When we were discussing XO laptops shipping with Fedora or Windows XP, but not Sugar, the signature user interface developed by One Laptop Per Child, I was amazed by the number of people who thought Sugar suboptimal. It seems they were not alone in their option. Nicholas Negroponte has expressed the same idea to the Boston Globe:
Some One Laptop critics said developing Sugar was like reinventing the wheel. Today, Negroponte admits the critics were right. "In retrospect, it wasn't necessary," he said. Last year, he disclosed an agreement with Microsoft to bring the Windows operating system to the XO laptop.
Getitng crazy with Sugar
I have to disagree with Negroponte. Sugar was an OLPC innovation up there with a dual-mode screen and mesh networking. By being child-centric, Sugar helped define the XO laptop as an educational device, a tool to empower children to learn learning, not a "$100 laptop". Sugar was almost exclusionary to adults used to windowing interfaces, an added bonus to keep the XO in a child's, not parent's hands.
Had the XO shipped with Windows XP from the start, it would've faced even more false criticism that it was "underpowered" or not "fully functional". In addition, while adult-centric software can be great for office automation, that's not the OLPC's goal. Its education, and as such OLPC need an educational experience for its end users. Or as Walter Bender says
"Just giving kids a traditional desktop, whether it's a Linux desktop or a Windows desktop or a Mac desktop, that's inadequate."
Now Sugar has its issues. Sugar is woefully underdeveloped, but that's not a fault of Sugar itself, only the resources and strategy applied to it. Sugar also could've had more administrator-friendly applications, the ability to print, and maybe even an Aquatic Sugar option for adults.
But are those drawbacks really enough to consider it a mistake?
Stay current with Sugar: subscribe to OLPC News via RSS Feed, Emails or Twitter.
My answer to criticism of the desktop metaphor is that when I was a kid I was introduced to the black & white Macintosh and found it to be very easy to use. Of course, with kids using the machines, various files tended to move around or disappear, but I don't see any reason why kids couldn't understand the desktop metaphor, even if they had never heard of "files" and "folders" before. The Mac, unlike Windows, has visual clues to what is happening, like when you double-click a folder icon and an expanding box emerges from that icon to become a window; I think these clues are sufficiently helpful for learning to use the computer.
It seems apparent that a UI for kids needs to have some safeguards to prevent the kids from permanently messing up the computer or accidentally deleting their own data, and some mechanisms to help kids find saved data (since kids lack organizational skills), but such mechanisms would help any computer-illiterate individual. Therefore, why not incorporate these features into an adult operating system so that not just kids, but geriatrics and other new users could benefit?
I just don't get why a UI for kids has to be totally different from an "adult" UI. Okay, so kids don't need "office automation", whatever that is (I've heard that term hundreds of times and I still don't get it.) But that just implies kids will run different programs than adults, not that the desktop/windows paradigm has to be completely replaced.
Could someone point me to a document that explains how sugar works (since I have no opportunity to try it) and/or justifies the differences in UI between it and adult systems?
David,
I consider Word and Excel to be great office automation tools. I can develop reports, calculations, and entire businesses based on those two tools. In fact, most of OLPC News was composed in Word.
But for a primary school child - think 4-10 year olds, a blank Excel sheet does not inspire excitement. TamTam, Moon, TuxPaint, and especially Speak and Record - now these are down right captivating.
Oh and you can always try Sugar on regular computers - from Mac to PC to Sugar on a Stick.
In theory Sugar was not a mistake. In practice, it wasn't nearly done by the crucial rollout of XO-1, and it was apparent months ahead of the release that it wouldn't be nearly done in time. Thus, a country would have had to be crazy to commit to a large-scale deployment. In practice, it has been a huge hindrance to OLPC.
However, given enough time and support, it could still be a great success in the future. The ideas are the right ones.
Yes, it was definitely a mistake.
I have never noticed kids having a difficulty learning computers or any other new technology. It is adults who have problems with this.
David, you are right that kids like the fun and games on the olpc more than Microsoft Excel, but this is an issue of what software programs are provided, not the desktop environment.
