Re: On Demand Nmap Build Testing

[Michael Pattrick <mpattrick () rhinovirus org>]

It's a good idea, especially with how practically any modern computer
can handle many simultaneous VMs with ease. But I would suggest
integrating tests into the mix; bugs that break builds are easily
identified. More useful would be if we could detect when a change
makes OS detection less accurate, or if FIN scan stops working on
NetBSD.

Maybe something for next SoC...

-M

On Tue, Jun 1, 2010 at 9:50 AM, jrf <jay.fink () gmail com> wrote:
> All,
>
> I thought I would share this with everyone in case anyone else was
> interested in doing something similar. Last week I setup some very
> simple scripts to do the following with virtual machines:
>
> start a screen session and logit
> power on a vm
> ssh to a vm and execute a script which then does:
>  svn up of nmap
>  rm -rf previous build (ran into problems running clean)
>  cp -R nmap nmap-build
>  cd nmap-build
>  configure && make
> power off vm
> mv the screenlog
>
> Then I repeat the process for the next vm. Right now I am only doing
> two, NetBSD-current (as of 2 weeks ago) and FreeBSD-current. I do
> periodic updates and builds manually on my physical debian and mac
> systems anyhow.
>
> At this point I am just about ready to automate building on NetBSD,
> FreeBSD and Linux (the mac not so much as I use it for work :)
>
> Anyhow, just something I thought everyone might be interested in. So
> far the only *problem* I have come across is the subshell on NetBSD is
> kind of a pain so I switched the build user to bash from pkgsrc.
>
> Regards,
>  j
> _______________________________________________
> Sent through the nmap-dev mailing list
> http://cgi.insecure.org/mailman/listinfo/nmap-dev
> Archived at http://seclists.org/nmap-dev/
_______________________________________________
Sent through the nmap-dev mailing list
http://cgi.insecure.org/mailman/listinfo/nmap-dev
Archived at http://seclists.org/nmap-dev/