Ouch, that streettalk stuff does sound painful.:-)
Anyway, the solution to the evaluation problem lies in what some other people have already posted, namely you make a seperate file with a custom grammar, and a perl|shell|whatever script that parses that and re-writes the real groups file as needed so it is just a static, normal format file. (Analogous to the way that the front page to slahdot is static, rewritten once every ten or so minutes by a daemon.)
I think there is a project underway to provide such a unversal DB interface analogous to DBI.pm in perl or JDBC in java or whatever. I want to say it was part of PHPLIB but I'm not sure. I do recall seeing mention of it on phpbuilder.net somewhere.
Ah, anyway, (vi|vim|emacs|notepad.exe) are probably the most popular web dev tools.:-) But then I'm a web progammer and not a designer so my perspective is no doubt a little different.
This is mentioned in the "20 years after" section of The Mythical Man Month, under user interfaces. There he (Brooks) gives references to research done into doing this. I don't have my copy handy or else I'd post them here.
That fuckhead spammer is going to look like the ``starring attraction'' at goatse.cx. I think millions of mail system administrators and mail users everywhere have just been avenged...
Note that this is supposed to be an in-joke for those that have already inadvertantly been to goatse.cx, I don't advocate going for those of you whose eyes are untarnished. You Have Been Warned.
Be wary of places that don't have very good quesion design. E.g. if you're looking for a web perl coder having the test ask them what <<= does isn't that useful, and questions that rely on the coder knowing bad coding habits are ones that are really inadvisable (e.g. obscure defualt settings and values that any wise employer would not want in their codebase). And yes, I thought Brainbench's perl test was a little suboptimal in this regard.:-)
Further, allow them to have whatever reference material they want and access to the programming environment they'll be using (could be as simple as an xterm). You're trying to see how well they will do on the job, so you need to try to replicate the job's conditions.
It does say the rom is user upgradeable. So if a project doesn't yet exist to munge that rom space, it should be at least possible. It looks like pretty standard hardware (VGA, strongARM, the modem is probably hardware b/c I doubt that machine has the horsepower to do realtime DSP that's needed for winmodems to work), so in theory looking at the specs it shouldn't be impossible.
Side thought: you know what would be cool? A pda/small-ass-pc-thingy that not only used a boot rom but had a TFTP bios-boot/running-update option (just slap a NIC in the PCMCIA slot and go, or use the modem if you're out-of-office). This would make wholesale corporate updates easy as pie and allow easy backups of of the rom images in case any rollbacks are needed. Plus this has been done many times before so it's a solved engineering problem (cisco routers spring to mind as one example).
Maybe I'm smoking crack but I'd think that's not a function of hardware (i.e. it is just an API that sits on top of the OS with the low-level shite being in kernel space). True, to get hw accel it has to talk to metal but then it does fall back to sw for a feature it can't find hw support for, just like OpenGL.
In other words I'd think that if windows ran then any windows program would too.
But then I haven't tried this (I keep a seperate machine for non-essential things like opening Office docs and playing Half-Life on.;-) ), so I could be talkin' out my ass.
Friend of mine works for (large Texas-based computer manufacturer who uses Ask Jeeve's engine as the basis of their online tech support thing, called Ask Dudley or some equally shitty name). According to him it would give "interesting" results sometimes when it was first brought online and it's vocab hadn't been trained/tuned much.
For example: searching for ATAPI.SYS would bring up as one of the options something like "I'd like to know more about EAT PUSSY." (Somehow ATA PI.SYS (space for empahsis) sounded like that to Ask Jeeve's engine.)
My fiance is just got her degree in journalism. Like the line in Pulp Fiction, that pretty much made me interested in journalism too.;-) (We did talk a lot about what she studied, pretty relevant stuff to this modern, media-heavy age.)
Anyway, having pull quotes from a business's competitors (NothernLosers and Astalavista, both of whose search engines I can't stand for the cruddiness of their result sets) is really, really stinky journalism. What a suprise both of them slammed Google! I'm shocked! Google must be a horrible company!:-/ This article must have been written by a high-school intern, as it obviously shows the person looked for The Other Viewpoint, but didn't try to get objective other viewpoints (like, at a guess, from stock analysts who are in theory neutral).
And yes, we did snicker about her getting a BJ on graduation day. (Bachelors in Journalism):-D She always was at the head of her class...
