 [ ]

Is that a good reason?  :)

Maybe this should be changed so that a default word is used instead 
...

if a lit-word is not supplied as an argument but a block

seems a reasonable suggestion.

FORALL and FORSKIP are faster than FOREACH in R3 because they don't 
have a local word that requires a BIND/copy of the code.

Wouldn't it be more efficient to just BIND instead of BIND/copy body 
blocks in loops by default, and additionnally provide a /copy refinement 
in the (very) rare cases where you don't want the body block to be 
modified? Is it a design decision or did I missed some obvious reason?

I agree, using recursion for walking through hierarchical structures 
is a good approach as these structures usually have a depth limit 
far below the stack limit (e.g. parsing XML data) . In the last years, 
I've found that using block parsing for block! trees walking was 
probably the most efficient approach (code size and speed wise).

even in that case, i prefer to code an emulated stack (using a block)

Here's an example: Come up with a recursive-parse-rule version of 
the ARRAY function, including the function value calls. Compare.

Wait, bad example. ARRAY is generative, so it is likely a bad candidate 
for PARSE.

afaik the rebol stack goes a few thousand funcs before failing... 
 but that depends on the amount of params you have on the functions, 
more args = more stack use.

rebol stack limit: 

>> i: 0 a: func [][i: i + 1 a] a
** Internal Error: Stack overflow
** Where: a
** Near: i: i + 1 a
>> i
== 14265

REBOL/Core 2.5.0.3.1
>>  i: 0 a: func [][i: i + 1 a] a
** Internal Error: Stack overflow
** Near: a
>> i
== 3151

Rebol/View 2.7.6, 64-bit Ubuntu Hardy
>>  i: 0 a: func [][i: i + 1 a] a                      
** Internal Error: Stack overflow
** Where: a
** Near: i: i + 1 a
>> i
== 1705

So WindowsXP is more recursion-friendly than Leopard...finally a 
good argument to not switch to MacOSX! ;-)

>> i: 0 a: func [z][i: i + 1 a 1] a 1                    == 1704
>> i: 0 a: func [z y][i: i + 1 a 1 2] a 1 2            == 1706
>> i: 0 a: func [z y x][i: i + 1 a 1 2 3] a 1 2 3    == 1705

possibly the stack includes a single payload block for calls... this 
would make sense, in preventing stack depth variance.

possibly, yes, by indirection.  push stack [func args-block]

and args-block is a mem copy of the args block with local values

This is why the R3 host code is separate, and will be open-sourced. 
The host code is non-portable. The VM can call API functions provided 
by the host code. This would be necessary for a variety of reasons, 
not the least of which is that memory management, tasking, networking, 
any number of things are implemented by different platforms differently.

http://www.openldap.org/lists/openldap-devel/200304/msg00123.html


Anyone made a REBOL version of this? It's a UTF-8 <-> ISO-8859-1 
converter in C.

hasn't peterwood just uploaded a pretty nice string en/decoder on 
rebol.org?

peterwood can I make a wish?


I really need a function which can detect if a file is binary or 
text... do you think its possible to have such capabilities in your 
module?

I have a printable? function that checks if a string has only ASCII 
printable characters. Would that meed your need, Maxim?

its already very usefull, its for detection of how to process files 
in distro-bot.  right now it just treats any non rebol script as 
a binary, but I'd like to add some processing to other text files.


maybe peter can extend the concept and figure out if there are only 
printable characters in other encodings  :-)

Doesn't seem a very suitable name for what it does ...

'alter was always a poor choice of name -- short for 'alternate apparently, 
but too short to be obvious.
Anton has suggested 'toggle as a better name for it.

toggle implies something is always there ... but in a different state. 
 Swap is to imply that something goes in and out :)

I've used ALTER once for user selection in a list, and it was even 
a special case. I think having the ability to just conditionally 
append to a block if the value does not exist is more useful. I.e. 
an ALTER, split in two functions.

>> a: [ b b c d ]
== [b b c d]
>> alter a 'b
== false
>> a
== [b c d]

'alter is a hard one to name. The operation is:

 *   if the target value exists in the series: remove the first instance 
 of it
 *   if it does not exist in the series: append it to the end.

