World: r3wp

Will, thx, never heard about this one.

Yeksoon, yes, web traffic. I heard about Googly Analytics but thought 
you need to add special stuff. Isn't this the case?

Tomc, AWStats is what I currently try to setup.

Google Analytics requires you to add a few lines of Javascript to 
every page that you want GA to track.
That limits the GA stats to user-agents that execute javascript.

Ah, ok, that was why I don't like it that much.

.

I can definitely see:
	none? port/locals/headers/Accept-Ranges
but is it ever true:
	"none" = port/locals/headers/Accept-Ranges
?

Well I shouldn't be so lazy. Let me start collecting empirical data 
...

foreach url [

 http://brokenbeatassault.com/depot/files/BBA_MusicShow_Podcast_001_March_Safire.mp3
	http://dlsvr05.asus.com/pub/ASUS/vga/nVidia/nv16921_Win2KXP.zip
	http://debeveiligingsupdate.nl/audio/bevupd_0003.mp3

 http://audio.wnyc.org/radiolab_podcast/radiolab_podcast501choice.mp3

 http://ftp.iinet.net.au/pub/ubuntu-releases/kubuntu/8.04.1/MD5SUMS

 http://www.ted.com/index.php/talks/dan_gilbert_researches_happiness.html
	http://events.ccc.de/congress/2008/wiki/Streaming

 http://sites.romkids.org/virgill/albums/Bestof PC Demoscene/like a childbirth.mp3
	http://rockfactory.us/files/rebol_song.mp3
][
	port: open/binary/direct url
	print [mold port/locals/headers/Content-Length url]
	close port
]

Only two hosts returned #[none]:
	www.ted.com
	events.ccc.de
and the rest returned strings of lengths, eg: "4279194".

I should test more urls to feel more certain. (Or read http specs... 
hmm.. which gives better results ?)

#[none] means that the header was not there. if it's there, then 
it's going to be a string.

Yes, that makes sense. I think HTTP/1.0 servers are likely to not 
bother sending Accept-Ranges (thus it remains #[none]), while HTTP/1.1 
servers are likely to send it (thus set to either "none" or "bytes").

Anyone know what the proper way to parse this style of URL is?  Opera 
spits it out.  


http://sitecheck2.opera.com/?host=www.rebol.net&hdn=SDbXzuPvK4tsiGnHOia5dA==

Steve, could you be a little more specific.  There are a lot ways 
to parse your parse question :)


For example, are you asking what the "?" and the "&" mean in the 
URL?  Or am I taking too literal?

is it the last "==" which is confusing?  or the encoded data?

The == looks like padding on encoded data to me.

Though I'm not sure:

>> to string!  debase/base "SDbXzuPvK4tsiGnHOia5dA==" 64

== "H6????+?l?i?:&?t"

I asked a broad question because I didn't know what answer I need 
... ;-)


Specifically, this comes out of Opera and when I parse it to HTTP, 
I don't know if I should make a Request of:

Host: sitecheck2.opera.com
Path: /
Target: ?host=www.rebol.net&hdn=SDbXzuPvK4tsiGnHOia5dA==


or some other way, like perhaps Path: none ? and Target: /?host=www.rebol.net&hdn=SDbXzuPvK4tsiGnHOia5dA==

...and at a qualitative level, I think this might be used as a referer...Opera 
does this from the Home Page button.  I don't get it from other browsers.

The message about tables in here at 31-Jan-2005 2:42:06 AM wrecks 
the AltME tables when it sends its output the the Web Browser.  I 
have seen this before. (press the Print icon in AltME and look at 
this message group in the browser, table lines are missing)

Youzer...re-installing Opera 9.63 fixed the problem.  Now it works 
perfectly.  My guess is that the sitecheck2.opera.com was revalidated 
by the new install process, so the site allowed a connection.  Dunno, 
just a guess.  A couple days of trying to fix my code went into that 
unsure lesson.

:)

Please post the Table bug bug to Feedback, thanks.

not escaping html tags is quite serious bug.

Anyone have experience using Oliverier Auverlot's Magic! I have tried 
several times to get it up and running but never got there...

Also, Is there any advantage to it over RSP?

you can ask on french BBS (even in english) http://www.digicamsoft.com/cgi-bin/rebelBB.cgi
Several users use Magic

Thanks, I've tried about 4 times over the last few years and keep 
giving up. The config files keep screwing me. tried under XP, 98, 
using WAMP and under Ubuntu...

