World: r3wp

Actually MS is even making situation worse with IE7 ... they will 
not backport to W2K (not to mention W98), still large install base 
of WinOS, so one other "platform" for developers to support ...

I read most discussion boards comments carefully, once some article 
about browser X, company Y is published somewhere ...

The web standards annoy me.  They are supposed to help make things 
more compatible but this isn't the case.  It is easy to build a webstandard 
complient web site that doesn't look ANYTHING like it should according 
to the standard in ANY browser.

yes, maybe it is impossible at all, with what HTML and related technology 
bloat evolved into ...

but there is still some common ground to support, where MS behaves 
like total morons ...

I use Opera and Mozilla each equally, and they work fine for me.

Firefox, I like the look of, but I had some trouble with -- but there 
are some good reports that the latest release fixes their main memory 
leak problems.

how, we have been trying to put Opera on Mac, and nothing but problems. 
 After 5 tries, have not even been able to download a full copy. 
 The website hangs, very odd.

Which is as good a reason as any to be happy that there is more than 
one browser to chose from.

mMesssages are coming in at about 1 per second.

Oops, sorry that was for a dif group.  I like choice, and I need 
to DL Opera so we can make sure we are compat, but it is really fighting 
us.

As a backstop, you could try getting some older versions of Opera 
from a browser vault:
http://browsers.evolt.org/?opera/mac

Maybe then an old version of Opera will update itself to the latest 
for you....

guys, you are unbelievable bashers of Mozilla
 I'm not! ;)


I've been testing four different browsers on my Mac (Safari, Opera, 
Firefox and Firefox PPC - http://www.furbism.com/firefoxmac/) and 
while the PPC build is 9.5MB compared to Opera's 5.5MB (which also 
includes M2 mail), it is noticeably faster than the other browsers 
and has not crashed once since I installed it 2 weeks ago. The only 
problem I've encountered is with my !@#$%^& bank's IE-only site (even 
with Opera I have to change spoof modes depending upon which particular 
page of the site I'm at, and Safari works fine except when the site 
tries to open a PDF statement within the browser using an Adobe Reader 
plugin – never mind the fact that Mac handles PDF natively ... !@#$%&).

Thanks Sunanda, I will try that. 
Ashley, yeah, we have been testing IE 7, same thing, Banks!

Go to agree with Petr .... Firefox works fine for me ... even my 
banking sites .... I cant remember tha last time I had to fire up 
IE.

I mostly use Safari on Mac these days. It works with my bank too. 
:-) When I'm on Windows, I mostly use Opera. I used to use Mozilla, 
and I still use Firefox from time to time, both under Windows and 
Mac. I very very rarely use IE. Safari can be used for 99+% of the 
sites, I visit. Today I had a problem, because I wanted to watch 
the 2 danish Superbowl updates, our reportes sent from the US. And 
a danish tv channel TV2 Sputnik require IE6 under Windows to run, 
and only that. Argh!

geomol, have you tried flip4mac yet? it works impressively with WMV 
video

though not TV2 Splutnik

Nope, haven't tried that one.

Has anybody experimented with emulating web continuations in Rebol 
? some info on ruby approach is here (http://www.rubygarden.org/ruby?Continuations) 
and Factor (http://factorcode.org/cont-responder-tutorial.txt)

http://lisp.tech.coop/Web%2FContinuation for lots of reference info 
on the topic

The technique came from Scheme. But for this technique to work you 
need that the language suports continuations them natively. REBOL1.0 
was able, but continuations were removed in 2.0, maybe 3.0 will have 
them again. We hope to be able to incorporate them in Orca, with 
addition of tail recursion, and other goodies ;-)

Yes,  we took them out. REBOL ran a lot faster as a result.  I used 
to be a huge fan of continuations 20 years ago. But, continuations 
do not provide enough benefit for the performance hit on evaluation 
speed and memory usage.  (Stop and think about what is required internally 
to hold in an object for any period of time the entire state of evaluation.) 
 It's more of a programmer play toy than a useful extension.

I believe this the paper with the original work. 'Modeling Web Interactions 
and Errors'

http://www.cs.brown.edu/~sk/Publications/Papers/Published/kfgf-model-web-inter-error/paper.pdf

Are continuations base for tasking/threading?

Do whatever you want with Orca, it is just that when I mentioned 
continuations few years ago on ml, Carl got on steroids and posted 
something like being-there, done-that, don't try to teach me how 
should I design language :-)

If they are not really much usefull, I am not sure I am the one who 
calls for "language purity" just because of the language purity itself 
...

Hm, to me continuations reminds of GOTOs, which can't be good!

I am not asking for native continuations but a way to emulate them 
in web applications.