The "remove first instance"  makes it more generic than rotate/swap/toggle 
when there is more than one instance of the target in the series.

'cull would be a great name for 'remove-each : )

cull deer herd [deer = not smart]

In theory, ALTER is a set function, so the series in question is 
not supposed to have duplicates.

You are quite right -- 'alter is still in R3 ..... My mistake.


Several of REBOL's set operations operate on series. Series are more 
like bags than sets, hence they can have duplicate values. That adds 
value for a programmer, but dilutes academic purity.


Perhaps 'alter could have a purer set defintion by becoming a no-op 
if the target value exists more than once in the series.

Is there a way to adjust computer time from rebol?

you are a real tease, Graham.

No matter what I change the rime zone to on the computer I get a 
rebol time with -4:00

Got it..thanks Graham...


Here are three lines that adjust the time zone and time of the local 
computer after checking time with a server with a daytime server 
running:


call "RunDLL32.exe shell32.dll,Control_RunDLL timedate.cpl,,/Z Eastern 
Standard Time"
a: to-date read daytime://myserver.com b: a/time
call rejoin ["time " b]

I guess it might be off by a second or less ...close enough

I originally hard coded one time server into my program as in Ladislav's 
example, but I had the misfortune of that time server being out of 
action for a few weeks which introduced a 30 second timeout when 
my program started up.  Now i adjust the timeout to a much shorter 
value, and randomize from a pool of time servers.  And then restore 
the timeout value.

What am I doing wrong here??
given a block of blocks: 
[

[67 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.1"]

[69 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.3"]

[68 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.2"]

[71 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.2"]

[268 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.2"]

[721 "FFP Vol.2 - BASIC ENGINE OPERATIONS" "Chapter 1 - INTRODUCTION" 
"1.2"]
]

Using alter to remove each block in a loop:
 foreach z a [alter a reduce [z] probe a wait 3]

with a being the blocks above

It seems to work for the first two but than skips a block and than 
fails...

same result using find-remove:

foreach z a [if find a reduce [z] [remove a reduce [z]]]

I get three left over blocks


i I run it again I get one left over block and f I run it a final 
time it reoves the final block

foreach will advance to the next index whether or not the series 
was modified:
>> D: [1 2 3 4 5 6 7 8 9]
== [1 2 3 4 5 6 7 8 9]
>> foreach X D [remove find D X]
== []
>> ? D
D is a block of value: [2 4 6 8]

Using remove-each is better for this:
>> D: [1 2 3 4 5 6 7 8 9]       
== [1 2 3 4 5 6 7 8 9]
>> remove-each X D [integer? X]
== []
>> ? D
D is a block of value: []

I'm a dope...Thanks for the clue...
I used another 'variable' to what I needed...it works now!

Modifying a series you're iterating over is always risky. Not only 
do you have to make sure you get it right, but there is always a 
possibility that iterators might change their behavior.

I was already thinking a little that you probably don't really need 
them since you can change/generate code structures at startup or 
"JIT" and "cache" them ... but it was just a hunch.. not something 
I could imagine all the way down ...

perf. penalty: REBOL, due to its "philosophy", is an iterpreted language, 
which is a "performance penalty" against compiled languages, yes. 
But who cares? I have seen many compiled programs much slower than 
what I was able to write in interpreted REBOL.

I sometimes have some stupid ideas of changing code and blocks into 
something else but I don't have a totally clear view yet if that 
is ok to do and what would be the right way to do that.

I have one example ... it was fun to make but all in all doesn't 
seem a good idea so I don't use it now (I think it's not really robust 
and it gives strange errors.. not 100%), I am not sure.. would classic 
macros make it more possible, this all make changes at runtime: 

any-is?: func [ 'PRE s ] [
	while [ not tail? s ] [ s: next insert s PRE ]
	any reduce head s
]
; usage
if any-is? empty? [ d-y d-m d-d title category ]
    [ print "Error some value is empty: " fail: true ]
if any-is? positive? [ apples oranges kiwis grapes ]
    [ print "There is some fruit in your fridge" ]
if any-is? [ not empty? ] [ a b ] [ print "yup" ]
if any-is? [ 5 < ] [ num1 num2 ] [ print "yup" ]
if any-is? [ 5 < length? ] [ str1 str2 ] [ print "yup" ]

janko, that's a great function!

