phpslash-devel — For developers of phpslash to communicate.

[Phpslash-devel] Re: Odd Problems with Comments

[Matthew L. <lei...@ma...>]

On Wed, 2003-10-22 at 12:24, Mike Gifford wrote:
> Hi Matthew,
> 
> the (array) function changed, but didn't resolve the problem with
> lib.resources.php
> 
> > > > This refers to the following foreach in the
> > > lib.resources.php
> > > > >
> > > > > >       foreach($requirements as $req) {
> > > > You might could change this to
> > > >           foreach((array)$requirements as $req) {
> > 
> > Made the change as outlined and get the warning shown below:-
> > ===================
> > Warning: in_array() [function.in-array]: Wrong datatype for second
> > argument in /home/harbour/class/lib.resources.php on line 349
> > 
> > Warning: load_resource(Array) [function.load-resource]: failed to create
> > stream: No such file or directory in /home/harbour/class/lib.resources.php
> > on line 350
> > 
> > Warning: load_resource() [function.load-resource]: Failed opening 'Array'
> > for inclusion (include_path='.:/usr/local/apache/php/lib/php') in
> > /home/harbour/class/lib.resources.php on line 350
> > ====================

I'm guessing we're referring to this block here:

 $includedFiles = get_included_files();
	    if (!in_array($req,$includedFiles)) { // line 349
	       if (!include_once($req)) { // line 350
		  pslError("Include failed: $req.");
		  $success = false;
	       }
	       else {
		  // debug($fs,"$req successfully loaded");
	       }

The first warning is about $includedFiles not being the right datatype. 
Why would that not be an array?  Is get_included_files not working?

As for the third, it looks like $req is an array.  Why would that
happen?  Not sure how that happened, either.

What does a dump of $_PSL['resources'] look like?

Sorry no help yet.

--Matt

-- 
Matthew Leingang
Harvard University Department of Mathematics
lei...@ma...

[Phpslash-devel] Odd Problems with Comments

[Mike G. <mi...@op...>]

Matthew, I think you hit the nail on the head here.  Thanks for seeing
what I missed.  

On Tue, 2003-10-21 at 11:57, Matthew Leingang wrote:
> On Tue, 2003-10-21 at 09:09, Mike Gifford wrote:
> > This is a Back-End problem using psl code.  I just had one of my users
> > check out psl 7.2 on their server and it is working just fine.  The BE
> > comment code is based on psl 0.7.1.
> > The problem is that in posting comments to the poll or to articles the
> > user gets a very, very long string of errors:
> You mean a very, very long string of warnings, don't you? :-)

Indeed..  I've got my warnings turned off, which is why I didn't see
them on my default install....

This wouldn't be visible if this were uncommented in the config.php
file, right:
	ini_set('error_reporting', 'E_ALL~E_NOTICE');

That is assuming that php scripts had the permission.

I had my php.ini set for so surprised I didn't see the errors:
	error_reporting  =  E_ALL

> > > > The comment function which gives
> > > > the following error on a loop with about 1138 entries
> > > > when previewing or posting:-
> > > > --------------
> > > > Warning: Invalid argument supplied for foreach() in
> > > > /home/harbour/class/lib.resources.php on line 343
> > > > --------------
> > This refers to the following foreach in the lib.resources.php
> > 
> > >       foreach($requirements as $req) {
> You might could change this to 
>           foreach((array)$requirements as $req) {

It doesn't give me an error and I'm not seeing the warnings either way.

Rob, does this resolve the problem for you?

> See if that works.

Thanks!

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

[Phpslash-devel] embracing the future

[Luis M <le...@ho...>]

Hello,

After seeing the phpslash-dev module lag (and the -ft also) for a few 
days/weeks, I gave careful thought to improving at least the -dev 
dramatically and then use that to do new releases. Here are few of the new 
dramatic ideas (which are not new per se, but needed in order to keep the 
project current with what other projects are offering):
(NOTE: order matters)

1. xml-rpc support: for things like transfering files from one computer to 
the server and for posting stories in a "blog" kind of way using nifty tools 
like gnome-blog and/or whatever other xml-rpc-compliant software the user 
might choose. It's a matter of deciding:
    a. what protocol to use
    b. how to implement it in a way that can be easily turned on or off
    c. having volunteers to work on this (Joe?, myself... etc...)
    d. try not to reinvent the wheel (copy other GPL projects' ideas if 
possible)

2. a consistent way to input text when creating new entries for stories, or 
posting comments, etc... We talked about using Phorum-like syntax: 
[url][/url] and the like, since that's very common nowadays on the internet.

