Work at SourceForge, help us to make it a better place! We have an immediate need for a Support Technician in our San Francisco or Denver office.

Close

#16 don't auto-block remailers addresses

closed-accepted
nobody
None
5
2002-12-14
2002-10-21
Trek Star
No

This patch is unfinished :(

to have a good autoblock, without changing too much
(without the confirmation feature), mixmaster in
rem.c:blockrequest() should create the exact regex of
the string (if pcre enabled), then check if this regex
can block some remailer and if not continue with the
normal process

the patch attached here don't create the regex and has
a bug:
i must use an additional buffer (copy_addr) in doblock
and not user directly the addr buffer because something
in the memory is broken!! (probably I don't understand
correctly the buffers library)

this patch was created on mixmaster 2.9b38
please help me to find where is(are) the error(s) :)

Discussion

  • Trek Star
    Trek Star
    2002-10-21

    patch to don't auto-block remailers on mixmaster 2.9b38

     
  • Logged In: YES
    user_id=1525

    The author of the patch reported some problems with it on
    IRC. waiting for further information. It looks ok otherwhise.

     
  • Logged In: YES
    user_id=1525

    it also doesn't free the buffers..

     
  • Trek Star
    Trek Star
    2002-11-26

    final patch to fix 621939# bug on mixmaster 2.9b40

     
  • Trek Star
    Trek Star
    2002-11-26

    Logged In: YES
    user_id=365839

    This patch fix the 621939# bug checking the remailer address
    list
    before to block a an email address.
    We should create the correct regex before to add it to
    dest.blk too.
    We should also free the buffers in blockrequest().

    Finally I've found a little bug in chain1.c:t1_rlist() at
    line 48:
    'n' should be initialized to the number of remailers already
    listed in
    'remailer' array, because the 'REMAILER' structure and the
    function
    support it.

     
  • Logged In: YES
    user_id=1525

    I don't get the comment with the n in chain1.c but I fixed
    the issue at hand. Thanks for your wori.

     
    • status: open --> closed-accepted