Jason Cooper

I will have to read up on the method, but from your description of it it does seem to be either only useful for specific applications or a waste of time. From skimming the paper I can see that the method requires that your server has enough bandwidth to the internet to cover the application DDOS as well as the increase in traffic from legit clients when they are asked to speak-up (Condition C2). This would suggest that the problem was with the power of the server to cope with the requests, I don't see how making your server deal with more pointless requests from legit clients as well as bots would make your server suddenly capable of dealing with the requests. (Note: I call the extra requests from the legit clients pointless as if they weren't pointless the clients would have been making them before regardless of being asked to speak-up or not.) A far better solution would be rather than putting you servers under extra strain by making you legit clients talk more, just increase the power/performance of your server set-up. Perhaps look at clustering and load balancing.

If you want to know more about the executable format commonly used by linux then check out My linkhttp://en.wikipedia.org/wiki/Executable_and_Linkable_Format for an overview.

Chances were that he was either following tutorials online or just coded part of it in a common way for viruses and one of Karpersky's signatures covers variations of that part. It would have been interesting to see if it identified it as a general virus or a specific one.

Go and speak to the librarians there and see if they can get books from other libraries in your area. If they can't get the book you can usually request that they purchase it, there may be a small charge for a purchase but nothing like the cost of buying the book yourself.

The answer here is to learn a bit about the layout of a Window PE file (Portable Executable). A very simple outline of a PE file is DOS Executable code PE Header Section Section Section The DOS executable code is there to tell the user that they can't use the program in DOS if they run it from DOS (Note that with windows XP and beyond not being built on top of DOS in the modern day it doesn't do anything) The PE header tells windows dynamic linker how to map the sections into memory and some other information to do with the program, like where to start. When you append another program onto the end of an existing one the file would look like DOS Executable code PE Header Section Section Section DOS Executable code PE Header Section Section Section Windows takes the program and uses the first PE Header to map it all into memory and starts running the code from where the PE header tells it. Windows doesn't see the second PE header so the sections for the second program don't get mapped into memory, and they don't get run.

All books can be free books if you know how to use a Library.

An in others it is illegal to knowingly distribute malicious software. One good thing though is that it usually isn't illegal to know how they work. I would suggest reading up old Virus Bulletin or old issues of 40Hex to learn about the basic principles of viruses and worms.

When it comes to IT books you can't go far wrong with any on the subject you are interested in from o'reilly.

These things aren't easy to create, but you do learn a lot from trying. Write a few and then try to break them when you find you can't break it stop and go read up on more about them and then try again. The chance of creating something secure without a lot of knowledge/ experience of the area are very slim, but the chance of learning things by trying is very high.

Assuming you are using a version of python beyond 2.5 you can, and should, use the subprocess.call instead of the os.system routine.

Definitely examine RC4 if you are wanting to learn about encryption as it is simple enough to get your head round. I wrote a tutorial about it a long time ago which might be useful to help understand how the algorithm works.

You can use find to do this. find /path/to/search/from -type d -empty Find will be on most Unix style systems and you can get a win32 version for windows.

Ask questions about what the employability of their students is like, after all you are doing this to get a job. If they won't tell you what percentage of students that graduated last year are currently employed then be a bit wary. Some colleges have a good reputation with employers while others have a bad reputation. This reputation is usually based upon previous graduates they have hired, so is usually a good sign of the quality of their courses.

If that is all you do on it then I doubt you would notice a difference. The main reason I can see for having a SSD in your netbook over a hard disk (and the reason I stick with my old eeePC 900 which has a SSD) is that it will take knocks a lot better. Mine still works fine despite being in my backpack when it has taken some serious knocks. If it had a hard disk I am pretty sure I would have bad sectors all over the place.

