Build your own home: why the user experience on the Edgeryders website is worse than on the commercial cloud, and why that's good

This text was written when the Edgeryders platform ran on a different, much clunkier and less usable technology. If you are reading this after 2017, you might think it’s not so bad, and I would agree with you. But that said, the argument still holds more or less in the same way.

As the unMonastery comes to life and smooth collaboration becomes a more pressing need, news has it that the unMonasterians are gravitating towards the use of proprietary software, which they prize for its superior user experience. The Edgeryders platform on which I write this is also in use, but using it, they report, is nowhere near as smooth as using Google Drive or Trello. They are right, of course.

[elf Pavlik], [Ben] and unMonasterians all: believe me, I know the beast. There is a trade-off between “anyone can use their tool of choice (more efficient and idiosyncratic)”, “we all are going to use X” (take full advantage of network externalities) or “it does not matter, we are going to make everything interoperable” (solves the contradiction between the first two but creates huge overhead, endless discussions in W3C etc.). There is no elegant solution, I’m afraid.

We decided to build our own home on the web from open end free software, improve it incrementally and turn the little frustrations into design and programming experience, as well as lessons in collaboration, mutual understanding. You know this – you were there. We made this decision in exactly the spirit in which you are building the unMonastery itself: you could just go out to IKEA and buy cheap furniture: it will work immediately, it will have been designed and manufactured by smart professionals with precise machines and user testing, it will be easy to install and replace. But you don’t do that: you insist, rightly, on starting from a blank slate.

As I understand it, you are in part doing this because you believe that hypercustomization will, in time, lead to better living than you could get if you just bought off-the-shelf furniture. But that’s just half of the story: the other half is that building the home builds the family, or the community, or whatever the social arrangement you are trying to build is. Humans will always sleep better in the bed they built themselves; communities will always be brought together by building things together (“barn raising”, remember?). This is just the way we are hardwired. When you look at things in this light, any discomfort in your chair on your bed is not just a quantum of bad vibes, it is information that you can and should absorb and reuse to move forward in a context of collaboration. So, when you encounter an obstacle, you don’t sneer at it and demand better service: you take it in, share it with your friends and partners, and mount an attack on it together with them.

Look, the Edgeryders website is no Google Drive. We cannot make it that smooth. No way open and free is going to compete with Google stuff in user experience. Not now, not for the foreseeable future, perhaps not ever. But it is our home; we are building it together; people do use it every day; and it will get better. What’s more, it will get better in interesting ways that uniquely suit us. It will co-evolve with the community – for example, we are discussing integration with the Economy App: try that with Drive! – if the community agrees to give it a chance, and to treat its many shortcomings as the gateway to our future as a sharing, collaborating community.

Of course, you will make your own decision. No one can or should press you into doing stuff that does not work for you. But I do ask you, in respect, whether the choice of evolving the Edgeryders website together is not more consistent with the unRule, whatever that turns out to be.


Thank you for sharing your thoughts…

sadly myself i may not find time to discuss it much further in next 2-3 weeks :frowning:

by then i would like that we all try to think how to avoid situation depicted below, just replacing linkedin, flickr, facebook with edgeryders, makesense, ouishare, okfn …


DataPortability - Connect, Control, Share, Remix from Smashcut on Vimeo.

really have no interest to contribute to systems which don’t address this issue!

1 Like

Sleepless Night.

This thread and surrounding conversation literally kept me awake for hours last night.

So caveats for everything that I say in this comment;

unMonastery is not migrating away from the EdgeRyders platform;

and this thread controvenes the idea of "the person who does the work

calls the shots" - which is the principle reason why previously I haven’t

raised issues with the EdgeRyders’ platform, because I haven’t had the time

or skillset needed to make the fixes/changes I desired.

So I want to break this up into 3 areas:

  1. unMonastery Matera vs unMonastery vs EdgeRyders

  2. Different Tools, Different Jobs. Unchartered Territory

  3. Baseline fixes/requirements for the EdgeRyders platform


1) What is unMonastery? What is EdgeRyders? What is unMonastery Matera? 

