[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [atlarge-discuss] Election Management
Dear Stephen, Abel,
I just proposed to carry some work while I have no time for that. If
someone is ready to do the job better and has the time an d dusto and want
to join the team of watchdog, just fine with me. I gave an aidea of what I
can do. Abel has said he has someone. May be us tree can work out what
Bruce expects from us; The only thing I know is that my old "postie"
freeware file works just fine to send thousands mails and I can easily
customize them in 10 minutes C programing.
Yes randomizing is no problem. You can even hash an MD5 thing. This really
depends on the tool you have which makes the code simple to whatchdog to
peruse. In C MD5 is a source library. Goes fast to link and comile and run.
But complex to check.
jfc
On 21:24 08/04/03, Stephen Waters said:
On Tue, 2003-04-08 at 13:18, J-F C. (Jefsey) Morfin wrote:
>
> The point is that a grep of the mailbox on "^@" gives the anonymous
votes file.
> Then the vote is in position 15 or 16, so very unlikly to be folded.
> It is obvious to remove spaces and to add xxxxnn00000 in 15th position.
> where 00000 is incremented.
> Then to sort by 15th position and to remove the duplicated votes.
> The to check the xxxxyyyy to make sure there is no intruders.
> then you sort by the 11th and you get the results.
> Can be done by anyone with any tool.
>
> Also, if we want to get a clear ballot, we need something not too heavily
> loaded with numbers.
>
> Ex.
> Beware: please do not change the "@--#--" sequence, you would waste you
vote.
>
> Please indicate who you vote for:
> @0123jklz#12 [ ] Doug Engelbart
> @0123jklz#13 [ ] Bill Gates
> @0123jklz#14 [ ] Jeff Williams
> @0123jklz#15 [ ] Jon Postel
>
> Do you agree for them to be given two cents
> @0123jklz#16 [ ] Yes
> @0123jklz#17 [ ] No
>
> Abel, Stephen and others, is there a flaw?
> I propose to test the sending of the mail in sending a personal
> letter to all the member to infrom them of the election. This would
> be a real life test of the mail generation. jfc
1) It scales linearly. That's fine for this election, but I'd like to
see a more robust solution in the future.
2) What happens if the line *is* folded? My Perl regexp is much better
than my grep regexp, so I'm not sure if command line grep can handle it.
What is the watchdog procedure for handling a chopped line?
3) Perhaps you should use a config file for the actual obfuscation
digits and unique nr. Or generate them randomly at run-time and write
out the digits to a file. That way you can release the source code
without fear.
4) Are you gonna parse the mail through MIME in case you get those pesky
"= 20" characters and such? I mean, obviously, if it's too weird the
watchdogs will have to manually check. How hard is manually checking?
5) I'm concerned that: swaters@no.info and mwaters@no.info don't have
enough entropy to make the ballot secret. I.e., let's say that since
these folks are on the same mail host with a broken MTA, they both get
their lines folded. Then, when the watchdog or checker looks at the
votes, it could very well be:
swaters@no.info: @5432pzat
mwaters@no.info: @2345pzat
jwaters@no.info: @1122pzat
fwaters@no.info: @3311pzat
Does this cause an alert for watchdogs? Will the watchdogs now have an
idea of who voted how (or is that already assumed? I forget.)?
Detail-oriented,
-s
---
Incoming mail is certified Virus Free.
Checked by AVG anti-virus system (http://www.grisoft.com).
Version: 6.0.463 / Virus Database: 262 - Release Date: 17/03/03
---------------------------------------------------------------------
To unsubscribe, e-mail: atlarge-discuss-unsubscribe@lists.fitug.de
For additional commands, e-mail: atlarge-discuss-help@lists.fitug.de