Submitted by “Red Hat has recently shared with the world the first ISO images of the system that is supposed to be installed on the OLPC laptops. I suddenly felt an irresistible temptation. I downloaded 291 MB ISO, burned it on a CD and started testing. Here is what I got.”
I agree with the main points of the little article, apart from the bit about installing KDE, that’s just daft on such a low spec machine.
Tool tips should not be too hard to do (nice clunky semi-transparent ones should do the trick, and a simple file manager as well.
I’m actually quite impressed with SugarUI–it’s something radically different, and judging from screenshots, it should suit its target audience very well when it’s finished.
I think a OLPC would run XFCE nicely. I could be completely wrong on that, but it is quite usable on an old 500 MHz Celeron running Xubuntu I have laying around.
If it winds up anything like the original BeOS, Haiku would probably be very snappy on a OLPC.
I could be wildly incorrect, but my impression is that they tried to eliminate text as much as possible because either they don’t expect a high level of literacy amongst the users and/or that internationalization would have been somewhere between arduous and impossible to accomplish across some parts of the developing world. I’m sure the kids are just as smart as in industrialized nations, but they might not be as capable of interacting with text-based interfaces. Once again, I might be drastically underestimating the literacy and language standardization in these parts of the world.
I wonder what constraints and guidance Red Hat’s designers might have been given by social anthropologists and regional experts. I’d like to see a paper on the design considerations. In fact, I’d be fascinated to read an analysis of how the OLPC believes the target audience would likely use these computers.
The OLPC team has accomplished the glamorous yet relatively easy part of bringing their vision to reality: manufacturing a capable portable computer than can be sold for roughly $100. The hard part is making these devices useful. Thanks to FOSS, the software can evolve organically to meet the diverse needs of its users. I’m not convinced that SugarUI is the best seed for this process, but I’m not convinced that it isn’t either. I think they need to introduce these into a few pilot communities and gauge the reaction.
http://wiki.laptop.org/go/OLPC_Human_Interface_Guidelines
And a glowing review of it from Mike Hearn (of Autopackage fame) here:
http://plan99.net/~mike/blog/2006/12/10/olpc-is-genius/
And a not so glowing review here:
http://www.90percentofeverything.com/2007/01/25/whinging-about-the-…
I knew it (the “Sugar UI”) would suck…why re-invent the wheel when there are other projects out there (KDE, *Box etc.) which have been doing it for over 10 years? This is wasting resources for nothing!
I think the OLPC developers believe that KDE, GNOME, et al. are too complicated and hard-to-learn for its target audience, third-world children that’ve probably never used a computer before.
IMHO, it’s well suited (and already impressive) for who it’s intended for, though probably not for your average OSNews reader.
Edited 2007-04-10 22:40
I think people underestimate these target children…
So instead of providing them with a standardized interface, we give them a custom one with “mystery meat” navigation (no tooltips to an icon’s function) and a crippled browser (no url entry field).
So what if a standardized interface is more complex, these kids are going to have these computers for months-years potentially, leaving them stuck with a dumbed down interface for the many years they want these computers to last isn’t going to help much.
Oh, and what’s with the GTK theme they’re using (in the abiword dialog) is that sucktastic 0.1 or something? GTK is system-wide themeable, why didn’t they spend the one afternoon to make a theme to tie the GTK pieces into the Sugar pieces? Hopefully they’ll do that before releasing it to the world… These things may be without internet connection for months to years at a time after deployment, the initial setup should be solid.
Since I’m griping, I’ll add one more: saving as Doc instead of ODF by default?! WTH? And no terminal for when the kid does advance to the point when they want to program…. And no file-browser whatsoever?!
The GTK theme is likely the way it is because of the screen’s special nature.
http://www.manucornet.net/pub/olpc/theme_and_display/
ODF would be nice but let’s face it, .doc is the most compatible format with the rest of the world at this point. And Abiword’s ODF support isn’t really up to snuff at this point, AFAIK. Maybe (hopefully) they’ll change it in future releases.
Edited 2007-04-11 02:49
Ok, and why not RTF, that EVERY word processor supports perfectly?
It’s a $100(ish) laptop. I think we can drop the R and just run straight up, plain text.
I think that’s the cop out answer. I started computers on a Commodore 64 and Windows 3.1. I was coding 8-bit assembly when I was 12. Children can absorb even the most complex interfaces and devices suprisingly easy. I know 7 year olds that have completed Doom.
For children to learn, you shouldn’t treat them like children
I think you guys are partially missing the point…these tools are not designed to teach them about computers, but to act as learning aids for general knowledge. The goal isn’t to create a generation of underage third-world hackers, but to help their education *overall*.
I wouldn’t worry about kids somehow being “warped” by using an alternate UI, either…they’ll probably have an easier time learning other UIs after this one than if they had never had any contact with a computer.
