Planet PHPhttp://planet-php.netPeople blogging about PHPen2004-04-01T18:53NNhourly12000-01-01T12:00+00:00ezSystems are classy folkshttp://www.schlossnagle.org/~george/blog/archives/243_ezSystems_are_classy_folks.html2004-04-01T18:53george schlossnagle's blogcover
Last week I helped the folks at ezSystems debug some APC problems they were having. The problems ended up being a 64bit architecture problem (they have uber-fast Opterons) and the bug is now fixed in 2.0.3.

Today I received Python & XML from them (off my Amazon wishlist). Thanks guys!

On a side note, my wishlist seems borked. The list I get when I search on my email address or name is not the same one I can edit when I log into the site. ]]>
PHP april fools...http://www.schlitt.info/applications/blog/index.php?/archives/145_PHP_april_fools.html2004-04-01T17:49Tobias Schlitt - Weblog
1. Changes to case sensitivity in PHP.
Alan Knowles announced that PHP will change to the studlyCase API and therefor will get everything broken by changing established functions.

2. IBM takes over Zend.
Myself hacked a little article about IBM taking over Zend to make PHP a compete of Java.

3. The first PHP virus has been seen.
Wasn't there one last year, too?

4. PHP has been overtaken by Micro$oft.
Mhhh... a little bit unreliable, if they had been taken over by IBM this morning... Maybe one should first look, what others wrote...

5. And finally, PHP4 and 5 showed their real faces...
Take a look at a phpinfo() output!

I guess I missed some, so feel free to comment on this entry, if you found another! ]]>
PHP Virus Attacking Web Hostshttp://www.sitepoint.com/blog-post-view.php?id=1614032004-04-01T12:19SitePoint's PHP Blog: Dynamically TypedDon Box sayshttp://www.infoworld.com/article/03/03/05/HNmanyspecs_1.html2004-04-01T08:00PHP Everywhere Don Box, co-author of SOAP, gives similar advice:
  • Read fewer specifications,
  • write more applications,
  • write less code by using tools that generate code automatically,
  • and remember that humans matter, so if you must write a specification, make it legible.

]]>IBM takes over Zendhttp://www.schlitt.info/applications/blog/index.php?/archives/144_IBM_takes_over_Zend.html2004-04-01T06:55Tobias Schlitt - Weblog
IBM stated, that PHP5 will become the main plattform for at first all web and network related producst. For that, PHP will be integrated into IBM Websphere until August 2003. Zeev Suraski from Zend: "IBM is an important strategic partner for us. Now we are able to compete with Java and .NET.". Irving Wladawsky-Berger sees even much more potential for PHP in IBM. For that, it will be integrated into the Lotus Domino/Notes series and will replace the proprietary script language used, whith Lotus Notes 7.0.

Shortly after this announcement, the PHP community announced some seriouse decisions on recently discussed problems. PHP will now replace all under_score_function names with studlyCaps writing. Andi Gutmans: "All major languages use studly caps for their naming convention, so PHP has to fit in this schema." The community itself introduced some more changes.

See Alan Knowles for more info
and the announcement of IBM and Zend here. ]]>
The DSO Myth (Part 3)http://www.schlossnagle.org/~george/blog/archives/242_The_DSO_Myth_Part_3.html2004-04-01T03:58george schlossnagle's blogThe stats results have been amended to reflect this. The downside of building your shared libraries without -fpic is that they end up not sharing a memory footprint, but you get the management benefit of being able to recompile specific extensions (or PHP itself) without recompiling Apache, which is nice. ]]>Changes to case sensitivity in PHPhttp://blog.akbkhome.com/archives/31_Changes_to_case_sensitivity_in_PHP.html2004-04-01T01:59Smoking toooo much PHP After rather heated discussions on IRC over the last week, concerning StudlyCaps naming and php core. The core php developers (from what I saw) appear to be close to agreeing on some changes for PHP5.1

  • Finally we will get Case sensitive functions / methods
  • the _ underscore naming convention for PHP functions will be gradually replaced with StudlyCaps for all objects,  methods and functions.

Alot of the discussion centered around the impact this may have on old code, In general it was the concensus that it would only break really bad code hosted on sourceforge, which wouldnt be a complete disaster..

I have to admit it was pretty rare to see such agreement in such an argumentative group, but they considered consitency to be the ultimate goal, and it would make the life fo developers considerably easier.

Some of the core deadlines in the strategy are
  • Changing the manual to add the extra 500+ functions, and make the old ones redirect to new ones, (eg. going to www.php.net/array_count would redirect automatically to www.php.net/arrayCount) this is expected to be complete by July
  • For adding the aliases in, Lukas Smith was unanomously nominated to undertake the code changes, normally a core PEAR developer this will be his first major expedition into the C source code. it is expected that this will be completed by November.

In the meantime, they do recommend that if you are using any of the existing under_scored methods, that you write short wrappers to ensure that when they are removed in PHP5.1, that your applications work:
eg.
if (!function_exists('array_count')) {
function array_count($a) { return arrayCount($a); }
}



View Extended Entry ]]>
Conditional Class Declaration - bad practice?http://www.sitepoint.com/blog-post-view.php?id=1612652004-03-31T18:58SitePoint's PHP Blog: Dynamically TypedThe DSO Myth (Part 2)http://www.schlossnagle.org/~george/blog/archives/241_The_DSO_Myth_Part_2.html2004-03-31T18:32george schlossnagle's bloghere was wrong.

Following Ilia's good advice, I just decided to get to the meat of the matter by benchmarking PHP scripts. Interestingly, the more complex the script, the higher the overhead. On a simple 'hello world' type page, ab showed around a 4 performance difference, close to what I would expect.  On a more complex page though (the Smarty demo page) I consistently got a 30 performance difference.

In retrospect I think the flaw with my original tests is that the functions called from within PHP are almost exclusively PIC code as well, so you actually incur a larger overhead the more frequently you call them.

For those interested, raw ab results are up here. ]]>
Planet PHP linked on PEARwebhttp://www.schlitt.info/applications/blog/index.php?/archives/143_Planet_PHP_linked_on_PEARweb.html2004-03-31T18:04Tobias Schlitt - WeblogPlanet PHP recently got a link on PEARweb. ]]>