3. having a way to input data by using JavaScript or whatever is standard 
nowadays for all main browsers. again, we will need a way to turn this off 
for those who don't care, but for those of us who do, it's just a good thing 
to have. it works for Phorum...


Comments anybody? Features I'm missing?

Some research should go into finding why to use certain protocols for 
blogging over the others, but, aside from that, I think these should be just 
one baby step towards keeping phpslash future-proof (for now that is).

P.S. as a side note, I will try to implement all of these and whatever else 
in phpslash-dev and either commit them to a special area, say "contrib" or 
allow it to be turned off by using the config file (turned off by default of 
course). However, with help it will be done faster and we can just move on 
to do a new release with these features.


----)(-----
Luis Mondesi
System Administrator
LatinoMixed.com

le...@ho...

"...The Mac does this so smoothly, it feels like an extension of your mind." 
- Paula Speer, MacWorld Magazine 2003-04

Public signature: http://www.latinomixed.com/lems1/public-a.asc

_________________________________________________________________
MSN Search, le moteur de recherche qui pense comme vous ! 
http://search.msn.fr/worldwide.asp

Re: [Phpslash-devel] Odd Problems with Comments

[Matthew L. <lei...@ma...>]

On Tue, 2003-10-21 at 09:09, Mike Gifford wrote:
> Hi Folks,
Hi!

> This is a Back-End problem using psl code.  I just had one of my users
> check out psl 7.2 on their server and it is working just fine.  The BE
> comment code is based on psl 0.7.1.
> 
> The problem is that in posting comments to the poll or to articles the
> user gets a very, very long string of errors:

You mean a very, very long string of warnings, don't you? :-)

> > > The comment function which gives
> > > the following error on a loop with about 1138 entries
> > > when previewing or posting:-
> > > --------------
> > > Warning: Invalid argument supplied for foreach() in
> > > /home/harbour/class/lib.resources.php on line 343
> > > --------------
> 
> This refers to the following foreach in the lib.resources.php
> 
> >       foreach($requirements as $req) {

You might could change this to 
          foreach((array)$requirements as $req) {

See if that works.

--Matt

-- 
Matthew Leingang
Harvard University Department of Mathematics
lei...@ma...

[Phpslash-devel] Odd Problems with Comments

[Mike G. <mi...@op...>]

Hi Folks,

This is a Back-End problem using psl code.  I just had one of my users
check out psl 7.2 on their server and it is working just fine.  The BE
comment code is based on psl 0.7.1.

The problem is that in posting comments to the poll or to articles the
user gets a very, very long string of errors:

> > The comment function which gives
> > the following error on a loop with about 1138 entries
> > when previewing or posting:-
> > --------------
> > Warning: Invalid argument supplied for foreach() in
> > /home/harbour/class/lib.resources.php on line 343
> > --------------

This refers to the following foreach in the lib.resources.php

>       foreach($requirements as $req) {
>        if (is_urn($req)) {
>           load_resource($req);
>        }
>        else {
>           $includedFiles = get_included_files();
>           if (!in_array($req,$includedFiles)) {
>              if (!include_once($req)) {
>                 pslError("Include failed: $req.");
>                 $success = false;
>              }
>              else {
>                 // debug($fs,"$req successfully loaded");
>              }
>           }
>           else {
>              // debug($fs,"$req already loaded");
>           }
>        }
>       }
>    }

I don't know why this appears on the users system, but not on the
cvs-demo of the code:
	http://cvs-demo.back-end.org/index.php//4

I thought that it might just be a problem with their firewall, but psl
works, so that doesn't make much sense..

Anyone have any ideas what might be up?

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

On Mon, Oct 20, 2003 at 02:34:51PM -0400, Mike Gifford wrote:
> On Sun, 2003-10-19 at 01:22, Joe Stewart wrote:
> > > I hadn't actually wondered around the phpslash-dev cvs before.  Wow, it
> > > looks a lot different.  Looks quite exciting where you are taking
> > > things.  Any plans on when you'll be able to release it.
> > No.  It needs wider testing.  I'm deploying it in a couple of ways.
> 
> Great..  Well, it looks encouraging..
> 
> The templates aren't presently separated in the module directories are
> they?  It tends to become a pretty long list of templates when you have
> them all in one directory..  I think finding what templates need to be
> edited is one of the bigger problems we're having.  
>

The templates can be in a "templates" subdirectory in a module's
directory.   This is lightly tested and was kind of unstable getting it to
work.  The phplib templates port to PEAR has support to look in multiple
directories for templates.  I haven't had time to backport and test
though.

