Actually, as someone who has recently sold an OSS based solution to a British local council body, I can tell you that most of them aren't that bad. They're mainly interested in getting the job done for as little outlay of public money as they can.
And in 10 years time all of the Newham councillors will have been swapped for new ones as an example of democracy in action, so they don't really care.
Yes. In fact, that's the entire idea. Increased response times -> reduced throughput. But as most people don't actually sit & wait for their e-mail to go out, it is acceptable for use in this kind of situation. You wouldn't be able to use it, though, on a web server...
Actually, I'm not sure that's true. Potential suppliers also have an interest in the case. The timing of this deal is interesting too -- I believe the government guidelines on OSS usage (which will require government bodies to favour open source solutions over proprietary ones where there is no clear advantage to the proprietary system) are due to be released in the very near future.
Re:Difference Between Bits and Bytes (with Chart)
on
NSLU2 Now More Useful
·
· Score: 1
I don't think most people here need a chart for that conversion. The OP was clearly talking about network speed, not HD disk; net speed is generally also measured in bits.
I don't get it. I understand the concept of what you would use one of these 'POW' tokens for (although the name is clearly ridiculous... acronym clash should be avoided whenever possible) -- e.g. to bypass a junk e-mail filter. But why would you want to be able to reuse them?
You may be right. I've never been entirely certain of which RAID level was which. I only ever use RAID 1.
If you're going to the trouble to enable software RAID, I'm sure you could enable it to mount more than 2 HD's (using a USB hub to provide the necessary ports).
That may be true. Another alternative that I just thought of is to set a cron job to do a regular backup of changed files to the second disk. That wouldn't require any kernel or core system updates, so could be achieved with no risk. Obviously it isn't quite as good, but it would probably be sufficient for most purposes. If you used compression (what's the processor spec on these things...?) or a larger 2nd disk you could keep a short revision history, too.
Re:GNAA Announces Achievement of 10 Millionth Post
on
NSLU2 Now More Useful
·
· Score: -1, Offtopic
1) It must be understood from day one that GNAA > j00, and that fristage postage is always ours.
Shouldn't that be frist-psotage?
2) GNAA software engineers have developed many wonderful automata for crapflooding blogs and various message boards. These will be provided in compiled binaries free of charge for your use.
And, presumably, will include trojans... free of charge for GNAA's use.
I might be getting my RAIDs mixed up, but I think RAID10 requires 3 disks, and according to the writeup on Linksys's site it only supports 2.
I suspect you would need to flash the firmware to achieve this. I'm not sure if any toolchains are available that would allow you to ensure you had a working kernel before you did this... i.e., you stand a high chance of writing off the device.
Re:Linux embedded integrators are lazy
on
NSLU2 Now More Useful
·
· Score: 4, Insightful
You are aware that the process for gaining access involves removing the hard disk from it, attaching it to another computer so you can modify certain files, and then returning the hard disk. Not the kind of thing you can do over the network.
If a hacker has physical access to your hard disk, you've got a lot more to worry about than this.
Re:Linux embedded integrators are lazy
on
NSLU2 Now More Useful
·
· Score: 4, Insightful
It shouldn't be this easy to let hackers break into the system
Why would you want to prevent them? It drives sales of your products.
the kernel is too tangled to successfully create a small distribution that is at the same time useful and feature-limited.
I don't agree. It is perfectly possible to do this, and know several people who _have_ done it. The issue is, it isn't worth the effort. It would take several weeks of developer time to determine exactly what is needed and what isn't, whereas there's actually no problem with including unnecessary features. So that's what happens.
Don't get me wrong, I think QNX et al are very cool systems, and there are many situations where they are more applicable than Linux. But I don't see anything wrong with Linux here.
I've gotta ask... what the hell is the point of that?
Benefits of a microkernel -- good software engineering practice, small enough to understand.
Benefits of a monolithic kernel -- faster performance.
Benefits of mixing the two -- as far as I can see, none of the above. You might get to keep the performance, but then you may as well stick with a monolithic kernel.
Re:How is this better than SSL?
on
Anti-Phishing Tools
·
· Score: 2, Insightful
SSL doesn't help against lookalike domain names. Of course, anyone with eyes and abrain ought to be able to spot that, but most people need something a little more blatant.
