AltME groups: search

I agree that TO [a | b] is very important in a parser. I was surprised 
when I found that R2 doesn't do it

Isn't it just a nice thing to have? You write less, yes, but you 
could just parse for
	[to a | to b]


Some might even argue, that if you have a lot of alternate rules, 
it is more readable to have the keyword, TO, together with the alternates.

Geomol,
    [to a | to b]
is not equivalent (not even nearly) to
    to [a | b]

ok Then it's a good argument.

However, it is questionable whether a user unable to replace
    to [a | b]
by an idiom like recursive
    rule: [a | b | skip rule]
can be actually able to use Parse...

I feel, you need to explain this. If I replace a with "abc" and b 
with "str", and I want to parse ["abc"] :

(From R3 prompt:)

>> parse ["abc"] [to ["abc" | "str"]]
== false

>> rule: ["abc" | "str" | skip rule]
== ["abc" | "str" | skip rule]

>> parse ["abc"] rule
== true

Or did you mean

rule: [to a | to b | skip rule]

?

Geomol - try to understand your example. Imagine html file for e.g. 
Your [to a | to b ...] is wrong ... imagine 'a is at the end of the 
file. Simply "to a" will find it, before "to b" is even given chance 
to be applied. What we want is - find what is first - 'a or 'b ....

Geomol, I certainly did not mean the idiom you suggested. I just 
used an idiom for THRU, which can be adapted as follows::

    a: [here: "abc" :here]
    b: [here: "str" :here]
    rule: [a | b | skip rule]

Idiom I long time wanted - to [a | b] might be more expensive imo, 
as it will do incremental searches, evaluate the lower index and 
apply the rule. Whereas recursive [a | b | skip rule] will skip one 
positition at a time, trying to match rules .... or that is how I 
imagine the difference :-)

But - it was mostly trial and error on my side.

 - yes, Pekr, agreed with that. Once it is just "trial and error", 
 that is what I call a question whether you actually are able to use 
 Parse.

What I don't agree with Doc and Gab is - if you are a guru, just 
don't use idioms! But plese - don't take them from those, who might 
find them usefull! I remember, during the parse history, that exactly 
such idiom was often requested and ppl were not able to use parse 
without it, because they were not able to understand the rule Ladislav 
provided. And - will we suggest them not to use parse? That's imo 
not rebolish ....

if you are a guru, just don't use idioms
 - did they really say something like that?

Ladislav - I don't care if I am able to use some guru system - it 
either works for me as a user, or it doesn't. Reserving something 
for gurus should not be the point. Not in Rebol imo ....

But - I have a suspicion, that Doc is not aware of most R3 enhancements. 
Proof is, few months back BrianH just pointed him to the R3 wiki 
enhancement parse doc. So I expect him to study it. Not all enhancements 
are imo bad ...

Ladislav, so simple question - so would you discard to [a | b]  from 
the R3? Do you find it as a performance penalty, or teaching ppl 
wrong habits, preventing them to actually understand, what is parse 
about, or something like that?

would you discard to [a | b]  from the R3?
 - no, I actually support it being there

But I see the benefit of TO [a | b].

Well, we will see, what is Red's parse going to be about. I expect 
Gab to implement his compile rules, or something like that. Actually 
I never investigated his system, so we might get a bit different 
stuff, who knows ...

Do you find it as a performance penalty, or teaching ppl wrong habits, 
preventing them to actually understand, what is parse about, or something 
like that?

 - The answer is "no" to all the above questions. It is neither a 
 performance penalty, nor teaching people wrong habits or prevent 
 them to actually understand...

I only have R2 parsing in World today, as my initial goal was just 
to get to a point, where my R2 programs could run. It would for sure 
be an idea to look at the extensions at some point. And then desicions 
has to be made, if it should all be with extensions, if there should 
be more than one way of parsing, if it should still be mezzanine 
or made in C, or maybe some JIT compilation. Many options. :)

If it's not a performance penalty (and it certainly might not, I 
haven't looked closely at it), then I see no reason not to have such 
good ideas as is found in R3 parsing.

I remember now, how I solved my TO [a | b] situations in R2. It's 
what I call 'positive' parsing, where I all the time look for a positive 
all the way to either a or b is reached. It's easier to just let 
it skip to either a or b, whatever comes first, yes, but I was able 
to parse about everything the other way.


I need to look through all the R3 extensions to parse some day, when 
parse needs an overhaul in World. I guess, having parse as a mezzanine 
is a good thing to port it to also Red or other languages?

