Nicholas, with all due respect, I think you're pretty seriously mischaracterizing the nature of One Laptop Per Child's problems. Laying the blame for OLPC's shortcomings at the feet of "open source fundamentalists" is misinformed at best and deliberately disingenuous at worst.
Now, to be clear: when you say that "Sugar needs to run on many platforms," I completely agree with you. I couldn't possibly agree more. But moving from that point, which is clearly correct, to an ad hominem attack on the open source community as a whole, is a frustrating and dangerous non-sequitor, and a slap in the face to the people who have been your most strident supporters for many years now.
Walter and Nicholas of OLPC
When a man like Walter Bender walks away from your shared dream because he feels like you are choosing the wrong path, then maybe you should consider being a bit more introspective, instead of lashing out at the big bad free software fundies.
Did Walter, your friend for 30 years, the guy with whom you built the MIT Media Lab, turn into a fundamentalist whack job over night? Really?
OLPC's goals have been extremely ambitious from the very beginning.
The possibility of failure has always been very real, even had you made all of the right moves from day one. Most of the issues you face are the issues that are inherent to solving really hard problems.
Fundamentally changing the computing metaphor from the noun-based "file" metaphor to the verb-based "activity" metaphor is a really hard problem. Building the only major networking stack using the 802.11s standard for grid networking is a really hard problem. And your reliance upon open source has, to date, been one of your most effective levers in solving those problems.
From my perspective, your biggest problem has been that you have not relied enough on open source principles to build Sugar.
First of all, your organization has been notoriously opaque. I'm absolutely certain that this hasn't been deliberate, but when you're running a community project, your first job -- and your second job, and your third, and your fourth -- is to make damned sure that when volunteers show up, you have something useful for them to do. Volunteers, in the open source world, are gold. For most of the history of the project, you haven't treated them that way -- not out of malice, but out of neglect. There were always "more important" things to do than to help a newbie contribute.
Second of all, until very recently, Sugar only ran reliably on the XO itself. From day one, it should have been a priority to have stable, checkpointed releases of Sugar running on Fedora and Debian and Ubuntu and every other distro, all installable with a simple yum or apt command.
"Release early, release often" -- have you heard that before?
Instead, if someone wanted to run Sugar on their own system, it involved running jhbuild, which involved installing a half-dozen different SCM clients and almost continually rebuilding from broken source repositories. Which made it devilishly difficult to write robust activities for Sugar. All understandable mistakes, to be sure. Mistakes common to young open source projects. Mistakes that are now being fixed.
Clasmmate PC - Sugarized!
The irony of your rant is that porting Sugar to many platforms, thereby increasing the number of potential users, thereby increasing the number of potential contributors, is an obviously correct move that will help you leverage open source more effectively.
To conflate that correct message with an attack on "open source fundamentalists" is misguided, and diminishes your ability to recruit community talent at the very time when your project most desperately needs it.
So cut it out already.
The folks at 1CC have enough problems to solve. They are really good, good quality people, and they're certainly not "fundamentalists", whatever that means. They've worked like dogs for you. They've sacrificed their personal lives to help make your noble vision a reality.
Maybe you should think a little harder next time before you, Great Leader, slap them in the face in such a public and mean-spirited way, because you can't figure out how to close deals. It's a shitty thing to do to people, and you ought to be better than that.
This Guest Post is from Greg DeKoenigsberg, an OLPC volunteer and a Fedora guy.
Hear hear and well said!
As one of those volunteer developers, the opacity of OLPC is something I constantly struggle against. Which is why my modest effort to provide an astronomy application for the XO has resulted in a program which has no i18n support and several UI bugs that are probably trivial to fix if somebody would just explain what the Activity class *is* instead of just telling me to subclass it.
StarChart is the most popular piece of free software I ever wrote -- it's being downloaded from my site around 40 times PER DAY. It's being included in the activity bundles for Peru and Nepal (even though it only speaks English). But it is certainly not the BEST QUALITY program I've ever produced. And some of the blame for that lies with the lack of how-to information for the potential developer.
