AltME groups: search

Sunanda - thanks a lot, did not know I could use more than one class 
for one particular element ...

hmm, now I read that browsers for PDAs do very little support of 
css, that is a pity ...

The biggest restriction of CSS is you have to work harder to build 
layouts.  Otherwise CSS gives you *all* the control that depricated 
HTML attributes give you -- *plus more*.  So perhaps there is a dominant 
web style used by those experimenting with CSS (understandable if 
designers are learning from the same source, whether that is Eric 
Meyer, A List Apart, or WPDFD); but to use CSS as the excuse for 
this is a copout...

Perhaps this is a better example http://www.meyerweb.com/eric/css/edge/curvelicious/demo.html

CSS Workflow -- easy: create a concept based on information needs; 
plan how to acheive this with the box model; create a base HTML template 
and build up styles around that, incorporating background images 
as required; then test and revise, test and revise, test and revise, 
etc.  Simplified somewhat.  Basically the same as any legacy HTML 
project, only easier.

Eric Meyer is a great resource for all technical aspects of CSS. 
 He is, by his own admission, not a designer.

Re. JS -- Netscape 4 used a technology called JS Style Sheets -- 
it was a horrific mess.  IE lets you evaluate JS within CSS -- it 
is a security risk, and of course does not work if JS is turned off. 
 Think of CSS as like an image format -- you don't evaluate PNGs, 
you just display them.

on Gab's Temple. 


pre-defining the 'id' is a way for developers and designer to communicate 
and agree on certain stuffs. That is how the separation of work is 
being done. Developer will code with 'address' in mind in their business 
logic..and designer just design the layout and only need to put in 
id 'address'.

I did not say I don't like how css look. It looks maybe even better, 
kind of a book design -more typographically correct, more blending 
of images and text, which was not easily possible with tables etc., 
simply different ...

re workflow - interesting - I will order Zeldman's book too .... 
Alistapart seems to be a good resource too. I never build large site, 
just few webpages, generated by script or using DreamWeaver 4 long 
time ago. My mine problem right now is - how to start to think of 
a design. From typographics area I know I should divide page into 
some sections. I did so using tables extensively, now I wonder, if 
I should use css columns, simply using "div" tags ...

try e.g. http://www.dxt.czand choose whatever category on the left 
... products are placed in a grid (table). I wonder if you would 
use table or go without it ...

personally, within the product category, I would not use table. But 
the whole left column (seach field, mailing list subscription, prodcut 
category ) will probably sit within a table

ID and class -- the "problem" (aka "advantage") of an ID is that 
you can only use it once on a page -- it has to be unique.

That's a good reason *not* to use it for the page design. So leave 
IDs to the template people. Then there's no clash.

For a product list -- a table -- then yes, I'd use a <table>. It's 
what they are for.

If it'sa single column table (just the product name, say) use a list 
-- <ol> or <ul>

That way the page is "semantically" marked up -- that helps accessibility 
aids.

The biggest advatnage you'll find in workflow is when the client 
starts making changes.


1000 pages all with <font color=blue> and they now want all <p>s 
to be green -- that's not a search and replace job -- you need to 
check the context of each <font> tag.. It might take hours. Next 
day, they ask for dark yellow.
In css:

p {color:blue} -- you can show them the whole site changed in 1 minute.


Though, of course, you need to think through the styles you need 
first. That's a big bit of the design.

Interesting that Firefox next release is intending to have a by-site 
user CSS feature. So I'' be able to personalise *any* site I view 
regularly in Firefox by simply adding some CSS.


I can see a market for customised style sheets (make MSN look good, 
or mold Google to my corporate colors, or just look funkier than 
the boring original). Afterall, websites are just about the only 
thing on my computer that I can't reskin.


In the future, sites that are easily reskinnable, may have a competitive 
advantage over those that are not. That means using CSS.  (CSS was 
always meant to make this possible. It's the browsers that have been 
slow).

[a href "http://www.rebol.com"] would match any link to that url... 
and so on

i use it mainly for [input] tages, when i want to match an input 
tag with a given name, etc.

yes petr, i add things as i need them. i just don't have the time 
to polish it up and design a dialect for it

on speed with respect to php (and probably other languages).


smarty itself can be further 'accelerated' with accelerators like 
ionCube 
http://www.php-accelerator.co.uk/


So, if one was to ever go down this lane of comparing speed, then 
there is a need to compare
1. tempate system alone
2. with accelerators on.

petr: Temple has a LOT of room for optimizations. :)

I would think FastCGI should provide good enough performance. 


