You don't need non-stop for trading. Non-stop is for things like ATC. Non-stop with tightly coupled processors tends also to lead to bug-compatability, i.e. a software fault on one is repeated upon the other.
What you need is minimal downtime (in the region of seconds) and zero loss of data. Eurex (the biggest electronic financial derivatives exchange) uses OpenVMS clusters. All critical hardware is doubled and unless someone makes an almighty boo-boo (like incorrectly sizing an order book), downtime is limited to a few seconds.
The time delay is caused during a host failure while the distributed lock manager remasters the locks. At this point, the market is held so nothing is lost.
The biggest Electronic Derivatives Exchange
on
NYSE Goes To Linux
·
· Score: 1
Runs on good old OpenVMS!!!!
Yes, Eurex uses OpenVMS at the host and middle layers whilst NT and Solaris are relegated to the members sites.
The reasoning is that VMS is very good at message handling and it very rarely falls over.
One of the largest p2p corporate networks in the eighties was Digital's Easynet. All you needed was an address and you could hook up. Most Digital staff had no idea of where nodes were and who ran them as addresses would be reused.
A guy I knew wrote a network mapper utility in the early eighties (maybe after the Parc worm so not the first). All it did was to go round looking at adjacent nodes, copying and submitting itself to each system as a network task running a shell script.
Of course, the corporate network was brought to its knees and default "Task" level access was turned off. A few days later, things were back to normal.
Management were not amused. Digital was already a well connected company and relying on the Easynet for day to day operations. The guy evenually admitted to it many years later and as he was by then a senior network consultant, there was't a lot they could do.
The thing worked because most people there were running a single OS, OpenVMS, which was left largely unprotected.
It was alway the principle of patents that an inventor was granted a limited monopoly on the sole condition that they disclosed the idea in full, so it can be studied and improved upon. The extension of this is that in the case of a program, the algorithm must be suitably and accurately described, or as you say: documented source code.
Recent case law in the US already has started to limit the automatic extensuion of patent cover from what was clearly described in the application.
At the same time, we have to be flexible enough to acknowledge that an algorithm in Java is an reimplementation of a patented algorithm in C. More difficult, unless you get a much better scheme for the vetting of patents.
That is a very good point, but Brussels is to lobbying in Europe, what Washington in th US. Many commercial lobbyists are already in place and the number is increasing, unfortunately we don't have so much in the way of non-commercial interests like an EFF for Europe.
Let us be explicit about this. ROT-13 or the cesarian cipher was used for data masking purposes and not for encryption in the Usenet. It was used to hide upsetting information from those who see it by accident, i.e. strong or obscene language. Those wanting to read the material would ROT-13 it back again.
The other point is that many Usenet readers have ROT-13 functions, are they all in breach of the DMCA too?
Write to Amnesty International. Explain that the guy is be held for breaking a cipher system used by Ceasar that is being protected by a an act of Congress (the DMCA).
He is a political prisoner being held illegally and should be supported as such.
China has 20 or so CSS-4 ICBMs targeted at US cities.
The article mentioned discusses range, it does not discuss the Chinese targetting lists. I do not believe that very many people have acess to such lists, particularly in the US.
China is more likely to have missiles tasked on Russia who they share a massive land border with, and with whome they have had far from ideal relations.
Lets come down to it, a country with the ability to create a single nuke wouldn't want to take on the US except through third-party terrorism. If a country attacks the US, then the US can retaliate massively, without using any nuclear weapons.
Oh and it will be BS about nuclear detection at borders. An assembled bomb is usually well-shielded and is easily handled at no risk. The only real chance of detecting a bomb (rather than just a pile of fissile materials) would be through other means such as with neutrons.
There are plenty of places that the US govt could put that money used for the NMD project that would genuinely reduce tension.
Sorry I don't know any "Rebirth Island". It is Vozrozhdeniye Island!!!!!
It is serious stuff and the area is considered to be hazardous. The only people allowed opn the island are the specialists and US consultants involved in the risk assesment. However, the area nearby remains relatively accessible.
I only worked in Uzbekistan for a while. I didn't work anywhere that part of the country and would not go there. I do know people who have though.
