For some context from a (small) country with more roundabouts (RAs) than traffic lights, please let me give some observations from personal experience...
Roundabouts are extremely efficient when there's light to moderate traffic levels, and no congestion in the "drains" (exits) from the RA. It is actually possible to cross an intersection while slowing only slightly.
The problems occur when the drains are full, so those in the RA can't exit, thus blocking those trying to enter. Ironically, this can occur when only one drain is congested and most who wish to pass through are exiting into an uncongested exit. This is because RAs set up the classic "Prisoner's dilemma" problem -- if I don't take the opportunity to enter the RA, even if my exit is blocked, you will enter, possibly blocking me.
I have proposed to the Government here (GoB) that one way to prevent this problem (or, at least, reduce it) is to pass (and enforce!!!) a law which says that one cannot enter a roundabout unless you will be able to exit it within 10 seconds. Or, set up a hybrid system, where traffic lights are installed which observe the current dynamic, and activate controls based on where traffic can currently exit when congestion is occurring.
However, these suggestions have not been implemented, and instead the GoB has spent tens of millions of dollars replacing several RAs with traffic lights. The congestion has actually become worse because of it....
To expand on this a little, VoIP is strictly *illegal* in many Caribbean countries. While the authorities tend to ignore individual users, Cybercafes and companies are often raided, and equipment seized and internet connections disconnected, for offering the service.
This policy make it explicitly legal for consumers (including businesses) to use VoIP in any manner they wished. The only restriction is providers of VoIP services require a telecommunications license if the transactions are conducted in Barbados. (Read: Skype, for example, does *not* require a license.)
Tsunayoshi, you have an opportunity here to inspire!
My advise would be to communicate to those listening that there is no such thing as Magic. *Everything* technical can be explained by and derived from fundamentals. By definition.
One thing I have always found startles an audience -- talk about the amazing empowerment that the HTTP protocol and the SMTP et al provide to we humans...
And then bring up a simple shell, telnet into the service port of a server (that you control), and then hand-type commands... Interact with the server directly.
The fundamental message is, simply: now a days things have been made to appear Magical; but fundamentally, it's all about agreed upon and open standards.
And, further... If one is willing, one can become a true "Wizard" -- knowing things and being able to do things which few others can.
A bit of a tangent, but this reminds me of a story a friend of mine had about a buddy who was doing research during the US nuclear bomb tests. He had a device taking a measurement near the bomb itself, and sending the signal down a wire for capture.
Immediately following the signal was a huge EMP burst. Immediately following that was the fireball, reducing the wire to plasma....
Needless to say, the engineering required to capture the data without blowing the hell out of the equipment was non-trivial. It also points out that there are many places where extremely high capture rates are required.
Some very interesting things happen very quickly....
Come on Peru, be a good little third-world country, and pay your Windows tax. How can the US economy hope to recover if M$ doesn't have additional markets to rape?
And you wouldn't want to risk those potential 15,000 jobs, now would you? No matter that any real employment will be exported to the US. No, don't bother thinking about that...
tcpdump shows a bunch of requests being opened with port 80, but never completing.
I'm not sure if this is a bug, or a feature of the worm. The DOS sessions seem to come on blocks of 16, the same number of requests made during the earlier infection period.
Last data-point of interest; I have access to networks 7000 km apart. The server off shore is seeing very little other than the 408's, while the other has yet to see them at all.
Anyone know of any actual products based on the HomePlug spec yet? Or at least, how the trials are going?
At 10Mbps over existing power-carrying copper, it could be a good solution for existing structures. Lets just hope it doesn't end up with the same types of problems 802.11b is now discovering.
Hopefully someone will start manufacturing home automation kit as well (light-switches with IP numbers, etc.) -- seems like the current primary focus is on providing TCP/IP between traditional computing and new multimedia devices.
It will be cool when the only wire going to the speakers is the power cord....
