Topic: Need help adding to trac contrib to allow incoming email to trac

[jptechnical]

This is what I need to do, but I am just getting my feet wet with SME. I don't know what the variables should be for mta_user and trac_user. Basically, it should allow an incoming email to be parsed into a trac ticket. I figure using the fetchmail contrib would be the simplest way to get the mail into trac.

Here is the url and the install instructions:

https://subtrac.sara.nl/oss/email2trac/wiki/Email2tracInstallation

Code: [Select]

Build process ¶

Warning Icon Image Pay attention to the MTA_USER and TRAC_USER, they are not something you can ignore!

Unpack the tarball. There are several methods to build the package.
No package manager ¶

The configure script build default is /usr/local

./configure (see --help)
    default values for:
       prefix=/usr/local        --> --prefix=<path>
       exec_prefix=$prefix/bin  --> --exec_prefix=<path>
       sysconfdir=$prefix/etc   --> --sysconfdir=<path>
       MTA_USER=nobody          --> --with-mta_user=<name>
       TRAC_USER=www-data       --> --with-trac_user=<name>
make
make install

If you want the python scripts to be included in the buildprocess add --enable-python
or...

Code: [Select]

RPM package ¶

cp email2trac.tar.gz /usr/src/redhat/SOURCES
rpmbuild -ba email2trac.spec

    * RPM package instructions seem incomplete. The above didn't work for me, I had to do the following (as root or using sudo). There may be an alternative way of setting TRAC_USER and MTA_USER than the one below but I couldn't find it. It's a shame it doesn't pull it from the config file. - Tolan Blundell slight_ at hotmail dot com.

      tar zxvf email2trac.tar.gz
      rm email2trac.tar.gz
      // Edit lines 37 and 51 of configure.in to alter the TRAC_USER and MTA_USER variables.
      mv  email2trac-0.13  email2trac-0.13.0
      tar zcvf email2trac.tar.gz  email2trac-0.13.0
      cp -R email2trac-0.13.0/ /usr/src/redhat/SOURCES
      cp -R email2trac.tar.gz /usr/src/redhat/SOURCES
      cd /usr/src/redhat/SOURCES
      rpmbuild -ba email2trac-0.13/email2trac.spec

I appreciate any help anyone can give. I am pretty handy with ctrl+c and ctrl+v 

[Stefano]

hi

what about reading and following this http://wiki.contribs.org/Trac?

and, of course, using search and looking in the wiki before posting?

HTH
Ciao
Stefano

[jptechnical]

I already installed the trac contrib, so that is fine.
I have installed the email2trac script on debian systems, but I don't understand the SME contrib mechanism for user authentication, who the wwwuser is, who should be the trac user (considering it get's it's auth from sme somehow).

I have read through how to make a contrib, but I can't code worth a damn... I am a hack. I am sure I can make this work with a ton of trial and error, but I would like to make it in such a way that my next software update doesn't undo my work, with the possibility of packaging it into a contrib or giving it to the original author as an addon.

Thank you for your condescension. I swear, you give my the riot act for supposedly not searching and you don't even read the subject to see I am trying to ADD to the trac contrib, I essentially asked for two pieces of information and guidance on the bast way to accomplish this. How about this, YOU show me where it is documented in the trac contrib page ANYTHING about what I am trying to accomplish. Your comment was rude and you are the one that didn't do his homework.

[jptechnical]

Sorry, that was harsh. I know you meant well Stefano. There is alot of documentation there, and on the forums, and I have in the past posted before fully researching the matter. However, this is not the case on this topic. Nevertheless, I apologize for my outburst.

[Stefano]

Hi..

my bad, I really didn't fully understand what you are doing

BTW, you need to install the dev tools packages on your sme (I suggest you to create a test sme on vmware).. then you can recompile/rebuild without any problems

HTH
Ciao
Stefano

[jptechnical]

noob question here, but please clarify so I understand.

If I compile that script, I have compiled it for the current kernel right? So, next update of the kernel I would need to re-compile the script. I can do the compiling on a parallel dev box (vm as you suggested), because the hardware is not so much at issue, it is the kernel you are compiling for. So, as long as my setup was the same on both (not 586 on one), and they are updated in parallel, then a compiled binary should work on both, even if one is a vm and the other metal.

Do I understand that right?

[Stefano]

hi

you are not compiling a kernel module.. so I think that it's not necessary to re-compile it everytime..

AFAIR I compiled freepops on a dev machine (sme 7...2 I guess) and I'm using it under a sme 7.3 without problems..

btw just compile and install it now.. then, when you'll upgrade your kernel, you'll see if everything is right

Ciao
Stefano

[jptechnical]

when you'll upgrade your kernel, you'll see if everything is right

good point