This is something like the new system that is being discussed on LKML at the moment. Effectively contributors must sign to the effect that they have rights over the code they have written and are able to donate it to a GPL project.
Behind one major system for trading financial securities (the largest for derivatives, at least) is a straightforward ISAM-type mechanism. Relational was too slow. The ISAM files certainly qualify for being called a database and the flavour of ISAM they use has ACID as well. However there are no constratints, no referential integrity and no space management. In short, there is a non-relational database without a management system.
I thought DEC had their own RDMS, RDB which also ran on Ultrix. In the early days at least, RDB shared its backend with DEC's CODASYL (Hierarchical) DBMS, DBM-32, but it certainly qualified as relational.
In theory, you need a CVS diff list at least. However, unless the commit comments are linked to a meaningful entry somewhere that shows where a change come from, you will have problems. It doesn't matter whether you use CVS or BK, you still need underlying mechanisms. One issue with Linux, is that it has a lot more contributors than *BSD, which tends to make things more complicated.
In the commercial world, you have change numbers which link to a documentation trail which shows who implemented something and why and who approved it. Linus is trying at least to improve the code provenance by looking at a certification chain between the patch generator, the maintainer and eventually Linus as release manager. Unfortunately, it still looks like a hunt through LKML for the documentation as you suggest.
Good point. However, the person who issues the certification should also be sure of who it came from. You are right, it isn't true anonymity but the contributors name need never go into the kernel. It may just be replaced as a John Doe.
So, if a person submits some source code from NT to one of Linus's trusted assistents, they should query it. Unless there is a chain of provenance going back to Bill Gates, there is no release. However, Bill gates can still opt to submit code but request that his name is kept out of the change log.
I think it is more about having someone say that they own the copyright of the work and have the right to GPL it. It doesn't mean no anonymous contribs, but somebody (one of the people with BK commit access would have to be able to certify it on behalf of the other person so they should knwo the person even if it is never identified publicly.
because PocketPC has PocketIE built in. You will pay for that ROM space whether or not you use it. It is best compared with the problems of separating IE from any other Windows OS, but less doable because of the integration on ROM. If you have a large program such as a browser, you don't really want it taking space in RAM. If you have a pocket-Moz it would be cool but I guess it would be better if you could build your own ROM image so you could lose IE. With a Pocket Linux you can.
Make it real easy to rent a PLB - even at the location. PLBs are great but they are cheap enough to think twice about unless you like regular back-country hiking.
Registering units to the renter makes it easy to determine who triggered the alert.
A friend runs a website specificially orientated towards foreigners working in a particular city. They meet once a week and enjoy some beers. They have a website run on a popular host whose catch-phrase is "50megs for free".
I'm in charge of a mailshot for the Ski Club in the same town and I usually give these other guys a plug.
Until recently. I get an aguished call from a very nice lady working at a central bank. She clicked on the link and was faced with porn. I work at a major bank under an up to date patched XP with the guy that runs the beers site. I had no problem from Mozilla but when my colleague tried with IE, it replaced his home page with porn and then lots of pop-ups. It also installed something that reinstalled itself whenever he tried to change. Ironically, it turns out that this was promoting a system-cleaning utility.
My colleague had not put this on the web-site and the hoster denies ever putting anything like that up. We don't know what happened and a couple of days later it was gone. The thing is that it went straight past the defences of two major banks and was very embarassing.
Not only the local cache but squid would have been fllled with these images. Nasty for everyone.
The point is that yes, if someone looked at the dates on the cache, it could be traced to a single incident but in many places, you would have been thrown out by then.
Actually, a DELL Inspiron 7500 notebook worked very well under RH8.0 for a suspend to disk (hibernate). It also worked ok under Windows 98SE and 2K but sucked under 95.
It can b done and slowly a lot of big companies want it for their desktops. Power and especially room heat are issues now, especially outside the US where energy prices are higher. If a business PC isn't working, particularly overnight or at weekends then it should sleep.
I get really confused by RH's PCMCIA support. I have one laptop with the Xircom combo LAN/Modem card. It didn't work in RH 8, it worked in RH 9, it had problems again with RH 9 and Fedora 1. Maybe the driver has fundemental problems (although it works ok under Knoppix) but my feelings is that there are some issues with PCMCIA support in general.
In the old days, photo journalists were trained to get their most controversial pictures out of the camera as soon as possible. 35 mill film cannisters are small and easy to conceal. I seem to remember one particularly striking sequence of someone being murdered for collaberation where the whole camera (a Nikon) was dropped into a bucket of muddy water for concealment. Both camera and film needed a thoroough clean but the film was published and the camera reused later.
CF, smart media and memory sticks are also easy to conceal but unfortunately they aren't quite as sturdy. Camera phones are interesting but it takes time to do an upload.