I have a pretty strong suspicion that you're just trolling but on the off-hand chance you're not I'll try to make a somewhat educated rebuttal. Note that I'm not a CS major (computational chemistry), but that I do program a lot and I have done so for money. In other words, I tend to have a very pragmatic outlook on computer science.
Object Orientation -- Leaving aside the large and very valid debate about whether or not OO is a panacea (it's not IMHO, grotty OO code is still grotty, as is grotty functional lang code, et al.), bondage-and-discipline languages that try to force developers into doing something are not a good thing. Why? People get pissed off about the restrictions and find ways around them. Example: #define private public in a C++ header file. If you subscribe to the ``efficient market'' theory (lots of smart people looking at the same problem will arrive at roughly the same answer (c.f. the stock market and share valuation)), and believe that strict programming languages are a good thing, then why isn't everything being written in Ada today? (Oh, and the ``Real programming languages'' bit is really idiotic. I guess Linus was just using those toy languages C and asm when he coded linux's kernel...)
Readabilty -- is a function of the programmer, not the language. Regexps are line noise but they are the most effective way of transmitting meta-lingual patterns, so it's just a sub-language like any other proigramming language (you're programming matches on text instead of bits on a chip). It's just as easy to write bad Java as it is bad perl, and it's _really_ easy to write bad C (don't forget the obfuscated C contest has been around longer than the obfuscated perl one).
Standardization -- Nothing is constant in the IT industry. Just ask the guys who wrote the DB code for the airlines back in the 60's before the whole relational db / SQL revolution. Standards for a langugae are of overblown importance anyway, just go and try to tell me that advanced (read: uses all the advantages of the environment) C++ code for VC++6 and g++ are the same, even if you use "Standard C++".:-) To use one of your examples against you (of a "real" programming langugae), Java isn't standardized. Sun could decide tomorrow that "well, we never did like the + operator so we're deprecating it in favor of $%$(*@ as the addition operator". Secondly, since when do businesses recode everything as soon as a new release of somethig comes out? 4 of the last 5 places I've done contract work for were still using RH 6.0 and perl 5.005. (The other place was using 6.2 and 5.005) Thirdly, the language is communaly designed so it is highly unlikely core language features are going to change night-and-day in one or even two releases.
Maintainability -- so in essence you're saying every programmer before and after you is a Forrest Gump? Uh, yeah. Basically this boils down to the same thing listed under readabilty: to get good code you need good (disciplined and technically proficient) coders. Try arriving on a consensus in your coding dept as to code standards (<-- this will be much more effective if the coders themselves come up with it that a dictate from on high), and make sure you hire good (see de above) people. The rest will work out. COBOL and VB are popular in business becuase they just want something that will get the job done, and since both of the langugaes were designed to fit into the "problem space" of business, they are a path of least resistance. (They are the "good enough" solutions, although I think and hope that Java is beginning to replace them.)
As to your original thesis that perl isn't used by real programmers, well, I have about 200Kb of genome analysis code that ties into some hefty vis libs written in OpenGL that proves it wrong.:-) I.e. I think that's outside of the realm of webdev and sysadmin (which perl is very good for but is hardly it's sole strong point)
This is not to say that everything should be done in perl. All true masters of their trades[1] realize that no one tool is the perfect one for all tasks, so each should be used in accordance with its strengths. I.e. I'm not going to be writing any kernel modules in perl any time soon, kOrbit notwithstanding.
[1] I'm not claiming I am one, just noting a fact. Michaelangelo used more than one brush painting the Cistine Chapel. (I think that's the guy, but even if I'm forgetting the name the point stands.)
I've never used Dxpc but my understanding of it is that it compresses all the X11 traffic on the fly. Used in combination with other tactics (e.g. using the lightweight apps/windowmanagers[1] like another poster recommended), should help your situation out.
I think there is some way to channel X11 over SSH, and further to finangle SSH into compressing the data stream, giving you compression as well as security. But I have absolutely no idea how to do that, so you're on your own in that regard.:-)
[1] if you want a lightweight windowmanager, definitely check out wm2, the rpm of which is 58Kb and at runtime (2x xterm, xdaliclock, xload, xsetroot -solid grey25, 1280x1024x16@100dpi) it looks to be taking up about 2.9 megs, of which a hair over a meg is actually shared. Note that wm2 is really minimal but then that's sorta the point.:-) Another low-usage window manager is aewm, also interesting as it's homepage includes links to several other ultra-lights. (<--becuase let's face it, twm is U*G*L*Y and should not be tolerated in polite society, hah)
I doubt it. It's pretty centric to his network, meaning that it's got a lot of sensitive info in it (not just ip addresses but comments like # ok, this next line indicates someone sniffing at the db server, telling you what IP the db server is running on) and that porting it to a generic network would be a bitch.
