[Mailman-Users] HTDIG patches! Arghh!

[Adam Lipson]

Thanks for the reply.

I tried to install from the source and somehow just kept on screwing up the install so the RPM worked and I used it.  Redoing it now would be a real pain.  I can definitely appreciate that installing from source is the best way to install it just was not working for me.  If you have a great suggestion as to how to reinstall this easily I am all ears, but I guess otherwise will wait for the next rpm to upgrade and suffer thru that upgrade process. 

as per the bcc I guess it is just a quirk of my install.  I had to approve the post, but then the footers did not appear.  Oh well, there are worse things in life.

Adam

-----Original Message-----
From: Richard Barrett [mailto:r.barrett at openinfo.co.uk]
Sent: Tuesday, June 03, 2003 11:29 AM
To: Adam Lipson
Cc: mailman-users at python.org
Subject: RE: [Mailman-Users] HTDIG patches! Arghh!


At 14:52 03/06/2003, Adam Lipson wrote:
>I do have to say thank you for writing and maintaining the patches as I am 
>trying to figure out how to integrate Mailman with HTdig as we speak.  The 
>problem I have run into is that I can't find documentation on how install 
>the patch with the RPM version of mailman.  I have searched thru the 
>archives on gmane.org and google groups to figure out how to use them with 
>the RPM and unfortunately can't find an answer.  Can someone point me to a 
>URL for this?

Assuming MM 2.1.2 in what follows.

It is nearly but not quite possible to construct a version of the #444884 
Mailman-htdig patch which can be applied to Mailman after it has been 
installed rather than applying the patch to the MM build directory before 
configuring and installing MM.

There are some minor bits of untidyness which would require hand fettling 
(path to python in the #! line of some added scripts, a directory to create 
and such) but nothing too difficult.

However, there is one difficulty which I cannot see a simple way around; 
the Mailman-htdig patch installs two additional cgi scripts, htdig.py and 
mmsearch.py, and I cannot see any simple way of generating the compiled C 
wrappers for these cgi scripts unless this is done when building MM from 
source.

While I am generally quite happy to use RPMs I find that installing MM from 
source is so easy that I always do that for preference. I certainly will 
not stick solely to RPMs if that gets in the way of achieving the 
objective. Can I tempt you to trying this approach?

>Also, another topic, but when one BCC's the list why don't the footers 
>appear on the bottom of the email? I am against BCCing a list, but someone 
>did and it appeared as if the footer was not there.  Is this a known issue 
>or is something in my config wrong?

I cannot reproduce this problem. If I Bcc one of my test lists and then 
approve the post (because the test list is not in the To or Cc header) the 
posting is distributed normally.

>Regards,
>Adam

------------------------------------------------------------------------------
Richard Barrett                                      http://www.openinfo.co.uk