Form a tiny non-profit company. Incorporation is simple and cheap. Lawyers are not required (although advised.) IANAL.
Said company files for the trademark -- meets commerce requirements. Company immediately opens the codebase, "business" as usual.
This is also handy if inforcement of the trademark is needed in the future, as the company already exists to act as the party being harmed for the complaint. Trademarks not enforced are not held.
SA is only one way the civilan signal is degraded.
Also, the military service (PPS) has two frequencies to work with, while civilians can only decode one. This gives at least 5 meters of better accuracy.
They've got a cheesy JavaScript interface you've got to get through the first time, but from then on you should be able to just select it from the RP history menu.
I've been monitoring the NASA channel in the background with RealVideo, and there's been a lot of traffic back and forth about computer system and networking problems they've been having this trip. They keep having to reboot servers and bring the network up and down.
As they each describe directory paths, it's clear they're using Windows based machines. It also seems like they're using a human (on the ground) doing single-file-at-a-time transfer for e-mail rather than a proper MTA, and this can't happen when the machine is being used "in certain ways".
Bill McArthur has been spending a lot of time fighting with the systems so far, and this isn't the first time software problems have cropped up during flights -- a few back a software program refused to write log files until someone figured out the directory limit for files had been reached.
I have to wonder -- why the hell are they flying with such flaky systems? And why are all the systems running Windows? I'm not suggesting everything suddenly switch to Linux, but it would seem to make sense that the servers (at least) run something more stable than the ones currently flying have demonstrated themselves to be.
I don't know if this has been caused by the loss of the Ku-band (high-bandwidth), but regardless, perhaps it's time to form an Open Source project for space-flight software?
It was thought up a hell of a lot longer than 20 years ago.. Early 1960s, in fact.
Actually, I was refering to it as a compromise as a computer output device. As soon as RGB, non-interlace devices became available, NTSC/TVs should have been dumped.
An important thing to remember with NTSC (and PAL, the european standard) is that the display is interlaced. This means each frame (30th of a second) is made up of two fields of 60th a second. The one field contains all the even lines, the other all the odd. This means hard (high-contrast) vertical transitions will jitter; look for this on the CG effects of news-casts. Makes text very difficult to do well on TV.
The other problem, of course, is all that information is packed into one wire. Unlike your monitor, where your R, G, B and timing signals are seperate from each other all the way from your video card to your monitor's phosphers. NTSC has everything multiplexed down a single channel. Color is defined as an angle around a color-wheel!
Basicly, NTSC sucks. It was a compromise 20 years ago, and should NEVER be thought of as a serious output device for anything above bad movies, talking heads, and low-detail games. On the other hand, I'm not going to be buying a HDTV receiver for a long time. 16Cx12C multiheaded does just fine. 9-)
It's well known that SETI@home have about three times as much processing power as they have data to process. And with the export restrictions lifted on encryption out of the US, the RC5/64 project is pretty much masterbation now. So what does one do with spare cycles?
May I suggested the new Optimal Golomb Ruler project over at Distributed.net. Searching this space can only be done exhastively, and is actually useful (in, admittedly, rarified areas). The Great Internet Mersenne Prime Search is another project where the work will actually have lasting use.
Not to take away from Seti@home -- interesting project. It just became TOO popular. Ditto RC5 -- the project probably helped force the restrictions to be lifted, but it's point has been made; why spend another two years or so on it?
It continues to blow my mind that NewTek still appear to have no plans in place to port their amazing LightWave 3D rendering system to Linux. I hope this latest move will perhaps light a fire under NewTek.
They have versions for NT, IRIX and Mac, and with a completely custom interface, porting would NOT be a problem. No "which GUI shall we use" type problems, since Lightware does all it's own UI.
I encourage everyone who loves Lightwave to take this opportunity to write to NewTek and politely request a port. If you're in a buying position, point this out. They've always ignored such requests in the past, but they can't hold out forever.