But basically all you need is a perl script that reads from stdin, filters those lines (regexps), and has a sort of function table or callback mechanism (e.g. if/someMatch/ then somefunc() ). Redirecting the output from ipmon to the script isn't hard (direct pipe in the case of running ipmon from the shell without the -d daemonize flag, the -f flag (iirc) that specifies log file combined with a named pipe pointing at the script, that sort of thing). I don't know about linux firewalls (iptable, ipchains, ipfwadm, whatever) and logging methods, so that may require a different solution.
The callback funcs are where the action is (i.e. pass the line on to syslog(open a filehandle to syslog on LOG then "print LOG $line;"), drop it on the floor, do something else; sort of a syslog firewall;-) ). The command to play the mp3 was something like
system("/usr/local/bin/mpg123","/path/to/thunk.mp3 "). (using the split arg form of system is more secure b/c IIRC that way you aren't subject to shell interpretation)
The mp3 itself is a pretty trivial matter. Buy a cantaloupe (I think that's what he used) and let it get more than overripe but not enough to be gross. Put mic next to cantaloupe, recording to a wav, then hit cantaloupe with board. Convert wav to mp3 using any number of tools. Viola.:-) If you didn't mind the extra disk space you could leave it a wav and use some other tool (or maybe still mpg123, I've never tried to play a wav with it so it might support that (much simpler) format). Of course if you don't want to DIY I'm sure there are plenty of funny wav files out on the internet to use (movie quotes, tv clips (beavis and butthead anyone?), random sounds, etc).
I posted about this (or at least it was a tangent in a post about something else) a few days ago. A friend of mine wrote a perl script to moniter the ipmon-generated output of his OpenBSD firewall. When teh script recognized common script kiddie scans, it would play an mp3 of a ripe melon hitting a board (or a board hitting a ripe melon, either way a rather satisfying, moist thunk), represent a kiddie fruitlessly hitting his/her head against the firewall.
(Common scans like the port 135-140 range of MSFT shit, that sort of thing.)
I mean, just think, the next time some 313370 nd00d script kiddie tries to hack a machine you own [1], you could call up his mamma and tell her to spank them 'till their ass fell off. DoS = DuS (Dial Up Spanking)
side note: friend of mine put a sound card in his openbsd firewall (sits in front of several semi-popular local websites), and hooked up ipmon's output to a perl script that looked for things that should be handled proactively (it also logged to syslog). Everytime a pattern occured that was probably a script kiddie, he had the script play a sound sample (system("/usr/bin/mpg123","./thunk.mp3") ) of a ripe melon hitting a board. The sound of script kiddies thunking into a wall was... gratifying.
>Oh, and those air conditioners? Lose them. Buy a fan and deal with it. Ceiling fans are probably the best.
Heh. You wanna come spend a week at my place here in Austin or my parents' places in Houston next summer? I think a good 5-7 days of the Texas summertime will change your tune. The fastest land animal is not the cheetah, its a no-ac-cuz-it-harms-the-poor-widdle-animals eco-friend running to turn the AC back on...
(Note that I am a liberal and an environmentalist, but let's face facts, summertime sucks hardcore in the south.;-) )
From what I know, mir translates as "world", "peace", and "community"/"community property." Sort of like, in the "mir" all is right/"mir" (peaceful) when everything is "mir" (belong by everybody). In the mir --> our town/place/world/existence. Very peasant-like.
Or something. At least that's what one of my prof's said. Like many russian-isms, it is a very compact lingusitic structure that sort of depends on context how it expands into English, which can be a much more verbose language at times I think.
Hmm. I think given that data that it may not be intentional that the Americans felt the Russians were hacked off at them. Russian body language and linguistic habits can seem really gruff and cold to people not familiar with them (e.g. Russians very, very, very, veryrarely smile in public, to them it's a subconcious sign of sugar-coated, goody-two-shoes insincerity (now think about how Americans tend to smile first and ask questions later, is it any wonder most Russians think Americans are a bunch of twits?)). So the Russians may have just been in normal operating mode and the Americans misinterpreted it (magnified by the fact they have the personal space of a veal calf up there). Russian culture and American culture have grown much more accustomed to each other than they were in the 1970s, though, so this may not be as big a deal as it might have been then.