I am not sure what was the problem .. it's a while since I used it 
.. I just thought this function can't be good ..

it works in general cases and I have it in some older code .. maybe 
errors got a little strange if datatype didn't match the PRE function 
... like empty? none ... I can take some time and test it again

the copy is a good idea though, cause if you use it for real data, 
some functions can mangle the content of the block.

I'd even do a copy/deep

copy could optionally be a refinement, if you know that the test 
isn't destructive... cause copy will be a performance hit .

maybe this behaviour was bugging me .. because I know I use the order 
of things in any or all a lot of times to avoid null errors and similar 
.. and any-is? always preprocessed all elements

very ;-)  I fixed a little issue with interdependant libraries last 
month, but a part from that it has remained the same for years.

I use slim and prebol together within the sdk, it works very well. 
 I should use %include.r though, its much better and would solve 
a few little complexities I have with using the prerebol.r

it is quite surprising for me, that Carl is not a fan of the PREBOL 
approach, taking into account, that he invented it. I had a version 
of INCLUDE before PREBOL came and adapted it to the PREBOL standard 
afterwards

...but INCLUDE (in essence) is just a combination of PREBOL and DO, 
so nothing really new

Not invented here
 may be a problem...

but now that he has a group of like-minded people, changes are much 
lower that we don't think in the same lines.

Use a get-word.

If you make an object with images, blocks and a lot of data and then 
set the object to none, and the data inside is not referenced elsewhere, 
will everything else inside be properly deallocated?

Perhaps not - 

>> stats
== 67680368

>> a: make object! [

[    b: make block!  1000000

[    c: make block!  1000000

[    d: make block!  1000000

[    e: make block!  1000000

[    f: make block!  1000000

[    ]

>> stats

== 167685587
>> a: none

== none
>> stats

== 167686079

>> recycle

>> stats

== 167685527

Which is probably to be expected as the values referenced by the 
fields in the blocks have not been de-referenced. Looks as though 
you'll have to do it yourself:

>> stats
== 167686051

>> a: make object! [                       
[    b: make block!  1000000                 

[    c: make block!  1000000                 

[    d: make block!  1000000                 

[    e: make block!  1000000                 

[    f: make block!  1000000                 

[    ]
>> stats

== 267690882
>> a/b: none

== none
>> a/c: none

== none
>> a/d: none

== none
>> a/e: none

== none

>> a/f: none

== none

>> recycle

>> stats
== 167687087

I'm sure there's a better way but this should work:


>> foreach word remove first a [do join "a/" [word ": none"]]  ;; 
provided your object is called a

in R3, 
>>set a none

in R2, that should work
>> set bind first a a none

hmm...
>>set a none
It works for both...

Peter, here's some R3 fun. To do the initial block allocation in 
R3:
    a: make object! 5
    foreach w [b c d e f] [extend a w make block! 1000000]
or maybe this way:
    a: object [b: c: d: e: f:]
    foreach w a [set w make block! 1000000]
To clear your way:
    foreach w a [set w none]
Of course Steeve's method works too, and better :)

henrik: the GC doesn't deallocate on demand... you can have several 
hundred megs of RAM allocated... and sundendly the RAM goes down. 
 and some things do not cause the rebol executables to shrink.


images, in my experience cause memory leaks, where they never get 
de-allocated from RAM under some circumstances.


in theory if you remove all references to a context, its content 
is supposed to deallocate, but when passing anything through view, 
all bets are off.

I've had 400MB pictures stay stuck in RAM, with absolutely no GC 
happening, not even going through view... just loading the image, 
without assigning it... its stuck  :-(

that's a nice improvement in R3

load %some-image.png


not even assigned once... causes a permanent memory use in the last 
version of which I had to use huge images.  I ended up using image 
magic which was able to use the equivalent of 20 GBs of image data 
in a single 75 minute composition.

yes, but that is a different matter

that's quite a problem, but is it possible to allocate space and 
load new image data into that location?