Re: Inline content modification?

----- Original Message -----
From: "Henrik Nordstrom" <hno@hem.passagen.se>
To: "Robert Collins" <robert.collins@itdomain.com.au>
Cc: "Joe Cooper" <joe@swelltech.com>; "Squid Dev" <squid-dev@squid-cache.org>
Sent: Thursday, January 18, 2001 7:27 AM
Subject: Re: Inline content modification?

> Robert Collins wrote:
>
> The kind I was thinking about is where you slightly modify the URL to
> get a modified content.
>
> The problem with content modification combined with URL rewriting is to
> make it consistent. What happens if the client sends a request for the
> original URL? If modification is done on the server connection then the
> client will get the modified content, which may not be what was
> expected. This also affects peerings. Peerings will be done on the
> "backend" URLs, where it might not be intended that the modifications
> should be applied.

Yes I agree - any changes made coming into squid _have_ to be changes that are wanted for all clients, for all operations. And
support has to be thought out for keeping the original metadata so that IMS requests upstream don't miss if things like
content-length change.

> > Filters are data filters, not HTTP protocol violators per se - they
> > can be used for that, but also to easily implement parts of the
> > spec : IMO some parts of the HTTP spec are best implemented as
> > filters - I.E. the TE code is about to become a filter.
>
> HTTP allows for non-transparent proxies to do much whatever they like,
> as long as some basic protocol parameters are fulfilled. Examples of
> this are image recoding, automatic translation, and much whatever else
> you can imagine.
>
> A transparent (in HTTP terms) proxy however cannot do any such
> modifications, not even on the URL.
>

I know :-]. What I was pointing out is that the filter is a data passing technique that may be more efficient for some of the
operations squid has to do and no less efficient for other operations than loop of functions currently in use in some spots. Filters
will also happen to make non-transparent operation much easier to create once things like old metadata support are in the support
code.

Rob
Received on Wed Jan 17 2001 - 14:50:59 MST