Or over-winter in Antartica? Generally the people you are with speak the same language and the bits like sockets are compatable, but if you run out of anything, you are even more isolated because of the weather.
Perhaps the astronauts should be sent there for part of their training?
Think damages. If they damage your computers, they should compen$ate you. If their employees don't know the difference between Windows and Solaris, then sue them.
Most free and commercial distributions are available in Russia. I can't say in detail, but the copies of commercial distributions are faithfully done and complete.
Some local dists are a litle more "Russian-aware" but last I heard, the tendancy is to go for Mandrake and then download extra software, i.e. fonts and so on from the net. The software seems to be open-licensed.
Sent to the lawyers by email. I am also writing from Germany so have something to say about this.
Dear Sirs,
I am writing to warn you that you are attempt to protect the interests of Adobe corporation has and will cause you much adverse publicity. Whilst I respect your position, the name in question is a) artificial and b) is based upon what the product does vis-a-vis document illustration. I realise that you are not an English firm and can theerfore not be aware of the normal use of words in the English language. You would be able to sucessfully argue the case in court but only with significant costs. However, Killustrator only works on a platform for which Adobe does not provide any support (Linux/KDE).
I am not sure whether you know about the OpenSource and Free software movements. The creator of the program has made the program available, free of charge, to anyone who has access to the Internet and there is no restriction upon secondary distribution. The end result is that copies may be present anywhere from New-Zealand to Uzbekistan. The creator did not perform any other action than to leave the files in a place that is acessible by the public. The act of copying was performed by the users.
I would strongly suggest that you request only a change of name, rather than the destruction of the programme. Otherwise you will find that your company becomes a laughing-stock on the Internet. The Open-Source/Free-Software movement is extremely popular in Germany and Europe and you *will* lose clients.
I am writing to you in my private capacity in the hope that you will avoid putting the law on intellectual property in Germany and the EU into disrepute. Incidentally, I do not use this particular program as it is intended for working in an environment that I do not use.
I use the word scheme advisedly and would use it in the case of any mechanism that an operator uses to charge for airtime and number rental (meaning network access). These are fairly abstract and the ways of recovery a cost and making a profit are also varied. The number of providers and resellers in Europe is a testimony to this.
Does the US force their mobile ideas onto other countries, well actually yes, and the passing off of second-hand equipment to go with it. The countries concerned, worry that their sales are too low. Some of them are using a hodge-podge of GSM technology but US-style charging. This offers technical advantages over straight AMPS or US style TDMA but no economic advantages (and all the issues with vulnerability to fraud).
I agree that there other plans that include transparent roaming at no-extra charge, but this isn't cheap.
As for the ability to use CLI to filter calls, it doesn't always work when you roam. Definitely not always with GSM either and in any case the penetration of CLI even from fixed to mobile in one European country is incomplete.
As for free local calls, this is another question, but a fixed-line local call is not the same as a mobile call. With a mobile call, you pay a premium to reach the person that you have called. With the local call, it is a matter of luck whether the called person is there. Does it reduce the number of calls? Well, no not really. From current studies, the European model seems to generate a greater degree of market penetration and greater income for both the provider and reseller. This is not a philisophical point, this is one of profitability.
Probably the greatest issue with roaming at the moment with the GSM model is the automatic call forwarding. that is, if I have a German telephone and am booked in a French network, why does a call from France have to go via Germany?
Please don't think that I am knocking the US on this one, but as with PAL vs NTSC, version 2 is generally an improvement on version 1.
No, the frequency isn't the problem. The issue is with receiver-pays and the lack of wide-area access to carrier without roaming. This scheme didn't just fail in the US, it failed in every country that they forced it on.
The point is that mobile numbers have a special area prefix, and then they are nationwide. The caller pays extra, but knows this because of the prefix. The receiver doesn't pay unless they are out of the country.
End result is that with the US-type system, private users like to keep their numbers quiet because of the cost issue whilst with the Europeanm type system, users often give out their mobile numbers in preference to their home numbers because of the ease of reachability.
This is an economics issue not a straight technical issue.