I have done some development on the OLPC and Sugar was a huge impediment to actually getting anything done. At one point to download a file from the Browse activity to actually use it in a program, you had to hop down to the command line, navigate to a certain directory, find a file with your download time and date, but some crazy hashed name, and unzip that file. These directions make no sense, for adults or for kids, and are reasons I stopped using the Sugar activities.
It is not the Sugar interface itself that is the problem, but the "Journal" providing some bizarre parallel file system. Sugar didn't reinvent an OS, but it did try to reinvent the filesystem and clipboard in ways that hindered development.
This is a problem for two reasons: 1, this platform was originally supposed to be friendly to software development by children, but the interface between the system where their files "lived" and that needed for programming is far too complex to tolerate. 2, make development difficult, and you drive away adult developers into other systems, which means they will spend less time developing in and for sugar apps.
Of course, XP is another mistake, they should have just used a good linux distribution, possibly with a somewhat specialized graphical shell or window manager.
Anyway I am probably one of the few North Americans who has used an OLPC in "the wild" for any significant period of time. I wanted to write a journal on it, but it was unclear how I could transfer any journal entries from the Write activity to another computer to upload them, so I ended up writing my journal all in vi. I think this is telling. I upgraded to ubuntu as soon as my five months outdoors were over and have not looked back.
You can read an account of my experiences using the OLPC at One laptop per camper. This was written early in my trip, but I had already been using the laptop for some time.
I love the hardware ideas in the computer, even the keyboard has grown on me (when no keys are sticking), but Sugar never did. There is still a good chance I will pick up an XO-2 if it comes out, but I expect I will again install ubuntu or another ordinary linux distro in place of SUgar.
As another North American who has used his XO for software development and also extensively on vacation trips for word processing and handling photographic images, I agree with your assessment: Sugar as the desktop isn't the problem; the Journal as the filesystem is. Journal needs an "escape hatch" to expose the linux filesystem -- this has been discussed/proposed but still nothing concrete exists.
Then again, having a chronological sort isn't a BAD thing, too.
I wind up dropping to terminal for everything to do with the filesystem. I use vi for code development. I start Totem from terminal. I run PySol from terminal. Otherwise, I use Sugar to launch apps and debug my code. And Abiword for word processing is fine by me. I've made really minimal changes to my G1G1 platform except to keep it up to the current official build. And I still like it.
The problem is not with kids being unable to "learn computers". The problem is, what they learn is wrong.
They learn that text files don't have strict sequential structure, that text has fonts, color and layout as inherent attributes.
They learn that everything you have stored on a computer can be "opened" (actually executed) by clicking on an icon, and there is no difference between programs and data.
They learn that computers are fussy little creatures given by almighty companies, and that they can be used safely if you follow some rules -- except nothing seems to work, and computer eventually becomes compromised.
They learn that it's OK to use compromised computer because it "may be a virus" if it is for some reason slower than expected.
They learn that computers can only be customized by going into a nice menu and selecting a background picture from a list.
They learn that computers have absolutely nothing to do with math or logic. Actually they don't even learn what math and logic are in the first place.
But they learn an IMPORTANT SKILL of printing pretty reports in MS Comic Sans font.
Hi Teapot,
From the way you opened your comment, I'm not sure to what extent you intend it as disagreement with my post. Most of the issues you name seem to apply to both Sugar and Windows, both of which I thought were flawed.
Changing the subject, I personally wish every child had a chance to learn programming on the C64 like I did. Great sound and video, BASIC built into the OS, and PEEKing and POKEing raw bytes to nearly any address in memory. Maybe I'm just showing my age, but that seemed like a great system for grokking what computers actually do.
The common mistake a lot of people make is thinking people care about programming. It's the "For us, by us" that's common in open source software. The vast majority of the world doesn't care. They are more interested in having a tool that isn't a steamy slow pile of crap(Sugar).
And even with Sugar, the "see every source" thing doesn't quite work, the code is awful and uncommented, development process is ridiculous, APIs keep changing.
Sugar is simply the essence of the XO.
I presented the XO to African children. Children who do not have had any contact with any forms of computing. I can confirm you that Sugar is really intuitive. Only it is not a question to initiate the pupils with data processing and with the programming as Teapot mentions it, the purpose of Sugar is to offer to the pupils a metaphor of participation where the collaboration and the creation of contents are prevalent.
