That article also contains the news that Give 1 Get 1 will be restarting in August or September.
Great, so the next iteration of the program will be starting up just as the last donor from late 2007 finally gets their long-overdue delivery!
After my experiences with the Grand Clusterfuck of G1G1 mk I, not to mention the Windows Compromise and the resignations of so much of OLPC's top talent, I'm discouraging everybody I know from contributing. Even if it's "Give 1/2, Get Three".
There seem to be three separate and distinct issues being conflated here:
1. India has laws that make it a crime to post "vulgar content" 2. Google provided information to Indian police in conformance with the law 3. Indian police are alleged to have badly mistreated a suspect
Be outraged about #1 and #3 if you wish, but I see no malfeasance inherent in Google's actions #2.
A lot of people, sadly, seem to have a sense of entitlement, a sense that they deserve the jobs or have somehow earned them through no action whatsoever.
Is this something that has been empirically observed, or is it just a stereotype you've cultivated anecdotally?
The average salary for a Sr. Software Engineer in the US is around $90k according to PayScale.com. Not bad, that's more than I make. In India, however, the wage is about 580,000 rupees, or around $13,500.
1. Is the Indian figure for "Senior Software Engineer", as well? 2. Are the skill sets and duties of a "Senior Software Engineer" comparable in both countries? (IT job titles, you will find, are often not very well-defined, even within a single geographic market.)
You could pay someone in India pretty well by Indian standards and still save a ton of money by American standards.
Assuming, as bean-counters too often do, that the work produced by the $90K American and by the $14K Indian will be equivalent in value to the American company. And there are plenty of reasons not rooted in racism or irrational xenophobia that would refute that assumption.
The fact is that while a good education makes a big difference, the real question of how good someone is depends on how well they learn and how open-minded they are
I don't disagree with this; however, it has been my non-empirical experience that the culture of learning in the United States does put a greater emphasis on open-mindedness, innovation, and exploration than do the learning cultures of other countries currently exporting programming talent.
If they're taught Java in an American university and O'Caml in an Indian university, the American is going to have a better immediate skillset
*sigh* A university education is not meant to be vocational training.
American universities can pump out idiots just as fast as Indian universities do, they just do it for a much higher price.
Maybe so. But a hiring manager stands a better chance of accurately evaluating whether a new grad would be a valuable contributor to their organization when he's sitting across a conference room table and speaking the same American English dialect as the candidate, as opposed to being a nameless resource listed on page five of a project bid document that's being discussed in an international teleconference, represented by a team lead who speaks with a thick accent.
Yes, I do want my programs to "work great", not "just work".
And it's not just the developers that want this for reasons of pride!
EVERY CLIENT would prefer, all else being equal, to have software that works great than software that just barely meets their needs. Unfortunately, clients usually don't discover how poorly their needs are being met until well after the costs of software acquisition have been sunk.
Mirror the database to a 2nd server and provide them read access to that.
Before doing that, present them with a proposal that itemizes all the costs they'll be responsible for it you do this: $XXXX to purchase the hardware, XX hours at a rate of $XXX/hr to have an admin perform the software install and configuration, ongoing support costs of $XXXX/month, and so on.
You'll find out how badly they want the level of access they've requested by how much they're willing to spend to have it.
The OLPC was clearly designed as a free R&D project from the beginning. Not free as in speech, but free as in, "hey, lets CALL it a charity. That way we don't have to fork out money for our R&D".
So where did the funding for the OLPC R&D come from? Somehow I don't think Negroponte, Bender, Jepsen, and everybody else worked tirelessly for years without being compensated for their time. I don't think Quantas would agree to divert their production lines to make little green budget PCs unless it was made worth their while.
When the OLPC was listed out at $100 I said it was way too expensive. I went on line and found all of the components to build a hand powered computer for $89. Single Unit Pricing.
Oh, please do post this list. I'd love to see how you managed to put together a rugged laptop more affordably than the XO-1.
You seem to have a fundamentally different perspective about the needs of a technology-driven education program than the OLPC Foundation does, and than the OLPC Foundation did at its inception for that matter. Perhaps you would care to explain your position in more detail?
