The Ubuntu architect: why it's important to attract all users
8th Oct 2012 | 12:00
How to get non-technical users to contribute
Allison Randal is such a brilliant speaker that she could keep us interested in anything, for hours.
Fortunately, when our sister title Linux Format met her she was talking about the best ways to engage the next wave of Ubuntu users to join in and become good citizens.
As technical architect of Ubuntu, it's her job to, as she puts it: "champion the community's vision for Ubuntu; to facilitate conversations as we integrate multiple perspectives and balance multiple needs; to ask good questions that help us find better solutions."
Who better to ask about Unity, the HUD and wanting to punch people?
Linux Format: Let's start with the work you're doing with Ubuntu. It's had really good forums for ages, and the Ubuntu community is one of the best things about it. How important is that in attracting non-techies?
Allison Randal: Ubuntu has focused on non-technical consumer users from the very beginning. So you may not be a developer, but you get recognition for your contributions to the project - Ubuntu membership recognises helping out at a booth, or answering questions at a forum.
Even though I'd say that, as that non-technical userbase gets bigger - there are 29 million users to 200 developers and 700 members - you can definitely feel the tension. For example, right now I'd say the bug queues are totally flooded with bugs that you can't really act on. It's not really a bug report that you can say "OK great, let's dive in, we'll fix this". It's a strain on the project that there are so many users who really don't get it.
So we've been doing a lot of thinking about how to help non-technical users. And I don't necessarily have the immediate answer; it's not a magic solution, it's just something that we're facing, and I know a lot of other projects are going to be dealing with soon.
LXF: I guess they don't have that kind of problem with Fedora or Gentoo. Every user's a developer, right?
AR: Not much with Debian either. Red Hat probably more so, but they're dealing with a lot of business users. That is one answer to the problem: that non-technical users get support, and then you have big support channels for them; business users do that quite a bit.
But I think that there's a big section of non-technical users who aren't going to pay for a support contract. They still have those very low-level questions that they need answered.
LXF: Do you think there's anything you could take from the Red Hat model of paid support? Because it's in their interests to make it easy for customers to contribute.
AR: Canonical does paid support as well, for business users. We'll give you a support contract for your entire company, or just the occasional one-off support contact.
So that's part of the answer: helping non-technical users see that the software is free, and there are people willing to help you, but we may not have the resources for volunteers to give you what you need.
LXF: You mentioned non-actionable bug queues. What kind of bug reports do you get from nontechnical users?
AR: We get a lot of them, unfortunately, and it's often things like: "Thunderbird crashed: fix it," or "I can't find my application anymore now that you rearranged the menus". It's kind of on that level.
The forums tend to get the more productive questions, you know, like "how do I make a presentation?" or "how do I connect my camera?" whereas bug reports are often not helpful. And, unfortunately, you have to keep going back to a standard answer, such as: "thank you for your bug. Could you give us a little more detail about when this happened, or what you were doing at the time?"
The worst thing is when we ask them "could you help us find a crash dump?" They can't, they just don't know how. I think part of the answer there is definitely better automated tools, so that when you have a crash it automatically catches the crash dump - it automatically submits it.
Firefox has some of this, where they actually collect all the information for you and just give you a nice little friendly screen that says "would you like to submit this to Mozilla? It could help us solve your problem". And a little "would you like to tell us about what you were doing?" Instead of having to submit bugs, it does it for them, and collects as much information as it can.
LXF: I know what you mean: we put a Linux distro on the coverdisc every month, and people used to ring us and say "what's my password?" and "where's Word"?
AR: And that's just going to get more and more. The more users that we have, percentage wise, only a very small percentage are going to be the technical folks who are helping out with the project itself.
From the beginning, free software was kind of skewed towards the thinking that all users will be contributors. All users have the potential to be contributors, and we're getting to the point where we have users who will never be contributors; that's not their interest, that's not their field of expertise.
And that's actually a good thing: if everyone were a programmer, that's not really a whole lot of diversity among your users, and it gives you a very small potential userbase. It's awesome if you know someone who's, like, a mathematician - brilliant but not a programmer. If they can get involved, or somebody who doesn't even use computers at all - if their very first computer is a Linux computer.
Or a Linux phone - there are a lot of those now: people who don't use computers at all but who have Android phones.
LXF: We keep coming back to this debate. Is it important for people to know their Android phone is running a kind of free software, or is it enough that they're using it, whether they're aware it's free software or not?
AR: I think the majority won't understand it at first, but it's a gateway. I certainly don't think that if you're using Android and don't know it's Linux, that it's bad that you're using Android. I think it's good that you're using Android, because people are using Linux, Linux is hitting the mainstream. The more people who use it, the more of them will gradually begin to understand it, because they'll get curious, they'll hear somebody talking about it.
But that's kind of our job, to keep talking about it and to keep it in the media. People can pick up on that and then gradually get into understanding what Linux is about and what free software means. Mozilla has done a lot of that.
