Text 134, 63 rader
Skriven 2005-02-18 18:28:22 av mark lewis (1:3634/12)
    Kommentar till text 107 av Maurice Kinal (1:153/401.1)
Ärende: MSGID
=============
shortened comments due to time...

 ml>> many tossers are limited by slightly less than 64k message
 ml>> size because they pull the entire message into memory and
 ml>> work with it there rather than spooling excess stuff to
 ml>> temp disk storage to get to the end of the message and the
 ml>> needed tossing data stored there...

 MK> I don't think that is a real problem to compensate for, and
 MK> might actually be a good thing <patent pending>.

actually, it is a problem but the limit on short messages has been "enforced"
for so long and by so much software (ie: user's offline QWK readers and such)
that it is very commonplace...

 MK> If pkt sizes (number of messages) aren't a limitation then
 MK> it makes it a tad easier to come up with a meaningful MSGID
 MK> but I still think a larger number of characters would make
 MK> all the difference as far as processing time goes.

why is processing time such a factor in your argument(s)??

 MK> As far as any machine here is concerned seconds are static
 MK> over more then just a few messages, especially with a byte
 MK> limitation as to their size, so having at least a couple
 MK> extra bytes to play with and keeping a meaningful hex
 MK> date/time stamp isn't such a bad idea methinks.  Personally
 MK> I prefer a year, day_of_the_year,hour,min,sec base hex
 MK> field that can be incremented by message number and a 10
 MK> character field happensto work out quite nicely.  It could
 MK> go the distance.  8 characters is also doable but then
 MK> increases the amount of steps to keep any meaningful
 MK> date/time information as well as increases the limitation
 MK> as far as how many years one can keep it unique.

isn't the time/date info in the message sufficient? at least the time info? ;)

 MK> I don't know why the three year limit especially with no
 MK> rhyme or reason behind how we all should be generating it.

ummm... i don't know if you are aware of it or not but some of those who
developed our fidonet technology also had a hand in developing the internet
technology many use today... remember, fidonet was developed for "those who
want to experiment with packet switching technology" so sayeth the history
files and quotes from the creator of fidonet ;)

 ml>> that's an old DOS-think limitation...

 MK> Only DOS-think for those who believed we all should be limited
 MK> by DOS-think.  I think it can be overcome and perhaps even be
 MK> exploited to all of our advantage.  Chances are that anyone
 MK> using software that old wouldn't even notice any alterations
 MK> at all.  They probably couldn't take advantage of any change
 MK> but it probably wouldn't hurt them either.  On the other hand
 MK> doing nothing might hurt everybody, including them, whoever
 MK> them are.

let's not forget, either, that many fidonet systems are still run on DOS based
systems ;)

)\/(ark

 * Origin: (1:3634/12)