nweaver (113078) writes "Many Slashdot readers are no doubt familiar with Netalyzr, our free, comprehensive network measurement and diagnostic tool that runs in the browser using Java. For those that aren't, its checks a ton of network properties and provides a handy report. At the same time, Slashdot readers also know that Java should probably be removed from the browser. We've been hard at work on a solution: a Android port of Netalyzr, which is both free and advertisement free. We implemented the full Netalyzr test suite, test run in the background (so you don't need to wait), and if your debugging someone else's network, you can have them run Netalyzr and share their results with you. Help us understand what works on the Internet, and what doesn't." Link to Original Source top
nweaver (113078) writes "Bad news for patent trolls: Someone is trying to patent being a patent troll. Yes, Clive Menendez, on behalf of Halliburton, is trying to patent "Patent Acquisition and Assertion by a (Non-Inventor) First Party Against a Second Party". Lets hope it gets approved, pity there is probably too much prior art out there already." Link to Original Source top
nweaver (113078) writes "Some Slashdot readers may already be familiar with our Netalyzr service, from this June story. For those who aren't, Netalyzr is a free network measurement and debugging applet designed to check for a wide range of network problems and neutrality violations, including unadvertised port filtering, DNS wildcarding, and hidden proxy servers. We are pleased to announce that Netalyzr is now out of beta. We've made many enhancements, user interface cleanups, and added a bevy of new tests such as enhanced DNS probing and checking for problems with fragmented traffic. Since the Internet is changing constantly, we would love it if Slashdot readers would (re-)run Netalyzr so we can see how things have evolved since June. More generally, the Netalyzr project aims to compile a comprehensive survey of the health of the Internet's edge. Your help in making the study a success is greatly appreciated — thanks!" Link to Original Source top
nweaver (113078) writes "How healthy is your Internet connection? Do you know if some
outbound services are blocked? Are there hidden proxies or HTTP
caches? Are there problems with your DNS server? We have developed a
free service to help answer these questions: the ICSI Netalyzr, now
publically available. This Java applet, developed by researchers at
the International Computer Science
Institute, allows you to probe your own network to discover
various properties and problems. The numerous tests include
discovering hidden HTTP proxies and caches, checks for port filtering,
IPv6 connectivity, latency, bandwidth, and buffer properties, and DNS
server health." Link to Original Source top
nweaver (113078) writes "At Defcon, probably the greatest hack was Alex Pilosov and Tony Kapela's demonstration on Stealing the Internet, or "All your Routes are belong to Them". This hack is simplicity itself. It has long been known how to hijack someone else's route on the Internet to block their traffic: just advertise a smaller block. It happened to YouTube when Pakistan decided YouTube was bad. But Pilosov and Kapela's attack is far more powerful, allowing the attacker to receive all traffic and forward it on to the victim, as a man in the middle. Yet it is so simple: The attacker just haves a second path to the Internet that he uses to return the traffic to his victim. So when the attacker advertises his hijacked route, he includes the return path to his victim. Now all packets except those on the return path go to the attacker, who can record, modify, and do what he pleases (such as reduce the TTL to evade traceroute detection) before returning them through the return path and onto the victim. Voila: all return traffic to Defcon's network was passing through New York! Yeup, A Man in the Middle on the Internet." top
nweaver writes "In a response to the LA Times editorial on copyright covered by Slashdot, a response was published in the LA Times arguing Copyright This!. Namely, a key observation: "If Intellectual Property is actually property, why isn't it covered by a property tax?" If copyright maintinence involved paying a fee and registration, this would keep Mickey Mouse safely protected by copyright, while ensuring that works that are no longer economically relevant to the copyright holder pass into the public domain, where the redisdual social value can serve the real purpose of copyright: to enhance the progress of science and useful arts.
nweaver (113078) writes "I purchased a XO laptop through the "Buy one, get one" program, as a
platform for experimentation, a toy to play with, and because of the 1
year T-mobile service. In many ways, the system is a disappointment.
Although the hardware is brilliant and (almost) flawless, the software
is so crippled that it is useless.
Also, a friend who works at Intel Research, when he discovered I was
writing a review of the XO, loaned me a classmate to play with.
First, the XO.
The XO Hardware:
The XO's hardware is remarkable in its simplicity, ruggedness, and
reliability. The convenient handle makes it easy to toss around (so
much that I've dropped it a couple of times, as it just invites casual
The hardware has just enough horsepower to be reasonable, with a 433
MHz x86 processor, 256 MB SDRAM, and 1 GB of flash memory. Battery
life appears good for a many hours if the software supports the power
management features, and the LiFeP battery can fully charge in about 2
hours with the computer active.