I would also need to be able to see where the "purchase upgrade" link points, to see whether it's a phishified URL.
After taking the test and getting 8/10, I realised that that information was available... but only if you use Internet Explorer. It has some broken javascript to set the link target url into the status bar, only it doesn't work right in Mozilla, probably because of the 'return document.MM_ReturnValue' (or something utterly meaningless like that) on the end of the onMouseOver action.
If you check the status bar the url is for a Brazilian site.
Hmmm...? I didn't get any status bar text on any of the links in that test, and when I used right click/properties to get the address, it linked back to the test page.
Just tested on IE 6.0 on Windows 2000. It allows you to click on such a link without any warnings, but the '...@' section disappears from the URL when it is displayed in the address bar, which ought to give you at least some feedback.
The problem is that with the variations of SQL syntax all over the place just using an API abstraction isn't enough. You'd pretty much need to use a standardised version of SQL that's translated by the access API to the local variant for the server in question. I don't think any of the APIs you mention do that.
MySQL is a major contributor to this problem, but it isn't the only one... you can't use the same syntax to create a table with an automatically produced index between MS SQL and Access, even!
If you have a clause in the license that says you may modify the license, you can probably get away with that. As long as the modifications are "reasonable" and you took "reasonable" steps to ensure all licensees were aware of the modifications.
Yes. The GPL is a bad thing whenever you need to interoperate with non-GPL software. This means that as integration gets more and more common, the GPL will become more and more annoying. Soon it will become clear that there are only two acceptable outcomes: all software under the GPL, or no software under the GPL. Anything inbetween and you'll find that everything you want to do with your software is illegal.
Slashdot Mirror
Actually, as someone who has recently sold an OSS based solution to a British local council body, I can tell you that most of them aren't that bad. They're mainly interested in getting the job done for as little outlay of public money as they can.
And in 10 years time all of the Newham councillors will have been swapped for new ones as an example of democracy in action, so they don't really care.
Well, I can dream, at least.
Yes. In fact, that's the entire idea. Increased response times -> reduced throughput. But as most people don't actually sit & wait for their e-mail to go out, it is acceptable for use in this kind of situation. You wouldn't be able to use it, though, on a web server...
Actually, I'm not sure that's true. Potential suppliers also have an interest in the case. The timing of this deal is interesting too -- I believe the government guidelines on OSS usage (which will require government bodies to favour open source solutions over proprietary ones where there is no clear advantage to the proprietary system) are due to be released in the very near future.
I don't think most people here need a chart for that conversion. The OP was clearly talking about network speed, not HD disk; net speed is generally also measured in bits.
According to the way I read the page, you're wrong. I quote:
It's a mult-step process to enable telnet. [...]
First, mount a NSLU2 initialized hard drive on a box that understands ext2/3 format.
But, whichever of us is right, you still need physical access. Anyone untrusted should not have physical access to your central network storage.
I don't get it. I understand the concept of what you would use one of these 'POW' tokens for (although the name is clearly ridiculous... acronym clash should be avoided whenever possible) -- e.g. to bypass a junk e-mail filter. But why would you want to be able to reuse them?
Ah! I get it!
I thought RAID 10 required 4 drives.
You may be right. I've never been entirely certain of which RAID level was which. I only ever use RAID 1.
If you're going to the trouble to enable software RAID, I'm sure you could enable it to mount more than 2 HD's (using a USB hub to provide the necessary ports).
That may be true. Another alternative that I just thought of is to set a cron job to do a regular backup of changed files to the second disk. That wouldn't require any kernel or core system updates, so could be achieved with no risk. Obviously it isn't quite as good, but it would probably be sufficient for most purposes. If you used compression (what's the processor spec on these things...?) or a larger 2nd disk you could keep a short revision history, too.
1) It must be understood from day one that GNAA > j00, and that fristage postage is always ours.
Shouldn't that be frist-psotage?
2) GNAA software engineers have developed many wonderful automata for crapflooding blogs and various message boards. These will be provided in compiled binaries free of charge for your use.
And, presumably, will include trojans... free of charge for GNAA's use.
