There are some situations where it might be impossible to do even an ISP-level restriction... of course those involve using proxies to get around restrictions on IRC by over-zealous network administrators.
But at any rate, even with an oper password, impact should have been minimal... (of course if the guy had/operserv raw enabled, I'm a little less certain... see below notes) it's with an SSH password to the box where unreal is running where things get troublesome (change password, then change the IRCd conf to your liking, you can't get control back until you get your server host to reset your password for you).
Theoretical scenario: Some jerk comes on and knows my oper password. First of all, my oper password is NOT the same as my SSH password. Let's say this jerk goes and/kill me, like in one quote I saw. First thing I do is hop back on and/gline him (or worse... I think/zline bans by IP, but I'm not a server root admin guy myself). I do believe with a/gline on him even with an oper pass he can't get on.
Of course if he/glines me I just SSH to my box, kill unreal, edit the conf to change my oper pass for good measure, and restart unreal (glines/klines/etc are volitile).
If you have a network you might have a bit of a problem, since you'd need to distribute the new conf to all admins, and glines tend to propagate along nodes until every single one gets shut down. But, if you can get in contact with the admins for the other servers usually it's trivial to get them the file and have them rehash and clear whatever damage (glines in this case).
As for the NickServ impersonation, it's important to use/nickserv or/ns INSTEAD of/msg to identify if your server supports it, since these aliases (on unreal at least) by default will only work if NickServ is a service. If your server doesn't, someone on this comments page mentioned that using NickServ@HOSTNAME will only msg NickServ if he's logged on from HOSTNAME... which, if it's a service, will always be the same. If someone impersonates NickServ, the host will change, and the/msg will fail (/whois NickServ to get the hostname).
One last thing occured to me... the services connect password... if someone had THAT, they could connect their own services (if the server was set up to allow services from any host, which is a poor configuration) or connect using an IRC client (not sure if that's possible, I'll have to experiment with that on my own server) and pose as genuine NickServ. Aliases would work, hostmask would show as legit services. Furthermore, since services get more powers than even the highest ranking opers, all hell could break loose. (Anyone who's fooled around with/operserv raw might know some of the powerful stuff you can do with it... and by the way, if you have it enabled on your server, disable it unless you're just running a server as a text or experiment and noone else uses it.)
I tried Slackware on my thumbdrive, but I could never get it to boot, the tutorial I used did cover replacing the boot sector, but the replacement boot sector could never find the operating system.
I then caved in and tried BartPE. I had similar problems with that, but I finally got it to work using a third-party tool.
Now I can boot from a thumb drive, which is pretty neat.
This brings up my main reason I don't use Linux though... it's near impossible to install on this computer without wiping my hard drive and starting from scratch. I can't find any tool to resize NTFS partitions (even the commercial Partition Magic fails to do it and has to roll back after about 3%), and that's all I have on my hard drives.
I'm not sure what algorythm, but I have Anope set up to use MySQL, I'm looking at the anope_ns_core table right now and passwords are stored as a 128-bit hash.
If I recall correctly, samy exploited MySpace (there's a link somewhere above if you never heard about it) by taking advantage of the fact that IE6 will execute Javascript: urls in CSS url() attributes (IE something like this:
background-image: url(javascript:codehere
Something like that at least. And of couse if you allow HTML tags with attributes anyone could stick a style="" on it and inject some javascript... in theory anyways.
I read somewhere, and I agree, that the best solution is to strip ALL HTML and use your own tag set (most web forums are way ahead in this department). If you do insist on allowing a subset of HTML, use whitelists to define allowed tags and attributes etc, instead of blacklists... because with a whitelist, if you leave something out, oh well someone can't use a tag they should be able to, it's more restrictive than it should be, they file a bug report and it's fixed. With a blacklist if you leave something out, it's a potential security hole.
I got a POS laptop from my church once (66mhz, MSDOS6, Win3.11). It had AOL 3.0 on it. I wondered if I could use the web browser piece of it to work on web pages, so I tried to load up one of my local HTML files in it.
I forget what it looked like exactly, but I recall it looked something like someone had taken the actual page I made and... exploded it.
I take it back then AOL wasn't terribly interested in delivering the broader WWW to their customers.
At college there are some serious limitations on p2p. I have managed to work around these since then (just gotta be careful not to generate any noticable traffic, and to encrypt everything) but before then I seriously considered using a free AOL trial. NetZero's free 10 hours per month just wasn't cutting it.
