The One Laptop Per Child, officially known as XO, is now appearing in people’s mailboxes. The unboxing photos are up on flickr. The OLPC buzz is hot!

But I’m a little sour about it. It feels like I have read nothing but breathless praise for the design and implementation of the devices, both the hardware and the software. Mixed with the kudos there have been some critiques of the methodology and pedagogy behind the whole project, questioning the idea of giving laptops to third-world kids in the first place and criticizing the designers for arrogantly avoiding user research and for not testing the device with real third-world kids. But even the harshest critics of the project seem to have nothing but praise for the design and even for the usability of the devices.

So why am I not excited? Well, to put it bluntly, I find the positive reviews of the UI design extremely hard to believe. From what I’ve seen, the UI bears all the hallmarks of a user interface disaster, a case study in designer-driven design. I don’t understand why the whole UX world isn’t awash in skepticism over an OS that looks all the world like a Microsoft BOB for the Wallpaper* set.

At some level I suspect there is a certain degree of reluctance on the part of user experience critics to stand up and say something bad about a project whose objectives seem so noble and generous. Maybe it’s a “if you can’t say something nice, don’t say anything at all” thing.

So I have a challenge for UX pundits and professionals who are also proud new owners of the XO: Say something nice about the Sugar UI. Or say something critical. But talk about the user interface for real, in detail, and don’t hold back.

Don’t just talk about how awesome the project itself is, about the great minds behind it, or about the clever hardware and the cool mesh network functionality. Talk about the usability of the software. Think of how the design might be different, how it might work better.

I’ve not actually used an OLPC yet (I hope to very soon). I have seen a lot of screenshots and videos, however, and have used the emulator a little bit. But even the screenshots give me a deep, gut feeling that something is very wrong with this user experience. To wit:

  • The game-like and oft-abused spatial metaphor, suggesting that the relative positions on the screen are where other people actually are in the real world.
  • The circular menu — a darling of academia, unproven in any real-world context. As with the spacial metaphor, I think this idea has promise, but seeing it on the XO tells me that the designers simply want to prove a point.
  • The idealistic and haphazard usage of language-agnostic iconography, which falls apart at every turn whenever words become unavoidable, defeating the whole point of using icons.
  • The frequent lapses into a menagerie of half-baked and crappy open source user interfaces.
  • The exposure of hard-core programming tools to extreme novice users (especially the choice of the ubergeek language Python!).

And, oh, those icons!


I can’t get over the creepy similarity between the Sugar UI’s icon for a person and the internationally-familiar “skull and crossbones” symbol, in particular its incarnation as the icon for minefield warning signage. Wealthy first-worlders might not see it this way, but if you live somewhere where minefields actually exist, and where children have been injured and killed by them, this might not be such an extreme connection. Not to push this too far, but the military term for a minefield/landmine is “UXO” (unexploded ordinance).

I hate to come across as bitter or petty here — I am actually quite sympathetic to the idea that technology can play a big part in the education of kids living in poverty around the world. I actually hope to be able to read some convincing arguments that the Sugar UI is great. In particular I would love to hear that it can and does work well for third-world kids.

The key word here is “convincing”. So far, much of the design commentary has been praise based on the pedigree of the team behind it — MIT Media Lab, Pentagram, Fuse Project etc. I want to get beyond that and talk about the UI itself and how people use it. Of course, this may take a while to emerge as the devices make their ways into the hands of children around the world. This is obviously a developing story.


13 responses to “Challenge: If You Can’t Say Something Nice about OLPC…”

  1. I thought the icon was a rip off of the cingular and IxDA logos. 😉

  2. Oh, Chris, come on, that’s totally unfair. I’ve been very public regarding my disappointment with the OLPC, my concerns about the project’s efficacy (on the utilitarian level) and arrogance (on the ethical level), my belief that Sugar is rife with incoherent thinking, and so on.

    I’ll admit that I’ve gone back and forth on OLPC’s prospects at least twice, as each new wave of data comes in, but surely you’ll grant me that I’m not one of those lining up to fellate Pentagram, Negroponte, and the Media Lab:

    This isn’t harsh enough criticism for you? This doesn’t budge the Fahey needle? Unless, of course, you don’t think I count as a UX pundit ; . )

  3. AG: You’re still one of the few. One of the only. I may be painting with a broad brush, but I still hear mostly a lot of “oooh”s and “aaah”s. As Fake Steve Jobs puts it, “we’re not supposed to scrutinize or evaluate this new machine, we’re just supposed to jump up and down and cheer”.

    I’m not so much concerned that there is too little criticism as I am concerned that there isn’t a deafening roar of criticism and even outright mockery. I still wonder how anyone can actually praise it at all. Thus my challenge.

    Your 24-hour critique, however, fits the bill.

  4. Two quotes on this subject by Jane Healy in her book “Failure to connect”:

    “Are the new technologies a magic bullet aimed straight at success and power? Or are we simply grasping at a technocentric “quick fix” for a multitude of problems we have failed to address?”

    “After surveying current attitudes for the nonprofit organization Learning in the Real World, William Ruckeyser told me, ‘The nearest thing I can draw a parallel to is a theological discussion. There’s so much an element of faith here that demanding evidence is almost a sign of heresy.’”

    Jane is actually talking about the introduction of computers in US classrooms in the 80s and 90s. Seems amazingly appropriate for this discussion about the OLPC though!

  5. Great criticisim. Love the quotes, Harry.

    It seems to me that the hardware design is much more intelligent than the software (neither of which I have used).

    I have a larger criticism of the OLPC religion, but I think that the hardware design is sound.

    Regarding the UI: As a developer I believe in fast iterations. Get the thing out there and see what works, then change it. But as a user experience researcher I can’t believe how little testing they have (apparently) done.