I'm curious as to why Windows on the XO is somehow considered as inhibiting the learning value of the XO.
I fully understand that XP is not open source, but why is it so fundamentally important that the OS be open source? It's not as if the hardware is open source, right? As long as there are no restrictions as to what can be put ON the OS, that's what ought to matter. Arguing about one OS vs another seems silly, and it strikes me that certain members of the OLPC community are letting their anti-MS bias get the better of them.
Oskar,
There are two main reasons why XP on the XO stirs up so much debate:
1. Sugar + XO is a targeted learning combination that is specifically designed to do one thing and one thing only: educate children. XP on the XO is a general purpose computer, one use case of which could be education.
2. Many of us invested our hearts and minds into OLPC because it was Open Source and not MS. To switch a fundamental aspect of the program this late in the game is alienating all of us who are here because Sugar + XO = Education + Open Source.
XP on the XO can be educational, its just not the OLPC we signed up for.
Personally, I'm fine with Windows on the OLPC, and trying to get Sugar working on Windows on the OLPC, and trying to get Sugar working on Windows everywhere else.
What I am *not* fine with is the way that NN chooses to express himself on these issues. The developers I know don't object to porting Sugar to Windows because they are "open source fundamentalists"; they object for the very pragmatic reason that the cost/benefit analysis doesn't make sense.
I *agree* with NN wanting broader usage for Sugar. I *agree* with NN wanting to disentangle software and hardware issues. I *disagree vehemently* with his assertion that these problems were caused by "too much open source", when the truth of the matter is that these problems were caused by "not enough open source". And his precious little soundbites are CHASING THE COMMUNITY AWAY.
Wayan,
"Many of us invested our hearts and minds into OLPC because it was Open Source and not MS. To switch a fundamental aspect of the program this late in the game is alienating all of us who are here because Sugar + XO = Education + Open Source."
The fundamental aspect of OLPC is, and has always been, that it's both open-source and open-platform. There's no "late switch". XP on XO is, as you very well know (as you blogged about it over a year ago), nothing new and both Walter and Kristic were defending that position in the past. From my point of view, the ultimate yardstick on the 'rightness' of this is the fact that Alan Kay and his team have arrived at exactly the same decision making Squeak (and comparing to Sugar it's a much more complete environment than Sugar) available on as many platforms as possible and that includes Windows. Kay, of course, has advanced the course of educational computing longer than just about anyone else and his authority on the subject is unquestionable. Ideological zealotry is never a good thing as it leads to, ultimately, loosing the sight of what matters most - and in this case it is to provide as many kids as possible with an exceptional educational tool that is an XO with the suite of software that is already available for it. The underlying OS is of secondary importance - a position, if you looked at the most popular and important open-source apps out there, most open-source developers support...
Greg,
"What I am *not* fine with is the way that NN chooses to express himself on these issues."
No doubt NN is a very outspoken guy and on many occasions what he said was rather silly (or appeared to be unless he did it on purpose to arrive at a particular outcome).
But I would be very careful to read too much, unless explicitly quoted and with the context provided, from various 'reports'. The blogging and 'reporting' sphere thrives on rumors and, from my point, the most amusing part is how often they quote each other - a bit of 'Chinese whispers' (and sometimes circular) in action really... :)
Delphi,
What you're talking about is making Sugar a set of educational apps - say Sugar Educational Bundle - that, like MS Office or Squeak, is designed to run on any platform. And an XO laptop that can run another OS. I think both are admirable goals, and I support each in general. But that's not the OLPC mission as I understood it.
OLPC's mission was constructionism personified as an Open Source educational software stack running on a rugged, efficient laptop. Change either aspect of that equation, and we have a different project. Again, the project could still be educational, but its not going to be the same OLPC.
The other excellent way to get more developers on board, in addition to getting Sugar on as many platforms as possible, is to get the XO Laptop itself in as many hands as possible. This means selling the XO Laptop outside of the developing world.