I imagine Doc will profile things if performance becomes his argument. 
I support what makes it easier to use PARSE to get the job done, 
though sometimes there may be confusion between those who understand 
parsing at a deep level and those who don't. I would rather have 
support for TO [a | b], and other rules, even at the cost of them 
having lower performance, versus people not being able to do the 
job at all. In docs, we can note the tradeoffs, and people can optimize 
if necessary.

Doc, is your plan for Red PARSE to use R2 dialect, or R3's?


R2 dialect plus a few of R3 additions that have no R2 equivalent 
or very complex R2 equivalent. Please don't ask me to make lists 
now, as it is not the current focus. When time comes to implement 
PARSE, we'll discuss every feature in detail if needed.

Is Gabriele still going to do the parse implementation?  Or was that 
just a rumour?

It was supposed to be a private info, but somehow private chats are 
leaking massively here these days. :-)

Is it a good think, that REBOL works on par with PEG, or are there 
more benefits in CFG parsing? Both methods have pros and cons, but 
maybe one of them would have advantages in relation to REBOL (or 
the REBOL way of thinking)?

#[[Geomol

Is it a good think, that REBOL works on par with PEG, or are there 
more benefits in CFG parsing? Both methods have pros and cons, but 
maybe one of them would have advantages in relation to REBOL (or 
the REBOL way of thinking)?
]]Geomol


- indeed, no need to reinvent the wheel, the arguments why the Parse 
expressions are the Rebol way have been summarized already

My view is, that if the first wheel wasn't round but oval or square, 
then it might be a good idea to reinvent it. And in programming, 
it's in general a good idea to rewrite every few years.


But the problem with REBOL is found in a different place. It's about 
making a solid foundation to build upon, and to make sure, each level 
of the tall building is solid before moving on to the next level. 
Much like it was described in "R3 Alpha Test Priorities":
http://www.rebol.net/wiki/R3_Alpha_Test_Priorities

Therefore it's cool to see a project like Red, that seems to be well 
thought trough, and where there is slow but (I feel) solid progress 
towards a goal.

As opposed to that, my opinion is that being pointed to a documentation 
you should read it before presenting opinions that "it may be oval 
or square" before even trying to read it.

Are you trying to shut me down?


I asked for Gerard's opinion about PARSE, if it should be more of 
a PEG or CFG parser. I read him, as he commented on R2 PARSE, that 
it works on par with PEG theory.

Is there a Blue language? If such a language was made, maybe it should 
be the exact opposite of Red? (Whatever that is.)

Blue is a system for teaching object-oriented programming, developed 
at the University of Sydney, Australia. It is an integrated development 
environment (IDE) and a programming language. Blue has been used 
for teaching since 1997. The development was stopped in 1999 when 
one of its principals, Michael Kšlling, began applying the IDE design 
to the Java programming language, resulting in BlueJ.


Sydney, Australia. Opposite side of the World, opposite of Red. :)

Pekr, it is just up to you for Red to be in that list as we already 
have a wikipedia page for Red. ;-)

Fortunately, github's helpdesk was kind enough to free that username 
after my request as it was dormant for more than a year. Cybersquatting 
names doesn't pay on Github!

Can anyone see a [Close] button on this page? https://github.com/dockimbel/Red/issues/484


Seems like a bug, or has Github changed once again their UI and hide 
the close action somewhere else?

It's part of Red/System runtime library which is undocumented so 
far. Some of the contributors have made some extraction tools to 
generate automatic docs, but none of the initiative has reached a 
usable state...

`allocate` function is currently just mapped to malloc(), my intention 
was to make a native fine-tuned implementation for Red/System, but 
as it's very low priority, it won't happen soon.

If you are not sure of your english being good enough, well go ahead 
and just do edit the site and we will tell you what can be approved. 
Nothing held against you! Happy to see a little more life (or is 
it live) on the site. together we can do more than on our own. It 
takes just some initiative now and then.

What about a youtube channel for Red? (Could also be a Youtube channel 
that is also for REBOL)

There are YT movies from the conferences. Someone could make a new 
playlist out of them

I could try to do something like that, but it will be a playlist 
under my account, not very findable/ related to REBOL/Red

That depends who links to it. Someone collected a Syllable playlist 
and I linked to it from our front page

There allready is a Red-playlist: http://www.youtube.com/playlist?list=PL595BBBBC4CD91581

I use a Linux/ARM image running on QEMU to help me debug the low-level 
parts of Android port (which is a really hostile environment for 
low-level coding).