(This is not to imply that Russians are actually cold and gruff, they just might seem that way to strangers. My experience with them personally has been 180 degrees opposite in that you couldn't hope for warmer friends once they get to know you and you become accustomed to their body language.)
... from the article's description the friction exists primarily between the ground controllers (American and Russian) and the ISS crew, not among the ISS crew as most previous posts seem to speculate. These three guys have had quite a while to train together on the ground and are working closely together under adverse circumstances, so I'd bet that they get along pretty well given where they are and what they're doing (how much stress would you be under in a tin can that makes a submarine look roomy, surrounded by hard vacuum, and you only have 48 hours worth of O2 upon setting foot therein?).
Now add to that stress some jerkoff piloting a chair on the ground being pissy at you. I'd swear and be sarcastic too. Note also that the Russian language, and particularly their military slang, has a rich oral tradition when it comes to profanity (English absolutely pales by comparison, from my experience as a native English speaker and university-educated Russian speaker[1]), so the two Russian crewmen no doubt have a large palette with which to paint the situation and ground crews unflatteringly.
[1] My fianceé is trilingual in English, German, and Russian, and in her opinion Russian is by far the most profane in terms of common usage. Just an additional point of reference...:-)
... but I was in a Home Despot, er, Depot last week and I could have sworn the little floor terminals and POS systems were HP pizza-box workstations. The GUI certainly didn't scream MS (more like (n)curses). I can imagine that unix thin clients (virtually dumb terms) make a lot of sense for this use (high reliability (hw and sw), low power consumption, ease of development). True, x86 hw is cheaper than proprietary unix hardware, and win32 probably costs less in bulk than HPUX or whatever, but then how much revenue do you lose when all the cash registers BSOD?
Now up steps linux, a POSIX environment than runs on cheap x86 hardware and is available for free with a full suite of programming tools. Switching is definitely not a hard decision to make if you're already using a Unix-terminal-based solution.
Well, there is linhardware.com which is a pretty good source of information regardling what works (and how well, links to driver info, etc. sort of like HCL++). Linux only, which is too bad. Maybe some enterprising person will generalize this out to (random name) openhw.org for all the major alterna-OSes (linux, *BSD, BeOS). I can see this as being pretty easy to do using Slashcode (OSes = categories) or similar. There would probably have to be some way to sub-cateogry as well (e.g. BSD:WAN Interfaces or BeOS:3d Accelerators), to allow a dual functionality from the site combining the current news format of/. and the heirarchical format of LhD.
--
1/2 answer Re:Can WINE talk directly to hardware?
on
Wine In New Skins
·
· Score: 3
I don't know about how to get a win32 bin to access linux hardware, but have you considered using perl? Perl has modules for accessing serial and parallel ports, and you can pick a UI of your choice (everything from (n)curses to Gtk to Tk to wxWindows to...). Perl works on win32 and *nix (and pretty much anything else with a power cord, I'm expexting a perl-driven toaster Real Soon Now).
Plus, it doesn't suck like VB.
There are of course other ways to do this (C anyone?), the other scripting language people are likely to mention is python, which is a perfectly vaid choice as well and may be more preferable to somebody with an OO background (or pseudo-OO like VB). I _think_ there are comm port APIs for python (that's just comparatively simple C code after all, and one thing all the *nix-originating scripting langs seem to be great at is glomming around C code (perl, tcl, python, et al)).
I was reading Aelieen Frisch's _Essential System Administration_ (2e) last night and ran across the tidbit that AIX systems have no need of kernel recompilation becuase literally everything can be reconfiured on the fly with CLI commands (analogous to sysctl or something, but applying to everything). Of course AIX only runs on IBM hardware to my knowledge so the kernel team for AIX has a much more limited set of hardware to accomodate than linux does. (Don't take this as a recommendation of AIX per sé, I have no hands-on experience with it.)
Slashdot Mirror
Thou art a Trolle, ye Foul Beaste! Get thine Asse Backe Unter Thine Foule Bridge, lest I Whaule Ooop and Kicke Thine Bootae!