> We've thought about setting up a javascript/css type screen where you
> can browse through and somehow determine graphically what templates
> generate what piece of the page you are looking at.  The view source
> method usually works, but is time consuming and confusing.
> 
> > > btw, in trying to actually run it (rather than scan the directories) I
> > > ran into a problem with the the sql slash-all.sql  I assume that there
> > > are changes in the cvs version of the sql and that this is what I should
> > > install to test it out.
> > definitely a different sql file ( lots more blocks).  There is no upgrade script yet either.
> 
> Is the new sql available somewhere?  A working one doesn't seem to be in
> the CVS, least last time I checked it out.
> 

The sql in the phpslash-dev should work.  

Joe


> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by OSDN developer relations
> Here's your chance to show off your extensive product knowledge
> We want to know what you know. Tell us and you have a chance to win $100
> http://www.zoomerang.com/survey.zgi?HRPT1X3RYQNC5V4MLNSV3E54
> _______________________________________________
> Phpslash-devel mailing list
> Php...@li...
> https://lists.sourceforge.net/lists/listinfo/phpslash-devel

Re: [Phpslash-devel] PHP Notice Errors

[Mike G. <mi...@op...>]

On Sun, 2003-10-19 at 01:22, Joe Stewart wrote:
> > I hadn't actually wondered around the phpslash-dev cvs before.  Wow, it
> > looks a lot different.  Looks quite exciting where you are taking
> > things.  Any plans on when you'll be able to release it.
> No.  It needs wider testing.  I'm deploying it in a couple of ways.

Great..  Well, it looks encouraging..

The templates aren't presently separated in the module directories are
they?  It tends to become a pretty long list of templates when you have
them all in one directory..  I think finding what templates need to be
edited is one of the bigger problems we're having.  

We've thought about setting up a javascript/css type screen where you
can browse through and somehow determine graphically what templates
generate what piece of the page you are looking at.  The view source
method usually works, but is time consuming and confusing.

> > btw, in trying to actually run it (rather than scan the directories) I
> > ran into a problem with the the sql slash-all.sql  I assume that there
> > are changes in the cvs version of the sql and that this is what I should
> > install to test it out.
> definitely a different sql file ( lots more blocks).  There is no upgrade script yet either.

Is the new sql available somewhere?  A working one doesn't seem to be in
the CVS, least last time I checked it out.

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] Useful Function Names

[Joe S. <joe...@us...>]

On Mon, Oct 20, 2003 at 02:27:54PM -0400, Mike Gifford wrote:
> Hi Joe,
> 
> On Sun, 2003-10-19 at 01:39, Joe Stewart wrote:

<snip>
> 
> > I see the code in question.  Yes, it needs documenting and rewriting.  k
> > and v are the key and value.
> 
> Yup..  Also pretty common..  But confusing if you aren't familiar with
> the convention.
> 
> In anycase, something else for the todo :)
> 

I had forgotten that is the variable names that the phplib perm class
uses.  This was written as a patch to it and submitted to phplib for
inclusion ( even though nothing has come of it).


Also - the practice of setting a variable to '' instead of null probably
is a leftover from php3.  Kind of a habit because of how much code we
already had doing this.

Joe

Re: [Phpslash-devel] Useful Function Names

[Mike G. <mi...@op...>]

Hi Joe,

On Sun, 2003-10-19 at 01:39, Joe Stewart wrote:
> On Sat, Oct 18, 2003 at 03:43:14PM -0400, Mike Gifford wrote:
> > It's hard to think of descriptive names some times.  You know what you
> > mean and what you want the function to do..  If you could do that it
> > would be great.  We'll certainly adopt it.
> been trying to use psl prefixes of late.

Some of our functions have be prefixes, but it isn't consistent.

> > > > Also in class/slashPerm.class:have_perm() - If single letter variable
> > > > names could be avoided or at least commented that would be good. 
> > > > Realize this is just a modified version of phplib, but it's always
> > > > easier to understand the code when it is written in human. 
> > > If it's just the $i integer increments, I'm not too worried about this.  This has been programming convention for a long time.
> > There are some $j's in there too..  I'm not all that concerned with
> > it..  Teh $i, $ii & $iii, $v, $k conventions are pretty normal.  Wasn't
> > sure what $j was referring to though.
> old fortran guy.  j is another integer increment.

Ok..

> I see the code in question.  Yes, it needs documenting and rewriting.  k
> and v are the key and value.

Yup..  Also pretty common..  But confusing if you aren't familiar with
the convention.

In anycase, something else for the todo :)

> > > As far as foreach loops instead of for loops.  This has been done as code
> > > is worked on and was one of the design goals. 
> > Yup.  Damn, I was hoping that http://www.phpslash.org was back up so I
> > could look at your developers guide so I could just refer folks there
> > (as we've done in the past).
> http://phpslash.sourceforge.net
> This is why I don't use the rooturl.  Both domains served with one setup.

