So why e.g. MS SQL Server does it exactly as GP wrote?
Also you can use much better algorithm for data aging than OS. You know your data and you know what is better to keep and what to let go to secondary storage.
But yes for applications where massive swapping is expected the strategy described might be useful.
I was always wondering if there couldn't be something similar done for code by compilers. So if you have application, that is swapped out and looping in wait loop, then this loop will ideally fit into very small memory footprint and leave RAM for better use of active processes.
Well and what about singing my own song in a public?
Yes, these extortionists are collecting money even from performance of your own stuff. Just in case you eventually join them.
100.000 ports is only for edge level leafs. You have to add additional ports for up-link, aggregation switches and of course core switches. Then you imagine how big Christmas tree it is.
That PMAC idea is really cool. But beyond that. Nothing special. Try to build something more large and you will find that your core layer switches have not enough ports as number of aggregation level switches will increase. And I am not mentioning problems with throughput when distant nodes will start communicating to each other. For me it looks like they are trying to make routers redundant. But building 100 000 node network with this topology will require really powerful core layer nodes. For large datacenters I would see some combination of tree/grid topology.
Even single disk from RAID array can provide lot of information. RAID 5 for example as also other variants uses block level striping. So as long as data someone is after does fit into such block, then it might be useful to have only one disk from raid array. And also if you have all disks from an array you have more recovery data (parity blocks) than if you use only one disk. So the guessing is much easier. But again are your data so critical for you to have such disk mechanically destroyed and not only data overwritten. Maybe few writes of random data. And if your data are so important not to fall into wrong hands (competitors, news reporters,...) why they are not encrypted in the first place? Not whole disk but the really sensitive data. Your disks might be stolen and then none of secure wiping will help you. Of course there are occasional events where disk dies and it is not possible to wipe data as someone wrote. In such case mechanical destruction can be handy. Once I put together computer pieces without any case and connected it to old 5 1/4 disk drive. I wrote a batch file for reading text of the diskette, which was ripped open so anyone was able to see it spinning, and put it into display. The next morning I came to work the program didn't work. When I looked closely I saw white line on the disk. It was because only small part of the disk was accessed and because it was open then dust settling on the disk helped to destroy it. As those days my intention was not to destroy it put together small program that read everything in memory and then moved heads randomly so not only a small part of the disk was brushed off. But the point is that modern hard drives are more sensitive to dust than was old diskette. So why not to open the drive and run secure wipe on it and watch how it gets overwritten while it is being simultaneously brushed and have a fun. How long you will leave it running is up to you.
Exactly. And how long it will take to update their SPAM filters? Are they really so naive or is there something more in the technology? Maybe the e-mails will be encrypted. But highly suspicious then.
Looking forward to see more details on this topic.
Smart cards are also insecure. Yes, they have cryptogrphic chip on it and you have to enter your PIN to get result out of it. But when your computer is compromised then you don't know who is reading your PIN and what data are sent to the card for encryption/signing. So if you use your smart card for login and also for access to your bank, then you can sing big check for some trojan during logon on compromised computer. There needs to be verification of the data that are being signed by smart card. Like with some two factor bank calculators. And also entering PIN directly into security device instead of compromised computer might be better.
Yes it come over my mind when I saw the picture where there is lot of space sacrificed to USB connector. And also the problems with usability of such excentric disc as mentioned earlier can render this as unusable. I am not a fan of RFID but this looks like ideal application for the technology if you just consider best technology for same thing.
When you consider what will happen with all your film library after several years, then any DRM scheme is bad because future generations will be unable to read what will be written on the DRM protected media.
I also remember something like this when SONY introduced their DV cameras. There was chip integrated in tape casette. Of coures it was used for other information (TOC and maybe other data). And didn't required to stick your casette in connector:).
Yes, this will make every copy traceable so you will more decide whether to provide a copy to your friend. But on the other hand it will not allow extortion schemas like pay per view and time limited view that are much desirable for record companies as they might generate more revenue.
The results heavily depends on the servers being monitored. It is very easy to say that on one server was 10% of requests made using Firefox and 90% of requests made using IE or the ratio could be completly different. But what can affect these results?
- web is interresting for specific community which tends to use specific browser.
- web is optimized for specific browser and then that browser will get more accesses on that server regardles of the number of browsers regularly used. In this case you can see that someone visited the server from Firefox and then immediatelly from IE.
How does the statistics handle these issues? Especially the latest one. It also explains the difference between the statistics provided by various analytics firms. Why/. have 60%? Because its readers are mostly Linux enthusiasts. My conclusion is that the higher numbers from statistics are reality due to deformations made to them described above. So it seems that WEB developers should start to consider Firefox as a brower that people realy use.
