RE: PEER_ERROR message
Johnson Walter-CWJ002 <[email protected]>
| Newsgroups | gmane.ietf.rserpool |
|---|---|
| Message-ID | <6F8DFFA2C996D711945800065BFC9E4A10D06ACF@il02exm11> |
Randall Stewart <> wrote: > Hmm.. that is a good point.. maybe it is > best just to leave the number.. > Agreed. > However the error name ENRP_ERROR or PEER_ERROR needs > to be the same throughout the enrp doucment.. > Agreed. > And if we have no use for it in ASAP.. then we may > want to get rid of it.. or if we do have a use then > lets write up how it is used :-D > Agreed. > I do like PEER_ERROR for both of them.. even though > it may mean something different :-o > Disagree, but really don't have a strong opinion. I know one can sort on the protocol and the message type to determine how to decode the message. But having separate names makes it clear to readers new to the ASAP and ENRP drafts that these are unique messages with unique uses. Within the drafts, the message names only act as tags for the real messages sent over the wire so why not make them as descriptive as possible. In this case, it might be best to make it ASAP_ERROR and ENRP_ERROR to identify the protocol error type. For instance, an ENRP namespace server can potentially receive both types of errors and there might be a need to describe a situation where both are received. This would be much easier to explain using unique names than describing both as PEER_ERROR. Just my thought. > R > > Walter