Menu
▾
▴
binarycloud-dev — development list for binarycloud
You can subscribe to this list here.
| 2001 |
Jan
|
Feb
|
Mar
|
Apr
(57) |
May
(287) |
Jun
(166) |
Jul
(286) |
Aug
(273) |
Sep
(254) |
Oct
(144) |
Nov
|
Dec
|
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2002 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Spiggy T. <th...@me...> - 2001-05-24 00:51:21
|
this is slightly off topic... but my boss and i disagree on the "preferred" coding style. he is an asp-guy and i am for php so maybe that makes adifference.. ill let you decide since i am hardly the expert: we have an asp application in use at work.. it uses a lot of document.write"<table><tr><td>some data here</td></tr></table>" type of coding, which works like the php echo (as far as i can understand). i dont like this style at all, it makes the code hard to read.. (plus playing with the quotes is a serious pain in the rear end). then in the code there is a part where the programmer actually did it "my way" which means close code tags and do html as plain html and open code tags when theres code to spit out. my boss says this is bad and slows it down. so which way is faster? i know echoing html is considered bad coding style (at least with php) paivi |
|
From: W.D.Sumilang <wa...@on...> - 2001-05-24 00:30:29
|
>actually, this win32 stuff beings up an interesting point.... later on >you'll start to see some fairly "deep" integration of the whole process with >makefiles, and those makefiles will be php, intended for use from a command >line. >does anyone have any experience using php scripts from a dos command line? Sorry, haven't tried this. >I'm sure its possible, but it's probably a little strange. >and along those lines, a number of the tools we have and will provide are >quite *nix specific (shell scripts, etc). >what are people's thoughts on this? Personally, I just haven't had enough time to put together a local *nix system for testing, but using linux for production... win95 is just something I've inherited locally... so it's not an issue for me... -W.D. __________________________________________________ FREE voicemail, email, and fax...all in one place. Sign Up Now! http://www.onebox.com |
|
From: Alex B. <en...@tu...> - 2001-05-23 22:40:00
|
> Hello: > All's tested well on win95/apache1.3.12/php4.0.4pl1 with TAO's patch... > anxiously awaiting future releases! can't wait to start coding some modules. good good. actually, this win32 stuff beings up an interesting point.... later on you'll start to see some fairly "deep" integration of the whole process with makefiles, and those makefiles will be php, intended for use from a command line. does anyone have any experience using php scripts from a dos command line? I'm sure its possible, but it's probably a little strange. and along those lines, a number of the tools we have and will provide are quite *nix specific (shell scripts, etc). what are people's thoughts on this? _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
|
From: W.D.Sumilang <wa...@on...> - 2001-05-23 21:58:09
|
Hello: All's tested well on win95/apache1.3.12/php4.0.4pl1 with TAO's patch... anxiously awaiting future releases! can't wait to start coding some modules. Thanx Alex... cheers, W.D. __________________________________________________ FREE voicemail, email, and fax...all in one place. Sign Up Now! http://www.onebox.com |
|
From: Alex B. <en...@tu...> - 2001-05-23 18:39:52
|
>> if anyone has any problems, let me know. > there is a bug in binarycloud.php with Win32, and I have attacted the > patched version. > > Ronald TAO > ron...@ho... hi ronald, thanks for the code, I've tested it on linux, it doesn't interfere, committed it. you're in CREDITS now :) _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
|
From: Alex B. <en...@tu...> - 2001-05-23 18:16:11
|
> exactly what I did, should be ok for now.
ah, good.
> [PATH_TRANSLATED] => /raidb4/iPlanet/server/docs/bc/index.php
agh! not quite :)
I don't see any Docroot or anything similar.
What I'll do with the "distro" version of r2 is include the current code in
binarycloud.php, but commented out.
