Below are projects that I have my eye on, ones I plan on following through on in my free time. If you have any comments, or are interested in working on one of them with or without me, or just giving me some ideas, let me know. Also let me know if you know of such a project already in the works, started by someone else; I hate to duplicate effort.
All of my projects are hosted by SourceForge. My developer account at SourceForgemay be of interest.
Note that I don't note any of my already-started software projects here; go to SourceForgefor information on them.
Without a doubt, we are going through a societal revolution; I
feel that this is an important time to help make sure that opinions about
changes that should be spoken outloud, for the public to hear. This is an
opportunity to help society change for the better. In short,
the time to act is now.
No matter which viewpoint on an issue one
takes, it is extremely important that one voice that advocacy now, to help
society make informed decisions in a confusing time. My realm of debate
lays in the concepts of free information and
Open Source. Many
important
articleshave been written by those affiliated with the
Free Software Foundation, an organization
with which I often side in debates. I am trying to do my part in advocacy
for those issues I feel most strongly about both through writing software
and
articles.
FreeBSD's syslogd isn't very good at having "mirroring" logging; that is, having the errors on two machines show up on the other. This is because it doesn't act (barely) any differently on forwarded messages (ones it received from a remote host than ones it generates. Linux's syslogd by default doesn't forward forwarded messages, and this can be turned off with the -h option. So, I intend to make a -H option (the opposite of -h) that turns this behavior on for FreeBSD's syslogd.
Conclave is a simple encrypted IRC protocol I'm working on. An overviewis available for reading. So far, I have pretty much finished the central-bot (written in Perl) mentioned in the overview, and am currently working on the client end. Right now, the client end is worked as a ircII client together with a Perl script.
It might be just me, but it seems one big problem with logging and detecting unusual and unauthorized system happenings is that either one ends up logging too little or too much. What is needed is statistic-based system analyzing to prevent DoS attacks, and detect intrusions. The system would be modular, with the base knowing how to detect various spikes or abnormalities in system progress. Then, the system could act as needed. This idea is still "way-out-there".
This one is way out there, and I've only really banged it around in my head a little. What I'm envisioning is a way to define a language for describing the security of networks of objects, and along with this, a way to graphically display this. Think of how you, as a programmer picture arrays or trees; I'd like to be able to do the same with objects which have security attributes.
I feel there is a great need for this sort of project given the hodge-podge or black-magic manner in which security analysis nowadays is performed. Having a well-defined approach, especially one which could be represented visually would be very exciting. If this project peaks your curiosity, I would especially love to hear from you!