So, I go to the sign up page. I fill out some of the stuff (it's a multi page form so I'm submitting as I go) but then I see they need a CC number. I'm not about to give them that (what if I forget to cancel? etc, not to mention my parents handle my accounting and they would want to know why I signed up for AOL when I had internet at college). So I cancel out of the form.
THEY SAVED THE ENTERED INFORMATION EVEN THOUGH I CANCELLED THE SIGN UP. I wasn't even aware of this until a few days later when a rep called me and tried to get me to reconsider and sign up anyways. Luckily it was a one time call and I made it clear I was no longer interested.
The difference is someone can go in and backport it to 9x/ME/NT if they want to. Just because there's no official build doesn't mean YOU can't make an unofficial one. And someone will. I can guarantee it.
But before you slashdot their servers, please, think of the poor college students trying to pay off college loans. Don't make them break into their savings for bandwidth!
"Many have speculated that if we knew exactly why PatTheGreat thought that, we should know a lot more about the nature of the universe than we do now."
Trillian is another popular choice, although if you want Google Talk/Jabber and plugin functionality you have to shell out a few extra $$$. Even though I have I'm still looking at Miranda... the latest alpha looks nice.
The most annoying thing about this though is that some of the protocols don't support buddy list groups, or at least Trillian doesn't, because if I connect to my accounts from another computer with Trillian or from another OS, my groups pretty much explode. It's very annoying. I don't think there's anything the Trillian dev team can really do about it tho.
There is no penalty for using unsigned drivers, other than the fact that you run the risk of them being unstable. Run dxdiag.exe and check the tabs, you might find that you're using an unsigned video driver right now.
The file nvd3dum.dll is not digitally signed, which means that it has not been tested by Microsoft's Windows Hardware Quality Labs (WHQL). You may be able to get a WHQL logo'd driver from the hardware manufacturer.
The only ones who would be REALLY inconvenienced by this are the OEM guys, who would have to make sure their products only contain green status drivers, because the status of the drivers could change. The hardware guys would also be under pressure to make sure their drivers make green and stay green. Of course we the consumers end up winning from this, in theory, with more stable drivers.
I do in Vista, because then MS will be able to easily note programs that don't work with Vista. They can then determine if they broke compatibility by accident or if it's the program's fault, and perhaps even alert the program vendor to the problem before the final version of Vista ships.
XP crash reports are fully viewable, if I recall. I turned them off in my XP because there's really no point, as far as I can see. With Vista however, it's beta software, so I can see the use of it. It already has been well established that the data sent can include documents or spreadsheets you were working on in the app that crashed at the time. This is old news and has already gotten it's 15 minutes of fame/complaining about.
I highly doubt MS archives the crash data they receive. They'd have to have more servers than Google to do so, and they have no legitimate reason so why bother to?
The NSA angle would work for a plot point in fiction, but realistically, I wouldn't even give it a second thought if it weren't for the whole phone log fiasco. But really... NSA collecting Windows error logs? Just think about how rediculous that sounds for a minute.
I don't think it would be worth any employee's time, much less their career, to try and mine sensative data from error reports. There's bound to be so many that have no information, and even those that do are a full memory dump... not easy to sift through, or code a program to sift through.
Some nifty things are possible if you hook functions in the WinAPI (so all calls to that function would go through your function first). Your app could then put whatever restrictions on access it you wanted (you could hook file open functions, registry open functions, etc).
Here's an interesting article I found, wasn't the one I was searching for tho: http://www.codeproject.com/system/hooksys.asp
Here is an article that shows how to prevent processes from launching: http://www.codeproject.com/system/soviet_protector .asp
This might not be precisely what you wanted (a bit hackish compared to what you seem to be looking for) but it would work.
Slashdot Mirror
There are some situations where it might be impossible to do even an ISP-level restriction... of course those involve using proxies to get around restrictions on IRC by over-zealous network administrators.
But at any rate, even with an oper password, impact should have been minimal... (of course if the guy had /operserv raw enabled, I'm a little less certain... see below notes) it's with an SSH password to the box where unreal is running where things get troublesome (change password, then change the IRCd conf to your liking, you can't get control back until you get your server host to reset your password for you).
Theoretical scenario: Some jerk comes on and knows my oper password. First of all, my oper password is NOT the same as my SSH password. Let's say this jerk goes and /kill me, like in one quote I saw. First thing I do is hop back on and /gline him (or worse... I think /zline bans by IP, but I'm not a server root admin guy myself). I do believe with a /gline on him even with an oper pass he can't get on.