The only code in the distro will be
define("BC_PATH","/blah/blah/blah/binarycloud/");
> I'm also getting JS errors and missing images, are there any other paths or
> switches that need setting?
cd /raidb4/iPlanet/server/docs/bc/
cp -R resources ../
reload
> Also, this: Unable to call _pear_call_destructors() - function does not
> exist in Unknown on line 0
Yes, after looking quite carefully at PEAR_Error, I am sad to say that it
isn't beefy enough.
So, we'e going to do an error class as well. Snarf!
> but I'm guessing you know about this one already.
yep.
_alex
--
alex black, ceo
en...@tu...
the turing studio, inc.
http://www.turingstudio.com
vox+510.666.0074
fax+510.666.0093
|
|
From: Faine, M. <Mar...@ms...> - 2001-05-23 16:58:36
|
in binarycloud.php, just hardcode the path for the moment
(i.e. delete all the code except "define(BC_PATH..." and just set the path.
exactly what I did, should be ok for now.
then, go to /path/to/bc/htdocs/index.php
at the bottom of that page, there's a bunch of commented out lines:
//$Debug->Output("globals");
is the one you want to un-comment.
that will give you a print_r of your globals...
Here they are:
Debug called with: 'globals'
= GLOBALS Array
==========================================================================
Array
(
[QUERY_STRING] =>
[REQUEST_LINE] => GET /bc/index.php HTTP/1.1
[REQUEST_METHOD] => GET
[SCRIPT_NAME] => /bc/index.php
[SCRIPT_PROTOCOL] => HTTP/1.1
[CONTENT_LENGTH] =>
[CONTENT_TYPE] =>
[HTTP_ACCEPT] => */*
[HTTP_ACCEPT_ENCODING] => gzip, deflate
[HTTP_ACCEPT_LANGUAGE] => en-us
[HTTP_AUTHORIZATION] =>
[HTTP_COOKIE] =>
[HTTP_IF_MODIFIED_SINCE] =>
[HTTP_REFERER] =>
[HTTP_USER_AGENT] => Mozilla/4.0 (compatible; MSIE 5.01; Windows NT 5.0)
[HTTP_USER_DEFINED] =>
[AUTH_TYPE] =>
[PATH_INFO] =>
[REMOTE_USER] =>
[HTTPS_KEYSIZE] =>
[HTTPS_SECRETSIZE] =>
[REMOTE_ADDR] => IP was here
[PHP_SELF] => /bc/index.php
[HTTP_POST_VARS] => Array
(
)
[HTTP_GET_VARS] => Array
(
)
[HTTP_COOKIE_VARS] => Array
(
)
[HTTP_SERVER_VARS] => Array
(
[QUERY_STRING] =>
[REQUEST_LINE] => GET /bc/index.php HTTP/1.1
[REQUEST_METHOD] => GET
[SCRIPT_NAME] => /bc/index.php
[SCRIPT_PROTOCOL] => HTTP/1.1
[CONTENT_LENGTH] =>
[CONTENT_TYPE] =>
[HTTP_ACCEPT] => */*
[HTTP_ACCEPT_ENCODING] => gzip, deflate
[HTTP_ACCEPT_LANGUAGE] => en-us
[HTTP_AUTHORIZATION] =>
[HTTP_COOKIE] =>
[HTTP_IF_MODIFIED_SINCE] =>
[HTTP_REFERER] =>
[HTTP_USER_AGENT] => Mozilla/4.0 (compatible; MSIE 5.01; Windows
NT 5.0)
[HTTP_USER_DEFINED] =>
[AUTH_TYPE] =>
[PATH_INFO] =>
[REMOTE_USER] =>
[HTTPS_KEYSIZE] =>
[HTTPS_SECRETSIZE] =>
[REMOTE_ADDR] => IP was here
[PHP_SELF] => /bc/index.php
)
[HTTP_ENV_VARS] => Array
(
)
[HTTP_POST_FILES] => Array
(
)
[SERVER_PORT] => 2500
[HTTPS] => OFF
[SERVER_NAME] => starbase
[REMOTE_HOST] =>
[SERVER_URL] => http://starbase:2500
[PATH_TRANSLATED] => /raidb4/iPlanet/server/docs/bc/index.php
There were more but didn't seem important let me know if I left anything you
needed out.
