If I'm connecting to a site with Lynx, I sure as heck don't want them to try to serve me some skeleton HTML that will be filled in with JS. Because my browser doesn't support JS, or only supports a subset of it.
User Agent being a completely free form field is the real mistake IMO. Having something more structured, like Perl's "use" directive, might have been better.
The problem with services using the user-agent to determine whether or not to allow a client access to a resource outweighs any benefit. I'm in the "it was a mistake to include this in the spec" camp.
If I'm connecting to a site with Lynx, I sure as heck don't want them to try to serve me some skeleton HTML that will be filled in with JS. Because my browser doesn't support JS, or only supports a subset of it.
User Agent being a completely free form field is the real mistake IMO. Having something more structured, like Perl's "use" directive, might have been better.