Slimelite BBS

An political question

Back to FREE25

From Pierre Phaneuf · 1996-04-07 01:51 · #73102 in FREE25

In reply to An political question by HECTOR SANTOS

 HS> Now that I am aware that it MAY NOT be right because of "follow ups",
 HS> the best one can do at this juncture is just make a "NOTE" of it to
 HS> the end user, either with a PROMPT or whatever.

I think I did my thing. This is what I had in mind. As long as you put the content of the MSGID in the REPLY even in follow ups, if you follow what you've written above, I see no point in continuing this argument.

 HS> If the editor package has a NODELIST for lookup, then its a mute
 HS> point and I think that a major reason why here we have a different
 HS> point of view of it - most people here are sysops, have a BBS or
 HS> point system and have a NODELIST available.  But I can assure you
 HS> that END-USERS, the majority of your email writers do not have access
 HS> to a nodelist to do verification.

I know it won't work all the time, but checking the other message in a mail packet could be a good way of getting a surefire address, if you can find a message from the right person... Using a combinaison of *both* ways would be best, IMHO...

 HS> Joho is one of the gifted young engineers and he has shown that threw
 HS> his technical writings more so than his software.  He is one that
 HS> likes to see the "entire picture".  In my opinion, I would think he
 HS> would agree with me but would submit to the current ill usage of FTS-9
 HS> because its now one of those things where its too late.  It does
 HS> effect the threading algorytms of many mail processors out there.

I agree with you about joho. I admire his work a lot, particularly his EMSI protocol. I learned a lot looking at this protocol. It is what made me think of my "infinite length data structure", which may either become one of the most used thing, may already be in use and I reinvented it, or it could stay unknown. Either way, I owe him much.

 HS> So the reason question, is there a compromise were we can keep the
 HS> integrity and important of addressing in the mail while still have
 HS> threading?

I have some difficulty about every side of addressing in echomail. I could rely on the MSGID, but it's optional. I could rely on the Origin line, but it's optional too (I know it is practically mandatory, but in the spec it says it's optional and I don't want to do software that will break if it gets mail without any). So much for getting a reliable reply address. Getting one for netmail follow ups is even harder, as we all learned in this argument.

 HS> I say its for BOTH. MESSAGE IDENTIFICATION to me implys the ABILITY
 HS> to "identify who" wrote the message and REPLY LINKING implys to me
 HS> the ability to MAP the "HISTORY" of the thread.  Its BOTH and it
 HS> works 100% perfectly for your more NATURAL reply system.  It breaks
 HS> for a FOLLOW UP and thats only because software writers literally
 HS> took the statement in FTS-9 to use the current MSGID for your reply
 HS> as their "natural" reply engine in their software.  What was not
 HS> clarified or "considered" was the "follow situation".

I think of it more as a *message* identification across the network (akin to the message number in the message base) more than an identification of who made it. I think the REPLY line should contain the content of the MSGID line, since it is a direct child of the message, it points to the one that is being quoted, even if it is directed to another person. If Mats Wallins see my message and follow up to you, the parent of his message should be this one, since it was the one that ticked him, not this message parent (as would do using the content of the REPLY line).

 HS> In my OPINION, this is a perfect example of where ADDRESSING is just
 HS> as important.  How many times has you read a REPLY and its to ALL. You
 HS> don't know who he is REPLYING too.  So yes, you are STUCK to use a
 HS> THREADING system to find out.  But this only WORKS for the small
 HS> amount of people, like sysops, who HAVE access the entire or lenghty
 HS> database of mail.  Most people do not.  So its just as important to
 HS> keep an "reply system" to BOTH parties in the CURRENT message.   Thats
 HS> just common sense to me.

Let's have a more sensible system than that, and add the feature in a formal manner... I know it was not meant that way, but having an INTL kludge even in echomail would be quite good in my opinion. Many (if not all) software use this line before anything else to get the address. I wonder what happen if you have an INTL kludge in echomail...

 HS> You sit down and analyze your natural REPLY engine logic and draw a
 HS> truth table with the BASIC assertion that ADDRESSING must stay intact
 HS> and you will conclude that ADDRESS is important and that isn't be
 HS> broken just so we can keep the same "serial no" for threading.

The point is that it is a threading device and that instead of using it in a sighly bended manner, we should define a proper way of doing it that would preserve *both* the threading of FTS-9 and the addressing that *is* needed.

 HS> But you sent me a NETMAIL now, where does the THREAD come into play?
 HS> How would your EDITOR now back track to find the original message
 HS> based on a THREAD that originated in a different FORUM?

I don't have the time nor the energy to do an editor right now, I want to finish off my mail processor first, but I agree with you that this is a bad part of most mail editor... I'd like to be able to follow a thread from echo to echo to netmail to echo as the subject changes and the discussion moves to stay in topic, and this is possible using existing MSGID/REPLY threading, just not implemented. It's an hard one too...

 HS> So what if you want to repl a message to me via netmail? You see Paul
 HS> write a message to me, and you want to instantly reply to ME via
 HS> netmail using the same message?  It may work for you with your sysop
 HS> editor having a nodelist, but for the majority of the end-users, you
 HS> deny them the right to use that feature and technically, there was NO
 HS> reason for it.  ALl I am hearing is that it will break the "SYSOPS"
 HS> threading logic on his local machine.

On my BBS, most (about 3/4) of my users are local one, and they benefit of the whole of the threading system, it being in heavy use. I don't agree that it's a "sysop only" feature. I have as many points as I have offline users, and threading works with Blue Wave (it's the most popular format on my BBS), as long as the message is in the same packet. Some of my users download their messages only every two or three day on weekdays, so they get pretty much all the relevant threading... I firmly agree that we should fix this addressing problem, but I oppose the idea of breaking the threading to achieve it.

Pierre
phaneuf@absint.login.qc.ca

--- FMail/386 1.02
 * Origin: Zaphod Beeblebrox for President! (1:167/136.14)