ml>  NB> but it's only to show you that whatever reader/editor you're 
ml>  NB> using
ml>  NB> is what messed the quote up. But it might very well not be the
ml>  NB> reader/editor you're using, either.
ml> 
ml> yes.. the one that caused the mess above (which i've removed for this
ml> reply) is not the one i use... it was already messed up when i wrote my
ml> reply... it is also messed up like that in the RAW PKT file that
ml> delivered it to my system...

NOTE: see the above? here it is again repaired... this repair took only two
DELete keystrokes... the first at the end of that short line and the second
at the end of another short line that appeared... thie reader/editor almost
has it perfect but this minor fixing is no big deal... not like trying to
repair chopped and/or missing words...

so anyway, continuing on...

ml>  NB> but it's only to show you that whatever reader/editor you're 
ml>  NB> using is what messed the quote up. But it might very well not 
ml>  NB> be the reader/editor you're using, either.
ml> 
ml> yes.. the one that caused the mess above (which i've removed for this
ml> reply) is not the one i use... it was already messed up when i wrote my
ml> reply... it is also messed up like that in the RAW PKT file that
ml> delivered it to my system...

 NB> Below you asked me about RAW format, this is what I was talking
 NB> about. 

ah... i wasn't sure what you meant... message base or pkt or maybe something else...

 NB> The raw packet file that's delivered to your system from a
 NB> Synchronet system, could very well contain the rest of the lines
 NB> that were originally cut off for viewing. 

yeah, no, they do not contain those missing parts... they wouldn't be
missing if they were in the PKT ;)

 NB> If that's the case, it's still wrong, of course, but at least 
 NB> moves us a little further in the case of fixing the issue. :| 

yes, definitely wrong but sorry to say the problem is and has always been
in the editors that do not wrap and flow quoted text... it has been this
way since the beginning of time for editors that do this...

ml> i will also note that i specifically did not attempt to manually fix the
ml> quote in question as i tend to do most of the time... there were just too
ml> many missing parts for me to remember them all and i don't have the
ml> capability on this point system to go back in another reader to view the
ml> original when i'm writing a reply...

 NB> You shouldn't have to fix anything. 

agreed but i like to try to ensure that everything is there that should be
there... just one missing word can completely reverse one's statement...
consider this below... the 'x' marks the margins for this example...

  x                                                                   x
  blah blah blah blah blah blah blah blah blah blah blah blah will not
  do blah blah blah blah... 

now, on a system that truncates, the entire word "not" is chopped
off and look at what the statement says when it is quoted back...

  x                                                                   x
   SO> blah blah blah blah blah blah blah blah blah blah blah blah will
   SO> do blah blah blah blah... 

in this case, both have the same margins... some have a shorter right
margin... the quote prefix is pretty normal... [space]xx>[space]... that
shoves the word "not" past the margin where it should be wrapped
to the next line but instead it is simply chopped off... now a statement
that was negative reads as if it were positive...

 NB> We have no advanced standards here, as they're all left as 
 NB> proposals.. 

sorry but not... there are proposals being worked on as well as standards
being updated...

 NB> BECAUSE, there's nothing we can do about older software that's not 
 NB> developed anymore, but still loved and used by many.

if the standard is updated, it doesnt effect that software anyway... no one
can be required to use any standards other than those that policy
mandates... this /should/ also include those which are required by normal
network operations (ie: the nodelist format) but that's been flushed down
the shitter by mailers that do not use the St Louis formatted nodelist...

ml>  NB> Mainly, this is so you have a little more knowledge in the argument
ml>  NB> involving you, Carol, and the others in another echo.
ml> 
ml> understood... and GoodOnYou for trying to understand what's going on
ml> and why...

 NB> Hey, I also mentioned the fact that SMB isn't supported in any of
 NB> their software. If it was, I'd bet more sysops would be inclined to
 NB> use it. 

yup, i saw your comment in that echo :)

 NB> What would be even nicer, is if it could be ran as an external
 NB> program (not necessarily a "door" but actually shelling out and
 NB> running the program, and when done, exits back to the BBS (ie: I
 NB> can run nano and vi from Synchronet, but cannot run GoldEd or any
 NB> of the other variants). 

you probably couldn't do it from remote but as a local only editor,
possibly... the problem is that the bbs generally creates a temporary file
for the editor to errmmm... edit... then when the editor exits, it saves
the file and the bbs imports it into the message base... if the file has
not changed, then the bbs may assume that the edit was aborted unless the
editor exits with a specific errorlevel for an abort as opposed to normal
termination... i rather doubt that nano and vi have such errorlevel exits
capabilities... they are from a completely different world and mode of
thinking... sysop/point editors are local editors that replace the bbs and
its message base interface... they are a lot more than just editors because
they do have to have the message base support built in...

ml>  NB> The original reply from Richard seemed to have cut the text off at
ml>  NB> the end of the lines. 
ml> 
ml> yup... that's due to the editor he used... it creates what is/was known
ml> as chainsaw replies... basically, it doesn't word wrap and reflow the
ml> existing quotes... instead it chops off text at its line limit and then
ml> quotes that... this is a very old problem and it indicative of what some
ml> term as "lazy coding"...

