NYCPHP Meetup

NYPHP.org

[nycphp-talk] upgrading PHP on CentOS 5 -- or should I?

David Mintz david at davidmintz.org
Wed Jul 30 16:24:53 EDT 2008


On Wed, Jul 30, 2008 at 1:04 PM, Justin Dearing <zippy1981 at gmail.com> wrote:

> David,
>
> So what exactly is the cause of the bug? Is different html of
> javascript being generated by different versions of PHP? Is it a
> matter of warningnotices and errors being raised in the code producing
> the javascript?



Ummm -- I don't know. That is, it's an Ajax thing and the JS is (mostly)
hard-coded client-side. With Firefox/Firebug of course I see exactly what's
going on with XHR. With IE6 I have no clue what's going on when it
fails/acts funny. That's the issue -- code that breaks ONLY with IE6 and as
far as I can tell, IE6 + PHP 5.1.6. Because try though I might, I can't
reproduce the IE6 bug on my dev box (PHP 5.2.6) but the bug does happen with
the same code on the production boxes (PHP 5.1.6).

Specifically, it's a situation where the user is posting a request for a
court interpreter for a court proceeding -- so there's a form with fields
for stuff like date, time, language. For the defendant name(s) however we
have thousands in the db already and don't need/want them trying to spell --
it's an ideal scenario for autocompletion. So they type something and select
a name from the results returned, and my callback inserts a node in the DOM
with a hidden field holding the record id, bla bla bla. Works great but in
IE6 sometimes there's an apparently random extra record gratuitously added
to the POST payload. At first I thought it might be caused by multiple users
doing stuff at the same time but that ain't it.

Oh yeah, in IE6 there is also the issue where they type something and
nothing happens the first time, type some more and then it works normally
(more or less). That's a known issue that was supposed to be fixed in the
release I am using but it ain't so.



> Also, why not install a seperate copy of apache and php in a seperate
> directory that you built from source? You can run apache on a port
> other than 81 and see if the problem still exists.
>

Jeez, I suppose I could but I have been there and done that:  wget and tar
xzf whatever, ./configure --etc --etc --etc, watch it fail,, chase down
dependencies, and dependencies upon dependencies, google obscure (to me)
configure errors, watch the compiler work for 11 minutes and then puke,
google obscure compile errors, feel manly once it's all over and lay back
and have a smoke... nah. I mean, I have never had much trouble compiling
apache but PHP has been too exciting.



>
> I have a hunch though that warning are being inserted into the output
> code that your other php configuration places in a log file. Check the
> PHP.INI settings.



I could be logging more verbosely and I will do that.  Thanks.


-- 
David Mintz
http://davidmintz.org/

The subtle source is clear and bright
The tributary streams flow through the darkness
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.nyphp.org/pipermail/talk/attachments/20080730/843cc6d2/attachment.html>


More information about the talk mailing list