+ 1 Ray > On Jul 21, 2017, at 10:32 AM, Andrew Sullivan <ajs@xxxxxxxxxxxxxxxxxx> wrote: > > Dear colleagues, > > On Tue, Jul 18, 2017 at 11:13:10AM +0200, Yoav Nir wrote: >> The new site is focused not on the existing people active in the IETF. It’s focused on the general public. The place where work gets done these days is datatracker. >> > > I have had a look at the beta site, and it seems to me that we have at > least three possible audiences: > > 1. People who wonder what an IETF is and how it might be useful > to them. That actually is an important function of a website, and > despite the arguments about how the IETF isn't about marketing > organization (and the xkcd observations about university websites) > we cannot ignore this function. To some extent, we are competing > with other ways of developing the Internet -- "living standards", > code as standard, multilateral standards bodies, &c -- and if we > think our way is good for some cases we do in fact need to market > that. I think the beta site is obviously better at that than our > current site, which appears to be designed with a MEMBERS ONLY > sign on the front. We are the hardest club to join given that we > don't have membership. I think the beta site is trying to make > that burden a little less, and I believe it is a good thing. > > 2. People who already are familiar with the IETF and are working > here. The beta site says it's supposed to be the "new front > door". I don't know about all of you, but in my own house I also > know how to enter by the other doors. Maybe we just need to use a > different entrance? > > 3. People whose connectivity makes the more graphically intense > and somewhat larger site less useful. I think it would be useful > to analyse the extent to which this is a real problem, and whether > the trade-off is adequate for the particular use case we have in > mind. After all, loading this (still not huge) web page is hardly > the most bandwidth-intensive thing a plausible IETF participant is > likely to do. Moreover, it's not clear whether the problem in > this case is slow backhaul links or slow local/last-mile links; if > it's the latter, more IXes (and CDNs) are likely good enough > mitigations. This seems like an issue that could use some > empirical data, but I am not sure how to get it. Perhaps the > participants of GAIA would hae some ideas. > > I appreciate the effort to make these changes in public with lots of > consultation: I know how much harder it is to run a project this way, > but I think it's a good sign that the tools are as usual being > developed this way. > > Best regards, > > A > > -- > Andrew Sullivan > ajs@xxxxxxxxxxxxxxxxxx >