 NB> It would be nice to know what editor he's using exactly. It could
 NB> just be a configuration option he hasn't enabled. Deuce's JSEditor
 NB> has the option to wrap lines, and the one I use on Synchronet
 NB> (Slyedit) has the same option. 

that information /might/ be known if synchronet played by the rules
?ID/tearline rules but it doesn't... at least not in richard's messages...
there's a PID (of the bbs which might be the editor if its internal line
editor [if it has one] is being used) and a tearline both of which are
indicating the same basic information... if there was either indicating the
name of the editor used, it has been wiped away and replaced... but the
problem is that both PID and tearline text are mutually exclusive... there
can be only one or the other but not both... synchronet supports the
PID/TID standard because it (and its tosser) inject those line into the
messages... so something (else) is goofed with the tearline text being in
the message, too...

ml> on important item to note is that one set of quotes has two leading
ml> spaces which throws off other quoting software such that they don't
ml> recognize the quote... this is another coding problem... one that can be
ml> handled by peeking at the characters specifically looking for the '>'
ml> character and adjusting the quote and reflow as needed... but this goes
ml> back to that "lazy coding" thing ;)

 NB> I'm actually in multiple emails with the developer of the external
 NB> editor I use for Synchronet. He might be on board, but has no
 NB> definitive action on how to do it. I've told him the only way for
 NB> him to see GoldEd or variants in action, would be to setup a VM
 NB> with a separate mailer, tosser, etc., since Synchronet's message
 NB> base format isn't supported. So we're kind of at a wall here.

the concept is really simple... starting at the top of the text to be
quoted... determine if this is already quoted text or if this is a new
quote and remember this (hint: boolean)... depending on which it is, inject
the proper quote prefix (existing one plus '>' or the defined one)...
when you reach the end of the line (ie: the right margin) if there's still
more characters, back up to the first space and inject the newline
character(s) plus the determined quote prefix... that backing up and
injecting the newline character(s) is what keeps the text from being
lost... so we now have the chainsaw look to get rid of so we peek at the
next line and see that it is not blank... that means that it is part of
this paragraph so we remove the new line character(s) at the end of the
current line... that brings the next line up and tacks it on to the end of
this one... if we are working with already quoted text, then we need to
remove the existing quote prefix from "middle" of this line since
we are reflowing it... now we just have to loop back up and check the end
of the line for characters past the margin and do this whole process for
each line in a loop until we get to a blank line indicating the end of the
paragraph...

there's more but that describes the basic wrap and (re)flow process... 

blank lines only consist of the newline character(s)... nothing else... a
single space on a line is /not/ a blank line... visually, it may look like
it is blank but it is not... it contains a space... handle it just like any
other line and update the existing quote prefix or add the new quote prefix
to it...

[trim]

ml>  NB> If this is the case, it's probably one of Synchronet's external 
ml>  NB> editors (can't really blame Synchronet, as there is no "default 
ml>  NB> editor" for it).
ml> 
ml> agreed... this was the point of contention a while back when the
ml> synchronet maintainer, rob swindel, was "brought" into the
FIDONEWS echo
ml> when i and others were discussing this specific problem in there... that
ml> discussion, as it were, then branched off to synchronet's MSGID stuff in
ml> which is creates a MSGID for messages that do not have a MSGID... sadly
ml> rob left before he saw my reply to him with the evidence of /that/
ml> particular problem... at least there was never any acknowledgement of it
ml> by anyone involved in that discussion... i've been keeping an eye on the
ml> synchronet CVS/SVN/GIT/whatever and have not really seen anything that
ml> would indicate a fix for either problem... not until maybe in the last
ml> day or so but i haven't followed up on that either...

 NB> I brought him there, so I know what it was all about. 

i know... i was being nice and trying not to stir it all up again ;)

 NB> I can't stand arguing for another, when the other is just as 
 NB> active, if not more, than I am, and especially still currently 
 NB> developing for his software.

i understand... and it is hard to be in the middle and explain what the
problem is when one may not understand it all themselves... rather like the
game of "gossip" where what's said at the end is no where near
what was said at the beginning ;)

 NB> I thought the MSGID thing was fixed. Granted, 90% of people don't
 NB> actually upgrade Synchronet every time something is fixed (since
 NB> it's 100% CVS, with no actual physical releases. I definitely DO
 NB> remember him doing something before he left the conversation.. and
 NB> I thought it was this. 

i dunno... there's two situations with them... one is the nonstandard
format of the originaddr field... i can't make a private reply to that
message based on the contents of that field as generated by synchronet...

the second situation is that of generating a MSGID for those messages
without a MSGID... nick andre's posts are a perfect example... they do not
have a MSGID because his system doesn't support FTS-0009... synchronet
creating a MSGID and adding it to messages without one is the worst of the
two situations because that message will not have the same MSGID on all
systems... each synchronet system will create its own originaddr field as
well as its own serial number and that blows duplicate detection all to
hell...