No, the takeup of Email has nothing to do with SMS popularity. EMail generally means finding somewhere to hookup to the Internet. WAP doesn't reall seem to cut the ice here so mobile Email is relatively painful. The business that SMS displaces si actually the pager. Pagers other than for certain special purpose uses (i.e. news pagers) are largely dead in most of Europe
SMS is extremely cheap from the network viewpoint It is as portable as the mobile telephone and it works better than voice when the coverage is poor or the network is congested. An example of this is the use of SMS for rescue calls, one the other side of the world.
My kids have mobiles and a fairly tight allowance for calls (intentionally so, otherwise they would bankrupt me!!!). This is not atypical of the situation with other teenagers. SMS stretches their budget and allows them to use the network in a more efficient way than with voice. The pricing scheme is such that they pay much more per network packet with SMS than with voice.
This is why I disagree with you about first the Email question and then the network economics question.
Unfortunately, I can't say the same. I was doing some work on a client site. They had routers, but they had been bought s/h and no support contract was yet in place for them. It wasn't even clear who could support them at one stage.
We did a setup at the client site and had a load of problems with the DECnet protocol stack. IP worked great, but we needed an IOS upgrade at a weekend. Despite even having a credit-card at hand, it was almost impossible to purchase the upgrade. We eventually had to buy the s/w in another country (it was not downloadable) and then the local distributor (who we eventually identified)was able to release their copy of the s/w to us.
Cisco themselves were extremely uncooperative, passing everything down the line to distributors and resellers who had neither the inclination nor the skills to provide 24x7 support.
The client was a securities trading house, not the best people to upset when you look to the share price and Cisco were definitely bad-mouthed over that one. As a contrast, another client of mine had a memory module go late at night. The system was supported only on a 9-5 contract. We had no problem to get an engineer and the part and we only paid for the labour cost, not that of the parts, as it could have failed 9-5.
Maybe Cisco works better in the US, but in Europe, I can't rely on them.
Not relevant. Van Gogh is not painting from fantasy but from his perception of the scene. He starts with activated dyes in the back of his eyeball, this starts with halide crystals (or even direct to CCD imagery).
To synthesise information in the way that Van Gogh does is not difficult. What is difficult is to come up with a new and unique method of interpretation.
Note that an improvement to this technique would be able to work from stereo-pairs, as theg artist is encoding depth information as well in the representation.
As a side point, it is thought in some quarters that Van Gogh's medical conditions and possible absinthe consumption may have contributed to his perception of colour.
Alpha will stay around for ten years or so.....
on
Alpha Up For Grabs?
·
· Score: 1
One of my clients is the systems division of an financial securities exchange. The host and intermediate layers run on OpenVMS/AXP clusters. This thing needs uptime and message passing (not byte streams). Porting would be a major headache, whether to Windows or to Unix. Incidentally the code base on the VMS side is Cobol, C, Fortran, Pascal.
Another major client is the US military, they also have a bag load of applications that just won't port easily.
The implication is that if Compaq pass the technology on, the customer (Intel) will have to continue to produce so that Compaq can fulfill its high-end systems contracts.
What attracted Compaq to Digital was not Alpha, it was the Digital service infrastructure and the client list for high end systems, who they hoped to be able to sell PCs to.
Compaq PCs had a number of problems that didn't endear themselves to corporate customers, so this never really worked. Even though these issues seem to have been largely addressed (however, my little Ipaq illustrates the bad QA that Compaq has suffered from).A good little beast but let down by a losy sound jack that is being echoed by many other users.
What you need is minimal downtime (in the region of seconds) and zero loss of data. Eurex (the biggest electronic financial derivatives exchange) uses OpenVMS clusters. All critical hardware is doubled and unless someone makes an almighty boo-boo (like incorrectly sizing an order book), downtime is limited to a few seconds.
The time delay is caused during a host failure while the distributed lock manager remasters the locks. At this point, the market is held so nothing is lost.
Yes, Eurex uses OpenVMS at the host and middle layers whilst NT and Solaris are relegated to the members sites.
The reasoning is that VMS is very good at message handling and it very rarely falls over.
Think of it as stage directions. I can image a number of quite legitimate uses for such a tool.