Of course if he /glines me I just SSH to my box, kill unreal, edit the conf to change my oper pass for good measure, and restart unreal (glines/klines/etc are volitile).
If you have a network you might have a bit of a problem, since you'd need to distribute the new conf to all admins, and glines tend to propagate along nodes until every single one gets shut down. But, if you can get in contact with the admins for the other servers usually it's trivial to get them the file and have them rehash and clear whatever damage (glines in this case).
As for the NickServ impersonation, it's important to use /nickserv or /ns INSTEAD of /msg to identify if your server supports it, since these aliases (on unreal at least) by default will only work if NickServ is a service. If your server doesn't, someone on this comments page mentioned that using NickServ@HOSTNAME will only msg NickServ if he's logged on from HOSTNAME... which, if it's a service, will always be the same. If someone impersonates NickServ, the host will change, and the /msg will fail (/whois NickServ to get the hostname).
One last thing occured to me... the services connect password... if someone had THAT, they could connect their own services (if the server was set up to allow services from any host, which is a poor configuration) or connect using an IRC client (not sure if that's possible, I'll have to experiment with that on my own server) and pose as genuine NickServ. Aliases would work, hostmask would show as legit services. Furthermore, since services get more powers than even the highest ranking opers, all hell could break loose. (Anyone who's fooled around with /operserv raw might know some of the powerful stuff you can do with it... and by the way, if you have it enabled on your server, disable it unless you're just running a server as a text or experiment and noone else uses it.)
I tried Slackware on my thumbdrive, but I could never get it to boot, the tutorial I used did cover replacing the boot sector, but the replacement boot sector could never find the operating system.
I then caved in and tried BartPE. I had similar problems with that, but I finally got it to work using a third-party tool.
Now I can boot from a thumb drive, which is pretty neat.
This brings up my main reason I don't use Linux though... it's near impossible to install on this computer without wiping my hard drive and starting from scratch. I can't find any tool to resize NTFS partitions (even the commercial Partition Magic fails to do it and has to roll back after about 3%), and that's all I have on my hard drives.
UnrealIRCd ships with sample configs to do this... they cause the commands ns and nickserv to send the text to NickServ ONLY IF it is a service.
I don't use Freenode myself but I imagine if it's Unreal it uses those commands.
I'm not sure what algorythm, but I have Anope set up to use MySQL, I'm looking at the anope_ns_core table right now and passwords are stored as a 128-bit hash.
47... what? Rupees? Yen?
Actually it's 13 on my Firefox... I set it to 14 and it looks alot better. Thanks for the suggestion! :)
If I recall correctly, samy exploited MySpace (there's a link somewhere above if you never heard about it) by taking advantage of the fact that IE6 will execute Javascript: urls in CSS url() attributes (IE something like this:
background-image: url(javascript:codehere
Something like that at least. And of couse if you allow HTML tags with attributes anyone could stick a style="" on it and inject some javascript... in theory anyways.
I read somewhere, and I agree, that the best solution is to strip ALL HTML and use your own tag set (most web forums are way ahead in this department). If you do insist on allowing a subset of HTML, use whitelists to define allowed tags and attributes etc, instead of blacklists... because with a whitelist, if you leave something out, oh well someone can't use a tag they should be able to, it's more restrictive than it should be, they file a bug report and it's fixed. With a blacklist if you leave something out, it's a potential security hole.
Get yourself a hotmail account and have PHP fire off e-mails to it. Tweak as needed until you get one through that's not marked as spam.
There are a handful of pages that proxy to google... for example.
I got a POS laptop from my church once (66mhz, MSDOS6, Win3.11). It had AOL 3.0 on it. I wondered if I could use the web browser piece of it to work on web pages, so I tried to load up one of my local HTML files in it.
I forget what it looked like exactly, but I recall it looked something like someone had taken the actual page I made and ... exploded it.
I take it back then AOL wasn't terribly interested in delivering the broader WWW to their customers.
At college there are some serious limitations on p2p. I have managed to work around these since then (just gotta be careful not to generate any noticable traffic, and to encrypt everything) but before then I seriously considered using a free AOL trial. NetZero's free 10 hours per month just wasn't cutting it.
So, I go to the sign up page. I fill out some of the stuff (it's a multi page form so I'm submitting as I go) but then I see they need a CC number. I'm not about to give them that (what if I forget to cancel? etc, not to mention my parents handle my accounting and they would want to know why I signed up for AOL when I had internet at college). So I cancel out of the form.
