Subj : Re: testing
To   : mark lewis
From : Ozz Nixon
Date : Mon Apr 29 2019 09:22:51

On 2019-04-23 10:47:10 +0000, mark lewis -> Wilfred van Velzen said:

>  On 2019 Apr 23 14:37:02, you wrote to Ozz Nixon:
> 
>  ON>> So 35 1/2 seconds to process 9,100+ msgs ... good/bad/descent?
> 
>  WV> That's hard to judge. There are so many factors of influence here.
> 
> that's true... also, "are there any shortcuts being taken" comes to mind...
> eg: zeroing seconds in the time stamps...

I understand shortcuts, none, not even buffering techniques... it is 
raw read/parse/write to JAM and the log. The log is probably the most 
sophisticated part of the whole tosser, deciding where, what, when to 
log... so it is useful, without being a dump of the PKT itself.

I was just cursious if I should spend time to make it faster - 
prefetching PKT bytes, buffering the log and JAM files. But, I like it 
having as small of a foot print, and the fact it is interpreted instead 
of native binary.

Thanks for feedback(s)...
O.

--- ExchangeBBS NNTP Server v3.1/Linux64
 * Origin:  (1:1/123)

.