Yup..  good to know it's there..  Pain about the .org though..

> > > Also don't use:
> > > for ($i = '0' ; $i < count($allSections_ary) ; $i++) {
> > > since the count function is call for each loop.  Instead calculate the
> > > count before the loop.
> > Interesting..  I didn't realize that..  That might be worth going back
> > and cleaning up to boost performance.
> I haven't noticed it actually helping significantly.  But better practice.

Good to know..
...
> > Yeah..  I figured that there must be a trade off for adding in the extra
> > code/variables...  I do think it is easier for folks who are used to
> > having the warnings/notices enabled to have an environment that isn't
> > flooded with harmless ones..  I've had a couple folks complain about it
> > this year.  
> Anybody getting them with 0.7.2, either submit a bug or patch please.

We've got to move over the rest of BE's code to 0.7.2..  It's still a
time consuming process though.

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] PHP Notice Errors

[Matthew L. <lei...@ma...>]

On Sat, 2003-10-18 at 16:14, Mike Gifford wrote:
> Howdy Joe,
> 
> On Fri, 2003-10-17 at 10:10, Joe Stewart wrote:
> > On Thu, Oct 16, 2003 at 12:34:42PM -0400, Mike Gifford wrote:
> > > Index: functions.inc
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
> > > retrieving revision 1.53
> > > diff -r1.53 functions.inc
> > > 617c617
> > > <       $templ->set_var('XSITEOBJECT', $metaobject['object']);
> > > ---
> > > >       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
> > > 811c811,812
> > 
> > I thought this one was fixed by testing if metaobject is an array.
> 
> Evan noticed it, so I'm assuming that it poped up somewhere as a warning
> or notice.  I've cc'd him..  Perhaps the object isn't defined on one of
> our pages (in which case it's our problem & not yours).
> 
> > > Index: lib.resources.php
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> > > retrieving revision 1.4
> > > diff -r1.4 lib.resources.php
> > > 444c444
> > > <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > > ---
> > > >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
> >
> > This is one of those screwy php things.  Anybody got a better way to get
> > rid of the warning?
> 
> Rather than just squish them?  

An idea which I think has been tossed around but rejected is to wrap
$_PSL lookups is a pslConfig() function.  pslConfig("resources.$urn") or
whatever could parse the string, check for existence of the keys and
levels of the multi-dimensional array, and return FALSE if it there's no
value without a warning (or maybe just a PSL-level warning that would go
away with $_PSL['debug'] off).  

Pros:

1.  No more global'ing $_PSL
2.  Can munge keys to map one level of $_PSL to another virtually
3.  More control of errors/warnings (as above)

Cons:

1.  Adds spatial complexity
2.  Probably adds time, too.  Joe says, "Remember: PHP is a scripted
interpreted language." :-P

--Matt
-- 
Matthew Leingang
Harvard University Department of Mathematics
lei...@ma...

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

On Sat, Oct 18, 2003 at 04:14:51PM -0400, Mike Gifford wrote:
> Howdy Joe,
> 
> On Fri, 2003-10-17 at 10:10, Joe Stewart wrote:
> > On Thu, Oct 16, 2003 at 12:34:42PM -0400, Mike Gifford wrote:
> > > Index: functions.inc
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
> > > retrieving revision 1.53
> > > diff -r1.53 functions.inc
> > > 617c617
> > > <       $templ->set_var('XSITEOBJECT', $metaobject['object']);
> > > ---
> > > >       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
> > > 811c811,812
> > 
> > I thought this one was fixed by testing if metaobject is an array.
> 
> Evan noticed it, so I'm assuming that it poped up somewhere as a warning
> or notice.  I've cc'd him..  Perhaps the object isn't defined on one of
> our pages (in which case it's our problem & not yours).
> 
> > > Index: lib.resources.php
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> > > retrieving revision 1.4
> > > diff -r1.4 lib.resources.php
> > > 444c444
> > > <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > > ---
> > > >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
> >
> > This is one of those screwy php things.  Anybody got a better way to get
> > rid of the warning?
> 
> Rather than just squish them?  
> 

Yes, rather than just ignore, I've been trying to make sure a variable is
 in use.

> I hadn't actually wondered around the phpslash-dev cvs before.  Wow, it
> looks a lot different.  Looks quite exciting where you are taking
> things.  Any plans on when you'll be able to release it.
> 

No.  It needs wider testing.  I'm deploying it in a couple of ways.

