AltME groups: search

Well, at least some refreshing new look after so long time :-))

Okay, i went ahead and changed "to-file join ..." to "rejoin [ ..."


Also threw in a little check for zero bit in calculating remaining 
time.

but at least spaces and digits could be used - it means charsets 
- which could be available during parse without need to define it 
all the time

To use the simpler of the CS terms:


Parse is a rule-based, recursive-descent string and structure parser 
with backtracking. It is not a parser generator (like Lex/Yacc) or 
compiler (like most regex engines) - the engine follows the rules 
directly. Since Parse is recursive-descent it can handle patterns 
that regular expressions wouldn't be able to. Since Parse backtracks 
it can handle patterns that ordinary recursive-descent parsers can't.


Basically, it puts the text and structure processing abilities of 
Perl 5 to shame, let alone those of the lesser regex engines.


In theory, Perl 6 has caught up with REBOL, but Perl 6 only exists 
in theory for now. By the time it becomes actual REBOL should surpass 
it (especially if I have anything to say about it).

I am also not an XML guru, but I will be by the time I'm done :)

You don't need to, but if you don't it will sit in memory until the 
next time you run the function.

I looked into rsp some time ago, and I liked it, especially as it 
was complete, with session support etc., but later on I found shlik.org 
being unavailable ...

to anyone not yet accustomed to 'PARSE, really do take the time to 
look it through and use it.

50% of the time its easier to match something which is not and then 
within that choice select something which is.

But you were not around at this time, were you ?

I tested sincerely, but to live without technology it is like living 
with the Middle Ages, one passes the three quarters of his time to 
produce his food and to make provisions for the winter .

I am having a hard time with using REMOVE on a parsed string.

reading? what is that? .... oh yeah... I remember ...  something 
people with time can do ... ;-)

I was thinking the same. I seem to remember, that at some time (some 
version of REBOL), -1 skip did work!? Hmm...

Aside from the one-time bind, repeat may be faster than loop with 
a self-incremented index.

my measurements show:

>> time-block [parse "a" ["a"]] 0.05
== 3.83615493774414E-7
>> time-block [parse "a" [#"a"]] 0.05
== 3.61204147338867E-7

, i.e. the opposite

use can use time-block to make sure (it's on my site)

Here's another benchmark:

>> data: head insert/dup make string! 10'000'000 #"a" 10'000'000

>> t0: now/time/precise loop 10 [parse data [some "a"]] now/time/precise 
- t0
== 0:00:06.078

>> t0: now/time/precise loop 10 [parse data [some #"a"]] now/time/precise 
- t0
== 0:00:04.296


Running this test several times shows that char! matching is, in 
average, 30 % faster than string! matching.

Well there you go. That's different numbers than last time, but more 
dramatic. It's just a #, easy fix :)

BrianH: "How does your reverse code handle "aa+", or "aa+a"" - damn... 
:)

Anyway, today I was thinking about different way how to do it, had 
no time to code it yet.
It's slow, I know...

>> t0: now/time/precise loop 10 [parse data [some #"a"]] now/time/precise 
- t0
== 0:00:06.389


>> t0: now/time/precise loop 10 [parse data [some [#"b" | #"a"]]] 
now/time/precise - t0
== 0:00:25.496

>> b: charset "ab"

>> t0: now/time/precise loop 10 [parse data [some b]] now/time/precise 
- t0
== 0:00:06.379

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?

; Try against this, the forskip code with the skip part taken out
forall: func [
    "Evaluates a block for every value in a series."
    [catch throw]

    'word [word!] {Word set to each position in series and changed as 
    a result}
    body [block!] "Block to evaluate each time"
    /local orig result
][
    if not any [
        series? get word
        port? get word

    ] [throw make error! {forall expected word argument to refer to a 
    series or port!}]
    orig: get word
    while [any [not tail? get word (set word orig false)]] [
        set/any 'result do body
        set word next get word
        get/any 'result
    ]
]

wrong, same problem ever, if it's the same sentence many time

parse data [all [rule1 | rule2 | rule3 | rule4]] 

all rules must apply 1 time and 1 time only  in any order, don't 
focus on what is inside rules (it contains sub-rules which contains 
sub-rules).

opt = match zero or one time

steve, we do that all the time, and we just take the last value when 
it is put more than once.

doc, integer! works for strings for some reason. it's the only one 
that works, and seems to match a sequence of digits (not sure if 
it does anything more than that). it's been there since a long time 
(probably from the beginning ;) but not documented.

