Subj : follow-up on: NCs who've updated to MakeNl_NG
To   : Marc Lewis
From : mark lewis
Date : Tue Jan 22 2013 14:16:11

 MarcL> Important to note here - it's NOT the file date that MakeNL is 
 MarcL> looking at, it's the julian extension which it apparently 
 MarcL> calculates on the fly based on the system date/time as to which 
 MarcL> extensions are acceptable.

 ml> ahh! that makes a difference, then... my configs do not use julian
 ml> dated extensions... i know about them and could use them but for
 ml> security reasons, do not... this also prevents the creation of
 ml> netseg diffs but that's never been a problem...

 MarcL> The julian date extension is the default for ALL diffs and lists. 

right... i thought i had noted that i use full filenames, including extensions, 
and that's why i apparently haven't seen this in my testing...

 MarcL> It, I don't think, would not show up anywhere in your normal 
 MarcL> config file - it's how the segments are normally transmitted 
 MarcL> after whichever program has produced them. E.g., Net9000.018 or
 MarcL> Net9000.242.  It's the intended distribution date; that's what
 MarcL> MakeNL is looking at in the Master directory and calculating the
 MarcL> file's relative age from that extension.

unless you use a specific extension (that your upstream *C knows about and 
configures for)... of course, that also means that historic files are not 
available as each new rendition of the specific filename would have to 
overwrite previous the one when it is hatched out to others in the region or 
net if one does that for those in their coverage area...

i had thought that it was looking at the file's date(s) instead of the 
extension... i can see where the extension in the julian day serialized 
methodology is ok but we also know that there can be problems... but the same 
is true with looking at the dates, as well...

i can also understand the software automatically saving the last one, two or 
three submittal files as noted in the history file and documentation but i have 
to wonder if something like that is better left up to the operator to do... 
back in the day, there weren't as many tools available as there are today but 
then again, having the software manage it is also easier and puts the onus on 
the coder(s) to get it right ;)

)\/(ark

--- 
 * Origin:  (1:3634/12.42)

.