I have some php gz compressing, caching (and more) code which does the next:
4 files are used:
1. http://laban.rs/test/index.html (gz, force cache, css external)
2. http://laban.rs/test/i1.html (gz, no cache, css external)
3. http://laban.rs/test/i2.html (gz, force cache, css internal)
4. http://laban.rs/test/i3.html (gz, new code, css internal)
Net headers (curl used):
http://laban.rs/test/-/a?i=index
http://laban.rs/test/-/a?i=i1
http://laban.rs/test/-/a?i=i2
http://laban.rs/test/-/a?i=i3
I have tried some online checkers and here are REDBot results.
Page index
https://redbot.org/?uri=http%3A%2F%2Flaban....st%2Findex.html
Page i1
https://redbot.org/?uri=http%3A%2F%2Flaban....Ftest%2Fi1.html
Page i2
https://redbot.org/?uri=http%3A%2F%2Flaban....Ftest%2Fi2.html
Page i3
https://redbot.org/?uri=http%3A%2F%2Flaban....Ftest%2Fi3.html
I would like to know:
1. Which of pages above gave the best results.
2. How to solve REDBot problems / warnings.
2.1. “The resource doesn't send Vary consistently.”
2.2. “The response body is different when content negotiation happens.”
2.3. “The ETag doesn't change between negotiated representations.”
3. How to gzcompress external CSS (or any other content)?
4. How to cache page-images?
Also:
1. Is pragma cache / no-cache depreciated?
2. Is connection keep-alive depresiated?
3. Do we still need content-length?
4. How I can simulate browser which does not support GZip compression to see the unzipped content?
Thanks,
Dag