Re: transparent output compression patch

At 06:42 PM 11/13/00 +0100, Thies C. Arntzen wrote:

On Mon, Nov 13, 2000 at 06:53:59PM +0200, Zeev Suraski wrote: At 18:45 13-11-00, Thies C. Arntzen wrote:

On Mon, Nov 13, 2000 at 05:52:37PM +0200, Zeev Suraski wrote:

Damn :) We'll have to merge code now. I'll first get your

original code

to work, and then you'd could merge in your changes. Zeev

    do we really think this belongs into a scripting language? i
    think that transparent output compression should be part of
    the web-server (like an apache 2.0 filter) or a proxy in
    front of the server.

I don't see any reason for this not to be in the scripting language. It's useful, there's no current way of doing this (we're still years away from wide Apache 2.0 acceptance), why not support it?

    we rejected a similar patch (for 3.0) for the very same
    reason around a year ago if i recall right. but that would be
    no reason to reject it now - i agree.

The reason was that in 3.0 it was terrible. It was a real PHP core hack but with the output buffering it can be done very cleanly and nicely.

There are things in PHP that are completely not useful for Web development and nobody cares about them - I wouldn't start rejecting useful functionality because it "doesn't belong in the scripting language"...

    i played around with the output-compression tools that are
    around and i'm not too impressed. but who knows maybe they
    had problems with my applications simply _because_ they
    weren't written by us and integrated into php;-)

    if we have a clean enough interface for this and it doesn't
    end up as a "hack" i agree -> let's get it in.

I guess I addressed this before reading the second part of your mail :) Andi --- Andi Gutmans <andi@zend.com> http://www.zend.com/