Got a simple HelloWorld text displayed in an Android native window 
using the emulator with an x86 image. Working on the ARM version 
now.

This APK was generated from Eclipse using a minimal Java loader class. 
Red will of course provide a simple solution for creating APK package 
without having to install Eclipse or the Android SDK (though the 
emulators are still usefull for testing).

JNI has a limit of 512 max object references.

Pekr: you can test it here:
https://www.manymo.com/emulators


But the Hello app exits immediatly, not sure it's a bug or related 
to their emulator.

@Graham, thank for reminding Red is a real global project now ;)

@Graham, sep/oct/nov, so when Doc writes probably before autumn, 
I hope it will be a cold summer and maybe concurrency will be here 
by august.

Arnold: the devcon in July will suck up a lot of my coding time, 
as I need to prepare the presentation and that's something I'm very 
slow at.

@GrahamC, I will probably focus on some Red internals like how to 
achieve symbolic code execution in a hybrid compiler/interpreter.

I also plan to bring a surprize, if I can make it in time.

* He had a lot of technical questions about Red.  He was wondering 
how Doc overcame (or planned to overcome) some technical hurdles, 
like how he would handle compiling symbolic Rebol code.  He said 
he had no doubt Doc could overcome those hurdles ...

Very slow in making presentations? http://www.rebol.org/view-script.r?script=presenter.r

Well I know all about that, making a decent presentation takes a 
lot of time.

I've added a few more wishes to the Red/System V2.0 wish list. If 
you have any comments please let me know.


Please remember these are wishes for Nenad to consider, not firm 
proposals.

A designer knows he has achieved perfection not when there is nothing 
left to add, but when there is nothing left to take away.
 -Antoine de Saint-Exupery

I will settle for less perfect but with a lot of possibilities ;)
perfection = not making errors , making no errors = do nothing

Congrats on the Android release Doc - very exciting!  (sorry, I've 
been away from AltMe for a while)

On my Nexus 7, the hello.apk comes up with a main screen that has 
"hello" in the top bar. Is that correct? To exit it, you have to 
swipe the app away. It does not respond to a return/escape.

@Gerard, you make me very curious to what you expected Hello.apk 
should do. I think it plays a Lionel Richie song! ;)

@Arnold and Pekr : What is the current behaviour of the Hello.apk 
? I thought it was displaying some welcome msg and then askin for 
some input before disappearing gracefully. Now on my machine a while 
screen is displayed with a "Hello" as its window name and nothing 
else is going on.... For the previous Red/System Hello, something 
almost similar was going on too. So it's not for the Red part fault 
but it is due to the lower layer ... I knew this but I can't diagnose 
what ios coming myself as I'm missing some vital knowlwdge here to 
be really helpful to debug anything ... on my own - sorry for having 
acquired some product in the first pioneers just to test how this 
kind of product was running on Android. I required some Flash and 
my Apple products just could'nt do it - but my tablet Android tablet 
was much slower on display than my iDevices. Hope that under Red 
and Red/System there will be some products running at full speed 
- if not I plan to let it go at a real bargain price very soon ... 
to get some more powerful device ASAP.

Now on my machine a while screen is displayed with a 
Hello" as its window name and nothing else is going on...."


Gerard, that's exactly what the current hello.apk is supposed to 
do.

Me too - thanks Andreas - So I regretfully missed the point basing 
my view on the previous Red/System Hello.apk. - Really Sorry Doc 
for this misconception. Just thinking that I should have looked at 
the source code instead of being like a blind - asking unuseful questions. 
I'll take note as to not disturbing anybody in this way in the future....

Just in case: I don't see the "Hello Android World from Red!" text 
on a Nexus 4.

@Doc : So the text msg is really missing - only the Window's name 
is appearing on my device - Phew !!! I was not disrturbing you for 
nothing at least! Hope the last log results could help a bit - but 
don't stress too much if your time is limited (rare) and I suppose 
it is .... thanks for the screen capture you sent.

This fact seems to show that Samsung really is doing things in a 
non standard way ... at a very low level may be !!!

Gerard, it's most probably a faulty apk, we'll see with the next 
demo.

SYS: It seems that not much is required to add to the PE file itself, 
I see no additional sections needed. However, it requires a proper 
checksum in the header. That checksum is either calculated using 
Windows API or the algorithm could be implemented directly in Rebol. 
In the first case, it shouldn't take more than an hour to add support 
to it, in second case a day or two. The main issue would be to find 
a simple way to test it once generated? Any idea?