Those few computer whiz kids in there will find ways to install another system on the machine (probably Linux, as I don’t see MS releasing a version of Windows for it…) and go from there.
Computers are part of any well-rounded educational curriculum int he 21st century, and these children should be taught the skills needed to use the standard GUI interfaces that thier peers are using. If not, they may be at a disadvantage if the only computer experience they have is with OLPCs.
I disagree. Learning an interface will not prevent you from learning another one, or even put you at a disadvantage when you do.
Case in point: computer games. They all have different interfaces – some quite complex – and yet people (kids, especially) navigate them without too much effort.
In any case, at the risk of repeating myself, computer skills are *not* the primary focus of the OLPC.
Here’s a novel idea: why don’t we wait until the project is actually launched and tried out before saying it’s a failure, hmm?
I never said it was a failure, I said that the UI is a mistake. I may be proven wrong. I may be proven right.
It is not when these children are growing up that the OLPC interface will be the problem, it is after they are grown up, and thrust into say, a university environment and all the other kids have used Windows/OS X/KDE/Gnome all their lives, they will have to take the time to learn the more complex and richer interfaces. This will distract from the task at hand, education. if the OLPC is the only computer a student has used, they will be at a disadvantage.
For 6 years I worked at a university in IT. We had an institute that had students from developing countries come and stay for a year at a time for various development, social, community and infrastructure building/planning curriculum. Most had trouble adjusting to the more advanced technology at their disposal in Canada. Most of them had used computers before, but the changes in interfaces and functionality led to some very big hurdles for them to overcome.
I don’t speak out of my ass on this particular topic, I have seen it first hand. All most all of the students caught up sooner or later, but they struggled with it. And these students were not competing for grades, scholarships or co-op placements. The same thing can happen here, and I believe it will
This is where we disagree…I think it’s still better to have used the OLPC with Sugar than no computer at all.
You say that students from developing countries had trouble adjusting to the more advanced technologies, but that doesn’t mean that those who had used other UIs (which would have to be DOS or Windows 3.x to represent a major difference) were at a disadvantage compared to those who had *never* used a computer.
I guess we’ll have to wait and see. I disagreed with the UI at first, but I now I think it can prove to be rather useful (not unlike the UI of a palmtop PC).
“This is where we disagree…I think it’s still better to have used the OLPC with Sugar than no computer at all. ”
You are certainly right there, any computer is better than none.
“it’s well suited (and already impressive) for who it’s intended for, though probably not for your average OSNews reader.”
I find this kind of thinking almost insulting to the target audience. Are kids in 3rd world countries stupider than 1st world kids and can’t handle Window/OSX/GNOME/KDE? That’s bullshit. Any street kid can learn to use a computer regardless of OS and I seriously doubt that it’s easier with a dumbed down interface. If it really is, why not target all kids?
I guess it’s a miracle we have computers at all today seeing as my generation learned computers using C64 basic, assembler and such, eh?
They are not stupider in the sense that they are less intelligent; however, they are less likely to be literate, and AFAIK, SugarUI focuses on removing text as much as possible.
Obviously, the lack of text makes it easier for someone illiterate to learn to use (and a lot of these kids may even learn to read on their OLPC, contrary to most first-world children who learn to read somewhat before using a computer or have mommy/daddy help them). Flooding a visual child with (verbal) text is difficult (not to mention discouraging) for them–it’s similar to giving a technology-ignorant person with no interest in computers one without an OS or documentation and saying “Figure it out”.
SugarUI’s textlessness also makes translating it into other languages much easier. Translating a UI into a multitude of developing languages (and not using any terms, but kid-friendly ones) would be horrendously difficult.
I guess it’s a miracle we have computers at all today seeing as my generation learned computers using C64 basic, assembler and such, eh?
It most likely took you a little while to learn BASIC and ASM, and I’m almost certain you knew how to read when you learned them. The OLPC is intended for its children to be able to “just use them” in no time.
People are missing the point of OLPC. It is not designed to be “just another computer”–it is designed to be a learning tool above all else. It’s true that it will likely foster an interest in computers amongst its users, and if these children are as intelligent as their first-world counterparts (and they are, like you said), then learning another, more conventional UI will be easy to them.
Edited 2007-04-11 07:52
They are not stupider in the sense that they are less intelligent; however, they are less likely to be literate, and AFAIK, SugarUI focuses on removing text as much as possible.
This makes no sense at all. If they are not literate, what’s the purpose of a web browser or a word processor on that machine?
If they are literate enough to browse the web and to write documents, they are surely literate enought to read text tooltips, for example.
Literacy is not required to use the camera, drawing, or ad-hoc social networking (because it uses the aforementioned), nor is literacy required to play the simple games that come with it.