Reichart's team is gunning for 0.30s processing for their scripts 
in QTask. That seems a reasonable limit.

Pekr - http://www.stopdesign.com/log/2003/11/13/nolimits.html- from 
a well-known designer and CSS proponent regarding workflow.

I have a problem with some layout.


Trying to align an output of 'processingtime' and some 'footer' on 
the same line.. but I always end up wiht the footer a line below. 
The left-right alignment is correct though.

---
.footer {
	text-align: right;
	font:10px verdana, arial, helvetica, sans-serif;
	color: #666;
	background: transparent;
}

.processingtime {
	text-align: left;
	font:10px verdana, arial, helvetica, sans-serif;
	color: #666;
	background: transparent;
}
--

--
<div class="processingtime">some processing time</div>
<div class="footer">some trademarks stuffs</div>
--

Span wil get you a similar effect, though there are differences.
Never heard of idiv -- is it a browser-specific thing?

.footer {width: 50%; float: right; text-align: right;}
.processingtime {width: 50%;}


Should be enough on the face of it.  Adding margins or padding will 
need a little recalculation...

a bit early for me to be thinking but  I know I have used  "no break" 
tags  aroud things I did not want new lines between

Tom, Yeksoon is trying one of those things that look simple, but 
take a little care -- two bits of text on the same line, with different 
justifications (left and right). 
Easy with a one-row, two-column table -- but that's overkill.

Easy with three divs -- one containing the other two, but that's 
close to overkill.
Otherwise, a bit of lateral thinking is called for.

Found a tool for css and firefox. Opens an editor in the sidebar 
with css for the main-page. and updates main-page on the fly on edit. 
https://addons.update.mozilla.org/extensions/moreinfo.php?application=firefox&version=1.0&os=MacOSX&category=DeveloperTools&numpg=10&id=179>EditCSS

I believe it is a simple alternative to %index.html being the default 
document...

<HTML>
<HEAD>
<TITLE>NEVIDITELNÝ PES</TITLE>
<FRAMESET rows="*,0" frameborder=0 framespacing=0 border=0>
<FRAME SRC="./index.php3">
</FRAME>
<FRAME scrolling=no SRC="./no.html">
</FRAME>
<NOFRAMES>

Your browser does not support frames, please click <a href="./index.php3">here</a> 
!
</NOFRAMES>
</FRAMESET>
</HEAD>
<BODY>
</BODY>
</HTML>

RSP strange error: possible, but didn't saw any of this kind for 
a while under IE.

Tested on a linux box with Firefox : no error.

rebooted, trying 4 example, it loads 3 sessions imediatelly, stops 
not displaying anything for fourth box, just got a busy indicator 
...

Maybe I should post a news on slashdot stating that the RAMBO is 
released and point to REBOL.com website ? ;-)) Looks like this method 
works quite well ;-)

Is there supposed to be a file called show.r? I can't find it.

pekr, I found a Uniserve task buried in the background :-)

Probably a stupid question, anyway, how to change the port to something 
else when running the cheyenne.exe ?

there's a -p command line option to set the listening port (80 by 
default). Example: cheyenne -p 8000

I see you plan on supporting SSL... is this in any way functional? 
or should we simply do a reverse-proxy through apache for now?

It could probably call php as a cgi, but I don't think it's designed 
for such a purpose. And about version - the latest available are 
probably the best.

Maxim: I plan to support SSL natively in REBOL. I'm waiting for Carl 
to send me a test version of /Command with server-side SSL enabled. 
In the meantime, I recommand using a SSL proxy rather that Apache. 
See STunnel for example.

The SSL support will be available only in Cheyenne binary releases. 
You'll need a /Command or /SDK license to make it work with the sources.

Scot: Cheyenne can interface with PHP using FactCGI. A FastCGI client 
alpha implementation is currently built in Cheyenne. All you need 
is a php executable compiled with FastCGI. I've already tested several 
php apps (even big ones like eGroupware), it works well and it's 
fast ! But, the configuration options are currently hardcoded in 
Cheyenne (need to export them in config file) and there's no dynamic 
php instance management (you have to launch each php-fastcgi server 
by hand). Once these 2 feature will be integrated in Cheyenne, you'll 
be able to run easily and php application.

is FastCGI.r usable separately as a free fastcgi protocol for REBOL, 
or does it work only in terms of Cheyenne?

