From darrellroot at mac.com Fri Oct 29 17:45:19 2021 From: darrellroot at mac.com (Darrell Root) Date: Fri, 29 Oct 2021 15:45:19 -0700 Subject: [netrek-dev] Netrek cited in my video on type-length-value encoding Message-ID: <530320B7-0F82-46D2-9346-2860743185DD@mac.com> Netrek-dev, I’m teaching a class on TCP/IP network programming and publishing my “practice” videos on YouTube. My most recent video covers type-length-value encoding. It starts with a review of the Netrek network protocol, which is type-value, and discusses how the lack of a length field makes adding new message types to the protocol difficult (because the recipient wouldn’t know how many bytes to advance after encountering an unknown message type). Apologies for highlighting a weakness of the Netrek network protocol, but I show proper respect to the grandfather of modern internet video games and was happy to incorporate Netrek into my class. Darrell https://www.youtube.com/watch?v=c4msYpc_uF4 -------------- next part -------------- An HTML attachment was scrubbed... URL: From quozl at laptop.org Fri Oct 29 18:06:32 2021 From: quozl at laptop.org (James Cameron) Date: Sat, 30 Oct 2021 10:06:32 +1100 Subject: [netrek-dev] Netrek cited in my video on type-length-value encoding In-Reply-To: <530320B7-0F82-46D2-9346-2860743185DD@mac.com> References: <530320B7-0F82-46D2-9346-2860743185DD@mac.com> Message-ID: <20211029230632.GA506636@laptop.org> Well done. At the time, the Netrek developers could not afford the bandwidth to include a length, and that's why the length is taken from a table held by both parties to the protocol. Arguably, this is a strength in efficiency, at the expense of requiring versioning. A version is included, but was not often used.