Re: Sending raw headers
| From: | Teodor Cimpoesu | Date: | Fri, 30 Jun 2000 17:19:45 +0000 |
| Subject: | Re: Sending raw headers | ||
| References: | 1 2 3 4 5 | Groups: | php.general |
| Request: | Send a blank email to php-general+get-4072@lists.php.net to get a copy of this message | ||
Hi Timo!
On Fri, 30 Jun 2000, Timo H wrote:
> Yes, webchat is exactly what I'm developing. It works nicely with
you too? :)
> redirections also but naturally is not as fast as it could be (and we're
I sow least NS on Linux doesn't send again Cookie headers on a redirected
location :(
> not satisfied until it's perfect, right?). I tried many different ways to
> do the headers, cgi version with -q parameter, etc. Doesn't work because -q
> denies _all_ headers (says "headers already sent in blabla.php in line 2).
my trick is:
on NS - output directly JavaScript, since it's interpreted by-fly
on IE - have a 0 size frame (NS starts Java for this, and scares
away any user) since it's allowed and have the generated JS in that page
on both cases (well,this should be least 70-80% of the users) I
update a display layer which has not flicking effect. [interested?]
> simply by echoing the Status:204 won't work. I wonder how hard it would be
> to patch the sources so php wouldn't send the content-header if special
> headers are sent before it.. You wouldn't know any workaround on this btw?
I stared for some time at mod_php4.c, but couldn't come up with a good solution.
Basically you need to test on Content-Type each time, and I guess there
was an efficiency issue not to do it (in 99% of the cases you will have one)
-- teodor