Geomol, the real advantage of continuations is for handing web forms 
and to ensure the users get a consistent experience. Check the paper 
Jaime points out

Joe I gues you could emulate continuations. But it is not an easy 
task. I have done some work on this direction by creatinga Monadic 
extension. But it is not yet complete.

The problem I trying to solve is strictly for web programming, e.g. 
ensuring there are no inconsistencies in a shopping cart, etc ...

The approach I have is that  every session has a cookie and disk 
storage associated to the cookie. When I define a web form, the action 
method gets a continuation id as a cgi parameter, so if at that point 
you clone the browser window, you as a user have to continuation 
ids

correction: two continuation ids

This  approach is not very scalable, it's just a start waiting for 
better ideas and input

When the user posts a form ,  the form cgi stores the continuation 
id and a rebol block with name-value pairs

If you post the second form also (something you would do e.g. when 
checking flights in a reservation engine, as Jaime's reference paper 
suggests) a second continuation id and rebol block would be stored 
for the same session

So basically the continuations are ensured by using both the cookie 
and associated storage and the continuation id that is added to the 
links as a cgi get parameter

I'll stop now so that I get more input from others. I imagine many 
of the gurus here have done something like this as this is the thorny 
issue with web apps

What you are doing Joe is what we old-timers call pseudoconversational 
processing.

Usually, you can kick much of the complexity upstairs if you have 
a TP monitor supervising the show. Sadly, most web apps don;t (a 
webserver doesn't quite count).

People have been doing this sort of thing for decades in languages 
without continuations support; so, though it's a nice-to-have feature, 
it is not a show-stopper.

Joe you are asking a question that finds its answer in a completely 
different model.  It reminds of the joke "What I meant to say, was, 
Mother, would you please pass the salt,' (look it up).


The answer is to throw away the brochure (page) model of the web, 
and move to web 2.0, where there is a cohesive (continuous) model.


The UI is complete separated from the backend, and the UI is a single 
entity, that is persistent during the session.  Everything else is 
simply a pain.


Most sites are horizontal (shallow) as opposed to vertical (deep). 
 And most are still modeling on the brochure (page) as opposed to 
the space (like a desktop).

I'm administrating some pages where is a lot of text articles published. 
And because 50% of the trafic is done by robots as Google crawler, 
I'm thinking about that I could give the content of the page in Rebol 
format (block). Robot will get the text for indexing and I will lower 
the data amount which is transfered with each robots request, because 
I don't need to generate designs and some webparts, which are not 
important for the robot. What do you think, should I include Rebol 
header?

That's a form of cloaking. Google does not like cloaking, even "white 
hat" cloaking of the sort you are suggesting:
http://www.google.com/support/bin/answer.py?answer=745


Better to respond to Google's if-modified-since header -- it may 
reduce total bandwith by a great deal:
http://www.google.com/webmasters/guidelines.html


Also consider supplying a Google Sitemap -- and that can have modification 
dates embedded in it too. It may reduce googlebot's visits to older 
pages
http://www.google.com/webmasters/sitemaps/login

But it's not just google who is crawling, at this moment I recognize 
11 crawlers who check my sites regularly.

Some of them are just bad -- ban them with a robots.txt


Some (like MSNbot) will respond to the (non-standard) crawl-delay 
in robots.txt: that at least keeps them coming at a reasonable speed.


Some are just evil and you need to ban their IP address by other 
means...Like flood control or .htaccess

REBOLorg has a fairly useful robots.txt
http://www.rebol.org/robots.txt

So you think I should not use different (not so rich) version of 
the page to robots.

Yoy could try that as a first step:

-- Create a robots.txt to ban the *unwelcome* bots who visit you 
regularly .

-- Many bots have a URL for help, and that'll tell you if they honour 
crawl-delay....If so, you can get some of the bots you like to pace 
their visits better.

If that doesn't work: you have to play tough with them.

I don't need to ban them:) I would prefere to play with them:) Never 
mind, I will probably make the Rebol formated output anyway. If I 
have RSS output why not to have REBOL output as well. Maybe it could 
be used in the furure, when Rebol will be able to display rich text.

Chceck if you can turn HTTP compression on with your webserver. It 
saves bandwidth with visitors who are served the compressed version.

The bandwidth is not such a problem now:) I was just thinking if 
it could be used somehow to make Rebol more visible.

Having REBOL formatted output is / can be a good idea: REBOL.org 
will supply its RSS that way if you ask it nicely:

http://www.rebol.org/cgi-bin/cgiwrap/rebol/rss-get-feed.r?format=rebol

But *automatically* supplying a different version to a bot than that 
you would show to a human is called cloaking and the search engines 
don't like it at all.

If they spot what you are doing, they may ban you from their indexes 
completely.