There a lot of questions about the British pictures and the faces of the perps aren't visible. However the US pictures show faces (including that insanely stupid woman giving a thumbs up). The perps in this case have admitted to the actions.
I understand the British pictures were on film rather than media and there are a number of technical inconsistencies - weapon-type, uniform and so on that suggest that they may have been staged.
First releases of Knoppix are actually quite good. What I like is that they remaster on a frequent basis so any minor problems and the latest fixes are scooped up very quickly.
Olympus do some quite nice little cases for their digital cameras. Water tight down to about 20M (so no wreck diving) or so and quite adequate for splash/dirt protection as well. The price of such cases is about $200 and it is fairly robust being made of plexiglass. Something like 5060WZ is a semi-pro camera with 5million
pixels. High-end housings are great under water becuse of buoyancy but above they are a PITA with their size and weight.
You only need a format like Quark Express or Adobe Illustrator iff the document needs much postprocessing. In fact Quark Express is a pretty hopeless format as things like fonts can't be embedded. Instead of a single file, you end up giving many, the quark file, the CMYK separated picture files and the fonts. A handful to manage.
Any good printshop can take and print a PDF. They can even tweak the colours if it is needed.
Not quite correct, hence the anti-copying measures. If you can bypass the copy-prevention, then you may potentially be responsible for distribution even if you have just made a backup copy. This is why explicitly breaking copyright protection is all you need for getting into trouble.
In our case we used a sodium lamp in high-school for Young's slit experiment. The idea was that it was monochromatic and several orders of magnitude cheaper than a low-end laser in those days. The slits weren't that small.
The main problem I have with biometrics is that it usually isn't even your attributes, it is just a syndrome of theose attributes. For example, your fingerprint is converted into a bit pattern thatr can be captured and replayed. At least with a password, it can be changed.
The problem is whether O_SYNC writes can be prioritised over reads. The scheme discussed here is orientated towards read-often/write-sometimes which is normally quite reasonable. However, if you really are waiting on that write, then it needs to be expedited.
One of the big things about databases is the reliable commit where all the crud you have done in a transaction either gets committed or backed out. The writes then become kind of important, and you really do want them to complete before continuing with anything else.
This messing with the I/O queue may make things interesting for the journalling process which is kind of vital to integrity. File placement could become even more important for this (and also the placing of journal/log files).
The rest seems to just effectively be a modified elevator (wait a bit before moving).
From an economic point of view, pure saving is probably not a good idea - a percentage of what you save must be held as cash and a certain amount paid to the Fed and from the rest, it costs money and time to reinvest. In other wotrds, ironically it is better for the economy when you spend $1000 than when you save it.
Another problem is whether one should pay consumption taxes on some kinds of savings such as equities (shares). It is arguable that they should be taxable in some way because the shares not only earn dividends, they also buy influence.
The issue with consumption taxes is that when they are applied at a single point, they are easy to circumvent. For example, to buy from out of state and import. VAT type systems are somewhat better as the tax is applied on the increase in value, each time the 'good' or service is resold, the tax is applied to the additional value. Harder to circumvent.
The problem in the US is the fact that everything works on a state by state basis. It is too easy for the guy to live in one state, enjoying the services there whilst buying from out of state and paying less taxes.
Slashdot Mirror
This is something like the new system that is being discussed on LKML at the moment. Effectively contributors must sign to the effect that they have rights over the code they have written and are able to donate it to a GPL project.
Behind one major system for trading financial securities (the largest for derivatives, at least) is a straightforward ISAM-type mechanism. Relational was too slow. The ISAM files certainly qualify for being called a database and the flavour of ISAM they use has ACID as well. However there are no constratints, no referential integrity and no space management. In short, there is a non-relational database without a management system.
I thought DEC had their own RDMS, RDB which also ran on Ultrix. In the early days at least, RDB shared its backend with DEC's CODASYL (Hierarchical) DBMS, DBM-32, but it certainly qualified as relational.
In the commercial world, you have change numbers which link to a documentation trail which shows who implemented something and why and who approved it. Linus is trying at least to improve the code provenance by looking at a certification chain between the patch generator, the maintainer and eventually Linus as release manager. Unfortunately, it still looks like a hunt through LKML for the documentation as you suggest.
So, if a person submits some source code from NT to one of Linus's trusted assistents, they should query it. Unless there is a chain of provenance going back to Bill Gates, there is no release. However, Bill gates can still opt to submit code but request that his name is kept out of the change log.
I think it is more about having someone say that they own the copyright of the work and have the right to GPL it. It doesn't mean no anonymous contribs, but somebody (one of the people with BK commit access would have to be able to certify it on behalf of the other person so they should knwo the person even if it is never identified publicly.