Do you think XP and a Linux distro will both fit side-by-side in the 1GB of NAND memory that the XO-1 currently boots from?
For that matter, do you think XP will ever run with acceptable performance and security on a computer that has a sub-600Mhz CPU, only 256MB of memory, and almost no "disk" to speak of? Even by 2001 standards, when XP was first released, that's a little short of XP's recommended hardware specs.
You know that feeling you got when Walter Bender left the project over a disagreement with Nicholas? That "Wozniak has left the building" feeling? Turns out we were right.
I got that feeling when Mary Lou Jepsen -- the OLPC CTO, the person most directly responsible for the hardware design of the XO-1, including the fantastic transflective 1200x900 screen -- left the organization in December. As with the Apple Computer, it was the hardware innovation that served as the foundation for everything else groundbreaking.
A $300 laptop that provides educational information for 5 or 10 years is a lot cheaper than the equivalent material in textbooks.
Only if the equivalent material is available via the $300 laptop.
So far, I have not heard of any educational textbook publishers volunteering to send electronic copies of their materials to developing nations gratis. Nor, as is necessary for most of the world, to translate them from their original language to those used in other locales.
So what are we left with? Wikipedia? As a replacement for pedagogically sound teaching materials?
The UI is a core issue. Why should it be materially different from what a billion computers already run?
Maybe -- probably -- there's something better than what those billion computers already run.
The OLPC leadership failed to understand that novel UI features were a nice-to-have, not a must-have, and failed to enforce a focus on the actual must-haves of the project.
To paraphrase them themselves: it's an education project, not a user interface project. Sugar's collaboration features, the Journal, the zoomable metaphor -- these are all good and promising concepts for Human Interface. But not one of them is core to the educational mission.
If the students are going to be able to go onward from OLPC, then their "language" must be "compatible" with the other "computers" they will see later.
Nonsense. I started computing at 8 years old running BASIC programs on a Tandy Color Computer, and I had no problem migrating later to using DOS 3.3 on an XT a couple years later, nor to each successive new OS and architecture.
I would sell the shares but it's my 401k and all of the available funds are managed by the same company.
So what you're saying is that there is a limit to what steps you are willing to take to defend your principles, and that limit is not uncorrelated to your finances.
Remember this: no matter how nice your office space is, if you're an "interactive agency" with an unspellable/unpronounceable name like "Tocquigny", you're going to be the first to go out of business when the Dot-Com Crash 2.0 happens.
moderation merits in Slashdot are hierarchical - the first moderators were wisemen chosen by the Mighty Taco Himself. Besides, anyone can metamoderate.
Which is not to say that Slashdot moderation is free from politicking -- despite having consistently excellent karma, the site stopped offering me moderation privs a couple years ago. I suspect this was related to my tendency to call a certain Slashdot Editor out on his habit of tacking his own personal commentary onto story submissions that he approved, instead of submitting a moderatable comment in response to the story like all the rest of us have to.
Surely CmdrTaco can pick whomever he wants to edit this site, and those editors can pick whomever they want to moderate. I just wish they'd be more transparent about the process.
It cost them a huge amount of money, sure, but Microsoft managed to spend its way into a significant share of the videogame console market while competing with two most consistently successful companies in the history of the industry. Xbox Live is the premiere online console gaming service, and something that Microsoft has made precious few missteps with.
I don't think the Xbox division can be considered a fiasco by any measure, and even in strictly financial terms it's at worst a short-term failure with strong potential for long-term successes.
Were it not for the crappy BIOS, programmers wouldn't have had to resort to writing directly to hardware to get an acceptable speed on the screen.
You seem to be implying that programs doing direct hardware access for performance was not the norm across all platforms in the early 1980s.
In those days, hard disk drives were optional, CPU clocks ran at less than 10 MHz, and unless you were a "power user" you probably didn't even have the full 640KiB of RAM that DOS could use to run programs. With resources that tight, there wasn't really a lot of room for implementing the abstracted multi-layer system models that we now consider good practice.