> btw, in trying to actually run it (rather than scan the directories) I
> ran into a problem with the the sql slash-all.sql  I assume that there
> are changes in the cvs version of the sql and that this is what I should
> install to test it out.
> 

definitely a different sql file ( lots more blocks).  There is no upgrade script yet either.

Joe

> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

On Sun, Oct 19, 2003 at 12:52:21AM -0400, Evan Hughes wrote:
<snip > 
> > > > Index: lib.resources.php
> > > > ===================================================================
> > > > RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> > > > retrieving revision 1.4
> > > > diff -r1.4 lib.resources.php
> > > > 444c444
> > > > <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > > > ---
> > > > >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > >
> > > This is one of those screwy php things.  Anybody got a better way to get
> > > rid of the warning?
> 
>   An isset (or a friend thereof) would probably work. 
> 
> e

uh duh.  Sorry I thought was one of those places where you are defining an
array value and it complains about the index.

Joe

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

On Fri, Oct 17, 2003 at 05:37:43PM -0400, Mike Gifford wrote:
> Hi Joe,
> 
> On Fri, 2003-10-17 at 10:20, Joe Stewart wrote:
> > If anyone hasn't seen the Back-End site lately, check it out.  Congrats to
> > Mike, Peter and everybody else:
> > 	http://www.back-end.org/
> 
> Thanks!  The old design was pretty darn ugly.  There's also a demo of
> the cvs for folks who want to check that out:
> 	http://cvs-demo.back-end.org/
> 
> We haven't changed the default psl username/password.
> 
> > gripe:  It doesn't work with konqueror for me.
> 
> Yes, we stumbled across that, but didn't have time to troubleshoot it. 
> It works in IE & Mozilla and though Konqueror is great, there just
> aren't a lot of folks using it.
> 
> You can log in and view it, the design just looks sub-standard.
>

better success on a different machine, I'll try upgrading the other konq.

> > Mike:  What version of phpSlash code is the current cvs BE using? 
> 
> Mostly it is based on phpSlash 0.7.1-2
> 
> Unfortunately due to some annoying problems I was having troubleshooting
> logins I updated login.php, slashPerms, etc... to 7.2
>

0.7.2 has almost all php warnings cleaned up from the public side.  There
is one annoying one that can generate a phplib warning ( this might be
only on the -dev tree).

I'd suggest you try using 0.7.2 if nothing was introduced to give you
problems. 

Joe

> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 

Re: [Phpslash-devel] Useful Function Names

[Joe S. <joe...@us...>]

On Sat, Oct 18, 2003 at 03:43:14PM -0400, Mike Gifford wrote:
> Howdy Joe,
> 
> On Fri, 2003-10-17 at 10:17, Joe Stewart wrote:
> > On Thu, Oct 16, 2003 at 09:23:50AM -0400, Mike Gifford wrote:
> > > One of our developers wasn't too keen on the function names "clean" and
> > > "dirty" in class/functions.inc
> > > In a future release could these be extended so that they are more
> > > descriptive?  cleanUserHTML perhaps?  He'd just like to have as much
> > > meaning built into it as possible.
> > yes.  I saw a note about name collision with getHeader in the BE cvs too.
> 
> It's hard to think of descriptive names some times.  You know what you
> mean and what you want the function to do..  If you could do that it
> would be great.  We'll certainly adopt it.
>

been trying to use psl prefixes of late.

> > > Also in class/slashPerm.class:have_perm() - If single letter variable
> > > names could be avoided or at least commented that would be good. 
> > > Realize this is just a modified version of phplib, but it's always
> > > easier to understand the code when it is written in human. 
> > If it's just the $i integer increments, I'm not too worried about this.  This has been programming convention for a long time.
> 
> There are some $j's in there too..  I'm not all that concerned with
> it..  Teh $i, $ii & $iii, $v, $k conventions are pretty normal.  Wasn't
> sure what $j was referring to though.
> 

old fortran guy.  j is another integer increment.

I see the code in question.  Yes, it needs documenting and rewriting.  k
and v are the key and value.

> > As far as foreach loops instead of for loops.  This has been done as code
> > is worked on and was one of the design goals.
> 
> Yup.  Damn, I was hoping that http://www.phpslash.org was back up so I
> could look at your developers guide so I could just refer folks there
> (as we've done in the past).
> 

http://phpslash.sourceforge.net

This is why I don't use the rooturl.  Both domains served with one setup.

> > Also don't use:
> > for ($i = '0' ; $i < count($allSections_ary) ; $i++) {
> > since the count function is call for each loop.  Instead calculate the
> > count before the loop.
> 
> Interesting..  I didn't realize that..  That might be worth going back
> and cleaning up to boost performance.
> 

I haven't noticed it actually helping significantly.  But better practice.