When open source developers actually have the system to play around with, it'll super charge Sugar development. Making an application on Sugar on Linux or Windows is maybe sort of cool, but it's so much more cool and interesting when you're actually running your application on the platform it's designed to run on, which is the XO Laptop.
It doesn't make sense to me to have XP + Sugar, what are the goals of have Windows + Sugar ??, I can't see the real points, port sugar and their required libraries to windows is wasting time...
Windows is an OS with a GUI (Graphical User Interface), so if we are replacing that GUI by Sugar... are we getting more features and stability than Linux ?, I don't think so... I would like to know the real reasons about this "supposed" decisions.
Its about time the market was flooded with XO laptops.
Why can't Quanta do what Asus did? They should be able to make them for under $100 if they produce enough. Maybe even outsell the EeePC.
Could they retail under $200 US?
Its obvious that OLPC has not lived up to their end of the manufacturing deal by their dismal failure to come up with orders.
Remember this from March 2007?
http://www.ft.com/cms/s/2/6d086a62-dcb3-11db-a21d-000b5df10621.html
...'Michael Wang, Quanta’s president, said on Tuesday that the concepts developed through the OLPC project could be applied to create commercially viable machines that are cheaper than anything on the market so far.
“We will definitely at the right time launch a commercialised product similar to the OLPC,” he said in an interview with the Financial Times, adding that several of Quanta’s customers were seeking to launch such a product...'
The XO is able to have just about any OS installed as is documented on the OLPCNews forum. Give anyone that wants an XO a chance to install edubuntu, Sugar or even XP.
I say Screw Nick!
Wayan,
"What you're talking about is making Sugar a set of educational apps - say Sugar Educational Bundle - that, like MS Office or Squeak, is designed to run on any platform."
Eh, no. Sugar apart from being a GUI can be view as a 'platform' wich also includes, in somewhat simplified view, both Python and Gtk environments on top of which (educational) apps can be built. Squeak, being practically a Smalltalk implementation, is a much powerful environment including a number of ready to use frameworks and used for higher still level environments perfectly suitable for kids to explore various ideas- both Etoys and Scratch are very good examples of those and, of course, run across many OSes (including Windows). The fact that Squeak/Etoys/Scratch can run on Windows in no way diminishes their value as open-source environments fully supporting constructionist learning - the host OS itself becomes, from that point of view, irrelevant. The situation with Sugar is totally analogous...
Quite frankly, even just having XO with with Squeak/Etoys/Scratch plus a AbiWord, Firefox and FBreader for textbooks would ( and regardless of the OS being used ) provide a fantastic toolset for Constructionist learning. Sugar takes it, of course, even further.
"OLPC's mission was constructionism personified as an Open Source educational software stack running on a rugged, efficient laptop.Change either aspect of that equation, and we have a different project "
And again. No "aspect" has changed - open-platform is (and of course this was supported by both Walter and Kristic) as important as open-source - that's how it was defined at the beginning of the project and that's how it remains. Do I think WinXP will be a better technical choice? No. Do I think having WinXP as another choice will make it more widely acceptable without sacrificing the OLPC mission objectives? Definitely! There was always talk about MS being able to provide their OS for XO if they wish and so is Apple. You are perfectly aware of this so I don't understand why you insist otherwise. As per my previous paragraph 'Open Source educational software stack ' can be fully supported regardless of the OS being used on XO.
Alan Kay and his team have been in this very situation well before OLPC happened - they chose kids rather than zealotry...
"I'm curious as to why Windows on the XO is somehow considered as inhibiting the learning value of the XO."
There are many reasons why a 6 year old, obsolete, proprietary OS developed for office work is not the best platform for education on small footprint hardware.
But I will pick another one.
Why is an open playground between a bussy higway and a deep canal, used at night by drug addicts and other "night life", not a good playground for small children? Because you will need too much oversight to let them play safe.