--
Ouch, that streettalk stuff does sound painful. :-)
Anyway, the solution to the evaluation problem lies in what some other people have already posted, namely you make a seperate file with a custom grammar, and a perl|shell|whatever script that parses that and re-writes the real groups file as needed so it is just a static, normal format file. (Analogous to the way that the front page to slahdot is static, rewritten once every ten or so minutes by a daemon.)
--
I think there is a project underway to provide such a unversal DB interface analogous to DBI.pm in perl or JDBC in java or whatever. I want to say it was part of PHPLIB but I'm not sure. I do recall seeing mention of it on phpbuilder.net somewhere.
Ah, anyway, (vi|vim|emacs|notepad.exe) are probably the most popular web dev tools. :-) But then I'm a web progammer and not a designer so my perspective is no doubt a little different.
--
This is mentioned in the "20 years after" section of The Mythical Man Month, under user interfaces. There he (Brooks) gives references to research done into doing this. I don't have my copy handy or else I'd post them here.
--
that auctions.yahoo.com will not be one of the ported areas... ;-)
--
That fuckhead spammer is going to look like the ``starring attraction'' at goatse.cx. I think millions of mail system administrators and mail users everywhere have just been avenged...
Note that this is supposed to be an in-joke for those that have already inadvertantly been to goatse.cx, I don't advocate going for those of you whose eyes are untarnished. You Have Been Warned.
--
Be wary of places that don't have very good quesion design. E.g. if you're looking for a web perl coder having the test ask them what <<= does isn't that useful, and questions that rely on the coder knowing bad coding habits are ones that are really inadvisable (e.g. obscure defualt settings and values that any wise employer would not want in their codebase). And yes, I thought Brainbench's perl test was a little suboptimal in this regard. :-)
Further, allow them to have whatever reference material they want and access to the programming environment they'll be using (could be as simple as an xterm). You're trying to see how well they will do on the job, so you need to try to replicate the job's conditions.
--
It does say the rom is user upgradeable. So if a project doesn't yet exist to munge that rom space, it should be at least possible. It looks like pretty standard hardware (VGA, strongARM, the modem is probably hardware b/c I doubt that machine has the horsepower to do realtime DSP that's needed for winmodems to work), so in theory looking at the specs it shouldn't be impossible.
Side thought: you know what would be cool? A pda/small-ass-pc-thingy that not only used a boot rom but had a TFTP bios-boot/running-update option (just slap a NIC in the PCMCIA slot and go, or use the modem if you're out-of-office). This would make wholesale corporate updates easy as pie and allow easy backups of of the rom images in case any rollbacks are needed. Plus this has been done many times before so it's a solved engineering problem (cisco routers spring to mind as one example).
--
Maybe I'm smoking crack but I'd think that's not a function of hardware (i.e. it is just an API that sits on top of the OS with the low-level shite being in kernel space). True, to get hw accel it has to talk to metal but then it does fall back to sw for a feature it can't find hw support for, just like OpenGL.
In other words I'd think that if windows ran then any windows program would too.
But then I haven't tried this (I keep a seperate machine for non-essential things like opening Office docs and playing Half-Life on. ;-) ), so I could be talkin' out my ass.
--
Friend of mine works for (large Texas-based computer manufacturer who uses Ask Jeeve's engine as the basis of their online tech support thing, called Ask Dudley or some equally shitty name). According to him it would give "interesting" results sometimes when it was first brought online and it's vocab hadn't been trained/tuned much.
For example: searching for ATAPI.SYS would bring up as one of the options something like "I'd like to know more about EAT PUSSY." (Somehow ATA PI.SYS (space for empahsis) sounded like that to Ask Jeeve's engine.)
Needless to say, that was changed real quick. :-)
--
My fiance is just got her degree in journalism. Like the line in Pulp Fiction, that pretty much made me interested in journalism too. ;-) (We did talk a lot about what she studied, pretty relevant stuff to this modern, media-heavy age.)
Anyway, having pull quotes from a business's competitors (NothernLosers and Astalavista, both of whose search engines I can't stand for the cruddiness of their result sets) is really, really stinky journalism. What a suprise both of them slammed Google! I'm shocked! Google must be a horrible company! :-/ This article must have been written by a high-school intern, as it obviously shows the person looked for The Other Viewpoint, but didn't try to get objective other viewpoints (like, at a guess, from stock analysts who are in theory neutral).