As I understand it, unMonastery is a concept that grew and was built from the

EdgeRyders community, which finds its physical incarnation in a prototype

based in Southern Italy, Matera - it has though a life of its own, EdgeRyders is its

launchpad, if it’s successful its good for both ER and unMon but in different ways.

unMonastery Matera is a project that is building its own autonomous way of existing

in the world - through a reimagining of what work and living mean in the

present moment. Each iteration of the unMonastery attempts to rebuild everyday

life from the ground up, sometimes it will adopt elements from previous incarnations, if

it makes sense but that means that it will pick, choose and build tools based on perspective

of its participants.

There weren’t any rules, structure or decision making processes in place when we arrived

in this location, because that would have undermined the ideas and desires of the individuals

who have given up their lives for a period of time in order to realise this brave experiment,

it would have also introduce a destructive hierachy into the group dynamics.

The online prescence of unMonastery Matera has some key requirements:

-The lowest possible bar to entry, due to the fact that people we’re trying to reach aren’t exactly

the digirati which makes up EdgeRyders.

-Have an at-a-glance quality, so projects can be understood quickly and allows individuals to

make instant contact via phone or email.

-Support visualisation and plugin data growing from the project - unMonastery Analytics etc.

-Be easy and fast to integrate and disgard things we think up on the spot, OSM street mapping, d3js etc.

-Have aesthetic autonomony

Right now the EdgeRyders platform doesn’t support any of these elements, to fix this one needs

to learn Drupal. I’ve spent the last 8 months project managing the development of a drupal based

website, we have 2 designers and 2 developers, the site is still a total mess. There is a dramatic

difference between learning markup, some js, using existing APIs and making commits on GIT,

to learning all these things then learning how to integrate them into drupal on top.

That’s not to say we’re not interested in doing that but I’ll get to that.

2) Different Tools, Different Jobs. Unchartered Territory. 

Picture the situation, you turn up in Matera there’s a constant downpour for days - you’re meeting

a group of people who you barely know from different backgrounds, with different skillsets, different intentions

and desires who for the foreseeable future you will be sharing a bedroom, eating with each day and working with daily.

Sounds fairly intense doesn’t it? Add to this, there’s no heating, no furniture, no internet - there’s

literally no infrastructure except you, your laptop and some unreliable WiFi spots dotted around the city.

How do you organise? - Forgive me if I’ve misunderstood but the impression I get from this post is that

we should have been organising around EdgeRyders based blogposts - creating tasks on the task tracker for;

fix the heating, reduce the humidity, get new keys cut for the building, find local food producers, document all receipts etc etc.

Wrong tool. You need something light and responsive, that doesn’t have Internal Server 500 errors, because

the environment feels bleak enough without additional technical problems.

This is the first time that EdgeRyders’ has attempted to do something that has a heavy and sustained

offline component that requires dealing with a difficult logistical environment with multiple stakeholders.

The bias of the platform towards long form blogposts doesn’t meet the very pressing needs we’re currently faced with.

Why are we using proprietry software? - This wasn’t my suggestion, this initially came from [elf Pavlik], who I’m

sure you’re aware is not exactly predisposed to using proprietry software, his reason; using expensive

to build software gives you the ability to work quickly together, identify the qualities and features needed to work

effectively. Once you know what you need, you can create a roadmap towards migrating and building open-source

solutions to solve the same problems.

In the projects I work on, my daily project management suite is made up of the following: SmartSheet, Twitter,

Pens and Notebooks, Remember the Milk, Google Calendar, Google Drive, Trello,, Popplet, Gephi,

Microsoft Excel, Evernote, Dropbox, Gmail Tasklist, iffft, Google Groups, a whiteboard, post-its.

I don’t ever anticipate that EdgeRyders’ will be capable of accomodating all these things and I wouldn’t want it to but I have a

workflow that allows these things to be relatively seamlessly connected to one another - in instances where there

is a limitation to connect the various services, I use them less.

  1. Baseline fixes/requirements for the EdgeRyders platform

An instant fix for all of the convergent desires outlined above would be for the EdgeRyders’ platform to

support data interoperability - it needs an API and its needs it’s underlying technical infrastructure to uphold the values it promotes, but it also needs some other things.