I have an example Windows driver that would be interesting to port 
to Red/System, so that could be a test case

I guess a 32 bit driver wouldn't work on 64 bit Windows?

I wonder how do you make the testing process simple and straighforward 
for a Windows driver.

The driver I inherited is a simple file filter driver, so it fits 
in the existing Windows infrastructure and it could potentially be 
almost empty

I'm not sure that you can just change a driver file "in place" after 
each recompilation. I'm wondering if you need to deinstall the old 
one and install the new one each time...

Haha, maybe not to that extent (well, I'm half-joking, for some kind 
of drivers, I would be surprised that is a requirement).

Yes, they use a stack-based allocation system, so each time a slot 
is popped, it becomes available for a new series. But, as series 
are not freed yet, slots are not popped.

Now how to time the script? Red/System and C, the executables have 
a start time in seconds in the terminal logs but that is imprecise 
and there is no endtime just a logout message.

On OS X, there should be a time command in the terminal

Did in a shell the following: 
time C-script
real	0m0.073s
user	0m0.028s
sys	0m0.004s
time Red/System script
real	0m0.059s
user	0m0.039s
sys	0m0.004s

It includes a bunch of display's but both the same.

Red/System seems to be only a third slower. Did you compile the C 
version with optimisation?

I will clear both scripts from the extra display's I built in during 
debugging.

Optimisation? Just hit build in XCode, happy it does something. You 
want the sources when I am ready cleaning them up? 

I thought about looping it a 1000 times each and see how it performs, 
but that is maybe only for more digits beyond the decimal point..

Looping is a good idea to get better timing accuracy

Having debugging code in there distorts the result, and a lot compared 
to the real code, since it takes so little time

Yes that is a lot, completely agreed. I'll have a third look. Seems 
the same stuff in the loops.

Knuth's code could compete at C obfuscated contests. Arnold seems 
to have done a literal translation to Red/System but keeping the 
same obfuscated symbols, so not easy to read. However, it seems at 
first look that the algorithm has been well preserved.

Here's a bug:

Yes, you were comparing

 - wrong - I was not comparing anything, nor complaining to anything 
 ;-) My question was more general, headed towards if in regards to 
 red/system architecture, the measure of being 8x slower than C (in 
 a concrete example guys were talking about), is good, or bad. I simply 
 don't remember outcome of prior discussions, that's all.

This is a more elaborate test, though

Here's a real algorithmic bug:

It's deceiving to write this in a form that looks like C. The Red 
evaluation rules compute this as

It should actually crash on what you have now, but if it doesn't, 
it pokes a value over an unitialised pointer, so any value in your 
program could be corrupted

Thanks for the compliment Doc, not really sure what you mean exactly 
by making it more like Red/System and less C: use more descriptive 
names? I will take a closer look at some ed/System examples out there.

Thanks Kaj for finding those and for the tips, the size of MM makes 
it the same in effect in this case, but it has to be <= then. Program 
not crashing, I was lucky then! off-by-one errors? My index goes 
from 1 up, where in C it is from 0 up, I had to debug this to make 
sure elements were swapped in the same way as in the original program. 
That is also why I declare KKP and LLP to as to save from adding 
1 to KK and LL over and over again. 


Knuth's algorythm was the first one I found, and I knew already of 
its existence, so it made sense to use what you have. Sure my Red/System 
code is not optimised.


Going to work on it now and tomorrow, and later I take on the Twister. 
It is a good exercise!

About odd, this solution managed to get all odd number from the even 
ones. Your solution is way more elegant, better fitting the language.

Base-1 Base-0, my personal view on this, it is the programmers choice 
on the level of the source code. What happens beneath the surface 
is compiler/linker sh*t. As a programmer and a human being I start 
to count at 1. 0 is not the new number 1 nor is 1 the new number 
2 etc. It is only an addressing issue, compare to the post. Houses 
in the street are numbered from 1 up to N. The first address a computer 
has in an array is the all 0 address, which is the first "pidgeon-hole" 
to be used. The computer doesn't know 0 as we understand it. Well 
you know all about it.

CPU are optimized for 0-based accesses. Using 1-base indexing will 
make Red/System a bit slower than it needs to be.

OTOH, you can always use pointer arithmetics to get a 0-base indexing 
model.

Funny

: #define odd? (x) etc as Kaj suggested gives a compile error on 
line where it is used: 
*** Compilation Error: undefined symbol: odd? 

*** in file: %/Users/Arnold/data/develop/red/testscripts/random.reds 
*** in function: ran_start
*** at line: 117 
*** near: [odd? ss [