--- "John P. Rouillard" <[EMAIL PROTECTED]> wrote:
>
> In message <[EMAIL PROTECTED]>,
> Eric Smith writes:
>
> >Risto Vaarandi wrote:
> >> David Vasil wrote:
> >>> Risto Vaarandi wrote:
> >>>> In order to overcome this ambiguity, I've been
> thinking about adding
> >>>> event2 action to SEC action set that would have
> the following syntax:
> >>>>
> >>>> event2 <time> [<event text>]
> >>>>
> >>>> 'event2' would be identical to 'event', except
> <time> parameter is
> >>>> mandatory and may contain other symbols than
> numerals (0-9) only.
> >>>>
> >>>> Note that there is also another option - drop
> the idea of having 'event2
> >>>> <variable_time> <event_text>', and use
> >>>>
> >>>> create <context> <variable_time> (event
> <event_text>)
> >>>>
> >>>> instead. In other words, when variable
> lifetimes are allowed for
> >>>> contexts, 'event2' can be actually expressed
> through 'create'.
> >>>>
> >>>> Any ideas?
> >>> I like the idea of an additional event, but the
> name itself 'event2' may
> >>> confuse some. Maybe something like 'tevent' for
> 'timed event' would
> >>> work.
> >>
> >> I had the same doubts about the naming -- adding
> a number doesn't make
> >> the name descriptive enough) --, but 'tevent'
> looks a nice choice ;)
> >
> >Why make it cryptic at all? Why not "timedevent",
> or "timed_event"?
> >"tevent" doesn't have much intrinsic meaning, to
> me. It's not like
> >there's a cost to making it readable.
>
> Risto,
>
> Why use positional arguments at all. I think the
> change between SEC
> V1's positional arguments to v2 labeled arguments
> was a major win.
> Keep event, but add a label --timeout.
>
> event --timeout %variable stuff
>
> This would work for all actions I think since a -
> sign (except for
> write) isn't a valid initial character. It can be
> optional if the
> event matches the form:
>
> event [number] stuff
>
> but allows for other future expansion of the
> functionality of all
> keywords should it be needed.
>
> -- rouilj
> John Rouillard
That's an interesting idea. I haven't actually thought
much about using key-value representation for action
lists as well - probably because most actions just
take 1 or 2 parameters, and thus the problem has not
been very urgent so far. It's true the idea is similar
to the shift from logsurfer-like positional rule
definition to a key-value based definition in sec v2 -
it certainly eases the addition of new parameters to
existing actions in the future. I am not sure whether
this idea will make it into the 2.4.2 minor version,
but it is clearly something to contemplate for the
next major version.
br,
risto
____________________________________________________________________________________
Never miss a thing. Make Yahoo your home page.
http://www.yahoo.com/r/hs
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2008.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Simple-evcorr-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/simple-evcorr-users