It’s designed to grow with them; eventually textual applications will be useful for them as they learn.
Edited 2007-04-11 10:31
There have been other reviews of Sugar, and I haven’t read one overly positive one yet.
While I do not believe any effort to reinvent is in vain, for if nothing else you learn what makes the originals work so well.
But with Sugar clearly lacking compared to almost every other WM, why is it seemingly being pushed… forced.
I’m afraid that if Sugar is clearly not good, it will still be used. Meanwhile, a few custom text files in any of the other WMs could achieve almost everything Sugar has without issues and serious lack of polish or features for that matter.
Most of these other WMs have been in heavy development for a decade or more. Solid as a rock, polished to a high shine, and easy to configure to meet almost any need.
I can see the need for customizing the applications, and tying it all together with a common theme and easy to understand interface. By the videos, the reviews and the details, Sugar lacks any feature that another, more stable, already developed, already tried-and-true and heavily tested WM doesn’t already or couldn’t easily offer.
Sugar, by any other name, would taste just as sweet? Not this Sugar.
Perhaps it’s cynic in me, but I think it’s less a case of “learning what made the originals work” than “learning the pragmatic reasons behind the compromises taken by the people who built the originals.”
They’re probably just throwing Sugar at the kid to see what comes out, maybe we’ll see some new generation that thinks totally different because of this. Who knows, if they get obsessed with their computers they might become like 500% more productive in Sugar than an average person is in Gnome or KDE…
“They’re probably just throwing Sugar at the kid to see what comes out, maybe we’ll see some new generation that thinks totally different because of this. Who knows, if they get obsessed with their computers they might become like 500% more productive in Sugar than an average person is in Gnome or KDE…”
Or it may be the other way round? Kids trying to get rid of what they may think is too dumb for them and installing Gnome or XFCE on the OLPC? 🙂
I’d like to see a GUI not being statically, but developing dynamically such as the intelligence, the experience and the education of the child is increasing.
why not simply have just used the matchbox window manager. would have been a better choice in my opinion
http://projects.o-hand.com/matchbox/
Ok – so I’ve used a few OS’s in my time including a number of Linux distributions and I’m 25, so I’m not one of these oldies that it takes me ages to get the hang of new fangled contraptions.
So – as I have a passing interest in the OLPC, I thought I’d take a look at what millions of children all over the world will be working with. And I for one am REALLY confused. How the hell do you quit an app without moving the mouse to the top to reveal some random circles that are not listed. The fonts in the painting application are too small. If the specs of this machine are so low, then why if you switch to regular Gnome are there desktop wall papers taking up disk space?
Oh I don’t know – perhaps these kids have more brain power than I…I just think the whole UI concept is strange.
So – as I have a passing interest in the OLPC, I thought I’d take a look at what millions of children all over the world will be working with. And I for one am REALLY confused. How the hell do you quit an app without moving the mouse to the top to reveal some random circles that are not listed.
I strongly suggest to check the OLPC keyboard so you can figure out how to do it:
http://wiki.laptop.org/go/OLPC_Keyboard_layouts
Doing on the actual XO-2 machine (yes, I have the B2test OLPC) is easy as (x) button is the equivalent of ESC and Alt-F4.
The fonts in the painting application are too small. If the specs of this machine are so low, then why if you switch to regular Gnome are there desktop wall papers taking up disk space?
Remember these activities are still in progress. XO-2 machine runs at 1200×960 AFAIK.
They should have gone with a modified exsiting window manager. I prefer Mezzo myself, since it has no menus in itelf, uses a developed WM, and is fairly intuitive.
Really kids could learn any UI if given enough time, its the machine specs that have to be devloped for.
The author compares the compatibility of SugarUI with advanced cell phones: “The capabilities of the applications that come with OLPC aren’t much better than those of modern cell phones.” Anyone can concur with the author on this point.
However: “In the not-so-poor countries like Mexico or Brazil, when the kids usually have some idea about computing and High Tech, OLPC may have problems with being usable enough.” I believe making a comparison in the Third World is misleading because it doesn’t take cost into consideration. It only takes features into consideration. Although OLPC is $100, a “modern cell phone” is probably around $500, not including the cost of cell phone bills. That makes advanced cell phones economically out of reach. Even if they weren’t, cell phones do not have as nice as a user interface as a laptop for word processing or web browsing.
Even if many of these kids might never actually earn enough to buy a high end smartphone, they’ll probably know what one looks like and what one can do, and use that as a comparison. And most high end smartphones come with a basic word processor and spreadsheet, web browser, mail client and a could of other basic productivity tools.
You have got to be kidding me! Have you ever walked through a shanty town? Any body who has electricity is king in a shanty town. Many of these kids don’t even know what a smart phone is, never mind what one does.