I'm also getting JS errors and missing images, are there any other paths or
switches that need setting?
Also, this: Unable to call _pear_call_destructors() - function does not
exist in Unknown on line 0
but I'm guessing you know about this one already.
-Mark
|
|
From: alex b. <en...@tu...> - 2001-05-23 16:37:12
|
> What would happen if you wanted a central install of BC to run more than
> one site? How would you tell it which binarycloud.php to include?
It doesn't have to.
If you're running, for example, a bunch of virtualhosts, the path to
binarycloud is set relative to the _virtualhosts_ document root, which means
you are using different binarycloud installs for each vHost.
> Also, is there any other permissible way of doing this - I could set up r1
> on free web hosts because I could use an .htaccess, but I don't think the
> above is supported, at least not with safe_mode enabled :-(
first: you are welcome to just prepend this:
<? define('BC_PATH','/path/to/binarycloud'); ?>
which will do the same job.
_but_ I would check to make absolutely sure that you can't do it with the
include, because it's really convenient to be able to run that with (x)
virtualhosts.
_alex
|
|
From: alex b. <en...@tu...> - 2001-05-23 16:33:45
|
hi mark,
try this:
in binarycloud.php, just hardcode the path for the moment
(i.e. delete all the code except "define(BC_PATH..." and just set the path.
that will get your install functional..
then, go to /path/to/bc/htdocs/index.php
at the bottom of that page, there's a bunch of commented out lines:
//$Debug->Output("globals");
is the one you want to un-comment.
that will give you a print_r of your globals...
can you copy that, and send it to me?
(you may want to clean parts of it that contain IPs etc)
that way I can see what the iPlanet env looks like.
thanks,
_alex
----- Original Message -----
From: "Faine, Mark" <Mar...@ms...>
To: <bin...@li...>
Sent: Wednesday, May 23, 2001 5:54 AM
Subject: RE: [binarycloud-dev] functioning download available
> There seems to be a problem getting it to install on iPlanet. I think the
> problem is with the code:
> $pieces = explode('/', $HTTP_SERVER_VARS[DOCUMENT_ROOT]);
>
> There is no $HTTP_SERVER_VARS[DOCUMENT_ROOT] or at least I can't find it.
> There still is doc_root from the php.ini, could I use that instead, if so
> how would I refer to it.
>
> -Mark
>
> -----Original Message-----
> From: Alex Black [mailto:en...@tu...]
> Sent: Tuesday, May 22, 2001 5:27 PM
> To: binarycloud-dev
> Subject: [binarycloud-dev] functioning download available
>
>
> http://www.binarycloud.com/download/
>
> under "binarycloud r2 pre-release downloads"
>
> the "binarycloud r2 shell" link gets you a tarball.
>
> note that you can see a functioning version at:
> http://www.binarycloud.com:8000/
> seems to be averaging about 2/thousanths of a second on that machine, my
dev
> machine is consistently coming in at ~.013 seconds :)
>
> once you have untarred that file, you'll have a directory: r2
> if you copy:
>
> r2/binarycloud.php to your php include path
> and
> r2/binarycloud to one level above htdocs
> and
> r2/htdocs/* to somewhere (anywhere) in your server's document root,
>
> you should have a functional install that renders pages.
>
> if anyone has any problems, let me know.
>
> btw, the above should be about as complex as installation gets for r2 :)
>
> _alex
>
>
>
> --
> alex black, ceo
> en...@tu...
>
> the turing studio, inc.
> http://www.turingstudio.com
>
> vox+510.666.0074
> fax+510.666.0093
>
>
>
> _______________________________________________
> binarycloud-dev mailing list
> bin...@li...