 NB> The reason he left the conversation so abruptly, is because his RC
 NB> was being changed at that exact time. He lost at least 3-4 messages
 NB> he posted to you (never made it to FIDONEWS), and by that time, he
 NB> wasn't willing to post them again, unfortunately.

i understand that... and that points out exactly why one should never use
their admin address for network operations... if one has an admin hat of
some sort, then they /should/ have at least two FTN addresses... their main
one that never changes (unless they change nets) and their admin AKA(s)...
*all* network processing is done with the main address so when the admin
address moves to someone else, only that address need be removed from the
configs and no one has to jump thru hoops changing feeds and such... admin
addresses should only be used for official messages... like today when i
wrote my RC about a situation... i used my NC address and sent it to his RC
address... but everything went thru our main addresses for processing...

ml> as for synchronet not having a "default" editor, is there
not something
ml> that does come with it for those system that want to run bare naked
ml> synchronet as delivered? all of the bbs software that i've even seen has
ml> had some sort of default editor, whether internal or external... it was
ml> there in the initial package and the editor used when there was no
ml> other... the bbs software i use, granted it is a beta that desperately
ml> needs an update that is unlikely to come forth, has a line editor as its
ml> default editor... this line editor is akin to the old DOS edlin tool...

 NB> I'm not sure what it comes with nowadays, as I've run it for quite
 NB> awhile, and have upgraded via CVS many times. If I remember right,
 NB> it only came with some old stuff like Pine, and a couple others.
 NB> Basically crappy text editors like you mention. JSEditor (made by
 NB> Deuce aka Stephen Hurd) is promoted, of course, and it might be
 NB> included now, but I can't answer that honestly. The one I use on my
 NB> Synchronet system, is non-default, but the developer made it
 NB> specifically to clone IceEdit and DCTEdit of old.. which it does
 NB> perfectly. Those two editors did NOT add initials to quotes. 

 NB> So now I'm trying to convince him to do that. :)

that's the easy part... just remember that there can be more than two
initials... i would use spaces and periods to locate them... not commas or
dashes, though... well, maybe dashes... depends on how they're handled by
other software for this task...

ie: fred stare             : fs>
    john k lemon           : jkl>
    john k lemon jr        : jklj>
    john k lemon, jr.      : jkl>
    john k. lemon          : jkl>
    john k. lemon jr       : jklj>
    john k. lemon, jr.     : jkl>
    steve vander horn      : svh>
    steve van der horn     : svdh>
    steve van.der horn     : svdh>
    steve van.der.horn     : svdh>
    steve van.der.horn jr  : svdhj>
    steve van.der.horn, jr : svdh>
    mary magdeline-jones   : mm>


ml> i don't have a problem with it... i am always trying to help others see
ml> and understand the problems that are visible out there... i've never
ml> touted anything i run as being completely fault-free or perfect,
ml> either... but that gets missed, too, and in some cases, like what
ml> happened previously, some try to use it as a weapon against me... oh
ml> well... they can fix their software whereas i cannot... not without a
ml> huge alteration in my production system which has been in operation since
ml> the late '80's or so...

 NB> I think it was only used as a weapon against you because you were
 NB> complaining about another software with the same issues. Yes, you
 NB> use a sysop reader/editor, but any of your users post, and there
 NB> will be the same issues coming from your system.

possible but it was also a cheap shot...

 NB> Basically, I do believe Rob is following more standards than
 NB> proposals. All we can do is politely ask for some of these
 NB> proposals to be added to the current development, which I'm slowly
 NB> doing, but I also don't want to be ignored in the process.

i know that feeling... one has to support proposals to get them raised to
standards, though... this is something we're seeing being taken advantage
of right now in that other echo with the recent post concerning "i
just took an accounting [...]" where sheer numbers are going to be
used to try to shove something down other's throats...

 NB> Oddly enough, Mystic is completely pascal, and seems to quote
 NB> pretty damn good as far as the new proposals. Oddly enough, this
 NB> software has been doing it since it's birth. :)

maybe because the author saw the quote problem(s) from way back and never
let it/them get into his software's editor(s) in the first place ;)

)\/(ark

---