Oldes: View can be a good choice for server is you need server-side 
dynamic image generation or manipulation. The drawback is for Unix 
servers, they'll require X libs to be installed.

Pekr: The protocol is not specific to Cheyenne, but you'll have to 
"extract" it from the module to be able to reuse it. But, we're talking 
about a FastCGI CLIENT implementation, and what you have in mind, 
I guess, is a SERVER implementation (like in /Command)...

It connect to a server application, like php-fastcgi or a REBOL FastCGI 
server built with /Command, or any other server accepting FastCGI 
clients. Then it sends the HTTP request and wait for a response back.

Doc: there is a SSL serverside command version (I know bcause I can 
see it...somewhere) so you should really push it.

On windows, a 2.6.2 Command version, no encap

maarten, thanks, I'm needing it too.  if cheyenne can handle big 
loads, and SSL, there is a good chance I'll try to use it.  The release 
commes out at a really good time for me.

Doc: why not bundle mysql:// with it as well, and add a couple of 
CGI/session/cookie libs?

Will "Mike, I started using Apache and rebol as cgi, this is not 
suited for performances as on every call to the cgi, a new instance 
of rebol is  initialized, run and closed.

I thought about using fastcgi, but never came to a working solution.
Now I use uniserve as main webserver, here some advantages:

-it is fast! On my local machine I get +- 600 req/sec for static 
pages and a max of 160req/sec for dynamic rsp pages

-it is written in rebol, I could easly(less than 10 lines code) add 
a rewrite engine

-child process are persistent, this mean you can keep state of your 
web applications, implement caching, keep a pool of connection to 
databases open (in apache + rebol/cgi you'd have to open and close 
the connection for every request)
-it is written by Dock whom I may be the biggest fan ;-)

btw I'm running an unreleased version (have bought commercial support) 
 that support http 1.1, stuff like If-Modified etc..

If you have more specific questions, I'll be glad to try and answer."

I did some bench with Apache 2.0 a few month ago: results were showing 
10% slower than Apache for small files, and 20-30% faster for > 64kb 
files. Theses results are very encouraging given the fact that Apache 
is compiled C while Cheyenne is interpreted REBOL.

Doing a quick ('n dirty) benchmark..

Marteen: I'll add a small Database abstraction layer for RSP: so 
I'll include both mysql and pgsql drivers once this feature added.

I'll add a remark module for it  :-)  its adds dialected tags to 
html, sort of a functional approach to web.

Running Apache and Cheyenne on my IPOD nano.. 

Using a simple rebol 'read'' loop (10,000 hits) to a 4kb static page 
(localhost).

Apache - 1:41
Cheyenne - 0:52

Graham: you did a very good job with your WebMail. I wish the new 
RSP API and features would allow you to finish the work and release 
it !

does cheyenne handle http 1.0 and 1.1 transparently (expecting yes, 
but need a specific answer)?

But a more elaborate benchmark may show closer results between these 
too. Size of files and OS have big impacts on the results.

No, it's a client implementation of my own (/Command implements a 
server-side FastCGI protocol)

thanks... so far, if everything works without bugs, I think you just 
made yourself a new user, I'll be testing it tomorrow.

HTTP support is stable, CGI are ok, RSP too, just avoid sessions 
for now. Next release will bring a complete rewrite of the session 
support for faster and more stable behaviour.

I should be able to make small releases each week (mainly bugfixes) 
and release monthly new features. It's the rate of release that I've 
doing since a year now (for a few ppl and customers).

Terry: Any php script would work, you just have to connect a php 
exe compiled with fastcgi support. I'll give more info on how to 
setup such config once the fastcgi module will go beta. I plan to 
bundle a php exe with the binary Cheyenne with no setup needs for 
end user.

'from' would be a better term

the ipod acts as a HD

Graham: yes it still has some issues, that's why I'm working on a 
new implementation that should fix all problems. That's my first 
priority for Cheyenne.

Storing session data in a DB might be an interesting solution, but 
it will be slower than the current native approach.

I've setup a Blog for Cheyenne by adapting Carl's one : http://softinnov.org/cheyenne/blog.cgi

Carl's blog needs a way to implement emoticons !

Memory DB: even in memory if you count the overhead of a SQL layer 
+ storage layer, it would still be much slower than direct storing 
of data in a REBOL block! or hash!.

Doc: I know, but I also read the RubyOnRails book where they measured 
all this. A database scales better, and for low usage the performance 
hit you tahe is little. Files/memory is faster but collapses pretty 
soon, MySQL runs with constant overhead (and little enough, I might 
add). It is more a question of balancing performance with different 
load scenarios.