> http://lists.sourceforge.net/lists/listinfo/binarycloud-dev
>
> _______________________________________________
> binarycloud-dev mailing list
> bin...@li...
> http://lists.sourceforge.net/lists/listinfo/binarycloud-dev
>
|
|
From: Faine, M. <Mar...@ms...> - 2001-05-23 12:54:56
|
There seems to be a problem getting it to install on iPlanet. I think the
problem is with the code:
$pieces = explode('/', $HTTP_SERVER_VARS[DOCUMENT_ROOT]);
There is no $HTTP_SERVER_VARS[DOCUMENT_ROOT] or at least I can't find it.
There still is doc_root from the php.ini, could I use that instead, if so
how would I refer to it.
-Mark
-----Original Message-----
From: Alex Black [mailto:en...@tu...]
Sent: Tuesday, May 22, 2001 5:27 PM
To: binarycloud-dev
Subject: [binarycloud-dev] functioning download available
http://www.binarycloud.com/download/
under "binarycloud r2 pre-release downloads"
the "binarycloud r2 shell" link gets you a tarball.
note that you can see a functioning version at:
http://www.binarycloud.com:8000/
seems to be averaging about 2/thousanths of a second on that machine, my dev
machine is consistently coming in at ~.013 seconds :)
once you have untarred that file, you'll have a directory: r2
if you copy:
r2/binarycloud.php to your php include path
and
r2/binarycloud to one level above htdocs
and
r2/htdocs/* to somewhere (anywhere) in your server's document root,
you should have a functional install that renders pages.
if anyone has any problems, let me know.
btw, the above should be about as complex as installation gets for r2 :)
_alex
--
alex black, ceo
en...@tu...
the turing studio, inc.
http://www.turingstudio.com
vox+510.666.0074
fax+510.666.0093
_______________________________________________
binarycloud-dev mailing list
bin...@li...
http://lists.sourceforge.net/lists/listinfo/binarycloud-dev
|
|
From: TAO R. <ron...@ho...> - 2001-05-23 12:26:42
|
>What would happen if you wanted a central install of BC to run more >than >one site? How would you tell it which binarycloud.php to include? if your binarycloud.php is not located in a public accessible dir, it should be no problem > >Also, is there any other permissible way of doing this - I could set >up r1 >on free web hosts because I could use an .htaccess, but I don't >think the >above is supported, at least not with safe_mode enabled :-( I have no problem with safe_mode enabled. Ronald TAO ron...@ho... _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. |
|
From: TAO R. <ron...@ho...> - 2001-05-23 12:14:00
|
>if anyone has any problems, let me know. there is a bug in binarycloud.php with Win32, and I have attacted the patched version. Ronald TAO ron...@ho... _________________________________________________________________________ Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com. |
|
From: Peter B. <re...@f2...> - 2001-05-23 07:00:30
|
At 03:27 PM 5/22/01 -0700, you wrote: > r2/binarycloud.php to your php include path What would happen if you wanted a central install of BC to run more than one site? How would you tell it which binarycloud.php to include? Also, is there any other permissible way of doing this - I could set up r1 on free web hosts because I could use an .htaccess, but I don't think the above is supported, at least not with safe_mode enabled :-( Peter. -oOo- Maple Design - web design, hosting, domain names http://www.mapledesign.co.uk -oOo- |
|
From: alex b. <en...@tu...> - 2001-05-23 06:18:24
|
> do you mean r2/binarycloud.php to the include path or to auto_prepend? > adding a file to a path statement sounds groovy though ;) I mean to your include path. That way there is literally zero apache/php etc setup, assuming you have all the basics :) _alex |
|
From: Nico G. <sc...@ac...> - 2001-05-23 06:03:17
|
