-----BEGIN PGP SIGNED MESSAGE-----
Date: Fri, 08 Sep 95 09:48:31 -0400
From: Steve Kent <kent(_at_)BBN(_dot_)COM>
ASN.1 is primarily a syntax and a concrete encoding for
transmission of data, not necessarily for storage. The intent, in
general, is to transform data from a local representation into ASN.1
for transmission, then back into a local form at the other end. In
than light, I'm not sure your comment about the overhead of storing
data in ASN.1 format is a generally fair criticism, although I agree
that one might choose to retain data in the ASN.1 encoded format as
a processing efficiency measure in some instances.
In practice, you're generally correct. However, there was one case
where there was a desire that an particular application's PDU, which
included a Kerberos ticket and authenticator, be smaller than the
Ethernet MTU, to avoid UDP fragmentation. This was possible using
Kerberos V4, but not possible using Kerberos V5, due to protocol bloat.
Not all of this bloat could be laid at the door of ASN.1, but certainly
a substantial portion was ASN.1's fault.
One can argue that this particular application was badly designed.
However, it doesn't change the observation that occasionally protocol
bloat has bad effects in real life.
- Ted
-----BEGIN PGP SIGNATURE-----
Version: 2.6.1
Comment: Processed by Mailcrypt 3.2, an Emacs/PGP interface
iQCVAwUBMFB7gkQVcM1Ga0KJAQEAbQP/e3/30ObYuHiRloJTmsEXLgTkgyirxZ2y
THokJ9albpORysmBc7ly+ZvyW47WgLPn0N6PXNUB/I+ov05mDtqJi/NIgLEhRj9W
tIJwech07FOQZZSeP6hkVo5EySoPwd8ONsJPurGZWJA5WaGLdy2vy14sBzV0oRxs
4AVCHX6qZb4=
=MAuM
-----END PGP SIGNATURE-----