Sure, they may have seen a poster advertising Nokia’s latest but that does not mean they can read the text, never mind the specs making any sense to them.
Dude, no offense, but what you said has no basis in reality when it comes to most third world countries.
I think it is important when discussing world economics that the third world is not the only place where poverty pervades. Due to the highly unequal economic distribution in the US, there are many places in this country where OLPC is a wonderful idea.
Well said, sir, and I congratulate you on your perspective. Thank you for putting the issue into a wider view 🙂
the easyest way is to download livepc, http://www.moka5.com/, then go to, http://www.moka5.com/livepc?page=1, and download one laptop per child.lpc file (it works like a torrent file),oh and dont forget to click the “let me use this livepc offline” button,i am uploading a screenshot now that shows you what to do. WARNING: the image file that livepc uses only works with livepc! and i made a quick tutorial to show what i mean,
http://img511.imageshack.us/my.php?image=livpcpd5.jpg
My son (8 years old) is perfectly at home on Windows, OSX and his personal pc runs Debian Linux and yes he even uses the shell (“the menu takes too long dad”). Most of you are basing your opinions of the Sugar UI on what YOU expect it to be like. Kids ‘just get it’ when it comes to new stuff like this. Coincidentally, some of the best people to teach about Linux (or other alternative OS’s) are those that do not have a pre-conceived notion of what a desktop should behave like. Ask a long-time Windows user how to stop a Windows machine. It’s pretty comical when they realize that the answer is “click Start”
Certainly we do not wish to keep pushing the same old boring desktop standards at the next generation when the reality is that Sugar is only the beginning of a radical change that’s long overdue in the desktop UI department. Bring it on!
“Most of you are basing your opinions of the Sugar UI on what YOU expect it to be like.”
So what you all are saying is, whether 3rd world kids like the UI or not, at least we’ll finally have a Linux UI that’s useable by the corporate types.
I had a really good read on this, very detail, and very useful information.
Thanks.
Flv to avi
http://www.avi-converter.net
Checking the extra screenshots, I notice one thing.
It use Squeak Smalltalk enviroment for some of its graphics programs!!!!
from http://www.squeak.org/ :
“Children’s Machine – The One Laptop Per Child (OLPC) initiative leverages Squeak to power the Etoys application shipped with every laptop around the world.”
http://wiki.laptop.org/go/The_Children‘s_Machine
Nice machine to hack
Basing my critique on the information presented by this article as well as the interface section of the OLPC wiki, my opinion of SugarUI as a user interface to be not quite satisfactory. I don’t agree with some of the design decisions that this project has chosen but it seems somewhat better than traditional Window, Icon, Menu, and Pointer (WIMP) GUI style that we are all used to.
One example is that I don’t agree with SugarUI’s pervasive use of a sole pictorial icon to represent a particular task, command or modal state. I understand the purpose of minimising the use of text (the intended audience may be illiterate) but I believe this to be bunk. The user would still require training for each icon so there may as well be text to convey more information for people that do know how to read.
I like how the two toolbars are at the top and bottom of the screen. This makes it really easy to target with a pointing device. One thing that I really like according to the wiki is that data will be automatically saved by the system. There aren’t many reasons in this day and age why the computer doesn’t automatically save the data that the user cares about.
This interface is intended for children for general purpose learning and may be “too simple” for “Real Work” (TM) but the paradigms that this project is proposing may very well be superior than what we have today. I don’t like this current implementation but I will keep my eye on this project as it looks interesting.
By the size (291 of however many limited MB this thing has) and inutility of this UI, it sounds like the OLPC product would be better served by using Damn Small Linux or Windows CE.
Edited 2007-04-11 17:53 UTC
I meant to make the title of the above post SugarUI = “Microsoft Bob” for Linux?. Apparently, post title editing isn’t quite working yet.
…. and it was great fun. A trully easy interface to use. Nice, clean and simple. It could definately use some color and sounds, but other than those 2 small things the interface is great. I can’t believe some people in this thread, saying that they are trying to “re-invent the wheel” with the SugarUI. Thats just nonsensical. The SugarUI was built from scratch with education in mind. It is designed to get out of the users way. There really is nothing to the UI at all… it is just a basic application launcher. There is no window management (well sort of… the small wheel in the centre of the screen is kind of a task manager/window chooser). The interface just moves completely out of the way and gives the user full interaction with the application/activity that they are using at that exact moment. Its perfect.
In my opinion, with a few modifications I would use SugarUI as a desktop replacement on my Linux Box, only because other window managers are becoming too concerned with features and less about ease-of-use.
SugarUI is simple, easy to use, easy to learn quickly and never gets in the way of what you are doing. I sugest everyone gran the LiveCD and a copy of VMWare Player (free) and try it out instead of reading critical reviews/previews and thinking you know all.