And yes, we did snicker about her getting a BJ on graduation day. (Bachelors in Journalism) :-D She always was at the head of her class...
--
I have a pretty strong suspicion that you're just trolling but on the off-hand chance you're not I'll try to make a somewhat educated rebuttal. Note that I'm not a CS major (computational chemistry), but that I do program a lot and I have done so for money. In other words, I tend to have a very pragmatic outlook on computer science.
Object Orientation -- Leaving aside the large and very valid debate about whether or not OO is a panacea (it's not IMHO, grotty OO code is still grotty, as is grotty functional lang code, et al.), bondage-and-discipline languages that try to force developers into doing something are not a good thing. Why? People get pissed off about the restrictions and find ways around them. Example: #define private public in a C++ header file. If you subscribe to the ``efficient market'' theory (lots of smart people looking at the same problem will arrive at roughly the same answer (c.f. the stock market and share valuation)), and believe that strict programming languages are a good thing, then why isn't everything being written in Ada today? (Oh, and the ``Real programming languages'' bit is really idiotic. I guess Linus was just using those toy languages C and asm when he coded linux's kernel...)
Readabilty -- is a function of the programmer, not the language. Regexps are line noise but they are the most effective way of transmitting meta-lingual patterns, so it's just a sub-language like any other proigramming language (you're programming matches on text instead of bits on a chip). It's just as easy to write bad Java as it is bad perl, and it's _really_ easy to write bad C (don't forget the obfuscated C contest has been around longer than the obfuscated perl one).
Standardization -- Nothing is constant in the IT industry. Just ask the guys who wrote the DB code for the airlines back in the 60's before the whole relational db / SQL revolution. Standards for a langugae are of overblown importance anyway, just go and try to tell me that advanced (read: uses all the advantages of the environment) C++ code for VC++6 and g++ are the same, even if you use "Standard C++". :-) To use one of your examples against you (of a "real" programming langugae), Java isn't standardized. Sun could decide tomorrow that "well, we never did like the + operator so we're deprecating it in favor of $%$(*@ as the addition operator". Secondly, since when do businesses recode everything as soon as a new release of somethig comes out? 4 of the last 5 places I've done contract work for were still using RH 6.0 and perl 5.005. (The other place was using 6.2 and 5.005) Thirdly, the language is communaly designed so it is highly unlikely core language features are going to change night-and-day in one or even two releases.
Maintainability -- so in essence you're saying every programmer before and after you is a Forrest Gump? Uh, yeah. Basically this boils down to the same thing listed under readabilty: to get good code you need good (disciplined and technically proficient) coders. Try arriving on a consensus in your coding dept as to code standards (<-- this will be much more effective if the coders themselves come up with it that a dictate from on high), and make sure you hire good (see de above) people. The rest will work out. COBOL and VB are popular in business becuase they just want something that will get the job done, and since both of the langugaes were designed to fit into the "problem space" of business, they are a path of least resistance. (They are the "good enough" solutions, although I think and hope that Java is beginning to replace them.)
As to your original thesis that perl isn't used by real programmers, well, I have about 200Kb of genome analysis code that ties into some hefty vis libs written in OpenGL that proves it wrong. :-) I.e. I think that's outside of the realm of webdev and sysadmin (which perl is very good for but is hardly it's sole strong point)
This is not to say that everything should be done in perl. All true masters of their trades[1] realize that no one tool is the perfect one for all tasks, so each should be used in accordance with its strengths. I.e. I'm not going to be writing any kernel modules in perl any time soon, kOrbit notwithstanding.
[1] I'm not claiming I am one, just noting a fact. Michaelangelo used more than one brush painting the Cistine Chapel. (I think that's the guy, but even if I'm forgetting the name the point stands.)
--
I've never used Dxpc but my understanding of it is that it compresses all the X11 traffic on the fly. Used in combination with other tactics (e.g. using the lightweight apps/windowmanagers[1] like another poster recommended), should help your situation out.