> > But in general, move to foreach as development continues.  One thing this
> > can do is generate more warnings instead of less in at least one instance.
> 
> Good..
> 
> > Also getting rid of the warnings by defining variables before their use
> > definitely slows down the code measurably.  But I have found a few bugs in
> > the code fixing.
> 
> Yeah..  I figured that there must be a trade off for adding in the extra
> code/variables...  I do think it is easier for folks who are used to
> having the warnings/notices enabled to have an environment that isn't
> flooded with harmless ones..  I've had a couple folks complain about it
> this year.  
> 

Anybody getting them with 0.7.2, either submit a bug or patch please.

Joe

> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 

Re: [Phpslash-devel] Useful Function Names

[Mike G. <mi...@op...>]

Howdy Joe,

On Fri, 2003-10-17 at 10:17, Joe Stewart wrote:
> On Thu, Oct 16, 2003 at 09:23:50AM -0400, Mike Gifford wrote:
> > One of our developers wasn't too keen on the function names "clean" and
> > "dirty" in class/functions.inc
> > In a future release could these be extended so that they are more
> > descriptive?  cleanUserHTML perhaps?  He'd just like to have as much
> > meaning built into it as possible.
> yes.  I saw a note about name collision with getHeader in the BE cvs too.

It's hard to think of descriptive names some times.  You know what you
mean and what you want the function to do..  If you could do that it
would be great.  We'll certainly adopt it.

> > Also in class/slashPerm.class:have_perm() - If single letter variable
> > names could be avoided or at least commented that would be good. 
> > Realize this is just a modified version of phplib, but it's always
> > easier to understand the code when it is written in human. 
> If it's just the $i integer increments, I'm not too worried about this.  This has been programming convention for a long time.

There are some $j's in there too..  I'm not all that concerned with
it..  Teh $i, $ii & $iii, $v, $k conventions are pretty normal.  Wasn't
sure what $j was referring to though.

> As far as foreach loops instead of for loops.  This has been done as code
> is worked on and was one of the design goals.

Yup.  Damn, I was hoping that http://www.phpslash.org was back up so I
could look at your developers guide so I could just refer folks there
(as we've done in the past).

> Also don't use:
> for ($i = '0' ; $i < count($allSections_ary) ; $i++) {
> since the count function is call for each loop.  Instead calculate the
> count before the loop.

Interesting..  I didn't realize that..  That might be worth going back
and cleaning up to boost performance.

> But in general, move to foreach as development continues.  One thing this
> can do is generate more warnings instead of less in at least one instance.

Good..

> Also getting rid of the warnings by defining variables before their use
> definitely slows down the code measurably.  But I have found a few bugs in
> the code fixing.

Yeah..  I figured that there must be a trade off for adding in the extra
code/variables...  I do think it is easier for folks who are used to
having the warnings/notices enabled to have an environment that isn't
flooded with harmless ones..  I've had a couple folks complain about it
this year.  

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] PHP Notice Errors

[Mike G. <mi...@op...>]

Howdy Joe,

On Fri, 2003-10-17 at 10:10, Joe Stewart wrote:
> On Thu, Oct 16, 2003 at 12:34:42PM -0400, Mike Gifford wrote:
> > Index: functions.inc
> > ===================================================================
> > RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
> > retrieving revision 1.53
> > diff -r1.53 functions.inc
> > 617c617
> > <       $templ->set_var('XSITEOBJECT', $metaobject['object']);
> > ---
> > >       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
> > 811c811,812
> 
> I thought this one was fixed by testing if metaobject is an array.

Evan noticed it, so I'm assuming that it poped up somewhere as a warning
or notice.  I've cc'd him..  Perhaps the object isn't defined on one of
our pages (in which case it's our problem & not yours).

> > Index: lib.resources.php
> > ===================================================================
> > RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> > retrieving revision 1.4
> > diff -r1.4 lib.resources.php
> > 444c444
> > <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > ---
> > >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
>
> This is one of those screwy php things.  Anybody got a better way to get
> rid of the warning?

Rather than just squish them?  

I hadn't actually wondered around the phpslash-dev cvs before.  Wow, it
looks a lot different.  Looks quite exciting where you are taking
things.  Any plans on when you'll be able to release it.

btw, in trying to actually run it (rather than scan the directories) I
ran into a problem with the the sql slash-all.sql  I assume that there
are changes in the cvs version of the sql and that this is what I should
install to test it out.