However after a few hundred thousand greylisting entries .. it started 
to collapse a little.  I should have gone db ...

Doc, I am also one of your fans. I've been going to your web site 
regularly to see if Cheyenne was released. I was excited to be able 
to download it. But I have a problem. Cheyenne loads fine, but when 
I try to test the links on the index.html page they fail. What am 
I doing wrong?

OK, http://localhostworks. I have a lot to learn, as you can see. 
Thanks for helping me through this. I was just clicking on the index.html 
file in the www folder.

Can you make the session storage server a RebService, called from 
the various processes that need to read/write the data? That overhead 
might be less than the SQL overhead.

I haven't had a chance to review the Cheyenne source yet. Are the 
subprocesses called by a central broker? Perhaps that broker can 
manage the session data. That way any in-process accesses could be 
almost as fast as straight block access, with only API overhead.

Well, if you package the session data in a manager API, you can export 
that API as a REBOL Service that subprocess can use as well.

Not sure. I remember that there was a version built on UniServe, 
same as Cheyenne, so we could use that.

Also, would be nice to have a link on your Cheyenne page to your 
Cheyenne blog.

Blog: I'll add it this WE, I'm still doing adding some features to 
the blog, I'd like to test it a few days more with a low traffic.

I'd love if there were an LNS implementation built on UniServe. This 
would let you turn Cheyenne into a general purpose app server.

Does Cheyenne have an option to limit which network adapters it will 
serve to? For example, I have 5 adapters on this computer (ethernet, 
wireless, 2 VMware, Hamachi), maybe 6 counting localhost, and I would 
like the web server to only be visible to the virtual adapters, not 
even opening a port on the physical ones.


I remember this being difficult in REBOL, that REBOL would always 
open server ports on all available adapters, so you would have to 
inspect the incoming connections and filter. Have you come up with 
a better workaround for this, or a way to do it properly?

IIRC via set-modes/get-modes, you can know what adapter is involved. 
I am afraid there is no other way how to prevent this. Opening a 
socket is OS thing imo, and I wonder if C socket functions allow 
you to choose your device?

Does each http request spawn a new instance ?

you mean a new REBOL process ?

HTTP requests for static ressource are handled in multiplexing mode 
by the main Cheyenne process. Only CGI or RSP can currently spawn 
a new process if there's no free process available to handle the 
job.

I haven't downloaded your latest versions yet .. is a console visible 
?

is is parse base, each rule consist of a match block!, a transform 
block! and a logic! to continue thru next rule or break

Thanks for your code, I'll give it a try. Btw, IE has 85% market 
share (according to a recent article on osnews.com).

But it's a little bit out of topic chat:-)

Doc,  I appreciate your time / work on this.  Just so I understand.... 
are you saying that the solution you would implement would only run 
on a Win machine and that solutions that run on other machines are 
not feasible  ... because they are basically cumbersome and unreliable? 
 e.g. http://drupal.org/node/44718And are schemes like NTLM Authorization 
Proxy Server unworkable?  refer http://www.geocities.com/rozmanov/ntlm/
    I can't see using Cheyenne in a work environment  unless it supports 
NTLM.

Nenad, seems there was a major uniserve rewrite with cheyenne.. are 
things pretty much the same? How do you install a new service now?

Terry: UniServe hasn't been rewritten, there are just a few features 
and new events added. The service API is exactly the same as before. 
Remember that Cheyenne and UniServe are two different products. Cheyenne 
is just using UniServe as low level networking layer.

MikeL: I've implemeted a cheap solution by using the win32 API. Implementing 
NTLM natively in REBOL  require a significant amount of time, that 
I don't have currently, or in a near future. There are some non trivial 
algorithms to implement in REBOL like MD4 hashing. If required, I 
may work on such lib on a contract basis, else I don't think that 
I'll have time to implement that before next year. Maybe someone 
else in the REBOL community would have the time to do the job ?

I've been cranking hard on Framewerks, and a beta is coming soon. 
We've completely blurred the line between desktop apps, and web apps, 
and I'm looking to EASILY sell 100,000+ copies of Cheyenne as soon 
as it's ready.

Doc is rewritting sessions currently and back on Cheyenne, at least 
that is what he told to me a week ago ...

about 100000 rsp requests in a couple hours, last time I checked 
it was related to accumulated connections not propery closed