I asked Carl at the first devcon for the rest of the datatypes when 
string parsing and he agreed at the time that if integer! was in 
there others should be as well.

There is some collation function which we need to wrap. I posted 
it here some time ago, but we were not succesfull in wrapping and 
utilising it. I also tried to look into IIRC Python sources, and 
it was not clear to me, how to specify it in REBOL level. IIRC it 
is callback type function ...

I've been playing.

login.rsp

<%
	user: "test"
	pass: "letmein"

        print [<html> <body> <pre>]
        help request
        help request/content
        help session
        help session/content
        print [</pre> </body> </html>]

	in-user: select request/content 'login
	in-pass: select request/content 'pass
	
	if all [user = in-user pass = in-pass][
		session/content/login?: yes
		response/redirect "/testapp"
	]
%>

REQUEST/CONTENT is a block of value: []

Everytime through...first time and after filling in the login form.

I was just going to try that... I started up cheyenne with -vvv this 
time  :)

Sorry gentlemen, I'm being dragged away for a bit.  Thanks for all 
the help.  Should be

back within the hour, but please don't waste any more time on my 
account.  I'll pester
people with more info shortly  :)

Cool.  I'll spend some more time on this tomorrow.  I know that it'll 
be a simple little
thing that I'm not seeing.  :)

It works okay most of the time ... but sometimes when I use the back 
arrow on my browser it crashes

don't loose your time with apache.. 8-)

The session cookie problem is related to incorrect timezone calculation 
in cookies expire time. I'm making a fix for Cheyenne, but due to 
a REBOL bug on Windows (http://www.rebol.net/cgi-bin/upnews.r?view=0005), 
I can't find an easy solution for that platform...

Actually, using..  do load %myfile  in my url-to-filename phase.. 
it's loading the functions each time, but I'll move them into main 
boot script when they're stable

RE: "How does Cheyenne's performance and security stack up against 
the other indians (Apache, Cherrokee, Hiawatha, ...)?"


Speed: close to Apache2.x for static files (tested a year ago, so 
new tests need to be conducted).


Security: not chroot-ed and no special protection against ddos attacks 
like Hiawatha, no bandwidth throttling (could be added), no CGI time 
limit (could be added), passes Nikto generic security tests flawlessly 
(last time check : few months ago)

Cheyenne is a key technology for my company, so we need to improve 
it rapidly, to be able to build higher levels frameworks, tools and 
applications (most of them will be open sourced). I currently have 
the opportunity to work most of my time on that project, that's why 
you see new releases coming often. I hope to be able to continue 
at the same rate for a couple more weeks, goal is to reach a v1 release 
candidate asap (some non-essential planned features might be kick 
out for v1.x in the process).

Time to reformat I guess...

I have a past project which *might* need to be replaced by cheyenne... 
but I will have tons of questions for sure... and without quick response... 
I am a bit wont to put time on this fix... since I am not totally 
sure it will do what I need....

Max: I'll be online in AltME from 9 jun to 13 jun most of the time, 
so, in this range, I should be able to answer your questions in  
12 hours max delay. But I guess other people, like Will, could also 
answer most of your questions. If you have a lot of questions,m aybe 
it would be better to send them on a private channel, or Chat channel.

No! I meant : do you mind logging in every time....

Terry: I may add an option to make sessions persistent on client-side, 
but that require cookies with expire time, means failing again in 
REBOL timezone issues...

Re: php / fastcgi, didn't had time to work on it this week, but it's 
high priority for my company, we need to put a few php apps onlin 
with Cheyenne, so I'll work on that in the next days.

Not accurately on Windows platforms due to REBOL inability to handle 
local summer time (daylight time saving period).

Terry: I'll provide an option in the next release to add session 
cookie expiration time, but will be "at your own risks" ;-).