Everything I have read and heard about Microsoft suggests that they are cowboys.
Code first, design later.
What about that story about how it took twenty people from six departments three months to reach a decision on whether to add some feature to the Start menu or not?
Maybe their developers have the cowboy mentality because there's a fundamental disconnect between them and the management bureaucracy that makes it impossible for anything to get accomplished otherwise...
The [WinCE] applications are pretty much there. When I'm on the road I need wifi enabled IE or Firefox to surf the web / do web enabled work. I need to view pictures, maybe edit a.doc or.xls. I need my calendar and the ability to queue up emails for my work mailbox (sync'ed with Outlook when I am anywhere near my work network.) That's about it - anything else is gravy.
To the road-warrior business traveler, maybe.
The platform is still pretty useless to the application developer, the artist, the musician, the scientific researcher, etc...
Not to dismiss whatever QA or shipping problems resulted in your Macs arriving with parts DOA, but... for what reason is it important for a business laptop to have working speakers in it? At least in my office, the sentiment is that the less noise your computer makes to disturb your officemates, the better.
And I wouldn't call your company particularly agile if they don't have a contingency plan in place for a single employee's primary computer being out of commission for a couple of days. No 'loaners' at all?
Apple has to realize if they want to compete, they need to open up a bit to their larger buyers.
I would have thought this was clear to everyone by now, but: Apple doesn't want to compete. They want to carve out a dedicated corner of market for themselves and own it thoroughly.
If proprietary stuff like "Flash" is required and Gnash isn't up to snuff (yet), doesn't it make more sense to as Adobe for a Flash port rather than throw the education deprived baby out with the bath water?
The standard Flash binary distribution for Linux can easily be run and installed on an OLPC laptop, if Gnash is considered insufficient.
The problem is that existing Flash content targeted to desktop computers with multi-GHz CPUs and a gigabyte-plus of RAM doesn't perform so hot on a 433MHz machine with 256MB. While I wouldn't be surprised to learn that the Adobe Flash + Windows software stack has certain optimizations which could improve performance slightly, we're still talking about going from two frames per second to three frames per second; still not tolerable for kids wanting to surf the Hannah Montana portal or adults to watch Homestar Runner toons.
They should find where their target demographic has commonalities with people in the more developed world. They should have been making something more like an Asus EEE PC to begin with.
So the XO laptop ought to have been just as rugged, but twice as powerful and also twice as expensive?
Would be great for us rich industrialized types, but it would have meant that the government could only afford to buy half as many, which would mean that only half as many children would be able to benefit.
Not to say there isn't some commonality between the needs of schoolchildren in the developing world and business travelers in the developed world -- for one, I'd love to see the XO's high-res transflective screen find its way into all kinds of hardware -- but in some aspects the requirements are entirely incompatible between the two.
If anything, the failure of OLPC so far has been a failure to create software yet that's as efficient and well-designed as the XO-1 laptop's hardware.
The leaders he mentions are not high-tech. Most people in those positions refer to the GUI on a computer as "Windows" whether it's Gnome, KDE or FVWM.
I don't know whether that's true or not.
What I do know is that if OLPC starts making hardware and software decisions based on what education ministry bureaucrats ask for, instead of what provides the best benefit to the students, they have already lost sight of their mission.
Honestly, I suspect that Windows XP/XO will never see a release. I think it's all just a ruse to keep OLPC distracted, and delay governments from making a purchase decision, long enough that a separate computing program for developing nations, centered on Microsoft Office for Vista no doubt, is ready for market. And what remains of the OLPC brain trust is falling for it.
If they're out of physical space (not just power and cooling) then the facility is way oversubscribed
Assuming that the new supercomputer has the same space needs as the one it's replacing, what you're telling us is that any machine room running at more than 50% volumetric capacity is oversubscribed. I can't agree with that.
Slashdot Mirror
That article also contains the news that Give 1 Get 1 will be restarting in August or September.
Great, so the next iteration of the program will be starting up just as the last donor from late 2007 finally gets their long-overdue delivery!
