Talk:Direct Internet Message Encapsulation

Size encoding oddity

The article states that “One difficulty here was that it could form an HTTP message of, essentially, any size (the limit being the size information for each chunk, which was 32 bits so 1 gigabit).”. In a protocol of this scope, encoding a length in bits seems highly improbable (bytes would be much more likely). In either case, however, the math appears to be wrong. A 32-bit unsigned integer can encode values up to 4,294,967,295, while a signed integer can encode values up to 2,147,483,647. Neither of these is 1 gigabit; they are 4 or 2 (respectively) gibi{bits,bytes}. I don't know the protocol and the documentation at Microsoft’s site appears to be gone, so I don’t know what the actual correct answer is. If it really does use a 32-bit integer and yet somehow have a size limit of 1 gibibit, an explanation of why such an unusual size cap exists would be appropriate. —Preceding unsigned comment added by Hawk777 (talk • contribs) 23:33, 9 May 2011 (UTC)

Most of this article needs review

I've also noted the 1 gigabit oddity stated above, but it is also stated later that "HTTP is connectionless". While I can agree that, by design, most HTTP servers are stateless are require artifices to simulate sessions, HTTP is by no mean connectionless. Connections/exchanges are inherent to any protocol... — Preceding unsigned comment added by 160.109.104.44 (talk) 12:32, 1 August 2012 (UTC)