XP is fundamentally unsafe and unreliable. I know very few adults who can secure their XP computer on the internet. I know no children that can do that. There is a billion dollar industry catering to secure the world's XP computers, and it is failing spectacularly. Why do you think an average developping world country could do better?
Sugar et al on the XO were designed from the ground up to secure the children and their data. Nothing is perfect, but there are no automated tools known that can infect Linux computers (viri etc.). And no exploit is known for Bitfrost.
Winter
No avoiding the security shortcomings of XP. How's that issue going to be addressed?
Delphi,
You're confusing two concepts
1. Bender, Krstic (and I) are okay with MS and Apple putting their OS on the XO at their choosing. Or porting Sugar or Activities to Windows or OSX. More openness, the better.
2. Negroponte promoting the idea of Windows as the sole OS for the XO, or even an OS that OLPC supports. That's not the same stance as the one that rejected OSX as the OS of choice because it was proprietary.
Two other points why to have XP on XO is not good:
- that will increase the price of the XO, for what I know XP is not free but perhaps XP will be "given" by Microsoft or the other entities of the same compagny "The bill gates fundation"
- You cannot have access to the code of XP. One good things with XO is that children can learn how a computer is working from high level stuff with python to low level with the system code.
Wayan,
" You're confusing two concepts"
I don't believe I do - but I admit that perhaps your own changing position is somewhat confusing.
"1. Bender, Krstic (and I) are okay with MS and Apple putting their OS on the XO at their choosing. Or porting Sugar or Activities to Windows or OSX. More openness, the better."
Well, I'm glad you finally agree with Bender and Kristic. Although not that long ago you were saying something rather different[1]:
"This is The End My Friend: Negroponte Says XP on XO in 60 Days:
...I say that XP on the XO is the end of One Laptop Per Child as an educational project. With a Microsoft operating system, an XO becomes a "$200 laptop", a cheap Toshiba replacement, not an educational learning tool for children."
"2. Negroponte promoting the idea of Windows as the sole OS for the XO"
I don't have any evidence that's what Negroponte is "promoting" at all - in fact his email you reposted here just yesterday says something opposite [2]:
"We have been engaged in discussions with Microsoft for several months, to explore a dual boot version of the XO."
And as wirzcat [3] has already tried to explain to you:
"No matter what MS does, the XO with a XP card in it will still be a dual boot device with the addition of the developer key. "
[1] This is The End My Friend: Negroponte Says XP on XO in 60 Days
( http://www.olpcnews.com/software/windows/xp_on_the_xo_in_60_days.html )
[2] Nicholas Negroponte on Sugar and One Laptop Per Child
( http://www.olpcnews.com/people/negroponte/nicholas_negroponte_sugar_olpc.html )
[3] Re: Windows XP on the XO - the technical discussion
( http://olpcnews.com/forum/index.php?topic=1367.msg10061#msg10061 )
Delphi,
OLPC promoting Windows on the XO is the end of Sugar as the default OS. Take the next sentence of Negroponte's quote, and combine it with his comments to AP, that in time, Windows would be the only OS.
"Some of you have seen what Microsoft developed on their own for the XO. It works well and now needs Sugar on top of it (so to speak)."
That is much different than the XO able to run Windows, OSX or any other OS, in the spirit of Open Hardware - a technical feat - and promoting Sugar as the OS for education as part of OLPC's mission.
Hi!
For me right now, I can not access various webpage videos. I need special codec files. I can not listen to my mp3's easily even after asking for assistance on the OLPC NEWS FORUM. My Sonata was linked to my SD card, but still will not work unless I go and reinstall my OS and then reinstall OPERA and Flash!
From what I know, Microsoft XP for the XO on my SD card when booted from that location will allow me to see those items now not seen. What choice do I have if what I need is not available? Dual booting XP and Sugar will give me what I can not now get from the machine!
If I get slower performance with XP, is that terrible if I get other items I want to use with my machine?
I read somewhere that people don't buy operating systems, they buy applications. Most people don't know (or care) about the cell-phone's operating system, they just want to call somebody. Open source is a wonderful thing but the customer (minster of education) doesn't seem to want it. The people who do want the XO (individuals) can't buy one.