I wouldn't use grep to do that I would use sed. #!/bin/sh echo "Gimme the URL boy!" read url curl https://www.googleapis.com/urlshortener/v1/url \ -H 'Content-Type: application/json' \ -d "{'longUrl': '$url'}" 2>/dev/null | sed -r -e '/"id"/ !d' -e '/"id"/ s|.*"id".*(http://.*)",|\1|' First the "2>/dev/null" part throws away some of the output that curl is putting out via stderr on my machine. Now the sed part has 3 parameters. The first ( -r ) sets sed to use more advanced regular expressions (On some OS's it is -E). The second parameter ( -e '/"id"/ !d' ) deletes any lines that don't have "id" in them. The third parameter ( -e '/"id"/ s|.*"id".*(http://.*)",|\1|' ) takes any line with "id" in it and pulls out the url. It then replaces the pattern with the url it has found. This can be a little confusing to begin with so read up on sed to get to grips with it. The second parameter could be replaced by piping the stream though grep first and grepping for "id". There are also many other ways that you could achieve the same sort of results (e.g. using awk).

As you have to size your antennas based on the target frequency and channel 6 is both the default channel and pretty near the middle of the available frequencies, it would make sense for the manufacture of those standard antennas to target that frequency. I don't know if the size difference of the antenna in the ranges we are talking about would make that much of a difference, but if they have similar choices to make elsewhere in the design of devices then the same logic would push them to aiming at channel 6. Of course channel 6 being the default it can get quite congested so you may find that you get a lower actual performance on that channel even though you get the best signal on it.

It isn't the API, its the shell. Or rather your use of the quotes in the shell. Try something like #!/bin/sh echo "Gimme the URL boy!" read url curl https://www.googleapis.com/urlshortener/v1/url \ -H 'Content-Type: application/json' \ -d "{'longUrl': '$url'}" and it should work fine, or it did when I tested it. The problem you had was that you were passing curl a JSON string to pass to the google's API. When you enclosed the JSON string in single quotes you were telling the shell to not expand any variables in it so '{"longUrl": "$url"}' will pass to the API {"longUrl": "$url"} When you enclosed the JSON string in double quotes you are telling the shell you want it to expand any variables it finds, so "{'longUrl': '$url'}" will pass to the api {'longUrl': 'Whatever URL you put in the variable'} Try playing with single and double quotes with variables in the shell using the echo command. After playing around a bit it should all start to make sense.

Sometimes spammers don't spam sites within certain times. Usually is is an attempt to avoid posting spam when moderators are about, so it will be visible for longer - i.e. till a moderator next visits the site.

I think the problem is that you have the $url within single quotes. Single quotes will not do variable replacements, it will be passing the longUrl as $url rather than the value contained in $url. Possible Solution, swap your double and single quotes round on the -d line.

It depends on what counts for a new user. If it is something like "less than 10 posts" then it seems reasonable to me.

We already have captchas on registration and as the spammers are registering I don't see how a captcha on posts would improve things. In my experience the harder it is to post the less real people post while the number of spam posts still increases (I am sure their AI can do captchas better than me now). Blocking IP ranges of known spammers can have a real impact on the spam numbers. On the last forum I was a moderator of we blocked a large chunk of China's IP range (not because the Chinese are spammers, just that was where we saw the largest quantity of spam coming from). The effect was amazing, our moderation times went from a 10-15 minute clean up each morning to a 2 minute clean up. The main reason that most sites don't like to block IP ranges is that they feel that it would be blocking legit members of their forum. I would hope that most of our members would be able to get round and their IP being blocked one way or another.

I would say run it for a good couple of hours and see how hot it gets before deciding if it needs additional cooling. If you can still hold after a couple of hours use then it is fine. I have seen some enterprise access points that can get really hot after a few hours of use, but even those didn't need any extra cooling.

The site could probably do with a couple more moderators in different time zones, just to erase the spam.

It might be worth doing a reset on your old router as well so that you are starting with the factory default settings rather than any odd settings that are left over from when it was the main router for the house.

That should do the job fine. You will need to switch off dhcp on your old router as you don't want two DHCP servers running at the same time, or you will find machines get given the same IP addresses or IP addresses on different subnets.