After my experiences with the Grand Clusterfuck of G1G1 mk I, not to mention the Windows Compromise and the resignations of so much of OLPC's top talent, I'm discouraging everybody I know from contributing. Even if it's "Give 1/2, Get Three".
There seem to be three separate and distinct issues being conflated here:
1. India has laws that make it a crime to post "vulgar content"
2. Google provided information to Indian police in conformance with the law
3. Indian police are alleged to have badly mistreated a suspect
Be outraged about #1 and #3 if you wish, but I see no malfeasance inherent in Google's actions #2.
Why is it that I'm not even considered because I don't have a grade
You just answered your own question.
If you truly can "out-develop any graduate you've met", prove it. Go get the little slip of paper that gives them the career advantage over you.
You may even learn something at Uni that you didn't realize you should need to know.
A lot of people, sadly, seem to have a sense of entitlement, a sense that they deserve the jobs or have somehow earned them through no action whatsoever.
Is this something that has been empirically observed, or is it just a stereotype you've cultivated anecdotally?
The average salary for a Sr. Software Engineer in the US is around $90k according to PayScale.com. Not bad, that's more than I make. In India, however, the wage is about 580,000 rupees, or around $13,500.
1. Is the Indian figure for "Senior Software Engineer", as well?
2. Are the skill sets and duties of a "Senior Software Engineer" comparable in both countries? (IT job titles, you will find, are often not very well-defined, even within a single geographic market.)
You could pay someone in India pretty well by Indian standards and still save a ton of money by American standards.
Assuming, as bean-counters too often do, that the work produced by the $90K American and by the $14K Indian will be equivalent in value to the American company. And there are plenty of reasons not rooted in racism or irrational xenophobia that would refute that assumption.
The fact is that while a good education makes a big difference, the real question of how good someone is depends on how well they learn and how open-minded they are
I don't disagree with this; however, it has been my non-empirical experience that the culture of learning in the United States does put a greater emphasis on open-mindedness, innovation, and exploration than do the learning cultures of other countries currently exporting programming talent.
If they're taught Java in an American university and O'Caml in an Indian university, the American is going to have a better immediate skillset
*sigh* A university education is not meant to be vocational training.
American universities can pump out idiots just as fast as Indian universities do, they just do it for a much higher price.
Maybe so. But a hiring manager stands a better chance of accurately evaluating whether a new grad would be a valuable contributor to their organization when he's sitting across a conference room table and speaking the same American English dialect as the candidate, as opposed to being a nameless resource listed on page five of a project bid document that's being discussed in an international teleconference, represented by a team lead who speaks with a thick accent.
Yes, I do want my programs to "work great", not "just work".
And it's not just the developers that want this for reasons of pride!
EVERY CLIENT would prefer, all else being equal, to have software that works great than software that just barely meets their needs. Unfortunately, clients usually don't discover how poorly their needs are being met until well after the costs of software acquisition have been sunk.
Mirror the database to a 2nd server and provide them read access to that.
Before doing that, present them with a proposal that itemizes all the costs they'll be responsible for it you do this: $XXXX to purchase the hardware, XX hours at a rate of $XXX/hr to have an admin perform the software install and configuration, ongoing support costs of $XXXX/month, and so on.
You'll find out how badly they want the level of access they've requested by how much they're willing to spend to have it.
The OLPC was clearly designed as a free R&D project from the beginning. Not free as in speech, but free as in, "hey, lets CALL it a charity. That way we don't have to fork out money for our R&D".
So where did the funding for the OLPC R&D come from? Somehow I don't think Negroponte, Bender, Jepsen, and everybody else worked tirelessly for years without being compensated for their time. I don't think Quantas would agree to divert their production lines to make little green budget PCs unless it was made worth their while.
When the OLPC was listed out at $100 I said it was way too expensive. I went on line and found all of the components to build a hand powered computer for $89. Single Unit Pricing.
Oh, please do post this list. I'd love to see how you managed to put together a rugged laptop more affordably than the XO-1.
You seem to have a fundamentally different perspective about the needs of a technology-driven education program than the OLPC Foundation does, and than the OLPC Foundation did at its inception for that matter. Perhaps you would care to explain your position in more detail?
