Text 5387, 105 rader
Skriven 2012-04-29 23:26:50 av Michiel van der Vlist (2:280/5555)
     Kommentar till en text av Maurice Kinal (1:261/38.9)
Ärende: blowing in the wind
===========================
Hello Maurice,

On Saturday April 28 2012 03:05, you wrote to me:

 MV>> You can't be serious! How long have you been in Fidonet?

 MK> Off and on for the past 20 years but at least 10 years doing the
 MK> pointy thing.

And all that time you did without ZIP?

 MK> I could call upon two nodelisted nodes that will back my statement up
 MK> about doing it raw (ie no zip).  As far as compression goes with
 MK> respect to transport I am in favour of it except have yet to
 MK> see a Fidonet node that can handle purely compressed pkt's.

I am not sure what you mean by "purely compressed packets"

 MK>   zip never, ever counted.

Like it or not, zip is the default compressor/archiver for fidonet. Fidonews is
distributed in zip format. The nodelist/nodelist diffs are distributed in zip
format. Some *Cs offer other compression formats as an extra, but zip is the
default.

 MK> If I want archiving then tar wins but I really don't see a need for
 MK> archiving msg's for transport and refuse to deploy DOS-think software
 MK> on my puritan Unixie boxes.

What's "DOS think" about zip?

 MK>   I *could* but I don't.  Another very good reason I don't have a
 MK> nodelisted ftn address.

Your choice.

 MV>> http://www.vlist.eu/pleaforutf8-1.txt
 MV>> http://www.vlist.eu/pleaforutf8-2.txt

 MK> Excellent!  Do you want to see them reformatted properly so that they
 MK> word wrap correctly dependent on display dimensions?

Thanks but no thanks. It was written as a Fidonews article and formatted
according to the Fidonews specs. Because you obviously missed it and could not
process the Fidonews archive, I make the original manuscript of the article
available a personal courtesy to you. Just so that you could read the content.
That's it. It has been published in Fidonews, it was not my intention to
republish it in modified form.

 MK>  Can you handle gzip compressed files directly?

No.

 MK>   Just gzipping them without reformat yeilds;

 MK> pleaforutf8-1.txt:       56.9% -- replaced with pleaforutf8-1.txt.gz
 MK> pleaforutf8-2.txt:       57.4% -- replaced with pleaforutf8-2.txt.gz

 MK> That dramatically cuts down the bandwidth.

So did the zipped version, but you could not handle that...

 MV>> That is not a matter for the FTSC.

 MK> Understood.  The utf-8 msg format could though.

Utf-8 is a character encoding scheme for unicode. It is nor a message format.
Two different things. Fidonet uses the message format as documented in FTS-001.
How to use UTF-8 in combination with this message format is documented in
FTS-5003.

 MK> As we speak it deploys a rfc 3339 datetime stamp which effectively
 MK> replaces the current obsoleted 2 digit year as well as makes the TZUTC
 MK> kludge redundant.  Also we have a pkt header in the works with
 MK> absolutely zero binary data.  So far it has proved itself as a
 MK> suitable transport pkt.

All very well, but come down to earth please. The type 2+ packet has been in
use in Fidonet for over two and a half decade. Almost all of the software that
can handle that is abandonware by now. There is just no way that all the
message processing software in Fidonet is going to be updated to accommodate a
new packet format. Forget it, it is no going to happen. Any attempt to force
the issue will kill Fidonet. For better or worse, we just have to live with
choices made in the past.

 MV>> UTF-8 and the Fidonet transport layer is covered in FTS-5003.

 MK> Check this reply's kludges.

I se nothing special in those kludges...

 MK> I think I am one of the few - as well as being first - to put it into
 MK> action.  I am WAY ahead of the game,

You run so fast, you left everyone behind and now you are on your own.

 MK> Nice article now let's see you put it into action.

I did my part of the action.


Cheers, Michiel

--- GoldED+/W32-MINGW 1.1.5-b20110320
 * Origin: http://www.vlist.org (2:280/5555)