Sugar is not a simple graphical interface user; it is a learning environment which gives access to the pupils to tools helping them to learn in authentic contexts and practices. They have at their disposal activities enabling them to record videos, to measure distances, to write texts in collaboration, to make research on the Internet, to program, etc.
Of course, the Journal must be improved but it remains nevertheless that it belongs to a metaphor of a learning environment. Sugar didn’t reinvent the wheel, it simply modified it to fit in a better context.
Had the XO shipped with Windows XP from the start, it would've faced even more false criticism that it was "underpowered" or not "fully functional"
Sugar running on a VM on a dual core machine runs worse, and uses more resources than WindowsXP on a 10 year old piece of crap computer. All sugar applications... sorry "Activities"... have an alpha feel. I have no idea how OLPC managed to create such a nice piece of hardware, but frig up the software so much.
I have to concur; Sugar is a drag on the performance of the machine and the kids here see it as a launcher and don't use the Journal very much. Would have been better to go with a proven OS and interface that had some better HCI.
Wayan, I think that "mistake" is too harsh and too broad a word to use. I think that given the timeframe required and the resources available, attempting to put Sugar on the XO was overly ambitious.
The XO was supposed to be an "Educational Appliance"... like one of those VTech toys but orders of magnitude more capable.
The fact that it was running Linux should've been a footnote for geeks...and something of a hook for them to tinker and explore.
But because the XO went out the door with an incomplete Sugar (stylus areas are still not supported to this day), it caused many to "peek under the covers" and have to deal with its Linux underpinnings.
All of this assumes that the Sugar UI actually solves a problem. I don't believe that it did, but reasonable people can disagree on that point. The Activities however are another story. I believe that they do a fantastic job for their target audience.
Sugar was the biggest mistake of OLPC and a catastrophic failure.
If OLPC simply had taken the approach of adapting a simple, desktop distro and write the collaborative applications in GTK or QT, they would have fared much better.
They wasted an awful lot of resources on Sugar. Their development cost could have been 1/4 without Sugar. They actually could have finished most of the initial work for the original G1G1 rollout, not to mention the G1G1/2008.
A lot of developer would have been able to collaborate with them. Instead, most of them didn't even start thinking about doing anything in Sugar seeing that they need to learn a completely new programming environment to help (Python + Sugar libs).
IF OLPC had launched the XO with _either_ functional XP or a Linux distro, at the initial ~$175 price point, they would have had the 10 million orders that Mr Negroponte was demanding in 2005--without field tests, without question.
Govts are hot to fund technology buys for schools (for reasons both licit and illicit). Sugar was/is a cool idea, it certainly was an impediment to large-scale adoption.
Developing-country governments are concerned with job skills and employability; it's very easy for policymakers and administrators to equate these with skills in the use of office-productivity software.
I have to agree with other commenters that Sugar was the biggest mistake of the One Laptop Per Adult effort. There's simply no reason why anyone over the age of about twelve would ever want to use it.
@Ashley Yakely:
"There's simply no reason why anyone over the age of about twelve would ever want to use it."
Is this really intended as sarcastic as it sounds?
Because the point of the XO was that it should be used by 6-16 yo. Which would make Sugar a good decision.
Anyhow, if you look at the "real" modern XP offerings on 7" netbooks, they are not that different from Sugar. No desktop etc.
XP only differs from Sugar in that it completely lacks any security, backup/journaling, and will completely trash the flash memory within a year.
Furthermore, XP was NOT AVAILABLE for the XO until MS saw that they were missing the boat. Remember, it was Vista or the highway.
So, with XP in mind, the XO could not have been developed in the first place. No netbook would have existed if MS would have had it their way.
Winter
@Winter: Good grief, so full of FUD.
Because the point of the XO was that it should be used by 6-16 yo. Which would make Sugar a good decision.
No. I'd say it's actually suitable for up to an ge of about 10. Unsuitable for 16 year olds.
Anyhow, if you look at the "real" modern XP offerings on 7" netbooks, they are not that different from Sugar. No desktop etc.