I think there is some way to channel X11 over SSH, and further to finangle SSH into compressing the data stream, giving you compression as well as security. But I have absolutely no idea how to do that, so you're on your own in that regard. :-)
[1] if you want a lightweight windowmanager, definitely check out wm2, the rpm of which is 58Kb and at runtime (2x xterm, xdaliclock, xload, xsetroot -solid grey25, 1280x1024x16@100dpi) it looks to be taking up about 2.9 megs, of which a hair over a meg is actually shared. Note that wm2 is really minimal but then that's sorta the point. :-) Another low-usage window manager is aewm, also interesting as it's homepage includes links to several other ultra-lights. (<--becuase let's face it, twm is U*G*L*Y and should not be tolerated in polite society, hah)
--
I doubt it. It's pretty centric to his network, meaning that it's got a lot of sensitive info in it (not just ip addresses but comments like # ok, this next line indicates someone sniffing at the db server, telling you what IP the db server is running on) and that porting it to a generic network would be a bitch.
But basically all you need is a perl script that reads from stdin, filters those lines (regexps), and has a sort of function table or callback mechanism (e.g. if /someMatch/ then somefunc() ). Redirecting the output from ipmon to the script isn't hard (direct pipe in the case of running ipmon from the shell without the -d daemonize flag, the -f flag (iirc) that specifies log file combined with a named pipe pointing at the script, that sort of thing). I don't know about linux firewalls (iptable, ipchains, ipfwadm, whatever) and logging methods, so that may require a different solution.
The callback funcs are where the action is (i.e. pass the line on to syslog(open a filehandle to syslog on LOG then "print LOG $line;"), drop it on the floor, do something else; sort of a syslog firewall ;-) ). The command to play the mp3 was something like
system("/usr/local/bin/mpg123","/path/to/thunk.mp3 "). (using the split arg form of system is more secure b/c IIRC that way you aren't subject to shell interpretation)
The mp3 itself is a pretty trivial matter. Buy a cantaloupe (I think that's what he used) and let it get more than overripe but not enough to be gross. Put mic next to cantaloupe, recording to a wav, then hit cantaloupe with board. Convert wav to mp3 using any number of tools. Viola. :-) If you didn't mind the extra disk space you could leave it a wav and use some other tool (or maybe still mpg123, I've never tried to play a wav with it so it might support that (much simpler) format). Of course if you don't want to DIY I'm sure there are plenty of funny wav files out on the internet to use (movie quotes, tv clips (beavis and butthead anyone?), random sounds, etc).
--
I posted about this (or at least it was a tangent in a post about something else) a few days ago. A friend of mine wrote a perl script to moniter the ipmon-generated output of his OpenBSD firewall. When teh script recognized common script kiddie scans, it would play an mp3 of a ripe melon hitting a board (or a board hitting a ripe melon, either way a rather satisfying, moist thunk), represent a kiddie fruitlessly hitting his/her head against the firewall.
(Common scans like the port 135-140 range of MSFT shit, that sort of thing.)
--
Of course Windows is #1 on your HD! POS won't boot from anywhere else. ;-)
--
I mean, just think, the next time some 313370 nd00d script kiddie tries to hack a machine you own [1], you could call up his mamma and tell her to spank them 'till their ass fell off. DoS = DuS (Dial Up Spanking)
side note: friend of mine put a sound card in his openbsd firewall (sits in front of several semi-popular local websites), and hooked up ipmon's output to a perl script that looked for things that should be handled proactively (it also logged to syslog). Everytime a pattern occured that was probably a script kiddie, he had the script play a sound sample (system("/usr/bin/mpg123","./thunk.mp3") ) of a ripe melon hitting a board. The sound of script kiddies thunking into a wall was ... gratifying.
--
>Oh, and those air conditioners? Lose them. Buy a fan and deal with it. Ceiling fans are probably the best.
Heh. You wanna come spend a week at my place here in Austin or my parents' places in Houston next summer? I think a good 5-7 days of the Texas summertime will change your tune. The fastest land animal is not the cheetah, its a no-ac-cuz-it-harms-the-poor-widdle-animals eco-friend running to turn the AC back on...
(Note that I am a liberal and an environmentalist, but let's face facts, summertime sucks hardcore in the south. ;-) )
--
From what I know, mir translates as "world", "peace", and "community"/"community property." Sort of like, in the "mir" all is right/"mir" (peaceful) when everything is "mir" (belong by everybody). In the mir --> our town/place/world/existence. Very peasant-like.
Or something. At least that's what one of my prof's said. Like many russian-isms, it is a very compact lingusitic structure that sort of depends on context how it expands into English, which can be a much more verbose language at times I think.
