I got thinking about the whole DDoS situation, and while I certainly have my
own opinions around all of this, there's nothing I can do about it.  What I
can do though is figure out ways I can improve the systems I'm working in.
 The place I think this starts is in our own Twitter libraries we work in
within our own language environments.  As Chad mentioned, HTTP protocol
dictates that libraries operating on the protocol respect 30* redirects when
requested.  It shouldn't even be an issue within our respective Twitter
libraries if they were using HTTP libraries that are fully HTTP-compliant.
 I was very impressed to learn that Perl's Net::Twitter uses LWP::UserAgent,
which is fully HTTP compliant, and I didn't have to do anything to adapt to
the new requests by Twitter.

Maybe it's time to start looking into each of our own respected Twitter
libraries and ensure they're utilizing fully HTTP compliant HTTP libraries
to access Twitter?  In this way it won't ever happen again, so long as
Twitter is following open standards and protocol.  I'm really surprised at
all the people having issues with 30* redirects when it's an HTTP standard
in the first place.  What other areas of our own code can we be fixing to
make our environments more efficient to work with the constantly changing
Twitter environment?


Reply via email to