When we were looking at what existing OSS license would allow us to release a project we were working on (a web-based portal system) while preventing someone from taking the codebase and modifying it and using it themselves to render services, without returning the changes.
In a brief discussion with RMS in Vancouver confirmed it: the GPL provides no such protection. In RMS' mind it doesn't matter, because it's not a matter of code losing its freedom. The fact someone else is gaining without contributing back doesn't bother RMS at all, but was not acceptable to the underwriters of our project.
A review of all the common OSS licenses do not provide protection in regards to this. For this reason, we ended up modifying the NPL slightly into the CommPort Public License (CPPL), which has explicit language about rendered services. If anyone's interested, have a look at http://www.tc.ca/commport/license.html.
This has been reviewed by several lawyers, although not yet by the OSI. Feedback from the OSI would be appreciated, BTW.
Assuming you want to see MSFT be held to task for their past actions, not having them broken up, and having Jackson hand down his Findings of Law instead of the DOJ/States/DoC being rushed with complicated legal language from MSFT, is quite possibly the best outcome we could have hoped for.
Sure it's going to be in appeal for years, but it will likely go straight to the high court and survive; Jackson's being VERY careful here. And a decision by a judge is a lot more powerful than some concent decree which can be ignored by future governments.
Lastly, by not breaking up MSFT, it means there's a single target for all the harmed parties to focus their (legal) efforts on. I expect MSFT's legal department will soon outnumber their SW development group.
Frankly, how the end-game plays out doesn't matter any more; the market has been able to recover somewhat with the beast under constant observation, and that isn't going to disappear any time soon. MSFT are having to compete on merit in a few areas, and as one might expect, failing terribly.
I'm hoping this might raise some interest at NewTek, who's Lightwave really needs to be ported to Linux. Those who've used the package know how amazing it would be to have this tool running natively under Linux. Even just the renderer would be a worthwhile first step...
My mistake. Please expand that bit to read "support cookies across sessions". Several sites I use maintain a lot of data in never-expiring cookies. I consider this a fairly major shortcoming, although as written it is wrong.
> And unless you install a browser as root, only one user can run it?
Generally, the bigger browsers with lots of support files do not work well when loaded into a users' account, or at least, permissions must be expanded beyond what is minimally required otherwise. We're trying to encourage our readers to install them as root because it's the norm, and generally easier.
Overall you seem pretty negative. Sorry you feel writing is so easy. A lot of work went into this and the previous article, but even so, mistakes will slip through.
Slashdot Mirror
Spreading the love... Sent.
Whoops. Replied to the wrong post above.
But invite received. Thank you!!!
Thank you very much!!!
I would be very appreciative of an invite. My username here on slashdot @chalsall.com
For some context from a (small) country with more roundabouts (RAs) than traffic lights, please let me give some observations from personal experience...
Roundabouts are extremely efficient when there's light to moderate traffic levels, and no congestion in the "drains" (exits) from the RA. It is actually possible to cross an intersection while slowing only slightly.
The problems occur when the drains are full, so those in the RA can't exit, thus blocking those trying to enter. Ironically, this can occur when only one drain is congested and most who wish to pass through are exiting into an uncongested exit. This is because RAs set up the classic "Prisoner's dilemma" problem -- if I don't take the opportunity to enter the RA, even if my exit is blocked, you will enter, possibly blocking me.
I have proposed to the Government here (GoB) that one way to prevent this problem (or, at least, reduce it) is to pass (and enforce!!!) a law which says that one cannot enter a roundabout unless you will be able to exit it within 10 seconds. Or, set up a hybrid system, where traffic lights are installed which observe the current dynamic, and activate controls based on where traffic can currently exit when congestion is occurring.
However, these suggestions have not been implemented, and instead the GoB has spent tens of millions of dollars replacing several RAs with traffic lights. The congestion has actually become worse because of it....
I developed this technique independently some time ago. So far none of the obscured addresses have been exposed.