> http://www.binarycloud.com/download/ looks very cool. i'll have a look at it tonight, after my exam. > once you have untarred that file, you'll have a directory: r2 > if you copy: >=20 > r2/binarycloud.php to your php include path > and > r2/binarycloud to one level above htdocs > and > r2/htdocs/* to somewhere (anywhere) in your server's document root, do you mean r2/binarycloud.php to the include path or to auto_prepend? adding a file to a path statement sounds groovy though ;) --nico -- nico galoppo - tremelo/leuven, belgium - erasmus/socrates student in grenoble, france - 4, rue b=E9ranger -- phone: +33-(0)76-85 23 19 --------------------------------------------------------------- [bash]:~$ man woman nico at crossbar dot net No manual entry for woman debian linux :: vim powered |
|
From: Alex B. <en...@tu...> - 2001-05-22 22:27:23
|
http://www.binarycloud.com/download/ under "binarycloud r2 pre-release downloads" the "binarycloud r2 shell" link gets you a tarball. note that you can see a functioning version at: http://www.binarycloud.com:8000/ seems to be averaging about 2/thousanths of a second on that machine, my dev machine is consistently coming in at ~.013 seconds :) once you have untarred that file, you'll have a directory: r2 if you copy: r2/binarycloud.php to your php include path and r2/binarycloud to one level above htdocs and r2/htdocs/* to somewhere (anywhere) in your server's document root, you should have a functional install that renders pages. if anyone has any problems, let me know. btw, the above should be about as complex as installation gets for r2 :) _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
|
From: Alex B. <en...@tu...> - 2001-05-22 18:51:36
|
hi all,
for the distro version of r2, I want to get to the point where a user can
download the system, untar it, point their browser to it, and have a
functioning installation.
the instructions should consist of:
-untar
-binarycloud/user/conf/datasources.xml
-cd binarycloud make
-./Makefile
point browser at:
/path/to/bc/files/in/htdocs/index.php
-----
along those lines,
I'm going to build a very simple little file called
binarycloud.php
which will be included from each page in htdocs:
include_once('binarycloud.php');
note the "pathlessness" - i.e. binarycloud.php will be in your include_path.
the idea behind this is to set the constant BC_PATH before anything else
happens, but to avoid using prepends, etc to get that working.
this also has a number of other advantages like the ability to have many
virtualhosts use the same cvs checkout of a binarycloud install, etc.
my question:
-does anyone know of a reason why
$HTTP_SERVER_VARS[DOCUMENT_ROOT]
would _not_ work as a basis for setting up BC_PATH?
I don't have access to iPlanet, etc, so I can't check that the above var
is useful on other server platforms besides apache.
i.e.
<?php
$pieces = explode('/', $HTTP_SERVER_VARS[DOCUMENT_ROOT]);
$count = count($pieces);
$cd_up = 1;
for ($i=1; $i <= $count-$cd_up; $i++) {
$path .= '/'.$pieces[$i];
}
$path .= '/binarycloud';
define('BC_PATH',$path);
echo BC_PATH;
?>
would be the contents of the binarycloud.php include.
_alex
--
alex black, ceo
en...@tu...
the turing studio, inc.
http://www.turingstudio.com
vox+510.666.0074
fax+510.666.0093
|
|
From: Alex B. <en...@tu...> - 2001-05-22 07:11:53
|
> next thing up is actually to build TemplateManager so I can do master
> templates and layouts in XSLT.
I got here and realized that was in-a-coding-stupor talk, and that's bull.
master templates, and module layouts, are normal php-in-html for a number of
reasons, not least of which we'd have to generate a huge amount of xml for
every page render, only to be hurt by using xsl.
anyway,
I have a completely functional Page class in cvs. The entire page render
pipeline is functional, I even have the ability to dynamically switch a
template based on a call to the name of the template in request_uri.
(woohoo!).
...
for those of you not familiar with the idea of layouts, it goes like this:
in r1, I had this itch, like sometimes I wanted to have a little bit of
extra fluff/formatting html in a particular page that I didn't necessarily
want to associate with a module.
so, layouts were born in r2, and they solve the only
production/implementation flaw with core_template: no "extra page specific
layout code"
quite convenient, and you can do nearly anything in a layout, cuz it's php.
