--- Log opened Mon Jan 28 00:00:17 2013
00:38 < mcallan> conseo: one suggestion: it took a long time for me to understand roughly what the page was about.  consider putting a single sentence atop the page and each major section: "This page/section is about BLAH" or "proposes BLAH".  don't expose the reader to details without context, or you immediately lose them
00:47 < mcallan> i think you are concerned about the problem of count "liveness", or up-to-datedness. you present a solution (fast parallel) in 2nd section.  but before looking at solution details, shouldn't we first discuss the problem?  what exactly is unacceptable that we are trying to correct?
00:54 < mcallan> you are always free to code a new count engine, ofc.  and others, too.  we must add general supports for that, especially for external engines for special purposes - any one of which might later be absorbed internally.  but in order to comment on design details of special engine, i must first understand its special purpose in terms of the problem it solves
01:09 < mcallan> that said, i'm also interested in how your solution would compare to the current method of updating the results via the couplet CountNodeW.uncast+cast, if we were to use *that* as a solution
03:37 < conseo> mcallan: ok, thx for the feedback
03:49 < mcallan> maybe we skype about problem (after exams), and open thread if you like
04:04 < conseo> maybe yes. i am most concerned about scalability, liveness is an outcome of it, i think
04:04 < conseo> i will focus the article on that
04:07 < conseo> i was concerned about having a flexible tree-building mechanism, which doesn't byte us later on and still allows control of counting mechanisms and resource flow by each candidate (node) and more elobarte counting mechanisms without any locking on the structure
04:07 < conseo> anyhow, you are worried about more important stuff atm. i just wanted to show you early
04:23 < mcallan> no i have time, lets talk about problem when u are free, e.g. after exams
04:39 < conseo> ok
--- Log closed Tue Jan 29 00:00:35 2013