Text 11884, 56 rader
Skriven 2008-10-12 02:26:20 av James Bradley (1:342/77.0)
     Kommentar till en text av Maurice Kinal
Ärende: sure
============
10-04-08 20:22: Maurice Kinal to James Bradley, "sure"

 MK> I am betting you already know the above and thus the above
 MK> is just idle chatter.

Something like that. <L>

 > I thought ASCII, as used to parlay our language,
 > is highly successful in compression ratios?

 MK> If the file is large, yes.  Text messaging?  With short text
 MK> messages the compressed version will often be larger than the

"Larger..." Imagine. I forgot, about how the quantity of words improves the
proportion, and I recall that compressing a compressed file often turns up
larger, but the above is news for my data-bank. Still, with the economy of
scale, forty, four-line messages should compress as well as four, forty-line
messages, no?


 MK> uncompressed original. One would have to combine many text
 MK> messages to an outbound file with a suitable delimiter to
 MK> differentiate individual messages to exploit on the fly
 MK> compression if the goal is to reduce bandwidth.  From my

A moot point now that everyone but a bushwacker, has access to highband. About
OLMR packets, I yearn for a *larger* overhead for the added functionality that
is now missing. Binary, or ASCII, they were, I assumed, files having no
"message" information but how to manipulate the mail door via uploaded mail
packages. As a sysop, I preferred using the mail door locally over Golded. If I
was calling in for mail at that time, OLMRing was still a no-brainer, plus
f-attaches and other stuff were in the mix.


 MK> observations, there was never that much traffic that on the fly
 MK> compression of text files was worth all the effort, Fidoally
 MK> speaking. However, taking a regular pkzipped echomail file and
 MK> stripping out everything that isn't ASCII produces greater
 MK> byte savings uncompressed than the original pkzipped
 MK> echomail file, AND all the important info is intact in the
 MK> end result.  Ergo if you really are concerned about Fido
 MK> bandwidth, which I noted you really aren't for very good
 MK> reasons, then my original statement about pure ASCII
 MK> transfers having a greater impact on reducing bandwidth
 MK> stands.

I'll just nod in agreement, because I haven't the foggiest what the hell you
were talking about in the first place. <L>



... James
___ MultiMail/Linux v0.49

--- Maximus 3.01
 * Origin: -=-= Calgary Organization CDN (403) 242-3221 (1:342/77)