It seems OLPC has two choices, give the customer what they want (XO with windows) or let the people who want one buy one.
Wayan,
"OLPC promoting Windows on the XO is the end of Sugar as the default OS."
Sugar is not an OS. It's, in simple terms, a GUI environment sitting on top of Fedora linux base. It'd be relatively simple to put it on top of any other linux distribution base or, in fact, any non-linux OS. And that's what *explicitly* Negroponte is saying should be the direction OLPC should take with future development of Sugar. Makes sense to me.
Ultimately, the choice will be customers' and, given the maturity and richness of Squeak/Etoys/Scratch environment, I wouldn't be surprised (but then I'm, as an old Smalltalker, probably bit based :) that's were plenty of development/usability will come from - the efforts of the team in Nepal is just one of those.
"Take the next sentence of Negroponte's quote, and combine it with his comments to AP, that in time, Windows would be the only OS."
I've read the AP article [1] you referring to (which also seems to be relying on you for the source of information :) and there are no Negroponte's quotes on this, only reporter's interpretation of what allegedly Negroponte has said - over the years I've read enough of nonsense/misinterpretations in the popular press to give it too much credit...
[1] AP - Low-cost laptop program sees a key leadership defection
( http://ap.google.com/article/ALeqM5hXa0O9XLMsWfaqt-sI9FqFy2IewgD9074MH82 )
Delphi,
Make that "Sugar/Linux as the UI/OS" and I spoke to the reporter only about Bender's departure. He talked with Negroponte directly - I may be good, but I'm not that good.
@wayan
"2. Many of us invested our hearts and minds into OLPC because it was Open Source and not MS. To switch a fundamental aspect of the program this late in the game is alienating all of us who are here because Sugar + XO = Education + Open Source."
Thereby making you open source fundementalists?
Greg Yohn: Maybe your IT needs indicates that you are not in user group to witch XO is targeted?
Wayan,
Another piece of info from devel list confirming 'dual boot', as per Negroponte's post, and, as I suspected, showing the AP reporter's ( 'Windows as the sole OS for the XO' ) interpretation being false. This is (although, typically somewhat boastful :) from Ivan Kristic' recent post on the subject (http://lists.laptop.org/pipermail/devel/2008-April/013361.html ):
"Dual-boot became the plan of record at OLPC and MS, and actual technical work began on this approach before I left."
Delphi,
And you think that, over the long run, it will stay a dual boot? I think that's remarkable optimism. I see it being a direct MS boot, with the option for Linux quite soon. And option that will not be easy to take with MS in the house.
Wayan,
"And you think that, over the long run, it will stay a dual boot? I think that's remarkable optimism."
No-one knows for certain what will happen in the "long run". Especially in this industry. But, given that my involvement in the software development pre-dates MS, my perspective on things is somewhat different than yours. Things change. We had 'No body ever got fired for buying. IBM' . Then IBM got replaced by MS. Now? Just last few weeks ago I helped an elderly friend of mine to move all his PC setup (using Linux btw, not that he cares) to using web-based Google suite for *all* his computer needs...
The fact that MS, after rejecting initial approaches from OLPC's at the beginning of the project, is now agreeing on a dual-boot setup is a remarkable victory for OLPC... As I said before, let the little green Trojan Horse roll on... ;)
@Delphi:
"But, given that my involvement in the software development pre-dates MS, my perspective on things is somewhat different than yours. Things change."
I must agree. Especially given that MS now have nothing to sell that will increase business desktop users productivity with respect to their years old XP. Vista actually DEcreases office productivity. And the uptake of Sharepoint must be stellar to compensate for Vista.
In general, companies that sell productivity "ware" that reduces productivity tend to go broke.
Personally, I guess that MS' monopoly will be over by the start of next year. (not MS, but their monopoly) In that case, the XO won't stay dual boot for long neither :-)
Winter