Considerable effort went into making the hardware rugged. By placing
the guts of the computer behind the screen, rather than in the base,
only the power, keyboard, and trackpad connections need to pass
through the hinge. This is in sharp contrast to a normal notebook,
which requires a wire intensive and failure-prone connection through
the hinge between the monitor and the motherboard. The hinge itself
also appears very strong and designed for years of abuse, while
allowing the screen to go into a tablet mode for closed-system
The rest of the packaging is also designed for longevity. The
membrane keyboard is splash-resistant and the plastic tough and
durable. Since there is no fan or hard drive, the components are
naturally shock and dust resistant. The stated 5 year lifespan goal
seems more than achievable, given the basic design.
The power requirements are thrifty. All it take is 16W at roughly 12
volts (with very wide tolerance) to run the circuitry and charge the
laptop. Even with a 15 Watt-Hour battery and power management
software almost completely broken it is still good for about 3 hours
There are only two "flaws" in the industrial design, one real and one a
matter of philosophy. The first is a lack of mouse control in eBook
mode. Without an additional 4-way switch and two mouse buttons, one
needs to reopen the laptop, rechange screen orientation, and use the
internal controls to navigate beyond the currently displayed document.
If there was mouse hardware on the display, the eBook mode would be
much more usable.
The second is the small keyboard. The keyboard has an awful feel,
but thats the price of sealed membrane switches and worth it for the
longevity. But the keys are unusable small. The XO is designed to be
used exclusively by children, rather than primarily by children. The
designers consider the unusually small keyboard a feature, not a bug,
as a way of deliberately discouraging adults from coopting a child's
In contrast, I believe that the device has so much potential that
I wonder why it should be deliberately limited to children? Shouldn't
adults be able to benefit from this powerful platform? Making a real
keyboard wouldn't make the laptop that much easier to steal, its just
that adults would no longer need to be either frustrated typists or
have to acquire a $10 USB keyboard.
The best part of the XO laptop has to be the display. Although
currently unique to the XO, expect this display technology to become
commonplace in eBook readers and many other low-power devices.
In passive mode (no backlight), the display is a high resolution
reflective 200 dpi display. In this mode, it is very crisp and an
excellent platform for reading text, even 2-column scientific papers.
Active mode (color with backlight) is still solid but not quite so
remarkable. In this mode, there are distinct fine diagonal bands even
when viewing black and white material due to how the pixels are
arranged on the display. But colors are crisp and bright, and it is
easily readable in variable light conditions. The tolerance for
conditions ranging from bright glare to pitch black is far greater
than any other notebook screen, although colors do wash out somewhat
in full sunlight.
I would expect this display technology to become commonplace in
many future devices. Between the low cost, very low power operation,
and ability to work anywhere from bright sun to pitch black should
make this basic design the de-facto standard for future devices.
The XO's Software:
While the hardware shines, the software groans. The first thing
which is striking upon turning the system on is how slow it boots. It
takes 1 minute and 45 seconds to boot, which is strange for an
embedded computer: a fixed platform with very fixed functionality.
Its not intollerably slow, but I expected faster.
While the hardware designers were creating a new and innovative design
out of largely standard components or processes, the software
designers decided to effectively "start from scratch".
You will hear that the XO is Linux. This is not true. Underneath the
hood, Sugar's system is Linux and X11, but the GUI (and therefore all
the high level GUI APIs) are new. The XO doesn't run Linux
applications, they need to be ported to Sugar.
Thus instead of starting with a high quality web browser (Firefox), a
solid email program (Thunderbird), and tolerable productivity
applications (Open Office), like every other Linux distribution, the
XO has a crippled web browser, a toy word processor, and no email
Not only does the new GUI throw out almost every existing application,
it even throws out the file system. Gone is the notion of saving
files. Instead, you "save" your work in the journal. Since the
journal only has time and application, you can easily load an old copy
of the file rather than the current version. Yes, its really that
Furthermore, the fileless model actually makes it easier for a student
to lose work, since not only is the model fileless, the application
model is stateless. So close an application (with no acknowledgment)
and the work simply goes bye-bye.
Thus if you want to use a bookmark in the web browser, you need to
save the state of the browser into the journal (explicitly saving the
current instance of application state), and instead of launching the
web browser again the next time, you have to dig through the journal
for the last saved copy.
And a further consequenced, since there is no notion of files exported
to the user, editing multiple documents or viewing multiple pages
creates separate application instances.