    There was supposed to be a 1000 student test in April, but I cant find the results.

    Anybody know about OLPC user testing results?

    FWIW: Here’s a video of the designer walking through the hardware design:

  6. chris, mine came in Friday, haven’t had a lot of time with it. One thing about the UI for us folks who have donated and received one, is that we won’t get to experience a lot of the collaborative nature of the software design. The neighborhood and group functionality is geared toward chatting, creating and sharing amongst your immediate olpc peers. It’s a different approach overall. The software is definitely dumbed down to the greatest degree, everything seems to revolve around activities, the unit keeps a streamlined history of everything you do. where you have browsed, what pdf you have viewed, what app you launched. It’s a very different approach.

    The screen is made for outdoor usage and that is probably why the icons are as simple as they are. I am not defending them, they do all seem very child-ish.

    I was disappointed that the unit did not have a touch screen, I though it did from the slate mode, but it’s just for ebook reading. I just feel the unit is dying for a tablet for the kids to take advantage of.

    It’s going to be interesting to see what apps show up on the device. Opera has been ported, it’s much more functional then the dumbed down mozilla code they have on it out of the box. It even has it’s own olpc theme. Skype also works, those are the only two apps I have installed it.

    2 things, it really reminds me of the emate design, and you can start it up in Pong mode. hold the gamepad up, and turn the unit on. Upon release two players can play pong, one using the arrows, the other using the x and o gamepad buttons. Hilarious…

    I need more time to give it a full once over. The lack of wpa out of the box is quite silly. you need to download a script put it on a thumb drive, pop it in, crack the terminal, run the script and put in your wpa wireless network name and password then reboot. Repeat for any other wpa network you need access too.

  7. Hey Chris,

    We’ve had two OLPCs for over two weeks now. I won’t say anything about subjective criteria, but I can tell you very objectively that the little bit of software it comes with simply is broken in many places, making this essentially an expensive brick reminiscent of a MagicLink or Newton.

    For instance: Nemo wanted to play a game of Memory, so we tried sharing that Activity. One of us would start it on one laptop, the other gets invited and we would see each other as participants, but the software hangs every time it’s the second player’s turn. How hard is it to get make this work? How hard to QA whether the only game on the laptop works? That’s like shipping Windows 95 with a broken Minesweeper.

    Another symptom of this “ship the shit” rollout is that the documentation (the Wiki on is a mess and important points are not mentioned or cannot be found. E.g., I think that when I allow our laptops to connect to our home network, they allow anyone else into it via the mesh network (and, as I understand it, the wireless chip used by the OLPC performs mesh routing even when the machine is sleeping). This is not mentioned anywhere in the documentation. Being reasonably proficient around computers and knowing how to use a keyboard, I shut down the mesh networking using the console, but not everyone else might know how to.

    I am a reasonably good programmer and would like to add some software to the collection of 20 or so programs currently available for this thing, but even there, the support is minimal. Setting up an emulator is incredibly difficult and the very narrow criteria for what kind of software can be included in the distribution leaves developers with only Python, but not Java or Mono (non-Microsoft .Net). Smart to exclude the largest developer communities.

    Someone who knows Negroponte claims that he is into the flash of things. It’s got to look super cool for the moment. Durability is not an essential criterion.

    I think the OLPC will be dead soon. Good product design or not.

    P.S.: Thanks for the Christmas card! It came today and made me check out your site. I haven’t been here since 1998 or so 😉

  8. The XO always seemed like a solution in search of a problem.

    Maybe that’s why they designed the Sugar UI as a problem in search of a solution.

  9. I thought this was a telling article from the BBC about some pretty basic failures. But of course people are excited about taking pictures …

  10. Alpchild Avatar

    The XO, most importantly, is a statement of support for a unique solution to global poverty: investment in knowledge.

  11. The skull and cross bones is one of the few (if only) pictograms that operate well in cross cultural settings. In surveys of people from various cultures (I cannot put my hands on the exact reference at the moment) it consistently conveyed negative connotations hence its widespread usage on warning signs, particulalry in developing countries where mines are common.
    Perhaps the pictograms on the OLPC screen mean, “Warning, interaction danger ahead!”

  12. Isn’t it more important to get potable water to children in third world countries than laptops? More drinkable water please! One third of the world does not have potable water for crying out loud.

  13. Glad I got to read this. Sad it took me so long.
    I’m no UI or UX expert. I’m an ethnographer (research on West African music) and teacher (university-level but influenced by constructivist pedagogy).
    Ever since the project gained widespread exposure, I have been looking for thoughtful comments about it. Critical thinking. Insight. Honest discussion. But everything I found seemed to be either an appeal to authority (as you note), praises for Negroponte’s vision, or some of those misinformed statements about “oh, if Africans could be more like us!” I probably became the most frustrated when I heard Negroponte say that criticizing (his brainchild) is silly since it’s a “non-profit effort.”
    Thankfully, not all architects think like he does.
    So, I was relieved when I first read comments on and elsewhere which gave the project a thorough analysis. Not analysis for its own sake. Not nitpicking. But calling a spade a spade. And debunking the “humanitarian” angle.
    I care deeply about some parts of the project. I think it’s quite sad that it should be imposed from the outside, but I think people do empower themselves through technology.
    About this blogpost specifically, I like:
    The mention that constructive criticism has long been silent (silenced?).
    The recognition that many praises are obvious appeals to authority.
    The acknowledgement of how arrogant a design project without user research may be.
    The cultural awareness displayed by the comment on signs for landmines.
    Belated thanks for a very useful post. Now that “OLPC as we knew it” is officially ending, it can serve as a warning for future projects.