Since the Googlebot doesn't appear to download referenced Javascript files, simple put the obscuring function into another file....
To expand on this a little, VoIP is strictly *illegal* in many Caribbean countries. While the authorities tend to ignore individual users, Cybercafes and companies are often raided, and equipment seized and internet connections disconnected, for offering the service.
Here in Barbados things changed only in August of 2007, when Cabinet passed the Barbados VoIP Policy. http://www.telecoms.gov.bb/Documents/Policies/PDF/voippolicy.pdf
This policy make it explicitly legal for consumers (including businesses) to use VoIP in any manner they wished. The only restriction is providers of VoIP services require a telecommunications license if the transactions are conducted in Barbados. (Read: Skype, for example, does *not* require a license.)
Tsunayoshi, you have an opportunity here to inspire!
My advise would be to communicate to those listening that there is no such thing as Magic. *Everything* technical can be explained by and derived from fundamentals. By definition.
One thing I have always found startles an audience -- talk about the amazing empowerment that the HTTP protocol and the SMTP et al provide to we humans...
And then bring up a simple shell, telnet into the service port of a server (that you control), and then hand-type commands... Interact with the server directly.
The fundamental message is, simply: now a days things have been made to appear Magical; but fundamentally, it's all about agreed upon and open standards.
And, further... If one is willing, one can become a true "Wizard" -- knowing things and being able to do things which few others can.
HELO!
Immediately following the signal was a huge EMP burst. Immediately following that was the fireball, reducing the wire to plasma....
Needless to say, the engineering required to capture the data without blowing the hell out of the equipment was non-trivial. It also points out that there are many places where extremely high capture rates are required.
Some very interesting things happen very quickly....
And you wouldn't want to risk those potential 15,000 jobs, now would you? No matter that any real employment will be exported to the US. No, don't bother thinking about that...
We're in sad, sad times.
It looks like Nimda may go DDOS after propagation.
access_log:
www.altered.com 139.50.200.221 - - [18/Sep/2001:20:55:51 -0700] "-" 408 - "-" "-"
tcpdump shows a bunch of requests being opened with port 80, but never completing.
I'm not sure if this is a bug, or a feature of the worm. The DOS sessions seem to come on blocks of 16, the same number of requests made during the earlier infection period.
Last data-point of interest; I have access to networks 7000 km apart. The server off shore is seeing very little other than the 408's, while the other has yet to see them at all.
At 10Mbps over existing power-carrying copper, it could be a good solution for existing structures. Lets just hope it doesn't end up with the same types of problems 802.11b is now discovering.
Hopefully someone will start manufacturing home automation kit as well (light-switches with IP numbers, etc.) -- seems like the current primary focus is on providing TCP/IP between traditional computing and new multimedia devices.
It will be cool when the only wire going to the speakers is the power cord....
Said company files for the trademark -- meets commerce requirements. Company immediately opens the codebase, "business" as usual.
This is also handy if inforcement of the trademark is needed in the future, as the company already exists to act as the party being harmed for the complaint. Trademarks not enforced are not held.
Also, the military service (PPS) has two frequencies to work with, while civilians can only decode one. This gives at least 5 meters of better accuracy.
See this article for more details.
The best source of a RP stream for the NASA channel is at Space.com. (URL: http://www.space.com/n ews /spaceagencies/nasatv_sched.html ) get you most of the way there.
They've got a cheesy JavaScript interface you've got to get through the first time, but from then on you should be able to just select it from the RP history menu.
Enjoy.
As they each describe directory paths, it's clear they're using Windows based machines. It also seems like they're using a human (on the ground) doing single-file-at-a-time transfer for e-mail rather than a proper MTA, and this can't happen when the machine is being used "in certain ways".
Bill McArthur has been spending a lot of time fighting with the systems so far, and this isn't the first time software problems have cropped up during flights -- a few back a software program refused to write log files until someone figured out the directory limit for files had been reached.