Just seeing first few lines of their javascrip seems that it is aimed at something different than to allow webmasters see what is happening on their websites. See:
var _uctm=1;// set campaign tracking module (1=on|0=off)
_uOsr[0]="google"; _uOkw[0]="q";
_uOsr[1]="yahoo"; _uOkw[1]="p";
_uOsr[2]="msn"; _uOkw[2]="q";
_uOsr[3]="aol"; _uOkw[3]="query"; ...
These are just variables and I didn't walk through code but it warns me.
Have anyone read EULA for this tool?
Where to start? Asking what of the functionality from Exchange your users are using. Remember Exchange is NOT e-mail server. It is groupware server with much more possibilities than regular e-mail servers. Who in the company is fed with Exchange? Users, BOSS or admins? I also hate "features" of Exchange like POP3 localization or its debility with handling e-mails with large attachements when I was unable to download such attachement over IMAP or POP3 but when I prove that I am also able to contact such server using Outlook then IMAP appeared to download the attachement. But with combination it is powerfull tool that could be hardly replaced. (contacts, calendar)
I know only one replacement for exchange. Lotus Notes/Domino.
Why RAID on notebooks?
Simply because for some people is daily backup procedure not enough. Just because their data are too volatile to simply restore last backup and then manually restore today's work. Answer yourself on another question: Why we use RAIDs in servers and not only make a daily backup?
Yes it couldn't be wrong. Especially when the google ignores anything outside US. Google is seen as global search engine. But it is no longer true. There are services on Google that are not available worldwide.
- Google Maps
- Google Ride
- Google Video
- Google Local
- Google Print
To be fair most of them are in beta stage so you may not expect such range of data available. In that situation there must be such projects as set by European libraries. At least to fill in the gap existing in Google's effort. And last Google Print doesn't make available content of the books. It only allows to search through them and it is big difference.
Or they will put robots.txt file at the root of their web and googlebot will turn away without downloading anything.:-)
BTW: It is OK to point out some resource without authors knowledge but if you want to copy any information on your server you must ask for permission.
You don't need to use different versions of M$ products to generate incompatible documents. Guys at M$ in their internalization effort translated Visual Basic too. It is nice to write any macro and then to try run it in translated version of Excel and vice versa.
SUN is hiring developers for their OS if they will work for free.
Yes, SUN wants you!
As long as there will be statement that any new release must be made avaliable under CDDL licence, the making available these patents makes no sense for Linux and any other software made available under GPL.
Does anybody know how much records (packets) can CISCO switches proces per second? This database seems to be nothing more than filtering machine which calculates some statistics from data it receives. And that is nothing new. If you are able to receive more than 140,000 packets, then you have faster database than one announced.:-)
Slashdot Mirror
So why e.g. MS SQL Server does it exactly as GP wrote?
Also you can use much better algorithm for data aging than OS. You know your data and you know what is better to keep and what to let go to secondary storage.
But yes for applications where massive swapping is expected the strategy described might be useful.
I was always wondering if there couldn't be something similar done for code by compilers. So if you have application, that is swapped out and looping in wait loop, then this loop will ideally fit into very small memory footprint and leave RAM for better use of active processes.
Well and what about singing my own song in a public? Yes, these extortionists are collecting money even from performance of your own stuff. Just in case you eventually join them.
100.000 ports is only for edge level leafs. You have to add additional ports for up-link, aggregation switches and of course core switches. Then you imagine how big Christmas tree it is.
That PMAC idea is really cool. But beyond that. Nothing special. Try to build something more large and you will find that your core layer switches have not enough ports as number of aggregation level switches will increase. And I am not mentioning problems with throughput when distant nodes will start communicating to each other.
For me it looks like they are trying to make routers redundant. But building 100 000 node network with this topology will require really powerful core layer nodes.
For large datacenters I would see some combination of tree/grid topology.
Even single disk from RAID array can provide lot of information. ...) why they are not encrypted in the first place? Not whole disk but the really sensitive data. Your disks might be stolen and then none of secure wiping will help you.
RAID 5 for example as also other variants uses block level striping. So as long as data someone is after does fit into such block, then it might be useful to have only one disk from raid array.
And also if you have all disks from an array you have more recovery data (parity blocks) than if you use only one disk. So the guessing is much easier.
But again are your data so critical for you to have such disk mechanically destroyed and not only data overwritten. Maybe few writes of random data.
And if your data are so important not to fall into wrong hands (competitors, news reporters,
Of course there are occasional events where disk dies and it is not possible to wipe data as someone wrote. In such case mechanical destruction can be handy.
Once I put together computer pieces without any case and connected it to old 5 1/4 disk drive. I wrote a batch file for reading text of the diskette, which was ripped open so anyone was able to see it spinning, and put it into display. The next morning I came to work the program didn't work. When I looked closely I saw white line on the disk. It was because only small part of the disk was accessed and because it was open then dust settling on the disk helped to destroy it. As those days my intention was not to destroy it put together small program that read everything in memory and then moved heads randomly so not only a small part of the disk was brushed off.