*nix n00b question, maybe, but why not just blank the [root] password out?
v /nulll l3 :sc50wKPq.zChw:2002:501:::/dev/null
Err... to stop anyone on your network from connecting and wiping all your data / nicking your pr0n collection?
One thing I have just noticed...
admin:sclzZZfodiRXY:502:501::/home/user/admin:/de
test_user:scEPG0VnVyqmE:2000:501:::/dev/nu
test2:scEPG0VnVyqmE:2001:501:::/dev/null
test
Its using the same salt for every password. This is horrendously insecure...
I might be getting my RAIDs mixed up, but I think RAID10 requires 3 disks, and according to the writeup on Linksys's site it only supports 2.
I suspect you would need to flash the firmware to achieve this. I'm not sure if any toolchains are available that would allow you to ensure you had a working kernel before you did this... i.e., you stand a high chance of writing off the device.
You are aware that the process for gaining access involves removing the hard disk from it, attaching it to another computer so you can modify certain files, and then returning the hard disk. Not the kind of thing you can do over the network.
If a hacker has physical access to your hard disk, you've got a lot more to worry about than this.
It shouldn't be this easy to let hackers break into the system
Why would you want to prevent them? It drives sales of your products.
the kernel is too tangled to successfully create a small distribution that is at the same time useful and feature-limited.
I don't agree. It is perfectly possible to do this, and know several people who _have_ done it. The issue is, it isn't worth the effort. It would take several weeks of developer time to determine exactly what is needed and what isn't, whereas there's actually no problem with including unnecessary features. So that's what happens.
Don't get me wrong, I think QNX et al are very cool systems, and there are many situations where they are more applicable than Linux. But I don't see anything wrong with Linux here.
But the 'layering' is actually a commingling.
I've gotta ask... what the hell is the point of that?
Benefits of a microkernel -- good software engineering practice, small enough to understand.
Benefits of a monolithic kernel -- faster performance.
Benefits of mixing the two -- as far as I can see, none of the above. You might get to keep the performance, but then you may as well stick with a monolithic kernel.
MOV 0, 1
re_progress.asm:1: illegal combination of opcode and operands
domo arigato
not domo arigoto
nihonho wo hanasu deshita.
May I be the first to say... huh?
SSL doesn't help against lookalike domain names. Of course, anyone with eyes and abrain ought to be able to spot that, but most people need something a little more blatant.
I would also need to be able to see where the "purchase upgrade" link points, to see whether it's a phishified URL.
After taking the test and getting 8/10, I realised that that information was available... but only if you use Internet Explorer. It has some broken javascript to set the link target url into the status bar, only it doesn't work right in Mozilla, probably because of the 'return document.MM_ReturnValue' (or something utterly meaningless like that) on the end of the onMouseOver action.
If you check the status bar the url is for a Brazilian site.
Hmmm...? I didn't get any status bar text on any of the links in that test, and when I used right click/properties to get the address, it linked back to the test page.
Just tested on IE 6.0 on Windows 2000. It allows you to click on such a link without any warnings, but the '...@' section disappears from the URL when it is displayed in the address bar, which ought to give you at least some feedback.
If you have two computers and one takes 1 minute to infect while the other takes 39 minutes, then the average is 20.
However, in this case the average is a completely and utterly useless figure and you'd be much better off expressing a range of times.
The problem is that with the variations of SQL syntax all over the place just using an API abstraction isn't enough. You'd pretty much need to use a standardised version of SQL that's translated by the access API to the local variant for the server in question. I don't think any of the APIs you mention do that.
MySQL is a major contributor to this problem, but it isn't the only one... you can't use the same syntax to create a table with an automatically produced index between MS SQL and Access, even!
If you have a clause in the license that says you may modify the license, you can probably get away with that. As long as the modifications are "reasonable" and you took "reasonable" steps to ensure all licensees were aware of the modifications.
Is this a case of the GPL being a bad thing?
Yes. The GPL is a bad thing whenever you need to interoperate with non-GPL software. This means that as integration gets more and more common, the GPL will become more and more annoying. Soon it will become clear that there are only two acceptable outcomes: all software under the GPL, or no software under the GPL. Anything inbetween and you'll find that everything you want to do with your software is illegal.