THEY SAVED THE ENTERED INFORMATION EVEN THOUGH I CANCELLED THE SIGN UP. I wasn't even aware of this until a few days later when a rep called me and tried to get me to reconsider and sign up anyways. Luckily it was a one time call and I made it clear I was no longer interested.
No, because then they'd just sneak DRM into that bill too!
The difference is someone can go in and backport it to 9x/ME/NT if they want to. Just because there's no official build doesn't mean YOU can't make an unofficial one. And someone will. I can guarantee it.
The top guys who know what they're doing KNOW it's a bad idea, but management says do it anyway.
Said management is definately looking a little pointy-haired.
I found this picture of the prototype: http://images.google.com/images?q=tbn:7SY6ib0nUY04 BM:battleteam.net/tech/fis/docs/images/halflife2_s canner2.png
...if they don't move fast enough for you:
http://students.guildhall.smu.edu/~weekdaywarrio r/SetupWeekdayWarrior.exee s/ShantytownSetup.exe
http://students.guildhall.smu.edu/~shantytown/fil
But before you slashdot their servers, please, think of the poor college students trying to pay off college loans. Don't make them break into their savings for bandwidth!
I'd have to say this is the first Slashdot article in YEARS that's given me stuff that matters! These'll be fun to play. :D
"Many have speculated that if we knew exactly why PatTheGreat thought that, we should know a lot more about the nature of the universe than we do now."
Trillian is another popular choice, although if you want Google Talk/Jabber and plugin functionality you have to shell out a few extra $$$. Even though I have I'm still looking at Miranda... the latest alpha looks nice.
The most annoying thing about this though is that some of the protocols don't support buddy list groups, or at least Trillian doesn't, because if I connect to my accounts from another computer with Trillian or from another OS, my groups pretty much explode. It's very annoying. I don't think there's anything the Trillian dev team can really do about it tho.
There is no penalty for using unsigned drivers, other than the fact that you run the risk of them being unstable. Run dxdiag.exe and check the tabs, you might find that you're using an unsigned video driver right now.
The only ones who would be REALLY inconvenienced by this are the OEM guys, who would have to make sure their products only contain green status drivers, because the status of the drivers could change. The hardware guys would also be under pressure to make sure their drivers make green and stay green. Of course we the consumers end up winning from this, in theory, with more stable drivers.
I do in Vista, because then MS will be able to easily note programs that don't work with Vista. They can then determine if they broke compatibility by accident or if it's the program's fault, and perhaps even alert the program vendor to the problem before the final version of Vista ships.
XP crash reports are fully viewable, if I recall. I turned them off in my XP because there's really no point, as far as I can see. With Vista however, it's beta software, so I can see the use of it. It already has been well established that the data sent can include documents or spreadsheets you were working on in the app that crashed at the time. This is old news and has already gotten it's 15 minutes of fame/complaining about.
I highly doubt MS archives the crash data they receive. They'd have to have more servers than Google to do so, and they have no legitimate reason so why bother to?
The NSA angle would work for a plot point in fiction, but realistically, I wouldn't even give it a second thought if it weren't for the whole phone log fiasco. But really... NSA collecting Windows error logs? Just think about how rediculous that sounds for a minute.
I don't think it would be worth any employee's time, much less their career, to try and mine sensative data from error reports. There's bound to be so many that have no information, and even those that do are a full memory dump... not easy to sift through, or code a program to sift through.
...
OK let me straighten some things out. Picture of your monitor != a photo, and it usually makes for a pretty #$%&y screenshot unless you have an LCD.
A scan is also definately not a photo, especially if you didn't scan a photo in.
The content of the images isn't going to change (degrade, maybe) by doing these things. C'mon, use your brain, gEvil!
telnet.exe is no longer included with Windows starting with Vista.
Luckily we still have putty.exe. :)
Some nifty things are possible if you hook functions in the WinAPI (so all calls to that function would go through your function first). Your app could then put whatever restrictions on access it you wanted (you could hook file open functions, registry open functions, etc). Here's an interesting article I found, wasn't the one I was searching for tho: http://www.codeproject.com/system/hooksys.asp Here is an article that shows how to prevent processes from launching: http://www.codeproject.com/system/soviet_protector .asp
This might not be precisely what you wanted (a bit hackish compared to what you seem to be looking for) but it would work.
9x/ME/NT support is dropped. Check the Bugzilla bug linked to in the article, it states it right in the title.