They get all these users who don't know what free software is at all - all they know is Firefox is great. And then they run programs, sort of like telling people what free software is. When you first install Firefox, you get a little pop-up box that talks about your freedoms and gives you the opportunity to click for more information on the Mozilla website about what it means to be free software.
LXF: How's Unity working out in terms of reaching new users?
AR: Unity is, I would say, an experiment, a pretty advanced experiment, specifically in usability. Unity was developed based on usability research with non-technical users, specifically non-technical users based on the old Gnome, with the little tweaks that Unity had.
We sat non-technical users in front of the old Gnome and the old Ubuntu, and the first version of Unity came out of that. The things that those users struggled with, we changed, and it's an iterative process from there. You develop a new interface, and you sit them down in front of that interface.
Now, how do you react to this? You find the things that they struggle with there, and you change them. It's a process of evolution, and it's not a completed process yet. It's a work in progress.
It's incredibly important that we have distros that are thinking about usability, thinking about how to really approach non-technical users, but I think it's also just important for technology in general.
Windows and Mac haven't been innovating for about a decade. They are now, which is interesting. Like, Gnome Shell and Unity launched, and then Windows 8 and the new release of Mac that's coming out…
LXF: The Lion view, with all the icons spread out over desktop…
AR: Yeah, exactly. All this time free software has been chasing proprietary software and now they're chasing us, which is kinda cool to think about.
LXF: It is awesome that Windows 8 has only just got the spinny desktop cube. That blew my tiny mind when I saw it first in Compiz, and now it's just 'meh'.
AR: Yeah, we've had that for a long time. Get with the program, Windows!
LXF: Is the HUD part of this process of chasing non-technical users?
AR: That's aimed at technical users. Unity was developed along doing usability testing for non-technical users, and we kind of lost some of the technical users along the way. So we started doing usability testing on technical users to see where they're struggling, because there are some things that we introduced that are really perfect for non-technical users but aren't for everyone.
One of the differences is that non-technical users may have one or two windows open; technical users may have 20. So there's stuff about window switching, there's some stuff about finding where you're going. The HUD is - I mean it's partly non-technical users in the sense that if you have someone who doesn't know the application they can just type 'glow' when they want something to glow, and it'll show them how to use that feature - so there are some benefits.
Non-technical users aren't very fond of text interfaces, like typing text, so I suspect it's going to be more appealing to the technical users, and then some non-technical users will kind of pick it up.
So here's where I'm not so sure: Google search. Non-technical users are fine with Google search - they get that. They get that they can type in a few words and it tells them where they want to go. So this whole idea, like with Unity, is to move away from clicking through menus to find my applications to, now, if I want a calculator I type 'calculator' and there it is. So it's a little bit of both; it's for both kinds of users.
LXF: Would you consider yourself a programmer?
AR: Yes, definitely. I started programming when I was about eight. My dad was a programmer, so I started really early on.
But in college, I did linguistics instead of computer science, and then probably about 10 years ago, when I was working heavily on the Perl Foundation and drafting the artistic licence, drafting the Perl contributor agreement, I considered going to law school. But I found that after a day of writing code I would be happy and relaxed, whereas after a day of writing legal text I would feel like punching someone. So I figured law wasn't the best career for me.
LXF: Isn't Larry Wall (another Perl polymath) also a linguistics guy?
AR: He is. In fact, Larry and I used to work in the same organisation. How I got involved with Perl is that I bumped into Larry and his wife Gloria, and we started talking about Perl and linguistics, and it was realised that we used to work for the same organisation. It was a little "ah-ha!" moment.
LXF: I remember reading an interview with Larry in which he was talking about all this cool linguistics stuff, which made so much sense, and the first time I looked at Perl code it scared the life out of me. Why all the brackets? Why all the Ctrl+Shift+weird character stuff?
AR: Larry would say that's not fluent Perl. Those features are there, but fluent Perl flows like text - you can write poetry in Perl. Actually, there's a lady who writes in Perl, called the Perl Poet.
LXF: We interviewed Mark Shuttleworth about three years ago, and he responded to some criticism about Ubuntu not contributing enough upstream, not contributing enough lines to the Linux kernel, and his response was that it's far more important to attract more people to get that system in place, like in Wayne's World II: "If you book them, they will come". Are you doing the same thing with contributors as opposed to users, to try to get more non-technical people contributing to Ubuntu?
AR: Definitely. One of the big focuses of Ubuntu membership is around building an identity in the project around any kind of contribution, so it doesn't have to be code. We have user groups around the world, and running a user group, or participating in a user group, counts towards your membership in the project. That's a really big deal.
It's a big deal not just for growing the Ubuntu userbase, I mean they're really important. When your neighbour tells you: "Hey, you should try this out", it's far more engaging than seeing it on a billboard. So that aspect is really important, but… it's just an important part of Ubuntu culture, accepting non-technical users. And that's not entirely necessarily natural for all the developers, so it's like a constant education process: learning a new culture.