As discussed before, I have now tagged the remains of Patrick McManus' work as te-20010110. I'm now commencing to get three ket
goals:
* RFC 2616 compliance for the te and transfer encoding headers. (this means just chunking support)
* a modular [de|en]coder system.
* gzip/deflate integration
the 1st one I expect to be trivial. The second should be fairly easy as Patrick wrote his code in a modular fashion.
The third one should be fairly easy again.
I'm hoping to convince you guys to merge in the 1st step once it's isolated and stable without the other two. The reason I think
that is a good idea is:
* It will increase the utilisation of persistent connections (don't have to drop the link on unknown object sizes)
* It's another step towards being rfc 2616 compliant (and that's a good thing right?)
* It should have little or no impact on stability & performance, whereas the second and third steps might.
* having it in place will allow Adrian and the other FS guys to see any issues it raises with their (re)work.
Henrik: I accidentally
cvs rtag -r Z-te_merge_HEAD Z-te-20010107_merge_HEAD squid
instead of the Z-te-20010110_merge... that I actually wanted - so both that command and
cvs rtag -r Z-te_merge_HEAD Z-te-20010110_merge_HEAD squid
have been run. What's the easiest way to get rid of the erroneous 20010110 one?
Rob
Received on Wed Jan 10 2001 - 05:47:43 MST
This archive was generated by hypermail pre-2.1.9 : Tue Dec 09 2003 - 16:13:16 MST