FTP server, a service that I wanted to add to UniServe since a long 
time...

Oldes, thanks, I've made a minimal url-encode function (due to lack 
of time). I'll wrote a better one for the next release. Not sure 
what approach is faster, changing the argument series or building 
a new one...will test that.

Starting from 1st july, I'll be full time on Cheyenne.

does something for a long time ... but will probably timeout ...

Last time I've checked, Ethereal can capture packets on localhost.

hopefully so. First time I run php-cgi firewall asked me if I want 
to allow the port

A little update about the work in progress on Cheyenne's new version 
: I'm a little late on schedule (one week), I should release the 
new version tomorrow with several big improvements. The PHP/FastCGI 
final support involves some redesigns of mod-fastcgi that take much 
more time than expected. The good news is that it will result in 
a simple, fast and reliable interfacing with PHP. The other new features 
includes : text localization framework for RSP, support for any CGI 
scripts (including non-REBOL) and a new file upload system using 
temp files on disk (up to 2Gb files supported!). Once PHP support 
will be finished, I'll declare the first 1.0 release candidate (with 
encapped versions released too). I'll write all the docs during this 
debugging phase, so the completed 1.0 should be ready by the end 
of this month. I'm full time of Cheyenne starting from today, so 
it's doable.

full time of => on

Cheyenne new version 0.9.16 is ready. The new FastCGI multiplexed 
support took me some time to debug, but the resulting PHP interfacing 
is now really stable. I just need to update the RSP API doc and run 
a few tests, so the download link will be published here in a couple 
of hours.

On windows platforms, you'll get the infamous DOS window flashing 
when executing an external CGI ! It's just a matter of 1 flag to 
correctly set in 'call C source code, if you're really annoyed by 
that, ask RT to fix it asap (for 2.7.6 that would be good)! ;-) I 
may reimplement completely call command in REBOL, but it would be 
a big waste of time and energy...it should be a 10 minutes fix for 
RT. Addind a time limit to 'call would be a good thing too, it would 
also avoid me the reimplementation of 'call to add such feature....

BTW, I didn't have time to test this new release on linux, especially 
the external application launcher...Will test that tomorrow.

I only have time in the weekends to do any development :(

It should be controled by a config file option, but I didn't had 
time yet to implement it.

Doc - so I just tried on my WinXP, and I have similar problem as 
on Vista. First time test.php is requested, it displays the table. 
Trying to reload page gives me "No input file specified".

We all spend tooooooooooooooo much time with these types of bugs...such 
a shame....

Reichart;   In defense of Doc et al, Cheyenne is still a Beta, someone's 
gotta spend the time  :)  might as well be us.   And Vista...XP, 
well it's MS and "good enough is good enough" seems to rule the day 
in Redmond.


And yes it is a shame.  If motorola had won an early lead in the 
chip wars, the 68K flat memory model would have saved an untold millions 
(billions?) of man hours for the PC industry.  If MS hadn't been 
allowed to overhype Chicago (for what, 3 years?) and then FUD  OS/2 
to death, we'd all have object oriented desktops that we could talk 
to by now. But we plug ahead, mostly oblivious.  :)  Sorry, this 
should have been in Vent, so I'll end with

Go Doc Go

Just to be clear, I'm lamenting the time Doc needs to waste.  


For example, I moved my WebCAm from one USB port on my PC to another 
on the same PC.  This apparently made Windows think it needed to 
reload the entire installer for the drivers..............................something 
like 50 megs, plus about 15 minutes.  WTF!?!


Earlier today I wanted to VNC into an office.  2 hours, and several 
IT people, and it was 5 seconds per frame update.  WTF!?!

I'm tired of working out simple problems over and over again…

I guess it should not. Currently, what takes me time is the web control 
panel that I'm adding to Cheyenne and the UI part (HTML) really takes 
a lot of time. Once it's ready, I'll release the 1.0 rc1.

R2 is not yet dead, it will takes time for R3 to mature and become 
as stable as R2 is now.

Btw, work on Cheyenne keeps advancing but slower, due to my lack 
of free time, customers have a higher priority...