Did you manage to make rebol cgi works?
I mean, is the webserver able to run .cgi script with rebol?

Rebol/core need to be installed on the server, chmoded and you must 
have put the right sheeband at the beginning of the .cgi script.

A simple .cgi script like this one is a good test :

#!/usr/bin/rebol -cs
rebol []

print "Content-type: text/html^/"
print {
<<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2 Final//EN">
<HTML>
<HEAD><TITLE>Rebol CGI test page</TITLE></HEAD>
<BODY>}

print ["Server date time is:" now]

print {</BODY></HTML>}

Then, when this script runs well, put the same sheeband at the beginning 
of the magic.cgi file.

Then I have a file test.rhtml (handler for .rhtml must be configured 
in the webserver) with this content to test magic :

<html><head>Test Magic</head>
<body>
Magic fonctionne t'il ?<P/>
<REBOL>
	print [

  "<b>Oui il fonctionne !</b><BR>Date et heure du serveur : <i>" now 
  </i> <P/>
		"Configuration actuelle :<BR>"
		"m-library-path : " m-library-path <BR>
		"m-session-path : " m-sessions-path <BR>
		"curdir : " curdir <BR>
		"home : " system/options/home <BR>
	]
</REBOL>
</body>
</html>

I'll try those suggestions. It's a bit more methodical than what 
I was doing...
Thanks, Dide.

Did you consider QuarterMaster instead? I used to want to go with 
Magic!, but QM is a much more modern design

I've looked at QM but I did not get it. I did not understand what 
it did. Magic seems to provide some basic gui elements that seemed 
like a quick way to build a website. I'll have another look at QM 
though. 

QM runs on Cheyenne?

Yes.

Chris will correct me if I´m wrong, but I think that´s basically 
the point. :-) QM is a generic framework implementing a Model/View/Controller 
architecture that leaves you free to plug in parts for each, whereas 
Magic! offers a basic set of rigid REBOL web widgets

With QM you can basically build any website, without having to reinvent 
the wheel on the lower level web server technology, while with Magic 
you would quickly run into the walls of the provided functionality

Magic! could theoretically be an easier way to start, learning even 
less HTML, but if you can´t get through its configuration, you might 
as well start with QM

I have to say, that I don't get that MVC concept at all. It seems 
to me, like some academic head came with that model some time back, 
while reality might be elsewhere. Do you really desing your apps, 
that its fits MVC model? E.g. that famous Ruby and other DB "automatic" 
schemas which are supposed be so cool have to be rudiculous to someone 
who actually knows SQL. Soon after you start reading docs you find 
out, that when you get to more complicated DB schema, they only provide 
you with excuses that you can't have everything. Sometimes frameworks 
work against you :-) Not that I would not understand their advantages 
otoh, hence I am asking, if QM is really so usefull? The thing is, 
that I tried to give it a try several times, and I did not know what 
to do actually, but maybe I am extremely stupid for such stuff, and 
am only able to work with plain cgi/fcgi ...

If there is some full example implementation of some site for e.g., 
I could give it a try ... last time I tried was some year ago or 
so ...

You´re mixing up MVC and Active Record, and I vaguely remember discussing 
this before

Not mixing - those relate. Does not have QM as a framework an Active 
record like data organisation?

I do agree it takes time to figure out MVC, and I think the terms 
are too much inspired by a technological point of view. It´s really 
about separating the concerns of the usual stakeholders working on 
websites, or applications in general: the programmer/database designer, 
the consultant designing the business logic, the graphics/UI designer 
and the secretary maintaining the content

Yes, but Active Record is just an implementation of the Model and 
can be swapped out for something else

Basically always when someone thinks this separation is too complicated, 
he is thinking from his own perspective assuming he would have to 
maintain the whole site and hop between tasks

Once you start thinking from the perspectives of very different people 
having to maintain very different aspects of systems, it makes perfect 
sense; and coincidentally also explains many of the problems with 
existing systems through lack of separation of concerns

Would you write 20.000 similar web pages for a web store, or would 
you write one template and store the properties of the goods in a 
database, for example?

I am always thinking from such perspective! My simple tagging system 
completly separates presentation from scripts. In such regards, concept 
like rebol server pages is something I can't absolutly agree with. 
So if QM uses RSP, then where is the separation?