NODE 2fbcd119Re: PGP
"David E. Smith" <dsmith@midwest.net>Wed, 6 Dec 95 11:03:26 PST
-----BEGIN PGP SIGNED MESSAGE-----
At 12:54 PM 12/6/95 -0500, you wrote:
> I've tried to find out the answers to my
>questions myself but I haven't been very successful.
>Actually, I haven't been successful at all. First of all,
>I've been using PGP on my computer but when I tried to port it
>over to my unix account it didn't work. I took the source code
>and tried to compile it using the unix g++ compiler but it gave
>me all of these errors. Where can I find PGP that is already
>compiled for unix or at the very least will compile for unix?
>Secondly, if I create a program that makes using PGP more
>user friendly can I let other people have it ( for FREE )?
>Lastly, what is considered overkill with cryptography?
>I don't believe anything is. My programs accept what the user
>inputs for how large prime numbers should be ( my public
>key programs ) but I make them able to accept values that go
>up to 30720 bits ( I don't believe a number can ever be large
>enough ). You can never be to safe...never.
Firstly - and don't take this personally - how much computer
knowledge do you have? The PGP sources use all sorts of
#ifdefs and other kludges. A fair amount of C coding
ability is needed to get the damn things to do much of
anything. (Hell, I can't read most of it - my precompiled
MSDOS version does what I need it to do.)
If you make PGP more user friendly... well, PGP 3.0 is still
coming Real Soon Now (TM) and it will include an API that
will make hooking into it ridiculously simple. Beyond that,
there are already a number of good DOS and Windows shells
for it, and nobody on unix-flavoured systems expects a clean
user interface anyway :) (Well, except for XWindows...)
But, if you write it - and I'm cautioning you to be sure
you don't reinvent the wheel - you can distribute it freely.
I suggest using the terms of the GNU GPL.
And the key values - well, you can give it 32k bits, but 1.
generating a key pair could take a prohibitively long time;
2. nobody else can use it (the default PGP distributions
are capped off at 2047 bits).
Good luck.
Dave
-----BEGIN PGP SIGNATURE-----
Version: 2.6.2
iQCVAwUBMMXoSwwyfvCScyE5AQEwtQQA0IHRhQiBej05B8KfFMJh95+XVvzW5ax1
SgQGrNABIdnyOkDSgPLA97vQ04agU8ytOyaDMW/X4IuI/zZgsqOzegsb57+TEcAz
sB7mvzmW0tTrEEdMRjBkaZRAnf0wTrf2EqtR3lshJCgzK1MB6szG3w4N8yb26YKD
2VyBrX10nRY=
=sRw4
-----END PGP SIGNATURE-----
----- David E. Smith, c/o Southeast Missouri State University
1210 Towers South, Cape Girardeau MO USA 63701-4745
+1(573)339-3814, "dsmith@midwest.net", PGP ID 0x92732139
http://www.midwest.net/scribers/dsmith/
NODE 338bf0c4Re: PGP
DANIEL CHARPENTIER <drcharpe@hamlet.uncg.edu>Wed, 6 Dec 95 11:26:46 PST
On Wed, 6 Dec 1995, David E. Smith wrote:
> Firstly - and don't take this personally - how much computer
> knowledge do you have? The PGP sources use all sorts of
> #ifdefs and other kludges. A fair amount of C coding
> ability is needed to get the damn things to do much of
> anything. (Hell, I can't read most of it - my precompiled
> MSDOS version does what I need it to do.)
Sorry...I was always taught that the user should not
be mandated to finish the job of the programmer. Please,
do not take me wrong, PGP is a remarkable tool...but I
was not aware I had iron out the bugs. That part is the
easiest to do compared with the genius of the program...
but I still should not have to do it.
> If you make PGP more user friendly... well, PGP 3.0 is still
> coming Real Soon Now (TM) and it will include an API that
> will make hooking into it ridiculously simple. Beyond that,
> there are already a number of good DOS and Windows shells
> for it, and nobody on unix-flavoured systems expects a clean
> user interface anyway :) (Well, except for XWindows...)
Well maybe they ( people on "unix-flavoured systems" ) should
expect a clean interface. If the interface were more sound then
maybe the common man ( person ) wouldn't be so scared of it. I have
a lot of friends that gave up on cryptography because they could not
get the darned programs to work ( at least the good ones anyway ).
> And the key values - well, you can give it 32k bits, but 1.
> generating a key pair could take a prohibitively long time;
> 2. nobody else can use it (the default PGP distributions
> are capped off at 2047 bits).
I was not using 32k bit keys with PGP...but I'll look at the
code. To tell you the truth I never thought about modifing PGP
to generate that large of a key. When I finish it I'll post the
source here. I'm sure someone here would like to at least have
the option.
Thank you for the input.
NODE fbcf431eRe: PGP
Derek Atkins <warlord@ATHENA.MIT.EDU>Wed, 6 Dec 95 11:42:08 PST
The PGP sources build on most UNIX platforms. Have you read the
setup.doc? Have you followed the instructions? What Unix platform
are you using? In general, all you need to do is:
1) build rsaref:
cd rsaref/install/unix; make
2) build pgp:
cd src; make <platform>
Modulo a few quirks, that should be all. Look at the PGP FAQ,
Buglist, Fixes, and Improvements Page for a list of known bugs+fixes
in PGP 2.6.2:
http://www.mit.edu:8001/people/warlord/pgp-faq.html
If you need more help, email me offline (i.e., do not cc: cypherpunks)
-derek