Are you talking about the "quick menu" on the Classmate? That's just a quick start menu still running on a standard windows file / start menu metaphor. You can still see the standard taskbar. Far different than Sugar which is a completely different. Other '7" ' netbooks with XP would be like the Eee 701 which has a bog standard XP install. Same run of the mill desktop.
XP only differs from Sugar in that it completely lacks any security, backup/journaling, and will completely trash the flash memory within a year.
Complete FUD. WinXP can be set up with limited access users, and ACL. That's security. Out of the box? No, but in an educational environment someone would have to deploy these machines, just like how people are supposed to be deploying XOs (which is another area OLPC is failing).
Do you mean lacks journaling in file system or the sugar journal? NTFS does have a journal, has had it before ext file system and NTFS is generally more robust against improper dismount than ext file systems.
Trashing flash in a year is also completely false. Lots of people with EeePCs running Windows XP for over a year without a problem.
Furthermore, XP was NOT AVAILABLE for the XO until MS saw that they were missing the boat. Remember, it was Vista or the highway.
XO there wasn't much of a boat to miss. Asus has shipped more "ULCPC extended license" XP machines than OLPC. By orders of magnitude.
@John Smith (alias Smith and Jones?):
FUD?
"No. I'd say it's actually suitable for up to an ge of about 10. Unsuitable for 16 year olds."
And you have what evidence from developing nations for that?
@John Smith:
"Are you talking about the "quick menu" on the Classmate? That's just a quick start menu still running on a standard windows file / start menu metaphor. You can still see the standard taskbar. "
No I am talking about this and all the other offerings that hide the desktop. And I do not see how a task bar is different from a ring under Sugar.
@John Smith:
"Far different than Sugar which is a completely different. Other '7" ' netbooks with XP would be like the Eee 701 which has a bog standard XP install. Same run of the mill desktop."
That is not the Eee I saw with a Linux without "run of the mill desktop". And none of these has been targeted at children.
@John Smit:
"Complete FUD. WinXP can be set up with limited access users, and ACL. That's security. Out of the box? No, but in an educational environment someone would have to deploy these machines, just like how people are supposed to be deploying XOs (which is another area OLPC is failing)."
A secure installation of Windows XP? Never seen nor heard of it. And if your solution is to remove all rights from the users, you just removed all of the flexibility from the notebook and added a horrible amount of support. This alone would double the price of deployment.
And before you even try to make a point about comparing security under XP and under sugar I would advice you to read about Bitfrost. Even MS cannot secure their own XP installations and networks.
@John Smith:
"Do you mean lacks journaling in file system or the sugar journal? NTFS does have a journal, has had it before ext file system and NTFS is generally more robust against improper dismount than ext file systems."
The Journal under Sugar has nothing at all in common with a "journaling file system". Nor am I very confident in NTFS' stability and robustness.
The Journal is a versioning file system. Such a file system keeps older versions of files. There is much more to it, but you can find that elsewhere.
@John Smith:
"Trashing flash in a year is also completely false. Lots of people with EeePCs running Windows XP for over a year without a problem."
Now that is new information if this is heavy use. However, the XO has bare flash, while the Eee has a controller running over it to handle the wear leveling. Running XP on bare flash will thrash it really fast. But if you can show me wrong, please do.
And while everyone and his dog are yelling that XP runs fine on flash memory/Eee, I cannot find a single XP-on-flash offering in the shops. Wherever I look, it is XP on 160GB HD. And even the XP-on-XO is on external flash with wear leveling build in. Not a native XP runing on the installed flash memory.
And all the XP offerings are 25-30% more expensive here too.
@John Smith:
"XO there wasn't much of a boat to miss. Asus has shipped more "ULCPC extended license" XP machines than OLPC. By orders of magnitude."
Yes, after the XO was a reality and Asus started to ship a lot of Linux machines. So only AFTER a market for netbooks was established on Linux was MS allowing (scrambling) to get XP on these machines.
Heck, Intel did not even HAVE a low power processor for netbooks when the OLPC started.
Winter
@John Smith (alias Smith and Jones?):
Only other name I go by here might be John if I foget the last name
And you have what evidence from developing nations for that?
What evidence do you have that sugar is well suited for 12-16 year olds?
