[Lincs] February Meeting?

[J]

> | Just a quick question... where and when will the RL meeting in February
> | be held? Any ideas on who is going to talk and what is going to be
> | talked about.
>
> I'm not sure if I'll make it to the next meeting. I'll try my best, but
> I can't promise to give a talk at this point.

Will try. Feburary is a long way away.

>
> | I notice that the volume of mails on the list has dropped to pretty 
> much
> | zero over the last month or so, I must admit I have been rather busy,
> | what has everyone else been up to?
>
> I've been playing with Debian Sarge and SuSE 9.1.

I've been taking over... I mean trying to set up an employee friendly 
and open source friendly company.

> I also gave some talks on the following topics:
> - - .htaccess management
> - - security audits and penetration tests for companies
> - - php based content management systems

I have a question actually about something related to this which I'm 
trying to work out for various purposes which breaks down into two parts 
One about security, one about CVS (or related).

1) How can I encrypt a folder structure through the use of a key 
mechanism so that I can use it with normal programs through a mount 
style mechanism : for instance, I know I can create a file and mount 
this as a directory, and then place files into this. I also know i could 
encrypt this file, is there any way of creating some sort of script that 
would encrypt the data being sent to the file and un-encrypt the return 
of the data. What I'm worried about here is that I dont really 
understand the way this works - I know I just do it ...

for instance if I send the command "open file" through the encryptor, 
this might turn out the other end in the block file as something 
completly wierd. If all instructions get interpreted in this manner then 
we're going to have serious problems.

On the other hand, if I have to encrypt individual files, this still 
gives people the ability to see what I'm working on through the file 
names (I give my files obvious names - its a company policy).

I dont want to have to un-encrypt the entire directory then re-encrypt 
it everytime I want to open / write to a file.

I want this to work at the OS level because I want to be able to put 
anything into these folders including applications and use any generic 
file based application on this directory (for instance I don't want to 
have to be forced into using emacs to edit the files because you can do 
it inside emacs or whatever).

How would changing my keys work for such a system or allowing multiple 
keys to be used to access it?

2) Supposing the above was a CVS structure (which part of it will be). I 
need to have two CVS structures - one "public" with various people 
having read-write to it, and one private where I will be doing my own 
development. There is a reason why I cant develop onto the same tree and 
must keep it private. In this situation, I will checkout the current 
release every now and then, develop my mods then move them back to the 
main tree. At the moment I am doing it without a second CVs. What I want 
to do is be able to use a "private" cvs to do my development on then 
merget mine back with the main tree.

Is there an "official" way of doing this? I know I can write some 
scripts to checkout into a temp directory the "public" tree, generate 
patches against when I checked out from my "private" tree then apply 
these patches to the "public" checked out and commit the lot (with 
consideration to checking for additions or removal of files etc.

> On Thursday I'm going to give a talk about how to hide data in unix
> filesystems at my local Linux User Group. For example: How to hide data
> in your journal or in your inode struct. Basically it's much about C
> programming.

Im not going to ask about this.... (neither if its a good idea or not).