The GUI model also doesn't support any sort of detailed menu, just a
strip across the top containing icons, and a strip underneath of text
saying which set of icons can be selected. Which is fine for a toy
application such as the video recorder or the music toys (which are
amusing and effective), but really suffers for real tools like a word
processor or a web browser.
On the plus side, the toys are cool. The flowchart-logo is
interesting, reminding me of ChipWits from my own youth. The music
games are cool, and the camera works well. They should work well in
attracting a child's attention.
But it brings up the question of whether Sugar can be saved. Although
one applauds the goal of simplifying the UI, simplification doesn't
mean crippling. There are plenty of simple UI examples, such as the
iPhone and Palm, or even the original Macintosh, which offer vastly
more usability, succesful state management, yet don't grossly
Between changing the menu structure, the file paradigm, how multiple
apps are launched, how the system handles evolving screen size, and
removing the numerous imperfections, is it better to soldier on with
Sugar or instead modify Gnome or KDE (or Windows XP for that matter)
and port the toys to a new environment?
Detailed XO Bugs:
The current version of Sugar also has serious performance and power
management issues. The hardware can support 10ms power management,
shutting down the CPU when it is unused, even between keypresses. Yet
the current environment doesn't even turn off the backlight when the
lid is closed, instead waiting for the sleep mode to kick in after
Even closed and asleep, the battery drains completely overnight, and
the mesh networking alone (I presume) is the source of an unexpected
drain where a fully off XO will drain the battery to nothing over the
weekend. This isn't that suprising, as with only 15 Watt-Hours, a
250mw drain will empty the battery in that time.
Performance suffers due to the heavy use of python. The web browser
often shows tearing when scrolling using the cursor keys. The rest of
the system is sluggish. The GUI has functional glitches,
including non-working cut and paste. And the system doesn't support
WPA-protected WiFi networks.
Finally, I haven't started investigating the security model, but it
seems to suffer from the iPhone/WinXP's "Everything running in the
same user who's root equivalent" problem. Open up the terminal, su to
root as no password, and do a/bin/rm -rf/*. Voila, a nuked XO. I'd
forgive the Windows XP model if the project did not toute security as
a primary objective of Sugar.
Fortunately, an Internet connection, a USB drive, and a working
computer is all it takes to happily reimage the system. Like so much
of the system, even the software is designed for easy "replacement" if
the system is corrupted.
Concluding thoughts on the XO:
The XO's hardware is simply brilliant, as is the low-level firmware
which allows trivial reimaging. But the software is a disaster, and
no kinder words are possible. Now there is the "this is beta
software" argument, but I think the problem is more fundamental (the
lack of files, the need to port every app to a crippled GUI) and thus
damages the overall system.
Hardware design is considered "hard", so it was handled by real
experts, who had a real budget, and who did a superb job. The only
new hardware technology was the screen, but even this was designed
around existing manufacturing according to Mary
Lou Jepsen, the CTO responsible for overseeing the design.
But software is "easy". Thus it did not receive the same development
budget or care as the hardware. Simply comparing the maturity between
the two makes it obvious: the hardware works as advertised, the
software is "it will be better at Update.2", but with nothing
addressing the fundimental limitations of the GUI.
This is probably worsened by the "not invented here" attitude among
software developers, and he "ecology will provide" open-source model.
The resulting Sugar software is a disaster.
While the XO was a carefully constructed hardware project, the Intel
Classmate was obviously thrown together as a basic ruggedized low-end
laptop. The hardware, although theoretically more capable, is vastly
inferior for the intended use in third world countries with several
weaknesses, including cost, power, cooling, and robustness.
The software (or lack thereof) is another problem.
The Classmate's Hardware:
Although physically similar, the Classmate is constructed very
differently. There is a screwed down cover over the laptop, which
holds the system closed with a magnetic clasp. It is almost 5mm
thicker than the XO but otherwise the dimensions and weight are almost
The magnetic clasp in the cover, although stylish, can be annoying as
it flaps around free on your lap when open and will also stick to the
trackpad. Additionally, the cover is screwed down, keeping a child
from accessing the battery bay, which is probably a good idea.
The keyboard is significantly nicer, with real keys and a real
keystroke. Although almost identical in dimensions, the keyboard is
vastly easier for an adult to type on. However, I'd worry about
longevity, as dirt could easily get stuck under the keys and jam them.
The ports are effectively the same, except the SD slot isn't sealed
(its just hidden under the cover) and the Classmate adds a 100 Mb
Ethernet while removing the camera and one of three USB slots.