Perhaps we should start by sending them Comer & Stevens as UDP (I wouldn't wish the RFCs on anyone) and follow it with a connect?
A guy I knew wrote a network mapper utility in the early eighties (maybe after the Parc worm so not the first). All it did was to go round looking at adjacent nodes, copying and submitting itself to each system as a network task running a shell script.
Of course, the corporate network was brought to its knees and default "Task" level access was turned off. A few days later, things were back to normal.
Management were not amused. Digital was already a well connected company and relying on the Easynet for day to day operations. The guy evenually admitted to it many years later and as he was by then a senior network consultant, there was't a lot they could do.
The thing worked because most people there were running a single OS, OpenVMS, which was left largely unprotected.
The cable modem is connected only to arouter on your premises. Everyone knows that we are just talking about a physical connection, don't they?
Sorry.
Recent case law in the US already has started to limit the automatic extensuion of patent cover from what was clearly described in the application.
At the same time, we have to be flexible enough to acknowledge that an algorithm in Java is an reimplementation of a patented algorithm in C. More difficult, unless you get a much better scheme for the vetting of patents.
That is a very good point, but Brussels is to lobbying in Europe, what Washington in th US. Many commercial lobbyists are already in place and the number is increasing, unfortunately we don't have so much in the way of non-commercial interests like an EFF for Europe.
The other point is that many Usenet readers have ROT-13 functions, are they all in breach of the DMCA too?
He is a political prisoner being held illegally and should be supported as such.
China is more likely to have missiles tasked on Russia who they share a massive land border with, and with whome they have had far from ideal relations.
Lets come down to it, a country with the ability to create a single nuke wouldn't want to take on the US except through third-party terrorism. If a country attacks the US, then the US can retaliate massively, without using any nuclear weapons.
Oh and it will be BS about nuclear detection at borders. An assembled bomb is usually well-shielded and is easily handled at no risk. The only real chance of detecting a bomb (rather than just a pile of fissile materials) would be through other means such as with neutrons.
There are plenty of places that the US govt could put that money used for the NMD project that would genuinely reduce tension.
It is serious stuff and the area is considered to be hazardous. The only people allowed opn the island are the specialists and US consultants involved in the risk assesment. However, the area nearby remains relatively accessible.
I only worked in Uzbekistan for a while. I didn't work anywhere that part of the country and would not go there. I do know people who have though.
Perhaps the astronauts should be sent there for part of their training?
You must agree that he believes in some aspects of object methodology though, i.e., inherited attributes!
Think damages. If they damage your computers, they should compen$ate you. If their employees don't know the difference between Windows and Solaris, then sue them.
Some local dists are a litle more "Russian-aware" but last I heard, the tendancy is to go for Mandrake and then download extra software, i.e. fonts and so on from the net. The software seems to be open-licensed.
I am writing to warn you that you are attempt to protect the interests of Adobe corporation has and will cause you much adverse publicity. Whilst I respect your position, the name in question is a) artificial and b) is based upon what the product does vis-a-vis document illustration. I realise that you are not an English firm and can theerfore not be aware of the normal use of words in the English language. You would be able to sucessfully argue the case in court but only with significant costs. However, Killustrator only works on a platform for which Adobe does not provide any support (Linux/KDE).
I am not sure whether you know about the OpenSource and Free software movements. The creator of the program has made the program available, free of charge, to anyone who has access to the Internet and there is no restriction upon secondary distribution. The end result is that copies may be present anywhere from New-Zealand to Uzbekistan. The creator did not perform any other action than to leave the files in a place that is acessible by the public. The act of copying was performed by the users.
I would strongly suggest that you request only a change of name, rather than the destruction of the programme. Otherwise you will find that your company becomes a laughing-stock on the Internet. The Open-Source/Free-Software movement is extremely popular in Germany and Europe and you *will* lose clients.
I am writing to you in my private capacity in the hope that you will avoid putting the law on intellectual property in Germany and the EU into disrepute. Incidentally, I do not use this particular program as it is intended for working in an environment that I do not use.