Why does dual boot require extra hardware??
Do you think XP and a Linux distro will both fit side-by-side in the 1GB of NAND memory that the XO-1 currently boots from?
For that matter, do you think XP will ever run with acceptable performance and security on a computer that has a sub-600Mhz CPU, only 256MB of memory, and almost no "disk" to speak of? Even by 2001 standards, when XP was first released, that's a little short of XP's recommended hardware specs.
You know that feeling you got when Walter Bender left the project over a disagreement with Nicholas? That "Wozniak has left the building" feeling? Turns out we were right.
I got that feeling when Mary Lou Jepsen -- the OLPC CTO, the person most directly responsible for the hardware design of the XO-1, including the fantastic transflective 1200x900 screen -- left the organization in December. As with the Apple Computer, it was the hardware innovation that served as the foundation for everything else groundbreaking.
80x25 white on black bash, baby.
You white-on-black elitists sicken me. What, amber-on-black not good enough for ya?
A $300 laptop that provides educational information for 5 or 10 years is a lot cheaper than the equivalent material in textbooks.
Only if the equivalent material is available via the $300 laptop.
So far, I have not heard of any educational textbook publishers volunteering to send electronic copies of their materials to developing nations gratis. Nor, as is necessary for most of the world, to translate them from their original language to those used in other locales.
So what are we left with? Wikipedia? As a replacement for pedagogically sound teaching materials?
The UI is a core issue. Why should it be materially different from what a billion computers already run?
Maybe -- probably -- there's something better than what those billion computers already run.
The OLPC leadership failed to understand that novel UI features were a nice-to-have, not a must-have, and failed to enforce a focus on the actual must-haves of the project.
To paraphrase them themselves: it's an education project, not a user interface project. Sugar's collaboration features, the Journal, the zoomable metaphor -- these are all good and promising concepts for Human Interface. But not one of them is core to the educational mission.
If the students are going to be able to go onward from OLPC, then their "language" must be "compatible" with the other "computers" they will see later.
Nonsense. I started computing at 8 years old running BASIC programs on a Tandy Color Computer, and I had no problem migrating later to using DOS 3.3 on an XT a couple years later, nor to each successive new OS and architecture.
Students, by definition, learn new things.
I would sell the shares but it's my 401k and all of the available funds are managed by the same company.
So what you're saying is that there is a limit to what steps you are willing to take to defend your principles, and that limit is not uncorrelated to your finances.
Remember this: no matter how nice your office space is, if you're an "interactive agency" with an unspellable/unpronounceable name like "Tocquigny", you're going to be the first to go out of business when the Dot-Com Crash 2.0 happens.
Enjoy the pretty scenery while it lasts.
moderation merits in Slashdot are hierarchical - the first moderators were wisemen chosen by the Mighty Taco Himself. Besides, anyone can metamoderate.
Which is not to say that Slashdot moderation is free from politicking -- despite having consistently excellent karma, the site stopped offering me moderation privs a couple years ago. I suspect this was related to my tendency to call a certain Slashdot Editor out on his habit of tacking his own personal commentary onto story submissions that he approved, instead of submitting a moderatable comment in response to the story like all the rest of us have to.
Surely CmdrTaco can pick whomever he wants to edit this site, and those editors can pick whomever they want to moderate. I just wish they'd be more transparent about the process.
* The 7+ billion dollar Xbox fiasco
Fiasco?
It cost them a huge amount of money, sure, but Microsoft managed to spend its way into a significant share of the videogame console market while competing with two most consistently successful companies in the history of the industry. Xbox Live is the premiere online console gaming service, and something that Microsoft has made precious few missteps with.
I don't think the Xbox division can be considered a fiasco by any measure, and even in strictly financial terms it's at worst a short-term failure with strong potential for long-term successes.
Were it not for the crappy BIOS, programmers wouldn't have had to resort to writing directly to hardware to get an acceptable speed on the screen.
You seem to be implying that programs doing direct hardware access for performance was not the norm across all platforms in the early 1980s.
