[Author Prev][Author Next][Thread Prev][Thread Next][Author Index][Thread Index]
[minion-cvs] Move open specification issues into their own document
Update of /home/minion/cvsroot/src/minion
In directory moria.mit.edu:/tmp/cvs-serv8522/src/minion
Modified Files:
ACKS HACKING TODO
Log Message:
Move open specification issues into their own document
Index: ACKS
===================================================================
RCS file: /home/minion/cvsroot/src/minion/ACKS,v
retrieving revision 1.3
retrieving revision 1.4
diff -u -d -r1.3 -r1.4
--- ACKS 7 Jan 2003 01:41:20 -0000 1.3
+++ ACKS 14 Jul 2003 22:16:00 -0000 1.4
@@ -10,6 +10,9 @@
Robyn Wagner
Bryce "Zooko" Wilcox-O'Hearn
+ Special thanks go to Christopher Porter, whose donation of
+ a copy of Visual Studio 6 made a native Windows port possible.
+
Finaly, thanks to George Danezis and the whole Mixminion-Dev team
for their work on the Type III remailer design.
Index: HACKING
===================================================================
RCS file: /home/minion/cvsroot/src/minion/HACKING,v
retrieving revision 1.19
retrieving revision 1.20
diff -u -d -r1.19 -r1.20
--- HACKING 26 Jun 2003 03:23:53 -0000 1.19
+++ HACKING 14 Jul 2003 22:16:00 -0000 1.20
@@ -9,21 +9,30 @@
A working /dev/urandom (see PORTING NOTES below)
+NON-HACKING PROJECTS:
+
+- We need documentation.
+
+- We need a man page.
+
+- See the spec for open issues.
+
+- We need a better web page.
+
+- We need a logo. (An alien with an eggbeater or an alien at a DJ's mixer
+ are two ideas.)
+
THINGS TO HACK: (Good introductory projects that I won't get to myself for
at least a version or two.)
-- Last time I checked, the client could run okay on Cygwin. Somebody should
- do a full port.
+- Last time I checked, the client and server could run okay on Cygwin. More
+ people should try it out, and see what needs to be done.
-- We need a native (non-Cygwin) windows port! (A client-only port would be a
- good start.) Most of the code is in Python, and what C we have is pretty
- standard, so I don't anticipate too many problems. Some known issues in
- the source are marked with comments labeled WIN32. Other areas of
- potential difficulty are:
- - The build process. (I don't know the first thing about how to build
- Python modules under windows. Python distutils should work, but I
- don't know what compiler is appropriate. Studying the Python
- webpages may be useful.)
+- We need a native (non-Cygwin) windows port! (Right now, I can make the
+ client work, but the server needs to be tested, and we need an installer.)
+ good start.) Most of the code is in Python, and what C we have builds
+ fine. Some known issues in the source are marked with comments labeled
+ WWWW. Other areas of potential difficulty are:
- Binaries and installer. (We should provide two versions: one that
comes with a Python interpreter, and one that doesn't. They should
both be easy to generate.)
@@ -32,11 +41,6 @@
client code uses SIGALRM to timeout dead connections.]
- The 'daemonize' code in ServerMain.py is almost certainly not
right on windows.
- - We assume the existence of /dev/urandom or some other similar entropy
- source; these do not exist on windows. OpenSSL's RAND_* functions
- may be helpful, but coming up with good seed information may be
- tricky. Clients can use RAND_event, but servers will need to get
- entropy from somewhere else.
- There will be other, as-yet-unexpected issues. :)
(Difficulty: moderate. Invasiveness: moderate)
@@ -88,16 +92,9 @@
THINGS TO THINK ABOUT AND HACK: (Introductory projects that will take some
specification work. Please, get your spec discussed on mixminion-dev and
- checked into CVS *before* you submit an implementation for any of these!)
+ checked into CVS *before* you submit an implementation for any of these!
-- There should be a way for clients to set some (but not all) RFC822 email
- headers. Likely candidates are 'Subject' and 'Content-Type'. Maybe there
- should be *limited* support for setting 'From'. [For example, if the user
- wants to set 'From' to "xyzzy", the From line would appear as "From:
- "Anonymous user claiming to be xyzzy" <nobody@your-remailer>". Maybe.]
- Whatever you choose, be sure to justify it. (Spec difficulty: easy, but
- you're likely to start a holy war and end up pleasing nobody.
- Implementation difficulty: easy. Invasiveness: slight.)
+ [See spec-issues.txt for more open issues].)
- Maybe, there should be support for multiple exit addresses (cc, bcc, etc.)
This is prone to abuse... (Spec difficulty: easy, once you've figured out
@@ -111,8 +108,8 @@
- Want a real challenge? We have an allusive description for how to do
K-of-N fragmentation in our E2E-spec documentation. Go flesh out the
- description and implement it. (Spec difficulty: moderate. Implementation
- difficulty: hard. Invasiveness: some.)
+ description. (Spec difficulty: moderate. Implementation difficulty: hard.
+ Invasiveness: some.)
- Right now, we never generate link padding or dummy messages. The code is
there, but it never gets triggered. Specify when it gets triggered (and
@@ -125,21 +122,6 @@
anonymity implications. If you can come up with an algorithm that
doesn't, the code should be pretty easy to do. (Spec difficulty: ????.
Implementation difficulty: easy. Invasiveness: some.)
-
-NON-HACKING PROJECTS:
-
-- We need documentation.
-
-- We need a man page.
-
-- See the spec for open issues.
-
-- We need a better web page.
-
-- We need a logo. (An alien with an eggbeater or an alien at a DJ's mixer
- are two ideas.)
-
-- We need a bugtracker.
HARD HACKING:
Index: TODO
===================================================================
RCS file: /home/minion/cvsroot/src/minion/TODO,v
retrieving revision 1.133
retrieving revision 1.134
diff -u -d -r1.133 -r1.134
--- TODO 13 Jul 2003 03:45:33 -0000 1.133
+++ TODO 14 Jul 2003 22:16:00 -0000 1.134
@@ -240,7 +240,15 @@
- Servers should use downloaded directories to print useful
nicknames for other servers rather than just IP addresses.
- Port to Twisted, if reasonable (see HACKING)
-
+ - MMTP / async
+ - Tests for all cases:
+ - Packet to server with bogus IP
+ - Junk
+ - Retry on bogus close.
+ - Multiple senders
+ - Bad senders
+ - Bad recipients
+ - Hunt down leaks
Features for "1.0" (some research/specification required):
- Nymservers
@@ -265,14 +273,6 @@
- Real PKI for end-to-end encryption
- MMTP / async
- Make MMTP bursty
- - Tests for all cases:
- - Packet to server with bogus IP
- - Junk
- - Retry on bogus close.
- - Multiple senders
- - Bad senders
- - Bad recipients
- - Hunt down leaks
- Directories
- Support for full-blown multiple-server agreement mechanism
- IPv6 support (must solve non-clique problem)