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] PHP Notice Errors

[Evan H. <eh...@ju...>]

On Sat, Oct 18, 2003 at 04:19:32PM -0400, Mike Gifford wrote:
> Howdy Joe,
> 
> On Fri, 2003-10-17 at 10:10, Joe Stewart wrote:
> > On Thu, Oct 16, 2003 at 12:34:42PM -0400, Mike Gifford wrote:
> > > Index: functions.inc
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
> > > retrieving revision 1.53
> > > diff -r1.53 functions.inc
> > > 617c617
> > > <       $templ->set_var('XSITEOBJECT', $metaobject['object']);
> > > ---
> > > >       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
> > > 811c811,812
> > 
> > I thought this one was fixed by testing if metaobject is an array.
> 
> Evan noticed it, so I'm assuming that it poped up somewhere as a warning
> or notice.  I've cc'd him..  Perhaps the object isn't defined on one of
> our pages (in which case it's our problem & not yours).

  I was getting a warning that 'object' wasn't a member of the array
(IIRC). This nixes that. 

> 
> > > Index: lib.resources.php
> > > ===================================================================
> > > RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> > > retrieving revision 1.4
> > > diff -r1.4 lib.resources.php
> > > 444c444
> > > <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> > > ---
> > > >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
> >
> > This is one of those screwy php things.  Anybody got a better way to get
> > rid of the warning?

  An isset (or a friend thereof) would probably work. 

e

Re: [Phpslash-devel] PHP Notice Errors

[Mike G. <mi...@op...>]

Hi Joe,

On Fri, 2003-10-17 at 10:20, Joe Stewart wrote:
> If anyone hasn't seen the Back-End site lately, check it out.  Congrats to
> Mike, Peter and everybody else:
> 	http://www.back-end.org/

Thanks!  The old design was pretty darn ugly.  There's also a demo of
the cvs for folks who want to check that out:
	http://cvs-demo.back-end.org/

We haven't changed the default psl username/password.

> gripe:  It doesn't work with konqueror for me.

Yes, we stumbled across that, but didn't have time to troubleshoot it. 
It works in IE & Mozilla and though Konqueror is great, there just
aren't a lot of folks using it.

You can log in and view it, the design just looks sub-standard.

> Mike:  What version of phpSlash code is the current cvs BE using? 

Mostly it is based on phpSlash 0.7.1-2

Unfortunately due to some annoying problems I was having troubleshooting
logins I updated login.php, slashPerms, etc... to 7.2

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

If anyone hasn't seen the Back-End site lately, check it out.  Congrats to
Mike, Peter and everybody else:

	http://www.back-end.org/

gripe:  It doesn't work with konqueror for me.

Mike:  What version of phpSlash code is the current cvs BE using? 

Re: [Phpslash-devel] Useful Function Names

[Joe S. <joe...@us...>]

On Thu, Oct 16, 2003 at 09:23:50AM -0400, Mike Gifford wrote:
> Howdy,
> 
> One of our developers wasn't too keen on the function names "clean" and
> "dirty" in class/functions.inc
> 
> In a future release could these be extended so that they are more
> descriptive?  cleanUserHTML perhaps?  He'd just like to have as much
> meaning built into it as possible.
> 

yes.  I saw a note about name collision with getHeader in the BE cvs too.

> Also in class/slashPerm.class:have_perm() - If single letter variable
> names could be avoided or at least commented that would be good. 
> Realize this is just a modified version of phplib, but it's always
> easier to understand the code when it is written in human. 
> 

If it's just the $i integer increments, I'm not too worried about this.  This has been programming convention for a long time.

As far as foreach loops instead of for loops.  This has been done as code
is worked on and was one of the design goals.

Also don't use:

for ($i = '0' ; $i < count($allSections_ary) ; $i++) {

since the count function is call for each loop.  Instead calculate the
count before the loop.

But in general, move to foreach as development continues.  One thing this
can do is generate more warnings instead of less in at least one instance.

Also getting rid of the warnings by defining variables before their use
definitely slows down the code measurably.  But I have found a few bugs in
the code fixing.

later,

Joe



> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 
> 
> 
> 
> -------------------------------------------------------
> This SF.net email is sponsored by: SF.net Giveback Program.
> SourceForge.net hosts over 70,000 Open Source Projects.
> See the people who have HELPED US provide better services:
> Click here: http://sourceforge.net/supporters.php
> _______________________________________________
> Phpslash-devel mailing list
> Php...@li...
> https://lists.sourceforge.net/lists/listinfo/phpslash-devel

Re: [Phpslash-devel] PHP Notice Errors

[Joe S. <joe...@us...>]

On Thu, Oct 16, 2003 at 12:34:42PM -0400, Mike Gifford wrote:
> Hello again,
> 
> Just thought I'd as a question about phpSlash's plans for handling php
> error messages.
> 
> Is the team working to eliminate php notice errors in upcoming releases?
> 
> One of our developers is dead set against not using notices.  We've done
> this with Back-End with some of the phpSlash code we've incorporated:
> 
> Index: functions.inc
> ===================================================================
> RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
> retrieving revision 1.53
> diff -r1.53 functions.inc
> 617c617
> <       $templ->set_var('XSITEOBJECT', $metaobject['object']);
> ---
> >       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
> 811c811,812
> <
> ---
> >
> >    $lastModified = null;

I thought this one was fixed by testing if metaobject is an array.

