[consulting] the Ultimate forum based Malinglist Manager

[Dan Robinson]

>>I know choice is good. But ATM we have only bads to choose from. I
>>therefore think we should work on one working solution. So here is my
>>proposal
>>:* team up with about 5 people on irc/jabber/skype and write down a good
>>: list
>>of demands. what do we want, how do we want it and what do we not want.
>>* team up with all the people we can get together. I am sure that with
>>Civicspace in this, we will get over a hundred people.
>>* invite one or two drupal coders.
>>* do a fundraise based on what thos coders think they need to get this
>>done. * Pay the coders fro+ that fundraise
>>* release this forum_ml.module (or maybe even bundle of modules) as The One
>>Solution.
>>* Write good solid documentation for debian/gentoo/?? to get that one
>>chosen ML manager configured to work with this module ....
>>* Write good solid documentation on how to configure your drupal system to
>>talk to that ml manager,
>>
>>    
>>
>
>Whoa, Nelly!
>
>While I like the idea of having a simple and easy way to do a forum <-> mail 
>list integration, including the setting up of the relevant software, I have a 
>strong suspicion that this route of implementation will not unite, but divide 
>around a set of scripts, modules and/or documentation.
>
>You're asking us to sign up to using a specific set of softwares just to get 
>the ease of a script or module to do the setup for us?  Personally, I would 
>rather invest time into learning how to do it with the software I want to use 
>and am familiar with, than to have to install and learn some other software, 
>especially with something like an MTA or MLM.
>
>Another point on this is that we all don't always have the luxury of 
>specifying the software we can use.  I have a project that I would like to 
>look at doing a forum <-> mail list integration, where the MLM and the MTA 
>have been chosen for me.  I already know, based on this thread so far, that 
>the proposed solution will not work for that project because of the software 
>choices.
>
>While it may not be the easiest or quickest path, my preference would be for 
>coding to standards and specifications rather than softwares.  If someone 
>wants to use software that doesn't conform to established standards, then 
>they can figure out a way to make it work.  Where as with coding to 
>standards, it should be no big deal to use whatever standards compliant 
>software we want.
>  
>
Your point is well taken - but the idea of having "established
standards" seems like a showstopper.  Also what you will wind up with
are some tools you can use to write some more code.  My personal opinion
is that there are no silver bullets here at all - just a lot of
choices.  We'll pursue a MLM specific solution (ezmlm) and leave open
the possibility of integrating other MLM's (or Drupal code) - however
there will be no guarantee that another MLM will be adaptable to the
code - because it may not have an adaptable interface (for example I
have no idea if Mailman will be able to easily "slot into" the
solution).  So what do you propose to actually produce?  What "standards
compliant" software are you referring to?