I have to wonder -- why the hell are they flying with such flaky systems? And why are all the systems running Windows? I'm not suggesting everything suddenly switch to Linux, but it would seem to make sense that the servers (at least) run something more stable than the ones currently flying have demonstrated themselves to be.
I don't know if this has been caused by the loss of the Ku-band (high-bandwidth), but regardless, perhaps it's time to form an Open Source project for space-flight software?
Actually, I was refering to it as a compromise as a computer output device. As soon as RGB, non-interlace devices became available, NTSC/TVs should have been dumped.
The other problem, of course, is all that information is packed into one wire. Unlike your monitor, where your R, G, B and timing signals are seperate from each other all the way from your video card to your monitor's phosphers. NTSC has everything multiplexed down a single channel. Color is defined as an angle around a color-wheel!
Basicly, NTSC sucks. It was a compromise 20 years ago, and should NEVER be thought of as a serious output device for anything above bad movies, talking heads, and low-detail games. On the other hand, I'm not going to be buying a HDTV receiver for a long time. 16Cx12C multiheaded does just fine. 9-)
May I suggested the new Optimal Golomb Ruler project over at Distributed.net. Searching this space can only be done exhastively, and is actually useful (in, admittedly, rarified areas). The Great Internet Mersenne Prime Search is another project where the work will actually have lasting use.
Not to take away from Seti@home -- interesting project. It just became TOO popular. Ditto RC5 -- the project probably helped force the restrictions to be lifted, but it's point has been made; why spend another two years or so on it?
They have versions for NT, IRIX and Mac, and with a completely custom interface, porting would NOT be a problem. No "which GUI shall we use" type problems, since Lightware does all it's own UI.
I encourage everyone who loves Lightwave to take this opportunity to write to NewTek and politely request a port. If you're in a buying position, point this out. They've always ignored such requests in the past, but they can't hold out forever.
In a brief discussion with RMS in Vancouver confirmed it: the GPL provides no such protection. In RMS' mind it doesn't matter, because it's not a matter of code losing its freedom. The fact someone else is gaining without contributing back doesn't bother RMS at all, but was not acceptable to the underwriters of our project.
A review of all the common OSS licenses do not provide protection in regards to this. For this reason, we ended up modifying the NPL slightly into the CommPort Public License (CPPL), which has explicit language about rendered services. If anyone's interested, have a look at http://www.tc.ca/commport/license.html.
This has been reviewed by several lawyers, although not yet by the OSI. Feedback from the OSI would be appreciated, BTW.
Sure it's going to be in appeal for years, but it will likely go straight to the high court and survive; Jackson's being VERY careful here. And a decision by a judge is a lot more powerful than some concent decree which can be ignored by future governments.
Lastly, by not breaking up MSFT, it means there's a single target for all the harmed parties to focus their (legal) efforts on. I expect MSFT's legal department will soon outnumber their SW development group.
Frankly, how the end-game plays out doesn't matter any more; the market has been able to recover somewhat with the beast under constant observation, and that isn't going to disappear any time soon. MSFT are having to compete on merit in a few areas, and as one might expect, failing terribly.
Actually, that would have been Einstein. Newton couldn't be humble if his life depended on it.
http://www.newtek.com/ -- write them and let them know you'd BUY a copy!
My mistake. Please expand that bit to read "support cookies across sessions". Several sites I use maintain a lot of data in never-expiring cookies. I consider this a fairly major shortcoming, although as written it is wrong.
> And unless you install a browser as root, only one user can run it?
Generally, the bigger browsers with lots of support files do not work well when loaded into a users' account, or at least, permissions must be expanded beyond what is minimally required otherwise. We're trying to encourage our readers to install them as root because it's the norm, and generally easier.
Overall you seem pretty negative. Sorry you feel writing is so easy. A lot of work went into this and the previous article, but even so, mistakes will slip through.