Hi, This sounds like a good plan to me. Let's keep the cookies that John and Jon have defined in syslog-sign and allow Rainer to make the definitions (with instructions to IANA) in syslog-international. This will clarify the use of cookies for other purposes and to differentiate between IETF cookies, vendor cookies, and any other types (e.g., experimental) cookies. Any objections?
Thanks, Chris On Thu, 4 Sep 2003, Rainer Gerhards wrote: > Hi, > > sorry for raising an additional suggest, but I am currently implementing > and as Albert said, questions arise while doing so ;) > > Cookies, as first appeared in syslog-sign, are a simple, efficient and > elegant way to add new functionality to syslog without the need to > change anything in older code. I assume this WG will make use of more > cookies as need arises. > > On the other hand, the same concept is also good for vendor-specific > extensions. For example, I am currently designing a way to let windows > event log data travel unaltered via syslog. I am very tempted to do this > with a cookie and will probably do so. > > This raises the question of standard vs. vendor cookies. I wonder if > some of the upcoming IDs (-international?) should formalize the cookie > as an optional field. If it starts with @#, then it is an IETF cookie > (IANA?), but if it starts with #@ (or @#X-)it is a vendor cookie. This > would remove any doubt, especially when more people begin to adopt this > scheme. > > Any thoughts? > > Rainer > > >
