From HECTOR SANTOS · 1996-04-09 02:42 · #73125 in FREE25
In reply to what is a serial number? by andrew clarke
AC> FTS-9 specifically mentions a "serial number" with regards to MSGID AC> generation, therefore two consecutive messages should contain MSGID AC> serial numbers that follow each other in series, surely? AC> (Is there anyone who would like to interpret the spec differently?) To serialize something, in practical terms, doesn't necessary mean that it must be in "series", "incremental in nature". It could mean that it progressive or a chain, "you serialize it so that you can follow a chain of events". Does serialization imply reversibility and accountability? How ever you want to intepret it, FTS-9 is just concern that the number is unique on the originating system for a period of 3 years. How do you achieve this is your goal in your design with the consideration of all the parameters and issues you have to deal with, namely, compatibility with other software not only running on the same machine, but on others as well. Surely, one can be an anal retentive semantic meaning and simplest use a "serial counter" to result in the a basic progressive incremental "counter" system. Well, first, there is no guarantee this "counter" will be unique across other systems. But lets just concern ourselves with what FTS-9 says "Unique on the original system" only. That is what it says. Right? This presents a multiple of problems since in order to guarantee uniqueness, all email generating processes in the system whether its coming from a multi-node, multi-threaded system, 3rd party door system, they all must have a "COMMON SERVER", if you will, to retrieve the last number used. Do you see the picture? Do you understand this problem? So currently, most systems use their own implementation because there is no standard mechanism for one to get the "current count" so that one can increment it. BTW, new Client/Server software do allow for this to happen now. Wildcat 5.0 requires that you use the SERVER to create or toss a message and it will always return a unique message id for its own dupe checking system. This is all relative to the installation. There is no way to retoss the same message. No way. Obviously, the vendor has factored in some "parameters" of the message into the "equation". In my opinion, this is ideal and is required. By using a serial number counter, you have effectively removed or excluded the parameters of the message itself and this increases the likelyhood that two SAME messages can be theoritically tossed into the system. Whatever the "equation" one uses, it must be a "Function" of the message parameters IMHO. But regardless, whatever the "function" is, lets look at one of the fallacies of FTS-9: It says that no two messages can have the same serial number on the same machine. Well, that implys that the SAME TWO MESSAGES can not have the same serial number on the ANOTHER MACHINE. Right or Wrong? If you say WRONG, you can stop right here. If you say right, then you do grasp the situation and then further analyze, that it also says that the generation of this unique serial number is "left to the implementor". Well, unless two systems USE the same "implementation" there is no way that one can guarantee that another implementation will NOT "REPRODUCE" a serial number that was already another system and sent to your system. So my point is, you can go directly to the specs and attempt to follow it and you will see that many things, there are issues that doesn't quick make sense. Remember, it is not the bible. The authors of the specs make great contributions but they are also not great sages with perfect vision into the future. Many times, it just doesn't quite work. If it did, we would be having many of the debates in this echo. You need to take the specs and do the best you can with them. If they are issues or questions about usage, bring it up. In my opinion, if there is one major problem with FIDONET mail formats, is the LOST of addressing. FTS-9, MSGID/REPLY offered an great opportunity to maintain this addressing which HELPED alot of BBSes and it certainly helped alot of users. In this echo, we get LOST of "serial numbers" and the chaining of the FTS-9 and it has hardly any impact in the fido market. Almost none whatsoever. We throw away the important of addressing. --- Platinum Xpress/386/Wildcat! v1.2j * Origin: Xpress Tech Support BBS * 305-248-7815 (1:135/382)