Fails to build / Any development? / Licences

Francesco Ferreri f.ferreri at namex.it
Sun Nov 6 10:52:01 UTC 2005


Hi all,
I would like to state my very humble opinion, as a software engineer 
more than a network specialist; at first sight re-coding in Perl seems 
to be an attractive solution: Perl is highly portable and a lot of 
useful modules and packages is currently available from CPAN archives.

But I think it's worth considering that IRRToolSet is quite an ambitious 
software project, with a rich set of features and tasks which may 
involve complex computations. So a possible re-implementation of the 
whole system is not merely a fact of choosing a different programming 
language but, most of all, it means spending a significant  effort in 
terms of preliminary analysis and design (which are, indeed, the main 
part of the game).
 From this viewpoint, it would be interesting to know what is the 
analysis and design work and documentation  which lies behind the 
current implementation, apart from the C++ code itself: maybe it could 
be positively reused or it may suggest alternative solutions to a full 
re-implementation.

Given that, I think that the availability of a robust, correct and 
reliable set of tools such as this is very important to promote and 
improve the use of routing registries. Maybe it's worth to spend some 
more effort on it!

Best regards,
Francesco Ferreri



James A. T. Rice wrote:
> Hi Folks,
> 
> I've delved into the pride-tools / RAToolSet / IRRToolSet every so often 
> over the years, with the usual result of them failing to build cleanly on 
> my systems despite some trying (I'm not a c++ coder, I don't pretend to 
> be), at which point I usually give up and wait a few more years to try 
> again. I'm at that point today.
> 
> This problem still exists. OpenBSD 3.6, OpenBSD 3.8, and FreeBSD 4.2 all 
> failing to compile IRRToolSet, despite having made many modifications to 
> the code to try to coax it into compiling - and this is under gcc 2.95.3, 
> 3.3.5, and 2.95.2.
> 
> RIPE NCC transferred maintainance of the code to ISC sometime last year, 
> but there doesn't seem to have been any code released since then, am I 
> looking in the wrong place? Do we have the ability to ask ISC or RIPE to 
> spend some time on IRRToolSet making it build cleanly across multiple 
> platforms?
> 
> We still evidentally have troubles with the non commercial use licence of 
> the tools. A licence which exists from the RAToolSet ISI days, and 
> probably can't be changed without a full reimplementation of the code.
> 
> Would it be worth finding / paying someone who could reimplement the whole 
> toolset in a cleaner manner under a licence compatible with peoples 
> business needs? Possibly in a language which might be more maintainable by 
> the community at large, and eaiser to implement new features in (eg 
> perl?).
> 
> The failure to do anything here just means one thing - people will 
> continue to not have the tools available which allow them to do sensible 
> policy enforcement on their peers / customers, and so fail to implement 
> such measures at all.
> 
> Thoughts, anyone?
> 
> Cheers
> James



More information about the irrtoolset mailing list