Re: Should IETF stop using GitHub?

[Keith Moore <moore@xxxxxxxxxxxxxxxxxxxx>]

On 8/2/19 8:50 PM, Bron Gondwana wrote:

One of the biggest challenges that the IETF has is bringing in new blood, and one of the biggest advantages of Github in particular is that it has a lot of mindshare right now, so there are many people who know how to use it and allowing them to interact with in-progress IETF documents using tools they understand is a good way to bring them in.

Understood.  And I'm very glad to see more interest in IETF from the open source community.

Having the process of joining the IETF be entirely learning a custom workflow and custom tooling which only the IETF uses is a recipe for isolation, and I would much rather see the IETF embracing the tooling that the rest of the world is using rather than running everything in isolation.

But this raises a problematic question - whose mindshare should be favored?   Should IETF favor github users over users of other tools?

What about 10-20 years from now when some other tool is in vogue?   Should IETF then favor the new tool, or the one that its then-established participants already know?

IETF originally had a history of trying to use tools that are accessible to a wide potential participant base, but has been departing from this for awhile:

• Email had the advantage of near-universal accessibility (at least by people on the Internet, and even by some not on the Internet).   And it was/is easy to use.
  
• The web had a similar characteristic, once high-quality web browsers became available on all major platforms.
• xml2rfc was and remains problematic.   Sure, anyone can use a web interface to the tools, but there's a significant learning curve and often a lot of fighting to get the XML to conform to the latest fad.
• Github not only imposes a learning curve for those who don't already use it, there's also the risk of lockin to a tool that we don't control, and potential privacy risks that at least need to be evaluated (now, and in the future).

I very much want to attract open source developers (among many other groups that could bring new energy).    But to me there's something very odd about IETF trying to favor github users, especially at the expense of existing IETF participants and at the expense of users of other systems.  It's not like everybody already uses git.   Really, it's not.   Lots of companies have investment in other ways of maintaining source code, that either pre-date or post-date git.

I'm not saying there's an easy or obvious answer, I'm just saying that we shouldn't presume that github is the solution just because it's currently popular, any more than we should presume that Word is the solution for document editing or that meetings should consist of Powerpoint presentations.

Keith