That is not the Eee I saw with a Linux without "run of the mill desktop". And none of these has been targeted at children.
Has nothing to do with your quote about 7” netbooks having no desktop. I’m telling you EeePC 70x with windows on them run bog standard windows. 7” netbooks running windows still have the desktop.
A secure installation of Windows XP? Never seen nor heard of it.
You must be the “oh I install random shit on my computer and blame other people like microsoft when it fucks up”
And if your solution is to remove all rights from the users, you just removed all of the flexibility from the notebook and added a horrible amount of support. This alone would double the price of deployment.
And Sugar is impervious to people breaking the software? Especially when it’s supposed to be designed to let people play with software? Is that just a feature for free software?
The Journal is a versioning file system.
Full stop, so that’s what you meant. That’s all I was asking, not an indepth discussion
Now that is new information if this is heavy use. However, the XO has bare flash, while the Eee has a controller running over it to handle the wear leveling. Running XP on bare flash will thrash it really fast. But if you can show me wrong, please do.
And while everyone and his dog are yelling that XP runs fine on flash memory/Eee, I cannot find a single XP-on-flash offering in the shops. Wherever I look, it is XP on 160GB HD. And even the XP-on-XO is on external flash with wear leveling build in. Not a native XP runing on the installed flash memory.
And all the XP offerings are 25-30% more expensive here too.
You’re making a big deal over wear leveling. So many flash devices use hardware based wear leveling... big deal... it’s not expensive to implement, and it gives you flexibility as to what you can run on it. Correct me if I’m wrong, but aside from the default sugar, every other GNU/linux distro running on the OLPC is via SD or USB. Which “ZOMG has hardware wear leveling!!!!!!”
As far as the Eee, I’ve been running XP on my 4GB flash 701 for 8 months no problems, and there’s tons of people over on Eeeuser with original Eees that have been running winxp on the SSD no problems. Many people have bought xandros versions and installed Windows on them (and vice versa bought windows versions and put GNU/Linux on it), why you’re focusing on what’s being sold by retailers I don’t know, since there’s actual users you could get data from.
Yes, after the XO was a reality and Asus started to ship a lot of Linux machines. So only AFTER a market for netbooks was established on Linux was MS allowing (scrambling) to get XP on these machines.
...But the actual demand is in netbooks (a la EeePC, Aspire One), and they have significantly outsold OLPC (in both linux and windows forms) because they sold to end users.
Heck, Intel did not even HAVE a low power processor for netbooks when the OLPC started.
And OLPC still isn’t meeting the initial gols of battery life in days not hours (even in ebook mode) with battery life comparable to other computers, and they have yet to release alternative charging devices. Meanwhile the underclocked Geode is so anemic it can’t get out of it’s own way for Sugar.
@John Smith
"What evidence do you have that sugar is well suited for 12-16 year olds?"
They can use it whenever offered?
@John Smith
"Has nothing to do with your quote about 7” netbooks having no desktop. I’m telling you EeePC 70x with windows on them run bog standard windows. 7” netbooks running windows still have the desktop."
Yes, but why would you want to? Is it actually productive? For novices?
I see a lot of offerings with no standard desktop on 7". So some people do like it.
@John Smith
"You must be the “oh I install random shit on my computer and blame other people like microsoft when it fucks up”"
Blaming the victim is not a valid security policy towards children.
However, I normally clean up XP for other people. I never have to clean up after I went on an installation spree on Linux, though.
@John Smith
"And Sugar is impervious to people breaking the software? Especially when it’s supposed to be designed to let people play with software? Is that just a feature for free software?"
Yes, Bitfrost makes Sugar impervious to people breaking the software.
Bitfrost has no counterpart in other software distributions (MS, Linux, nor BSD).
Before you reply you might inform yourself about Bitfrost on, eg, wikipedia:
http://en.wikipedia.org/wiki/Bitfrost
@John Smith
"Full stop, so that’s what you meant. That’s all I was asking, not an indepth discussion"
You might have known before you started to complain?
@John Smith
"You’re making a big deal over wear leveling. So many flash devices use hardware based wear leveling... big deal... it’s not expensive to implement, and it gives you flexibility as to what you can run on it. Correct me if I’m wrong, but aside from the default sugar, every other GNU/linux distro running on the OLPC is via SD or USB. Which “ZOMG has hardware wear leveling!!!!!!”"