In those days, hard disk drives were optional, CPU clocks ran at less than 10 MHz, and unless you were a "power user" you probably didn't even have the full 640KiB of RAM that DOS could use to run programs. With resources that tight, there wasn't really a lot of room for implementing the abstracted multi-layer system models that we now consider good practice.
Everything I have read and heard about Microsoft suggests that they are cowboys.
Code first, design later.
What about that story about how it took twenty people from six departments three months to reach a decision on whether to add some feature to the Start menu or not?
Maybe their developers have the cowboy mentality because there's a fundamental disconnect between them and the management bureaucracy that makes it impossible for anything to get accomplished otherwise...
The [WinCE] applications are pretty much there. When I'm on the road I need wifi enabled IE or Firefox to surf the web / do web enabled work. I need to view pictures, maybe edit a .doc or .xls. I need my calendar and the ability to queue up emails for my work mailbox (sync'ed with Outlook when I am anywhere near my work network.) That's about it - anything else is gravy.
To the road-warrior business traveler, maybe.
The platform is still pretty useless to the application developer, the artist, the musician, the scientific researcher, etc...
Not to dismiss whatever QA or shipping problems resulted in your Macs arriving with parts DOA, but... for what reason is it important for a business laptop to have working speakers in it? At least in my office, the sentiment is that the less noise your computer makes to disturb your officemates, the better.
And I wouldn't call your company particularly agile if they don't have a contingency plan in place for a single employee's primary computer being out of commission for a couple of days. No 'loaners' at all?
Apple has to realize if they want to compete, they need to open up a bit to their larger buyers.
I would have thought this was clear to everyone by now, but: Apple doesn't want to compete. They want to carve out a dedicated corner of market for themselves and own it thoroughly.
If proprietary stuff like "Flash" is required and Gnash isn't up to snuff (yet), doesn't it make more sense to as Adobe for a Flash port rather than throw the education deprived baby out with the bath water?
The standard Flash binary distribution for Linux can easily be run and installed on an OLPC laptop, if Gnash is considered insufficient.
The problem is that existing Flash content targeted to desktop computers with multi-GHz CPUs and a gigabyte-plus of RAM doesn't perform so hot on a 433MHz machine with 256MB. While I wouldn't be surprised to learn that the Adobe Flash + Windows software stack has certain optimizations which could improve performance slightly, we're still talking about going from two frames per second to three frames per second; still not tolerable for kids wanting to surf the Hannah Montana portal or adults to watch Homestar Runner toons.
They should find where their target demographic has commonalities
with people in the more developed world. They should have been making something
more like an Asus EEE PC to begin with.
So the XO laptop ought to have been just as rugged, but twice as powerful and also twice as expensive?
Would be great for us rich industrialized types, but it would have meant that the government could only afford to buy half as many, which would mean that only half as many children would be able to benefit.
Not to say there isn't some commonality between the needs of schoolchildren in the developing world and business travelers in the developed world -- for one, I'd love to see the XO's high-res transflective screen find its way into all kinds of hardware -- but in some aspects the requirements are entirely incompatible between the two.
If anything, the failure of OLPC so far has been a failure to create software yet that's as efficient and well-designed as the XO-1 laptop's hardware.
The leaders he mentions are not high-tech. Most people in those positions refer to the GUI on a computer as "Windows" whether it's Gnome, KDE or FVWM.
I don't know whether that's true or not.
What I do know is that if OLPC starts making hardware and software decisions based on what education ministry bureaucrats ask for, instead of what provides the best benefit to the students, they have already lost sight of their mission.
Honestly, I suspect that Windows XP/XO will never see a release. I think it's all just a ruse to keep OLPC distracted, and delay governments from making a purchase decision, long enough that a separate computing program for developing nations, centered on Microsoft Office for Vista no doubt, is ready for market. And what remains of the OLPC brain trust is falling for it.
If they're out of physical space (not just power and cooling) then the facility is way oversubscribed
Assuming that the new supercomputer has the same space needs as the one it's replacing, what you're telling us is that any machine room running at more than 50% volumetric capacity is oversubscribed. I can't agree with that.