:)
...
some of the methods in Page are a little dense for my taste, although anyone
can gain a cursory understanding of the class by reading Page::BuildPage
Where possible, I will attempt to simplify the methods, but then again no
one will need to putz with (or even know much about) this code so long as
they are aware of how it acts under varying conditions.
some examples:
-you can entirely omit the $bc_page[templates] array (you don't have to
have a master template.. if you don't the module groups will just be printed
in sequence.
-you can change the load order of module::Init methods, just as in r1.
-etc.
this will all be documented.
time for sleep.
_a
an update:
> Init (done, cooked)
> Page (done, cooked)
> Auth (kneading the dough)
> Perm (ditto)
> DB (ah, thanks manuel)
> Err (need to putz with PEAR_Error to see if it jives)
> Lang (this is simple, and will come last)
> Sess (done)
> Request (done, but needs poking and prodding before going into CVS)
>
> soon(er!)! soon(er!)!
--
alex black, ceo
en...@tu...
the turing studio, inc.
http://www.turingstudio.com
vox+510.666.0074
fax+510.666.0093
|
|
From: <gr...@i7...> - 2001-05-22 05:10:22
|
Typo3 ie yet another. (http://www.typo3.com/) This is pretty seriously complex system with a demo at http://demo.typo3.com/ . The admin interface is pretty darn cool and it has it's own 'language' and so on. The templating stuff is a little wierd to figure out tho. I'm quite surprised that it's still relatively unknown. One interesting note though, is that the guy has basically done the entire thing by himself which is pretty darn impressive. Much like young Alex :) Cheers, Graeme "Andreas Aderhold" <a.a...@th...> To: <bin...@li...> Sent by: cc: bin...@li...urc Subject: RE: [binarycloud-dev] binarycloud vs. other web app frameworks eforge.net 21/05/2001 08:46 PM Please respond to binarycloud-dev Hi, I found another framework/cms. It's called eZ publish and can be found here: http://developer.ez.no/ It's quite interesting an feauture-rich. I did not try it out - did someone here? It looks ok. Nevertheless, bc will be far better :-)) Andi _______________________________________________ binarycloud-dev mailing list bin...@li... http://lists.sourceforge.net/lists/listinfo/binarycloud-dev |
|
From: Alex B. <en...@tu...> - 2001-05-22 04:51:44
|
> I'm a little confused about the processing of xsl. > Obviously there must be some kind of program that > parses the xsl and xml file to produce the end > product, and from what I've seen IE is the only > browser that fully supports it. Mozilla has a > processor it it as of May 11 (TransforMiiX ), but it > seems a little hackish. It is. > At any rate, wouldn't it make more sense to have the > xslt be done on the server, instead of in the > uncontrollable browser? Possibly with things from the > xml.apache.org projects (such as combinations of > xerces, xalan, cocoon). yes! which is what all this noise about xsl is about. IE5 supports client side xslt, but I really have no use for it. The very idea is frightening :) > I don't know a lot about what I'm saying here, so just > give me some direction (and I would consider "read > more" a valid suggestion, just tell me what to read) > as perhaps I am missing some of the big picture. no, you are not missing anything. you made up what made sense to you, and what you made up was right :) I think that's the best test of a system: " if I was going to do this, how would it work" and comparing the implementation to the expectation. _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
|
From: Curtis C. <c_c...@ya...> - 2001-05-22 04:18:21
|
I'm a little confused about the processing of xsl. Obviously there must be some kind of program that parses the xsl and xml file to produce the end product, and from what I've seen IE is the only browser that fully supports it. Mozilla has a processor it it as of May 11 (TransforMiiX ), but it seems a little hackish. At any rate, wouldn't it make more sense to have the xslt be done on the server, instead of in the uncontrollable browser? Possibly with things from the xml.apache.org projects (such as combinations of xerces, xalan, cocoon). I don't know a lot about what I'm saying here, so just give me some direction (and I would consider "read more" a valid suggestion, just tell me what to read) as perhaps I am missing some of the big picture. Thanks, Curtis. __________________________________________________ Do You Yahoo!? Yahoo! Auctions - buy the things you want at great prices http://auctions.yahoo.com/ |
|
From: Alex B. <en...@tu...> - 2001-05-22 01:11:38
|
hi all,
I have a functioning Page class, which (though it's a little, uh, dense at
the moment) works just fine.
