Locked Re: Community management of main fd.io website

Neal Hartsell

IMO, web platform / update mechanics are important, but it's the context of how we got here that I find more interesting... 

  1. What is the problem FD.io faces and what should be done about it?
    1. Plain and simple, great tech story, not enough market awareness or adoption
      1. The data plane is down stack, so we are but a component to a much larger story of digital networking infrastructure
      2. As a result, enterprise developers have largely never heard of it
        1. If they have, they don’t know why the need it
      3. There ought to be a much broader base of contributors to VPP et al
        1. We don’t yet have a virtuous cycle of “more developers -> more adoption -> more developers..."
    2. The marketing strategy seems highly focused on events, which are showing little payoff
      1. We appear at super large events, but wind up in a room speaking to ourselves - valuable for sure, but not moving the needle
  2. The above two issues have given rise to a “marketing rethink”
    1. Part of that thought process has been….
      1. With limited budget (couple of hundred k?, IDK), events are expensive, maybe more digital marketing would drive better results
      2. If so, let’s step up the web message
      3. What would we do?
        1. A more compelling home page experience
        2. User stores / Use cases that talk less about FD.io as a project (who cares) and more about the value of what FD.io tech can do / how it speeds TTM, jacks up packet processing throughput, saves dollars, makes it easier for CNCF de elopers to get to the promised land, etc.
        3. Stronger press/web/blog/SM (and yes some sort of physical venue presence / participation / megaphone) effort
  3. That then gave rise to a set of issues (that I hear about, but don’t have enough committee tenure to support or refute…)
    1. There is low / no budget to get digital marketing infrastructure changes done
    2. It takes too long to get simple site changes done
  4. If Ed says Github+Hugo+Netlify is the way to go, let’s do it - especially if it solves the two issues in #3

We need to get past the mechanics and focus on the exact business goal and marketing effort that will achieve that.

I work in a company that is very open-source minded, but a few of us tend to keep a tight leash on the web properties.  So, it will be a new experience (for me) to see how messaging ideas get approved by the “community”, but, I expect it's solvable.

To that end, I do plan to make a submission that I believe strengthens the story / home page experience here. I personally like the look and feel of the new site - whoever did it (kudos), just think the messaging, story and CTAs need a brush up.

BTW - on the top fold, "flexable" should be "flexible” :-) 

If the project succeeds, it lifts all boats.

My 2 cents (you get what you pay for) :-)


On Jun 5, 2019, at 5:52 PM, Edward Warnicke <hagbard@...> wrote:


On Wed, Jun 5, 2019 at 3:28 PM St Leger, Jim <jim.st.leger@...> wrote:

Ed/Ray/FD.io Community:


I haven’t been directly involved in this discussion.  But my observations are that there is an effort in FD.io to improve some of the content and documentation (which is always commendable) but also some level of frustration with how to get that content updated and published on the website and/or wiki pages.

Beyond frustration.  The community has been requesting for a year proper linkage to docs... and its still not done.

The website hasn't been updated meaningfully since launch three year ago.  Our events are multiple multiple links deep in the site.  The slightest attempt to change anything involves insane efforts and high latency via staff.  What we have now is simply not functional: for anyone.  It's an impediment to the community, and it's not really helpful for marketing either.

The fundamental issue is the extremely old fashioned way the site is maintained.


My past experience both here and in other projects has me conclude:

- Use a marketing team for the “pretty” public facing content. This would be all the fd.io facing material.  FD.io has a marketing committee that should own and drive that effort.  They should also be working with the broader LFN marketing advisory committee for support, ideas, etc.

- Use the technical community for technical content that predominantly shows up on the wiki, developer targeted pages.  If req’d the FD.io TSC could ask the LFN TAC for help (though I don’t suspect that’s needed.)

