World: r3wp

Check
	?? do-events

WAIT [ ] === WAIT NONE

13-Sep-2006 Anton Rolls

WAIT without also waiting for events:

view layout [
	button "wait 2" [
		remove find system/ports/wait-list system/view/event-port 
		
		wait 2 ; wait two seconds
		
		insert system/ports/wait-list system/view/event-port
	]
]

Oops, this is Rebol3 group. Sorry, I'm in Rebol2 head-space.

Thanks

I was hoping it would do more or less what you programmed: servicing 
system events but without waiting for all windows to close

How works the preallocation in R3? str: make string! 10000 length? 
str -> 0 dtto for blocks etc? why it happens in r3 program? prepared 
for future use?

Preallocation allocates the space needed to store the required number 
of characters into string. Nevertheless, the length of the string 
is zero, since the string initially does not contain any characters.

it's there primarily to help the garbage collector.

So the space IS already reserved?

So the space IS already reserved?

So the space IS already reserved?

sorry Altme hangs for me

Here is an example showing a difference:


>> time-block [str: make string! 0 loop 10000 [append str "a"]] 0,05
== 0.0023125


>> time-block [str: make string! 0 loop 10000 [append str "a"]] 0,05
== 0.0024375

>> time-block [str: make string! 10000 loop 10000 [append str "a"]] 
0,05
== 0.002265625


>> time-block [str: make string! 10000 loop 10000 [append str "a"]] 
0,05
== 0.002265625

yes, when doing a make string! 10000, the space is preallocated. 
this helps the garbage collector to collect quicker and also speeds 
up operations on the string when its expanding in size.

its expanding in size, within the preallocated size, that is.

I would not say that it "helps the garbage collector to collect quicker", 
but it does not require to reallocate/recollect the space when expanding 
the string to 10000 characters length

and, of course, when no reallocation occurs, no content movement 
is necessary

Does it also means that it allocates consecutive space?

REBOL strings as well as REBOL blocks (parens, paths, the majority 
of series,...) use consecutive space

Is it already supported to get the output of a CLI programm executed 
via CALL?

no

there were several discussions about the topic. Carl stated, that 
R2 code was very complex, and is willing to provide source for adaptation. 
After some complaint, we got /wait at least. I think that for now 
you have to use call/wait, output to file, and read the file ....

Is there any reason why proxy is not implemented? Should it be part 
of http scheme or should be intermediate scheme? ie http -> proxy 
-> tcp. Where is oplaned to save proxy settings in R3?

The HTTP scheme in R3 needs a lot of work (hopefully not a full rewrite). 
That's why.

For what it's worth, the cURL binding supports proxies

Currently only through environment variables. I don't know if that 
works on Windows

Kaj full respect to your effort, in the other hand the proxy seems 
not to be overcomplicated. it may be a good training task, in R2 
it has maybe 20 lines. question is if there is some architectual 
restriction (for example synchronous / asynchronous etc.)

It sure would be nice if you implemented it

Remembering the function naming discussion from the !REBOL3 GUI group 
and seeing the 

http://curecode.org/rebol3/ticket.rsp?id=667&cursor=1#comments

I could not help but point out:


As I see it, not using the question mark *is* violating a naming 
principle that was explicitly stated. I know, that in REBOL we don't 
have to be that rigid, but, when we have explicitly stated a principle, 
we *should* stick to it. ( http://www.rebol.com/r3/docs/concepts/scripts-style.html#section-10
)

http://curecode.org/rebol3/ticket.rsp?id=1818&cursor=1

The we should stick to principles. But I am not sure even Carl himself 
is strictly following the rules. In his doc he claims, that 'quit 
is as clear, as quit-system. Well, we have 'do, and we have 'do-browser, 
'do-service, where we are breaking on encapsulation rules, with excuses 
to not polutu 'do's name-space (not complicating it - because in 
other words, the proper way is to use refinements, as do-browser 
could be do/browser as well)


What is a bit tricky about question marks is, that the meaning is 
not clear enough,e.g. - modified? Does it stand for the logic value, 
returning the true or false, or does it stand for the return of modification 
date? How should user know?


That is just my opinion on this topic - sometimes things are not 
easy to sort-out. Rules are rules, and we should probably stick to 
them ... the other thing is, if we are not forgetting another rebol 
"rule" (or at least principle) - make things pop-out to your mind 
at first sight, if possible. So - what is more self-explanatory - 
faces?, or get-faces (or what was the suggestion alternative)?

Take my notes as just another point-of-view,not a counterclaim to 
what you stated ...

You second ticket is interesting indeed ....

The points you made are intelligent and need a discussion, so, here 
goes:

'quit vs. 'quit-system and 'do vs. 'do-browse: Carl just pointed 
out, that if the -system part was unnecessary, it shouldn't have 
been used. That is clearly not the case of 'do vs. 'do-browse, where 
the second part cannot be seen as unnecessary

Whether to prefer 'do-browser or do/browser - such a principle was 
not stated explicitly, so, we do not have any "guide" which one shall 
be preferred

Well, I think I know why we went with do-*, open-* - simply to not 
overload those functions with refinements, and hence slowing them 
down ...

Carl should comment. The question is, if you can get his attention 
- he is not much active in R3 user-land last 2-3 months ....

I would suggest you going with faces? then, we can always change, 
while in alpha mode ....

modified? Does it stand for the logic value, returning the true or 
false, or does it stand for the return of modification date? How 
should user know?

 - the function name rarely suffices to inform the user what exactly 
 the function does. In such a case it is the task for the doc string 
 to inform the user, or, we should use two words, as described in 
 the doc.

(even the doc string may be insufficient, and an online documentation 
should be consulted in some cases)

As far as the 'faces? vs. 'get-faces names are considered, they both 
adhere to the function naming convention. The only advantage of the 
'faces? name is, that it is just one-word, so it is "more in the 
line" with the function naming convention in my opinion)

just a note - actually, it was not get-faces, but faces-of, IIRC?

'get-faces was your original proposal, AFAIR. Nevertheless, the 'faces-of 
name, which looked as preferred by the majority, is clearly violating 
the above function naming convention.

for me it's more important how it looks in code.. so it's for example:
	foreach face faces? window [ ... ]
vs.
	foreach face faces-of window [ ... ]

And, in fact, it is a C naming convention, except for the fact, that 
in C it would need to be faces_of

One of the possible solutions is to just add the *-of as an alternative 
convention, and have the option to choose. The only problem remaining 
is, that it uses two words instead of one.

What is interesting (and surprising me), is the fact, that, not reading/remembering 
the REBOL function naming convention, lots of people immediately 
were able to define any kinds of "ad hoc rules" which (purportedly) 
were in effect in Rebol for function naming, and used that as their 
argument why their preferred name was in accordance with the REBOL 
function naming convention.

I always had the impression that the "word names" section in the 
style guide is more descriptive than prescriptive. To that effect, 
I'm very much in favour of documenting *-of as an alternate convention.