The XO does not have hardware wear leveling on its internal memory. Therefore, currently only Sugar can be installed natively. Period.
Which obviously means that XP was never a valid alternative. Period. Which was the point from the start.
@John Smith
"As far as the Eee, I’ve been running XP on my 4GB flash 701 for 8 months no problems, and there’s tons of people over on Eeeuser with original Eees that have been running winxp on the SSD no problems. Many people have bought xandros versions and installed Windows on them (and vice versa bought windows versions and put GNU/Linux on it), why you’re focusing on what’s being sold by retailers I don’t know, since there’s actual users you could get data from."
The XO is slated to be used for 6 years. If knowledgeable people write XP will not last that long on controlled flash memory, I believe them.
If you show me you do not even understand how Sugar works, and also tell me XP can be made secure, I very much doubt your expertise.
@John Smith
"...But the actual demand is in netbooks (a la EeePC, Aspire One), and they have significantly outsold OLPC (in both linux and windows forms) because they sold to end users."
How does that affect the point that the OLPC was right in developing Sugar?
@John Smith
"And OLPC still isn’t meeting the initial gols of battery life in days not hours (even in ebook mode) with battery life comparable to other computers, and they have yet to release alternative charging devices. Meanwhile the underclocked Geode is so anemic it can’t get out of it’s own way for Sugar."
Which is relevant to the above post in what ways?
The fact that the OLPC was unable to get all their planned features in Sugar is in itself not a reason to question their decision. Especially as there would not have been netbooks if they hadn't developed and produced the XO.
So the fact that there are now, five years later, more expensive netbooks that can run XP on flash for over 8 months is hardly a reason to criticize the OLPC for their design decisions 5 years ago.
Winter
That's my EXACT point: no one over the age of about twelve should ever want to use Sugar.
Its designed for children - not adults. Its designed for those who have not seen a computer before, not those with X years of windowing experience. Its designed to motivate children to learn, not for them to work in an office.
So its very concept is one of OLPC's "clock stopping hot" innovations.
If there is a mistake, I say its in the implementation. No, make that - its implementation was a catastrophic disaster, as evidenced by all reasons given by those who say it was a mistake.
Had it been cooked longer & better, we wouldn't even be having this conversation.
Sugar is not a mistake! As the rest of OLPC is an academic project! Repeating and improving work of others (other OS/UIs) is for 3rd class academics. So a new OS/UI as different from the rest as possible is the only thing that makes sense.
As of Python/libraries, you must use the leading edge tools and possible “improve” on them to be a respectable scientist. The journal is also on the leading thinking on archiving and searching (see timeline, tagging). If you think of it, the whole project from hardware to software is full of real innovation, as any top-gear academic project should be.
However, for mass production and market appeal, is the quality of the “manufactured product” that is important for the end user. Not just the design. That is where OLPC and Sugar failed. To use the ever-familiar car analogy, they tried to sell a concept car to the masses. Looks great but is not working on your neighborhood road… With time some concept cars or elements of it, are polished and go into mass-production to benefit us all. This might also happen to Sugar if industrial approaches and thinking are implemented.
In short, Sugar is a great concept with lousy implementation, because of time and resources constrains and possibly attitude... We can shoot it down, dismiss it and forget about it or fit it to the real world needs. I’ll go for the second. Spreadsheets and formatted text can be easily implemented in Sugar, but imagine the benefits that can spill over to your favored OS, if this can be done in conjunction with an effective ad hoc collaboration.
Python and Sugar libraries is a nightmare, but imagine if all the applications in your favorite OS where self-contained and you didn’t need to hourly update your multiple pieces of security software. Polished interfaces need just that, but imagine if you can press a button see under the hood and make your own without predetermined oprions. Journal does not allow you to categorize your files as you see fit, but honestly how often do you use file search utilities to find your “categorized files” in a machine that you use for more than a year?...
I do not know how many people/organizations were put off by Sugar in buying the XO, and I do not think that anybody has any evidence for that either. We are only postulating, and if we go that way I can think of MANY more things that went wrong with the implementation of the OLPC experiment…
Sugar was not a mistake. It will certainly outlive the XO. In the long run, it may have more impact on education worldwide than the XO.