-a heavily optimized wysiwyg editor which strips out all unneeded HTML/CSS

-autosave / drafting functionality

-a spellchecker

-better file upload support


-The platform has begun to move at a snails pace in recent months, it needs a clean up and the optimisation of

various scripts.

-Remove error reporting upon posting at the top of the page.

-Investigation into the server logs, Internal 500 Server Errors have started to become very prevalent.

-A staging environment for experimentation and testing, so things can be properly tested before going live.

-A stripping back and removal of all unused pieces of functionality.



User support:

-Decent documentation in an obvious place

-Single video tutorial on using the platform

-Help notes in relevant locations in the GUI

-User journeys that take a service design approach to EdgeRyders; which account for on platform, off platform

and offline interaction - so that the platform can serve the way people work rather than restricting it.

Fundamentally, it needs a technical development roadmap and a rigorous scoping exercise to assess priorities and

needs. It also needs a team, it needs a designer, developer and a good teacher who has enough time to teach

those who are less able but willing.

These are my thoughts and feelings, and now that I’ve said all this I don’t feel comfortable just leaving it here

expecting someone else to fix the issues.

So here’s my offer, I’ll spend half a day a week whilst I’m in Matera scoping, researching and mocking

up functionality/fixes for the platform - this though is the limit of my skillset I can’t build it, I’m

not a drupal developer and neither is anyone else at the unMonastery.

Next steps?


No need to lose your sleep!

No worries, [Ben]. Your plight is understood. And if you do need to move your project somewhere else, so be it.

View it from a positive angle: Edgeryders (not just teh website, but the community and the company) is here for the long haul (we will announce the details probably in the coming week, but you get the idea). It is a resource. You can use it, or not, whatever works for you: since you are doing the work, you are free to call the shots. By staying close to us, you lose flexibility, but you gain stability: together, we were able to keep the fire going in the long months between when you sketched up (December 2012) and when the Matera prototype went live. Nobody guarantees there won’t be other dry spells, with fewer hands on deck, though I dearly hope it there will not be any.

From my perspective, a patch has emerged: you do your things on Trello and Drive (hey, we use Drive too occasionally) and what have you, but there has been constant conversation and updating on Edgeryders. This makes it easier for the hundreds of edgeryders who are not in Matera to stay close to the unMonastery, promotes a constant in- and outflow of people into the unMonastery, and capitalizes your great experience for everyone here. Not elegant, but it works, sort of. But if you now go out and build another workspace using community developers, I am very afraid we will get two technically subpar platforms instead of one, each with half the traffic and conversation! Don’t underestimate the ability of code writing to syphon up workdays: [elf Pavlik] is known to be a brilliant programmer (I have not had the pleasure of working with him yet), but [Matthias] is brilliant too, and he has logged more than 500 hours on the damn thing. If the platform is underperforming, it is likely to be because building good platform is very expensive and hard. This means it will probably be hard for you too.

according to our ISP, 500 errors are due to a DDoS attack that also involved – believe it or not. Typically these things don’t last long, and tonight ER seems to be working for me, despite my finnicky Thai connectivity and frequent brownouts.

I think we’re on the same page.

But there was initial misunderstanding.

By (re)building an additional site our interest is not to double the workload - the absolute ideal situation would be a setup that allows parts of ER to act as CMS for the site. (e.g a wiki page would serve up blocks of content but redact comments, in a way that is stylistically more accessible). Technically, we could just write a page scrapper with something like python but I’d say this is less than ideal - some technical understanding from [elf Pavlik] and [Matthias] of what options are available would be useful. RSS, JSON, etc?

I don’t see creating an additional site for unMonastery as a move away, I see it as a lightweight portal, through to EdgeRyders, individual websites, social network profiles and the networks of individual’s. Similar to the way that twitter is utilised by ER and other organisations to drive traffic in different directions.

“Edgeryders (not just teh website, but the community and the company)”

I agree but I felt in someways this was lost in the points that were being made - and this is what I was gesturing towards when I said there needs to be a service design approach to user journeys. It’s worth saying that some of the most valuable members of the ER community don’t post or read very often/ever, but I don’t believe that they feel any less a part of the network and when it comes to the moment when they’re needed, our workflow seems to bring them in at the right time.