Last I heard, TV transmitters are up in the megawatt range. It isn't that unlikely.
because PocketPC has PocketIE built in. You will pay for that ROM space whether or not you use it. It is best compared with the problems of separating IE from any other Windows OS, but less doable because of the integration on ROM. If you have a large program such as a browser, you don't really want it taking space in RAM. If you have a pocket-Moz it would be cool but I guess it would be better if you could build your own ROM image so you could lose IE. With a Pocket Linux you can.
Actually IBM these days is primarily a services based company. This fits in pretty well with the Linux and FOSS in general business model.
Registering units to the renter makes it easy to determine who triggered the alert.
Mobile phones are also useful for emergency use.
I'm in charge of a mailshot for the Ski Club in the same town and I usually give these other guys a plug.
Until recently. I get an aguished call from a very nice lady working at a central bank. She clicked on the link and was faced with porn. I work at a major bank under an up to date patched XP with the guy that runs the beers site. I had no problem from Mozilla but when my colleague tried with IE, it replaced his home page with porn and then lots of pop-ups. It also installed something that reinstalled itself whenever he tried to change. Ironically, it turns out that this was promoting a system-cleaning utility.
My colleague had not put this on the web-site and the hoster denies ever putting anything like that up. We don't know what happened and a couple of days later it was gone. The thing is that it went straight past the defences of two major banks and was very embarassing.
Not only the local cache but squid would have been fllled with these images. Nasty for everyone.
The point is that yes, if someone looked at the dates on the cache, it could be traced to a single incident but in many places, you would have been thrown out by then.
It can b done and slowly a lot of big companies want it for their desktops. Power and especially room heat are issues now, especially outside the US where energy prices are higher. If a business PC isn't working, particularly overnight or at weekends then it should sleep.
I get really confused by RH's PCMCIA support. I have one laptop with the Xircom combo LAN/Modem card. It didn't work in RH 8, it worked in RH 9, it had problems again with RH 9 and Fedora 1. Maybe the driver has fundemental problems (although it works ok under Knoppix) but my feelings is that there are some issues with PCMCIA support in general.
CF, smart media and memory sticks are also easy to conceal but unfortunately they aren't quite as sturdy. Camera phones are interesting but it takes time to do an upload.
I understand the British pictures were on film rather than media and there are a number of technical inconsistencies - weapon-type, uniform and so on that suggest that they may have been staged.
First releases of Knoppix are actually quite good. What I like is that they remaster on a frequent basis so any minor problems and the latest fixes are scooped up very quickly.
Olympus do some quite nice little cases for their digital cameras. Water tight down to about 20M (so no wreck diving) or so and quite adequate for splash/dirt protection as well. The price of such cases is about $200 and it is fairly robust being made of plexiglass. Something like 5060WZ is a semi-pro camera with 5million pixels. High-end housings are great under water becuse of buoyancy but above they are a PITA with their size and weight.
Any good printshop can take and print a PDF. They can even tweak the colours if it is needed.
Not quite correct, hence the anti-copying measures. If you can bypass the copy-prevention, then you may potentially be responsible for distribution even if you have just made a backup copy. This is why explicitly breaking copyright protection is all you need for getting into trouble.
On the other hand, we should be glad it isn't a hole in the hull, otherwise it may be a case of someone putting their finger in.
In our case we used a sodium lamp in high-school for Young's slit experiment. The idea was that it was monochromatic and several orders of magnitude cheaper than a low-end laser in those days. The slits weren't that small.
The main problem I have with biometrics is that it usually isn't even your attributes, it is just a syndrome of theose attributes. For example, your fingerprint is converted into a bit pattern thatr can be captured and replayed. At least with a password, it can be changed.
The problem is whether O_SYNC writes can be prioritised over reads. The scheme discussed here is orientated towards read-often/write-sometimes which is normally quite reasonable. However, if you really are waiting on that write, then it needs to be expedited.
This messing with the I/O queue may make things interesting for the journalling process which is kind of vital to integrity. File placement could become even more important for this (and also the placing of journal/log files).
The rest seems to just effectively be a modified elevator (wait a bit before moving).
Another problem is whether one should pay consumption taxes on some kinds of savings such as equities (shares). It is arguable that they should be taxable in some way because the shares not only earn dividends, they also buy influence.
The issue with consumption taxes is that when they are applied at a single point, they are easy to circumvent. For example, to buy from out of state and import. VAT type systems are somewhat better as the tax is applied on the increase in value, each time the 'good' or service is resold, the tax is applied to the additional value. Harder to circumvent.
The problem in the US is the fact that everything works on a state by state basis. It is too easy for the guy to live in one state, enjoying the services there whilst buying from out of state and paying less taxes.