All modern communities tend towards the Github+Hugo+Netlify style approach (though some are still on Jekyl) and community manage their sites.  Examples:

Outside of LFN projects or DPDK, I am aware of few open source projects who's main website *isn't* community managed via Github+${Static generator}+{Authhoster}

That said... I know for a fact that there is marketing participation in that community management of those sites... marketing contribution, review, etc.  Marketing is  very *valuable*.  But communities generally *do* manage their own sites.

I took an AI from last weeks marketing committee meeting to ask that the marketing committee get 1-2 reps (like other fd.io projects) to the process so they can participate in the review/management with the rest of the community.
I think that's a very good idea, and one I hope the TSC adopts.  


I don’t know why we’d ever want to get the communities technical experts working on non-wiki tech pages.  And if there’s some gaps wrt support from the marketing team then perhaps the topic should be brought up there and addressed head-on.

Because it's how most open source communities do it, including some of the more successful ones out there.  Its the modern way of handling open source community web sites. :)


Hope this perspective helps a bit. Let me know what “we already tried that” or other barriers exist. Then let’s get the right folks to step up, own them, and resolve them. Our tech experts can then focus on creating the best technical documentation and not worry about website layout, etc.

Yeah... I think what we've hit is "All attempts over multiple years to do it this way have failed" combined with "Most modern open source projects have had success with this Github+Hugo+Netlify approach".

I think the thing we *must* do is get our marketing community plugged into the process appropriately, which hopefully we can do tomorrow :)





P.S. I’d join the discussion tomorrow, but I’ll be in a plane.


From: main@... [mailto:main@...] On Behalf Of Edward Warnicke
Sent: Wednesday, June 5, 2019 11:00 AM
To: Kinsella, Ray <ray.kinsella@...>
Cc: main@...
Subject: Re: [FD.io] Community management of main fd.io website


Totally.  There was some discussion after you left about whether to take a vote or wait a week.  The consensus was that you had seemed very supportive, and so we felt comfortable proceeding.

Our intention was not to preclude any further discussion you think might be productive.  Lets talk about it tomorrow :)




On Wed, Jun 5, 2019 at 12:13 PM Kinsella, Ray <ray.kinsella@...> wrote:

Hi Ed,


Looks like this vote occurred after I left the call, so I missed it.


I very much want to commend the work that John DeNisco in particular has done here, it is really impressive.

John has been working tirelessly on the FD.io documentation and the website since this time last year, he deserves a lot of recognition for this.


However I think this bears a little further discussion at the TSC.

For instance, I understand that this will make technical contributors such as PTL’s lives easier.

I am 100% on that page, however do we have input from the FD.io marketing committee on the change?

What happens to Linux Foundation marketing support after this change is made?


Could we schedule it as an agenda item tomorrow?




Ray K


From: main@... [mailto:main@...] On Behalf Of Edward Warnicke
Sent: Thursday 30 May 2019 18:04
To: main@...
Subject: [FD.io] Community management of main fd.io website


In todays FD.io TSC meeting, the TSC voted to take steps to allow the FD.io community to manage the main fd.io website.


As part of the discussion, John DeNisco showed off a Github+Hugo+Netlify version of the fd.io site, which you can see here:



It is driven by PRs to a github repo.  The repo for the above demo is:


https://github.com/jadenisco/fdiomain if you want to take a look at how it works.


We will be moving this to fdio/site in Github as part of this process.


Anyone can push a PR to alter the website (example).  The PRs generate preview site that can be examined prior to merge.  Deployment of merged content to the main site is automatic.


The other thing the TSC decided today was to ask the PTL of each project to propose 1-2 representatives to get access to review site changes in Github and have access to the Netlify org.


So PTLs, please respond to this email indicating your nominees for those roles.  Please cc them so I have emails to use to invite them to various things.  Bonus points if you can include their Github ids :)


In particular there is a place for each project to list content about themselves here:



There are some mockups for vpp and CSIT, we would like all active projects to provide content.

Join {main@lists.fd.io to automatically receive all group messages.