The real problems with OLPC were organizational.
When I started in computing (Univac), I was taught never to create/update hardware and software at the same time. Specifically, you must build new software on the current platform and introduce new hardware running the current software.
In the case of OLPC, Sugar should have been developed and introduced as a Linux distribution to run on current hardware (what SugarLabs is now doing). The XO should have been introduced with an appropriate existing distribution.
However, the three really big mistakes were (1) not organizing to support one school at a time pilot deployments, (2) not distributing the first G1G1 through Amazon (or other effective channel) and (3) not staffing OLPC to support the community effectively.
Pilot schools would have provided necessary feedback to the development process and helped develop the infrastructure to support wider rollout. The 'give one' could have been used to do this. Effective distribution of the 'get' one would have rapidly expanded the available development community. Developers rightfully do not like to be interrupted. You must have mentors (technical marketing) who meet regularly with (not interrupt) the developers. These mentors then welcome the newbies to the community and help them become productive developers. (Mike Fletcher attempted to do this, but could not succeed because the XO/Sugar developers viewed him as an interruption.)
Tony
My 6 year old daughter uses Windows on our other computers, but the Sugar interface is WAY less confusing for her. For one thing, it's a lot harder to accidentally lose focus, yet learning how to task switch wasn't that hard for her. Why do you think games take over the screen on Windows? For that exact reason - a cluttered windowing environment is confusing!
Even great Linux environments like Ubuntu still share many of the same downsides.
Sugar interface is not a mistake per se. It needs polishing and probably better documentation. Lets not forget Sugar was nothing down than trimmed Fedora distribution and porting to others distros was straightforward.
What Sugar did is exposing issues on many applications that abusively use memories.
Since Sugarlabs is mainly made for volunteers , why keep whining instead of going to check Sugarlabs staff and ask them what we can do to improve Sugar? Debating whether it is a mistake only turns to be a waste of time.
When I first indicated to colleagues that I was participating in the G1G1 program in Nov 2007 I got two reactions. The second one was to attack the arrogance of giving computers to starving children. While somewhat off the mark on the false dichotomy of food vs technology it probably was more accurate on the general arrogance of many well intentioned projects.
I would include among them:
Inventing an "educational model" that required unnecessary nonstandarization.
Presuming to know what 10 year olds in Rwanda find stimulating or useful, and assuming that applies equally to 16 year olds in Peru and 6 year olds in Poughkeepsie.
Maintaining purity on Open source at the cost of making basic things not work.
Elevating Mesh networking - admittedly a cool idea - above the basic connectivity model for an increasingly pervasive internet.
Assuming that 12 year olds have no need for or capacity to organize coherent directories and would be better served by a list of stupid icons.
So sugar is a different model. Great. Was it needed? What was wrong with just a tweaked Linux desktop? About 3 years ago we reached the point where various Linux Distros like Mandriva or Ubuntu were easy and familiar enough to be installed and used by non expert users, and could run on an old Pentium II.
I have been using computers since I was about 7 years old and my father let me play moon lander on a DEC PDP. I had an Apple II in Junior High, and IBM PC 3 months after it came out. I have worked in Software for over 10 years. So I have a little experience with a lot of stages and levels of computer use. I am no great fan of windows but for non expert users it kinda works.
Instead of Sugar why not just check the revolutionary educational model hubris and just cook up a completely functional lean Linux that would not only work for starving 10 year olds but might also prepare them for a future in technology not ghettoized for 10 year olds. I was really surprised when i got my XO because I assumed Sugar was just some layer sitting on top of Redhat that you could ignore. No such luck
And jesus if you wanted to appeal to kids, then the platform the won't run youtube videos was just not the way to go.
I wanted to do several things via G1G1. First I wanted to participate in an effort to seed technology in developing or rural areas. Second I wanted to support the development of simple, lean, rugged computers that could perform a number of non traditional uses. Third I wanted a computer I could give to my three year old that had a keyboard he wouldn't destroy, and no moving parts to drop and break.
Well, 1 for 3 i guess