> Index: lib.resources.php
> ===================================================================
> RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
> retrieving revision 1.4
> diff -r1.4 lib.resources.php
> 444c444
> <    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
> ---
> >    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;
> 

This is one of those screwy php things.  Anybody got a better way to get
rid of the warning?


> Note this may have already been squished in the cvs version.
> 
> Mike
> -- 
> Mike Gifford, OpenConcept Consulting
> Free Software for Social Change -> http://www.openconcept.ca
> Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
> Nonconformity is the highest evolutionary attainment of social animals -
> Aldo Leopold
> 

[Phpslash-devel] PHP Notice Errors

[Mike G. <mi...@op...>]

Hello again,

Just thought I'd as a question about phpSlash's plans for handling php
error messages.

Is the team working to eliminate php notice errors in upcoming releases?

One of our developers is dead set against not using notices.  We've done
this with Back-End with some of the phpSlash code we've incorporated:

Index: functions.inc
===================================================================
RCS file: /cvsroot/back-end/back-end0.5.x/class/functions.inc,v
retrieving revision 1.53
diff -r1.53 functions.inc
617c617
<       $templ->set_var('XSITEOBJECT', $metaobject['object']);
---
>       $templ->set_var('XSITEOBJECT', @$metaobject['object']);
811c811,812
<
---
>
>    $lastModified = null;
Index: lib.resources.php
===================================================================
RCS file: /cvsroot/back-end/back-end0.5.x/class/lib.resources.php,v
retrieving revision 1.4
diff -r1.4 lib.resources.php
444c444
<    $ans = $GLOBALS['_PSL']['resources'][$urn] ? true : false;
---
>    $ans = @$GLOBALS['_PSL']['resources'][$urn] ? true : false;

Note this may have already been squished in the cvs version.

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

[Phpslash-devel] Useful Function Names

[Mike G. <mi...@op...>]

Howdy,

One of our developers wasn't too keen on the function names "clean" and
"dirty" in class/functions.inc

In a future release could these be extended so that they are more
descriptive?  cleanUserHTML perhaps?  He'd just like to have as much
meaning built into it as possible.

Also in class/slashPerm.class:have_perm() - If single letter variable
names could be avoided or at least commented that would be good. 
Realize this is just a modified version of phplib, but it's always
easier to understand the code when it is written in human. 

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

[Phpslash-devel] ERROR: Error in source file:not well-formed (invalid token)

[Mike G. <mi...@op...>]

Howdy folks,

I'm hoping to upgrade a site of ours which is producing this error:
	ERROR: Error in source file:not well-formed (invalid token)

But when I did a search for it I found that both early Back-End and some
phpSlash sites came up with this error in google.  

I had thought that this was a server issue, but since there's a lot of
phpSlash sites with the error, it occurs to me that it was a bug which
was found & fixed at some point.

If anyone has run into this and knows the context I'd be interested.

Mike
-- 
Mike Gifford, OpenConcept Consulting
Free Software for Social Change -> http://www.openconcept.ca
Beads of Hope Campaign - http://uc.openconcept.ca/petition.php
Nonconformity is the highest evolutionary attainment of social animals -
Aldo Leopold

Re: [Phpslash-devel] Blocks Everywere, Navbar control

[Luis M <le...@ho...>]

[snip]

>Is there a wget one-liner that slurps a PHPSlash site to static html?  I
>tried it once but got lost.

straight from Marcel Gagné column at Linux Journal for October 2003:

wget -m http://localhost

-m stands for "mirror" and it encompasses "-r -N -l inf" and " -nr ".

[snip]

----)(-----
Luis Mondesi
System Administrator
LatinoMixed.com

le...@ho...

"...The Mac does this so smoothly, it feels like an extension of your mind." 
- Paula Speer, MacWorld Magazine 2003-04

Public signature: http://www.latinomixed.com/lems1/public-a.asc

_________________________________________________________________
MSN Messenger : discutez en direct avec vos amis ! 
http://www.msn.fr/msger/default.asp