“If the platform is underperforming, it is likely to be because building good platform is very expensive and hard. This means it will probably be hard for you too.”

It is hard but there are varying gradients of difficulty, drupal development is really hard in the context of platform utilised by a diverse group, a (relatively) straightforward presentation platform less so, we’re not aiming to construct any kind of workspace - more similar to what the EdgeRyders business site is aiming for, which as a front of house requirement, I think it is essential.

Every tool has its place

Some remarks from my understanding of what is and is not (both present and potentially):

  • Of course there is lots of things to do, rework and redesign in this Drupal site. "My list is long", and as I said before, this is the messiest website I have ever cared for, which was so far simply due to a lack of volunteer resources to put into this compared to the resources needed by a website this big ... . But it's going to get better, and I will catch up with important things over the next month.
  • When you do it right, it takes a lot of time. Ben mentioned SSL for example. I started to implement that over the last days, and it turns out this is how to do it right. Which took about 8 hours to research and write alone. SSL "as usual" with all the three-letter agencies listening in is frankly just snake oil, esp. given that we have a platform with public content and our users are technically inclined enough to not reuse a password from something important ... .

    (And because “doing things right” takes so much time, it is better to use available tools that have been “done right” for their purpose, such as Trello, instead of using lots of dirty hacks to make do that all. The best we can realistically expect is to be the right tool for a limited, confined purpose in the future. What would that purpose be? That’s what I want to discuss next:)

  • is for inter-project collaboration. Let's keep expectations reasonable: cannot become a project management site that everyone will be happy with for managing their internal little tasks. Drupal is an old system with little support for AJAX for example, so it is hard to impossibe to hack it for making it speedy enough to manage lots of little tasks. Any specialized open source or commercial task management application is the better tool for this, and we're not going to invest hundreds of hours of programming to reinvent that for the Drupal context. Plus, task management requirements are very different between projects, as Ben outlined for unMonastery. In my view, I do not even want this to become a hosted task management platforml, or any other tool that is critical for a project's mission. The task this platform can fill well is inter-project collaboration: showing intermediate project results and getting feedback for them, and coming together to organize "joint venture projects" such as FormStorm, and to found new collaborations and find collaborators. (In my view, it's like the Github for non-source projects: projects on Github use that as their public collaboration interface, but almost always have additional internal tools for micro-management, such as mailinglists, own website with a wiki etc..)
  • is not for website hosting. This is a difficult decision, but I think it fits best into the software and content architecture to not offer website hosting features inside this Drupal website. Means, if Ben is looking to create a self-contained, independent-looking little public facing, static content website for unMonastery, a little WordPress install (on Edgeryders' server if he wants) is a way to go. Content on on the other hand will (1) always be accessible with the domain only, and (2) keep the same site-wide navigation structure and graphical design, to promote website coherence and provide intuitive navigation. This is how Github does it (in contrast to SourceForge, where you can have own websites on And Github does everything right :-) ... except the "teams" feature, that is. So you can host your project on without needing another web presence (as many do with their open little source projects on Github), but once a project grows so big that it needs more of an own identity or more specialized collaboration tools, it will need to use external web resources as well.
  • How to support further development? Ben is welcome to help with Drupal development half a day a week of course. The overheads of context switching and learning how to contribute to a specific project are quite high for anyone not working on a project permanently. That's why we discussed that synergistic mutual support with network barter tech would be a solution. We will offer this solution to the community once it's released. (After some major UI rework, Economy App is now undergoing debugging and lots of little fixes, nearing its first public release ...).
1 Like

Like that

That’s profound, [Matthias]: “the GitHub of nonsource project, optimized for interproject collaboration”. I like it a lot, though I am still missing the human factor: both in the sense that humans with projects need other humans as much as they need elegant tools; and in the sense that humans are plastic (communities are plastic by self-selection, and individual humans are plastic because they can learn new stuff) which makes it kind of pointless to talk about stuff like the technical level or the aptitude to share “of the average user”. I will give it some more thinking and expand on it soon (well, soonish, seen as I am on holiday now).