Regards,
Does the US force their mobile ideas onto other countries, well actually yes, and the passing off of second-hand equipment to go with it. The countries concerned, worry that their sales are too low. Some of them are using a hodge-podge of GSM technology but US-style charging. This offers technical advantages over straight AMPS or US style TDMA but no economic advantages (and all the issues with vulnerability to fraud).
I agree that there other plans that include transparent roaming at no-extra charge, but this isn't cheap.
As for the ability to use CLI to filter calls, it doesn't always work when you roam. Definitely not always with GSM either and in any case the penetration of CLI even from fixed to mobile in one European country is incomplete.
As for free local calls, this is another question, but a fixed-line local call is not the same as a mobile call. With a mobile call, you pay a premium to reach the person that you have called. With the local call, it is a matter of luck whether the called person is there. Does it reduce the number of calls? Well, no not really. From current studies, the European model seems to generate a greater degree of market penetration and greater income for both the provider and reseller. This is not a philisophical point, this is one of profitability.
Probably the greatest issue with roaming at the moment with the GSM model is the automatic call forwarding. that is, if I have a German telephone and am booked in a French network, why does a call from France have to go via Germany?
Please don't think that I am knocking the US on this one, but as with PAL vs NTSC, version 2 is generally an improvement on version 1.
The point is that mobile numbers have a special area prefix, and then they are nationwide. The caller pays extra, but knows this because of the prefix. The receiver doesn't pay unless they are out of the country.
End result is that with the US-type system, private users like to keep their numbers quiet because of the cost issue whilst with the Europeanm type system, users often give out their mobile numbers in preference to their home numbers because of the ease of reachability.
This is an economics issue not a straight technical issue.
SMS is extremely cheap from the network viewpoint It is as portable as the mobile telephone and it works better than voice when the coverage is poor or the network is congested. An example of this is the use of SMS for rescue calls, one the other side of the world.
My kids have mobiles and a fairly tight allowance for calls (intentionally so, otherwise they would bankrupt me!!!). This is not atypical of the situation with other teenagers. SMS stretches their budget and allows them to use the network in a more efficient way than with voice. The pricing scheme is such that they pay much more per network packet with SMS than with voice.
This is why I disagree with you about first the Email question and then the network economics question.
If that was from Kubrik's Dr. Strangelove, you forgot the combination bible and Russian phrasebook!
We did a setup at the client site and had a load of problems with the DECnet protocol stack. IP worked great, but we needed an IOS upgrade at a weekend. Despite even having a credit-card at hand, it was almost impossible to purchase the upgrade. We eventually had to buy the s/w in another country (it was not downloadable) and then the local distributor (who we eventually identified)was able to release their copy of the s/w to us.
Cisco themselves were extremely uncooperative, passing everything down the line to distributors and resellers who had neither the inclination nor the skills to provide 24x7 support.
The client was a securities trading house, not the best people to upset when you look to the share price and Cisco were definitely bad-mouthed over that one. As a contrast, another client of mine had a memory module go late at night. The system was supported only on a 9-5 contract. We had no problem to get an engineer and the part and we only paid for the labour cost, not that of the parts, as it could have failed 9-5.
Maybe Cisco works better in the US, but in Europe, I can't rely on them.
To synthesise information in the way that Van Gogh does is not difficult. What is difficult is to come up with a new and unique method of interpretation.
Note that an improvement to this technique would be able to work from stereo-pairs, as theg artist is encoding depth information as well in the representation.
As a side point, it is thought in some quarters that Van Gogh's medical conditions and possible absinthe consumption may have contributed to his perception of colour.
Another major client is the US military, they also have a bag load of applications that just won't port easily.
The implication is that if Compaq pass the technology on, the customer (Intel) will have to continue to produce so that Compaq can fulfill its high-end systems contracts.
What attracted Compaq to Digital was not Alpha, it was the Digital service infrastructure and the client list for high end systems, who they hoped to be able to sell PCs to.
Compaq PCs had a number of problems that didn't endear themselves to corporate customers, so this never really worked. Even though these issues seem to have been largely addressed (however, my little Ipaq illustrates the bad QA that Compaq has suffered from).A good little beast but let down by a losy sound jack that is being echoed by many other users.