| Ken Schaefer 2005-12-01, 8:55 pm |
| "rusga" <reply@newsgroup.plz> wrote in message
news:op.s04o02o5umizor@localhost...
: a) While the server is up and there's no change in the content it provides
: the ETag value remains the same, *tagging* (and I hate this word) the
: server for a long period of time. This scenario is the majority of web
: servers around.
There's a separate ETag for each page. It's an MD5 hash isn't it? of the
page content?
: b) Think. Why should it be an unique ID? For the purpouse it serves, a
: "Changed: True/False" (or the like) header would sufice.
How would that work? How does the server know whether the page has changed
compared to the browser's copy? When the page changes, a new MD5 hash is
generated. The browser can compare that to the MD5 hash it had before.
: c) Also I consider it redundant at least with already existing cache
: control directives.
It's not about caching AFAIK. It's about knowing whether the page content
has changed
: There's no subversive intention on the will of remaining anonymous and
: preserving privacy. Only on the intention of stoping or controling this
: will, there is.
Where is the privacy risk? Who's privacy are we talking about here?
Cheers
Ken
|