It's a french expression meaning : someday ...(just kidding) ;-). 
I need a couple of days of work on Cheyenne to reach the rc1, but 
I only have an hour or 2 to work on it, here and there, so it's hard 
to schedule a release date...the customer project I'm leading is 
taking almost all my time.

So how much money do we need to raise to buy you enough time to finish 
it?

This is a serious remark - Cheyenne is a killer app and if a lot 
of the rebolers donate someweher between $20-$50 you mighht be able 
to work dedicated on it for some time. The community would get a 
great product....

so Apache invokes the rebol interpreter each time .. whereas with 
Cheyenne the interpreter is always resident

My brain hurts..I'm using Rebol to make my life easier. I do not 
have time to figure out a new wiki or  whatever every other day.. 
Nothing against you smarter guys but Iove rebol because I'm finally 
able to see a problem or need and solve it with a simple rebol app. 
!Cheyenne is my hope for an easy rebol path to the web without Java 
script, php, perl, AJAX etc.

This conversation has made me decide to pursue getting vanilla running 
under .9.16  So close to Cheyenne 1.0rc1 though.   I think it'll 
be time well wasted.  


Alan...still need something delivering off of port 80 though no? 
 Maybe not HTML in the future but Cheyenne is an awesome http delivery 
system.  imho.

Yeah, QM conflicts with  request  and Vanilla conflicts with  session 
 (at the quick glance I've taken so far).  These types of problems 
are pretty easy to fix given the motivation.  REBOL is eminently 
readable; all it requires is a little motivation, time and judicious 
use of context or global find and replace.  Collisions are always 
hit or miss, but the web related scripts usually collide just by 
nature of the higher level words being the perfect words for the 
concept at hand.   People running Apache would never see these collisions.


R3 holds a lot of promise in allowing us rebols the freedom for independent 
development that others can combine in fun and magical ways, worry 
free of this issue.  Another year and this type of complaint should 
be a thing of distant memory.

Yes, that is my hope.  I've been pining for the namespaces for some 
time (instead of building up a succession of leaky objects).

not yet but he is working on a new release, probably a good time 
to send in bugs (if any!) and feature requests http://softinnov.org:8000/curecode/

RSP is the best way and the faster inexecution/response time (instead 
of Apache cgi, it doesn't need to load rebol and your init library 
at each request), to do dynamic stuff with Cheyenne. The API is very 
sleek and you can read about it in the downloadable documentation 
part of Cheyeene download.

Considered yes, have time to do it, no.

You may not have the time,  I might though :)

ahh - if it happens again let me know and ill try to track it down 
- theres a check on zooming cards that only lets it happen at certain 
time

LOL :-)

I had some good sleeps this week-end, as I had a fever and a bad 
throat. I also had time to play Assassins Creed and the old Outcast 
some, so I do other things than coding. :-) I've been working on 
many of the things, I post here in REBOL3, for some time, before 
I post it.

I'm thinking of adapting Rebtris for a very basic network-based competitive 
play.  The idea is several people log in (to a minimal CGI server), 
everyone starts at the same time and scores and game-over times of 
all players are constantly refreshed and displayed.  Ultimately determining 
who a) had the highest score and b) lasted the longest.  The game 
would be capped so that no one could realistically survive past 15-20mins 
(I have my reasons).

i don't know what the Wii is using so no idea on what the best way 
would be. using opera might be overkill, otoh maybe it'll save you 
a huge amount of dev time.

Thanks Reichart, I will :)  I've played a little with the OGRE engine 
in Purebasic - I think it's time more of these standard tools start 
getting wrapped in REBOL, now that the command potential of REBOL 
has been opened up.  I personally think that the pay-for-api access 
was one of the main reasons that REBOL never became as popular as 
it should've.  Wrapping cool tools should really help open more eyes 
as to what REBOL can do, at least on initial look'n'see.

(Geomol - it's a complex subject , I need some more time to reply 
:)

good game.

took me some time to figure out lvl 10 as well

yep, and i think this was second time (i've seen it when it was released, 
maybe not first position that time though.)

what do you mean by it being clumsy? Did you have exp. with any other 
3d engines at the time? (I was so far only making 2d games, but I 
wasted a lot of time looking and trying various 3d engines)

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.