But the point is that modern hard drives are more sensitive to dust than was old diskette. So why not to open the drive and run secure wipe on it and watch how it gets overwritten while it is being simultaneously brushed and have a fun.
How long you will leave it running is up to you.
Exactly.
And how long it will take to update their SPAM filters?
Are they really so naive or is there something more in the technology?
Maybe the e-mails will be encrypted. But highly suspicious then.
Looking forward to see more details on this topic.
Just sort posts in different order :-).
But it will probably not help you with your problems.
Smart cards are also insecure. Yes, they have cryptogrphic chip on it and you have to enter your PIN to get result out of it. But when your computer is compromised then you don't know who is reading your PIN and what data are sent to the card for encryption/signing. So if you use your smart card for login and also for access to your bank, then you can sing big check for some trojan during logon on compromised computer.
There needs to be verification of the data that are being signed by smart card. Like with some two factor bank calculators. And also entering PIN directly into security device instead of compromised computer might be better.
When you consider what will happen with all your film library after several years, then any DRM scheme is bad because future generations will be unable to read what will be written on the DRM protected media.
I also remember something like this when SONY introduced their DV cameras. There was chip integrated in tape casette. Of coures it was used for other information (TOC and maybe other data). And didn't required to stick your casette in connector :).
Yes, this will make every copy traceable so you will more decide whether to provide a copy to your friend. But on the other hand it will not allow extortion schemas like pay per view and time limited view that are much desirable for record companies as they might generate more revenue.
The results heavily depends on the servers being monitored. It is very easy to say that on one server was 10% of requests made using Firefox and 90% of requests made using IE or the ratio could be completly different. But what can affect these results? /. have 60%? Because its readers are mostly Linux enthusiasts. My conclusion is that the higher numbers from statistics are reality due to deformations made to them described above. So it seems that WEB developers should start to consider Firefox as a brower that people realy use.
- web is interresting for specific community which tends to use specific browser.
- web is optimized for specific browser and then that browser will get more accesses on that server regardles of the number of browsers regularly used. In this case you can see that someone visited the server from Firefox and then immediatelly from IE.
How does the statistics handle these issues? Especially the latest one. It also explains the difference between the statistics provided by various analytics firms. Why
Just seeing first few lines of their javascrip seems that it is aimed at something different than to allow webmasters see what is happening on their websites. See: // set campaign tracking module (1=on|0=off)
...
var _uctm=1;
_uOsr[0]="google"; _uOkw[0]="q";
_uOsr[1]="yahoo"; _uOkw[1]="p";
_uOsr[2]="msn"; _uOkw[2]="q";
_uOsr[3]="aol"; _uOkw[3]="query";
These are just variables and I didn't walk through code but it warns me.
Have anyone read EULA for this tool?
Where to start? Asking what of the functionality from Exchange your users are using. Remember Exchange is NOT e-mail server. It is groupware server with much more possibilities than regular e-mail servers. Who in the company is fed with Exchange? Users, BOSS or admins? I also hate "features" of Exchange like POP3 localization or its debility with handling e-mails with large attachements when I was unable to download such attachement over IMAP or POP3 but when I prove that I am also able to contact such server using Outlook then IMAP appeared to download the attachement. But with combination it is powerfull tool that could be hardly replaced. (contacts, calendar)
I know only one replacement for exchange. Lotus Notes/Domino.
Why RAID on notebooks?
Simply because for some people is daily backup procedure not enough. Just because their data are too volatile to simply restore last backup and then manually restore today's work. Answer yourself on another question: Why we use RAIDs in servers and not only make a daily backup?
Yes it couldn't be wrong. Especially when the google ignores anything outside US. Google is seen as global search engine. But it is no longer true. There are services on Google that are not available worldwide.
- Google Maps
- Google Ride
- Google Video
- Google Local
- Google Print
To be fair most of them are in beta stage so you may not expect such range of data available. In that situation there must be such projects as set by European libraries. At least to fill in the gap existing in Google's effort. And last Google Print doesn't make available content of the books. It only allows to search through them and it is big difference.
Or they will put robots.txt file at the root of their web and googlebot will turn away without downloading anything. :-)
BTW: It is OK to point out some resource without authors knowledge but if you want to copy any information on your server you must ask for permission.
You don't need to use different versions of M$ products to generate incompatible documents. Guys at M$ in their internalization effort translated Visual Basic too. It is nice to write any macro and then to try run it in translated version of Excel and vice versa.
Hmm, your scrit doesn't work on SUN Solaris. :-(
SUN is hiring developers for their OS if they will work for free.
Yes, SUN wants you!
As long as there will be statement that any new release must be made avaliable under CDDL licence, the making available these patents makes no sense for Linux and any other software made available under GPL.
Does anybody know how much records (packets) can CISCO switches proces per second? :-)
This database seems to be nothing more than filtering machine which calculates some statistics from data it receives. And that is nothing new. If you are able to receive more than 140,000 packets, then you have faster database than one announced.