[rtg] Traffic is wrapping

[bill fumerola]

On Tue, Apr 08, 2008 at 10:38:41AM -0600, Leech, Jonathan wrote:
> My first thought was to improve and contribute to the JRTG codebase, but
> with no disrepect to the original developers and design, that just
> wasn't feasible. 

i understand that position.

> Second, I would like to point out that in the example from my code there
> is no Java or OOP, other than Java String comparisons. And that
> particular example is of features I made for JRTG that RTG just doesn't
> have, the ability to change its configuration on the fly, and to send
> the polling results to multiple delegates.

i don't want to get into a language war[1], but the line i quoted is
enough logic in one line to hurt my brain.

it's funny that for all the OOP and C++-like features java provides,
it's still just string comparisons to determine if it's time to instantiate
something new.

defining multiple database servers was one thing i was adding to the
lex/yacc code, first as a fallback, later to point hosts or OIDs

> I don't have any experience with Ruby, as I know it best, my colleagues
> now it best, there are a tremendous amount of libraries written for it
> (including SNMP4J which is the basis for JRTG), and is the most stable
> platform around for writing server-side code. Java's concurrent
> threading packages make writing multithreaded programs like JRTG a
> breeze.

java may make writing programs like that a breeze, but reading them
sure is a different story.


but as i said, i don't know java. if i did, i might have done my near
rewrite in java.  if php had anything close to real thread support, i'd
have considered rewriting the whole thing in php. it doesn't so i didn't.

i didn't know ruby at the time (and still am just taking baby steps). i
merely was musing that ruby has a nice balance of OOP, thread
primitives/harvesting/etc, and ease of entry over C that could help
attract more developers to it.

i guess the point i was trying to make is that, code-wise, i can see the
original rtg code in JRTG... it just happens to be surrounded by 3x-5x
times more code to support the magic that all the java classes, DB
agnostic layer, thread libraries, etc. that's neither here nor there.
it's well documented in javadoc. it's a quick read to get through all
the com.virtella.poller package. i'm not knocking it.

do you have any intention on updating/maintaining/supporting the code
as a public project or is what's on the sourceforge site just a onetime
import?

-- bill

1. http://www.joelonsoftware.com/articles/ThePerilsofJavaSchools.html