I helped some friends about the same time, that were starting a game 
developing company. I had a little experience there first with Torque 
before they changed to Reality Engine, which used C# and C++. I left 
after a few months, because I saw, it wouldn't happen. 2 years later, 
they stopped the development.


Unity had a development tool with lists of content and parameters 
to set. At first it looked good, but in practical work, it fast became 
a huge mess. It was difficult to find and set parameters, so a lot 
of time was wasted in the process of creating a game. The porting, 
I was doing, ended up being a totally recreate in some areas, like 
the animations. The data couldn't be kept and converted from the 
old version to the new, but had to be recreated. I even tried to 
figure out the binary files using REBOL, but was very hard, as much 
of it wasn't documented. That a bad sign for me, when we talk going 
from one version to a newer one with the same engine.


But it could have changed in the last years, I don't know. Haven't 
looked.

I did start writing a Rogue like game with Rebol today though :) 
I did try to make a jump and run game half a year ago when I first 
started to learn Rebol, but the keyboard events made that impossible 
(at least for me because I don't know if you can improve the event 
system). It takes a few seconds until the keyboard event is sent 
continually when you press and hold a key, like for walking left 
or right. For the rogue like game this isn't a problem since you 
always move one step at a time which also acts as a turn in the game.

That's the irk I have with PC gaming: That you have to upgrade your 
hardware every 6 months and wrestle with drivers. I advised my nephew 
again to get into PC gaming for those reasons as I don't want to 
change the diapers on his PC, every time something doesn't play.

a big step in GFx occured at about the same time, you could actually 
call it a leap.   Hardware shaders, better illuminations and gfx 
ram jumped to be 256MB at minimum on most cards.


but most games of today allow very low settings which allow you to 
work on older machines... I've got a 4 year old 1.5Ghz core duo with 
an old 8600M nvidia card.


and SC2 is quite responsive on it.  though I would like to play with 
a bit more graphics quality... its not the game's fault.  4-5 years 
ago 3D game graphics quality was still pretty deficient.  but with 
any average 100$ card today,  you'll get pretty amazing graphics.

Finished it today too, some time this morning. Played the whole damn 
night yesterday, I was shocked when the sun rose :) Yeah the end 
left a lot to be desired but Starcraft 2 is a trilogy, so that was 
to be expected. I only hope Blizzard will really manage to get the 
second game out in just one year. I can't really see that happening. 
It'll probably be 3 or 4 years. Did you get to play all missions? 
My mission counter says 25 of 26 missions completed. Where's the 
26th mission and why wasn't I able to play it? ^^ And if I do count 
by myself I get to 28 completed missions in the mission archive anyhow 
so maybe their mission counter is just one big bug. The missions 
are very short but I'd say it took me between 30 and 50 minutes for 
each mission and with 28 missions + 4 bonus protoss missions, this 
is a whole lot of content. Most games only give you 7 to 12 hours 
of gameplay (some a lot less, some a lot more). Starcraft 2 took 
me over 20 hours to play through. And then you still got challenge 
missions, multiplayer, custom maps Vs A.I, user contributed missions 
thanks to the level editor, etc.

Right, that could be the case. I guess I'll just try it out some 
time :)

minecraft, still in alpha, probably the best indie game of 2010 (and 
before)  http://www.youtube.com/watch?v=ogA9Bhh4luA
It's stealing all my free time (and some work time too)

We just released one smaller project. It's Flash again, and almost 
no REBOL used this time.. only just a few file manipulations and 
find/replace while moving from AS2 to AS3 code :/
http://amanita-design.net/games/osada.html

Not free?? I played it for a long time :)

Kaj, .... Gab pretty much nailed it.


Endo, .... Indeed, so are you saying you can't beat this go program 
every single time in just a few moves?  I seem to be able to beat 
it every single time almost instantly, which I think was Ladislav's 
point.


I'm just confirming I did not simply find a way to beat it as a sort 
of trick.  I used to design AI for games, so I tend to get around 
it pretty quickly.


If you would like I will send you a video of me plaything this game 
several times.