Performance wise, the Classmate is substantially more advanced, with 2
GB of flash (instead of the XOs 1), a 256 or 512 MB DIMM, and a more
agressive processor running at 900 MHz. But this has a price, namely
$300 or more (depending on memory) instead of $200.
This additional computing affects the power budget. Rather than
requiring 16W at a (world standard) 12V, the Classmate needs 65W at
20V. I suspect this is vastly over specificed (my Macbook uses a 60W
supply, and I can't believe that the Classmate uses more power).
Instead, Intel probably went with an off-the-shelf power supply and
the resulting PSU was 60W. But even if the power supply only required
30W, the use of 20V is problematic, because the Classmate can't run on
car batteries or similar sources.
A further concern is the effect on reliability. To dissipate this
heat, the Classmate has an internal cooling fan and openings for air
to flow in and out, over what appears to be a largely exposed
motherboard as well as internal heatsinks. Even in the land of paved
roads, windows, and Roombas, I've lost far too many computers to dust
over the years. Fortunatly the fan generally stays off most of the
time, which reduces the dust issue, but it would be a more robust
system if it was passively cooled without venting.
It is also not clear how the Classmate can be repaired, while the XO
has reparability built into the design (such as the replaceable
backlight and color-coded screws). The Classmate certainly has more
screws, just when examined. As this is a loaner system, I can't
attempt a dissection.
The Classmate's Display:
The classmate's display is slightly nicer than the XO in color mode,
because the standard LCD doesn't have the grainyness of the XO's
transflective screen. However, for reading text, the XO is superior
thanks to the greater number of pixels (1200x900 rather than 800x400)
when the XO is in reflective grayscale mode.
The Classmate's display is also constructed like a normal laptop:
a thin display with a complex cable running through the hinge, to a
heavy motherboard in the base. This makes the laptop balance more
nicely, but will probably limit robustness (laptop hinges are a common
failure point, often with the monitor display wires breaking). The
display is also conventional, so there is no eBook or Tablet mode.
The Classmate's Software:
What software? Instead, its a very vanilla if stripped down XP
installation with office and a single groupware app which I can't seem
to run (but as
this system has been passed around, this may have been disabled by a
It lacks the cool toys of the XO: no making videos, no playing music.
Thus it will have more difficulty grabbing a child's attetion. But
the real applications (web browser, office application) are far
superior, as they actually work. It also takes about 20 seconds longer than the XO to fully
However, although XP runs well on the Classmate, there are a couple of
glitches. In typical manner, the logged in user is the Administrator,
so it is equally easy to nuke the system as the XO, and I'm unfamiliar
with reimaging the system (so I did not attempt this).
Windows XP doesn't like the 800x400 display. Instead, it wants an
800x480 display, so software either squishes the display or you have
the "scroll around the screen" effect, neither of which are
I believe the XP interface is far too rich to start with for children,
but a simplified shell would probably work wonders for the initial
usability. Yet once a child is comfortable with the system, XP is far
Hardware wise, the Classmate pales in comparison to the XO. It
costs more, eats more power, and can never be as robust. The
Classmate is a conventional laptop, while the XO really is the
promised revolution in laptop design. The Classmate simply is not a
viable competitor in its current form.
Software is another story. The XO has better toys, but is so crippled
by the user interface that Sugar only seems suitable as a toy.
The Classmate has enough software to be a real tool, with real
applications and a notion of files. But it lacks a "simplicity mode"
for young children, and Windows XP is not happy with such a small
I'm hoping that Microsoft's interest in the XO is real. For all its
annoyances, XP is vastly more usable than Sugar. The XO running
XP could be a real tool, either as a starting point for a
child-friendly layer on top, or in a "Start with Sugar, graduate to
But as is, the Classmate doesn't have the robustness or the cost. But the XO currently doesn't have a working environment which does justice to the hardware. Hopefully both systems will evolve towards workable solutions." top
nweaver writes "Popular Science Magazine has announced its Innovation of the year, which are Nano Solar's Print on Solar Cells. Unlike conventional solar cells, these are printed onto sheets of flexible aluminum, with the company claiming a cost of $.30/W for solar cells. Nano Solar's Factory for producing mile long rolls of solar cells is almost online. The potential is staggering. Even assuming that the completed cells, in a household system, cost $2000/kW to produce, this will easily undercut electricity as even at just $.10/kwh and producing for just 8 hours per day and 300 days in a year, a solar installation with such cells would have a 12%/year return on investment. We may be only a few years away from the Solar Age." Link to Original Source