It includes modules, and executed Init() and Output() for each module within
each module group in the correct order.
It appends the output of each module to:
$this->output[group_name]
so...
next thing up is actually to build TemplateManager so I can do master
templates and layouts in XSLT.
I could be lazy and just make a shell templatemanager that includes files,
but egh :)
---
So, on the list of core classes:
Init (done, cooked)
Page (baking, but you can smell it in the living room)
Auth (kneading the dough)
Perm (ditto)
DB (ah, thanks manuel)
Err (need to putz with PEAR_Error to see if it jives)
Lang (this is simple, and will come last)
Sess (done)
Request (done, but needs poking and prodding before going into CVS)
soon! soon!
_alex
--
alex black, ceo
en...@tu...
the turing studio, inc.
http://www.turingstudio.com
vox+510.666.0074
fax+510.666.0093
|
|
From: Alex B. <en...@tu...> - 2001-05-22 00:21:37
|
we have our own function in php now: get_defined_constants(); thanks to Sean for the implementation, we should see this go out with PHP 4.0.6. :) _alex -- alex black, ceo en...@tu... the turing studio, inc. http://www.turingstudio.com vox+510.666.0074 fax+510.666.0093 |
|
From: alex b. <en...@tu...> - 2001-05-21 15:22:07
|
> > Don't agree. I think user groups are essential. Imagine a module called > "forum". It's a generalized module implementing the forum functionality. You > have db tabels for lets's say "forum_posts", "forum_categories", > "forum_logs", "forum_topics" and "forums". So having usergroups (that > holding the permission pattern) you can use them for assigning some forums > to the group "developer" other to "user" some to "anonymous" so that they > can't be accessed by the other group. We don't need _midguard's_ user groups, as bc r1 already has that, and r2 will have an expanded/groovier version of it. So I'm not saying "user groups are dumb" which would be, well, dumb. I'm saying that we already have this and don't need to go looking elsewhere for it. > If you need more, take Workspaces. Workspaces can combine usergroups or > users. So a workspace called "Customers" contains the usergroups "ad > clients", "normal clients", "web clients" or whatever, another "intranet" > contains your intranet users. So forums belonging to the workspace > "Customers" are available for all users beloning to "Customers". This > applies also for documens etc. The functional elements in r2 auth/perm are users/roles. We could easily add a faux hierarchy to that, i.e. be able to classify users in groups simply to make them easy to deal with. I like that idea, it will probably be added as part of the entityEditor. > Groups may also be interesting in mapping unix groups to the database. Don't > especially know a application for that, but I think sourceforge does this > way. At an even lower level, there are pam modules for doing db auth in unix machines, SF uses ldap because of the number of users - which is nice. I think it's called pam_ldap > Maybe this is all a bit complicated and narrow-minded. But I think there is > a real usage for usergroups. Workspaces maybe interesting for the CMS > module. User groups are absolutely necessary. But we don't need midguard's :) _a |
|
From: alex b. <en...@tu...> - 2001-05-21 15:17:09
|
> The only problem with Midgard is that you are virtually building your > Apache/PHP/MySQL install around it. I think we may be able to use some of > the code, but could do better. And save people having to install apache > modules just to run the CMS. you have to run an apache module? that's a bit much. yes, so it sounds like midguard would be useful to look at, but not necessarily use it wholesale. _alex |
5 messages has been excluded from this view by a project administrator.
