On 1/18/15 5:53 AM, Claude Warren wrote: > I prefer the mailing list because it pushes new concepts to me. Git and > such requires that I work harder to get the information. Most of the > Apache mailing lists have a high signal to noise ratio. And even the > signals I am not interested in don't take that long to dispose of.
+1 - even the "noisy lists" which are, IMO actually a good sign. As others have pointed out already, there are two reasons that we insist that projects have dev lists and adhere to the principle "if it did not happen on the list, it did not happen" (meaning no decisions can be made elsewhere). 1. The list provides a "single point of transparency" for the project. It is easy for anyone - regardless of time zone, employment affiliation, cool tools, secret handshakes, etc - to see what is going on now and how we got to where we are. 2. The list serves as an archive of project decisions. There are some significant benefits of mailing lists that alternatives will have to deliver: 3. We can host and archive them on ASF infrastructure, effectively guaranteeing durability and independence. This is very important. The ASF has already outlived quite a few cool collaboration vendors / quasi-OSS watering holes. 4. At base, they traffic in plain ASCII text with headers capturing the critical information necessary to reconstruct discussions. There are lots of tools available to analyze archives. A good question to ask about any suggested alternatives is do they meet the requirements 1. and 2. and share the advantages 3. and 4. or have other advantages. One thing that I agree is a shame is the deterioration of mail clients and archive search tools. Gmail is a disaster. I personally use Thunderbird for all ASF lists. I also use markmail a lot for searching, as it is superior to what we have internally. Phil > > Claude > > On Sun, Jan 18, 2015 at 12:34 PM, Benedikt Ritter <brit...@apache.org> > wrote: > >> Hi all, >> >> over at the Apache Commons Project, we have a long discussion about our >> mailing lists. Are they to noisy? Should they be splitted up into sublists? >> Should individual components go TLP? >> IMHO Ben McCann summed up the core problem pretty well [1]. Mailing lists >> are simply a outdated tool from the 90s. They can not compete with tools >> like github/gitlab that integrate the code with the possibility to do code >> reviews, disucssions and bugtracking. >> >> Now I'm curious: Does anybody here really like the use of mailing lists? Or >> do we all simply go through the struggle of setting up filters etc. just >> because this is the way it has always been? >> >> Regards, >> Benedikt >> >> [1] http://markmail.org/message/iizay3mmf2msvaf2 >> >> -- >> http://people.apache.org/~britter/ >> http://www.systemoutprintln.de/ >> http://twitter.com/BenediktRitter >> http://github.com/britter >> > >