--
Hmm. I think given that data that it may not be intentional that the Americans felt the Russians were hacked off at them. Russian body language and linguistic habits can seem really gruff and cold to people not familiar with them (e.g. Russians very, very, very, veryrarely smile in public, to them it's a subconcious sign of sugar-coated, goody-two-shoes insincerity (now think about how Americans tend to smile first and ask questions later, is it any wonder most Russians think Americans are a bunch of twits?)). So the Russians may have just been in normal operating mode and the Americans misinterpreted it (magnified by the fact they have the personal space of a veal calf up there). Russian culture and American culture have grown much more accustomed to each other than they were in the 1970s, though, so this may not be as big a deal as it might have been then.
(This is not to imply that Russians are actually cold and gruff, they just might seem that way to strangers. My experience with them personally has been 180 degrees opposite in that you couldn't hope for warmer friends once they get to know you and you become accustomed to their body language.)
--
... from the article's description the friction exists primarily between the ground controllers (American and Russian) and the ISS crew, not among the ISS crew as most previous posts seem to speculate. These three guys have had quite a while to train together on the ground and are working closely together under adverse circumstances, so I'd bet that they get along pretty well given where they are and what they're doing (how much stress would you be under in a tin can that makes a submarine look roomy, surrounded by hard vacuum, and you only have 48 hours worth of O2 upon setting foot therein?).
Now add to that stress some jerkoff piloting a chair on the ground being pissy at you. I'd swear and be sarcastic too. Note also that the Russian language, and particularly their military slang, has a rich oral tradition when it comes to profanity (English absolutely pales by comparison, from my experience as a native English speaker and university-educated Russian speaker[1]), so the two Russian crewmen no doubt have a large palette with which to paint the situation and ground crews unflatteringly.
[1] My fianceé is trilingual in English, German, and Russian, and in her opinion Russian is by far the most profane in terms of common usage. Just an additional point of reference... :-)
--
... but I was in a Home Despot, er, Depot last week and I could have sworn the little floor terminals and POS systems were HP pizza-box workstations. The GUI certainly didn't scream MS (more like (n)curses). I can imagine that unix thin clients (virtually dumb terms) make a lot of sense for this use (high reliability (hw and sw), low power consumption, ease of development). True, x86 hw is cheaper than proprietary unix hardware, and win32 probably costs less in bulk than HPUX or whatever, but then how much revenue do you lose when all the cash registers BSOD?
Now up steps linux, a POSIX environment than runs on cheap x86 hardware and is available for free with a full suite of programming tools. Switching is definitely not a hard decision to make if you're already using a Unix-terminal-based solution.
--
Well, there is linhardware.com which is a pretty good source of information regardling what works (and how well, links to driver info, etc. sort of like HCL++). Linux only, which is too bad. Maybe some enterprising person will generalize this out to (random name) openhw.org for all the major alterna-OSes (linux, *BSD, BeOS). I can see this as being pretty easy to do using Slashcode (OSes = categories) or similar. There would probably have to be some way to sub-cateogry as well (e.g. BSD:WAN Interfaces or BeOS:3d Accelerators), to allow a dual functionality from the site combining the current news format of /. and the heirarchical format of LhD.
--
I don't know about how to get a win32 bin to access linux hardware, but have you considered using perl? Perl has modules for accessing serial and parallel ports, and you can pick a UI of your choice (everything from (n)curses to Gtk to Tk to wxWindows to ...). Perl works on win32 and *nix (and pretty much anything else with a power cord, I'm expexting a perl-driven toaster Real Soon Now).
Plus, it doesn't suck like VB.
There are of course other ways to do this (C anyone?), the other scripting language people are likely to mention is python, which is a perfectly vaid choice as well and may be more preferable to somebody with an OO background (or pseudo-OO like VB). I _think_ there are comm port APIs for python (that's just comparatively simple C code after all, and one thing all the *nix-originating scripting langs seem to be great at is glomming around C code (perl, tcl, python, et al)).
--
I was reading Aelieen Frisch's _Essential System Administration_ (2e) last night and ran across the tidbit that AIX systems have no need of kernel recompilation becuase literally everything can be reconfiured on the fly with CLI commands (analogous to sysctl or something, but applying to everything). Of course AIX only runs on IBM hardware to my knowledge so the kernel team for AIX has a much more limited set of hardware to accomodate than linux does. (Don't take this as a recommendation of AIX per sé, I have no hands-on experience with it.)
--