Is it just me or is the pace of kernel releases accelerating dramatically. Just a quick look at the number, 2.6.6 seems to indicate that we are now 3/5 through the lifespan of the 2.6 kernel.
Version numbers are not usually decimal, they're usually a sequence of arbitrary dot-separated integers - don't think of decimal fractions, think of books with chapter 1, section 1.2, subsection 1.2.34, paragraph 1.2.34.5 (this scheme is common in maths and computer science textbooks).
Linux 2.4 is up to 2.4.26, 2.2 is up to 2.2.20ish, and I'm sure 2.6 will go well into the 20s too. (Updates are also more frequent in the early stages of a release kernel - I seem to remember 2.4 settled down to infrequent, minor updates somewhere around.17 or.18).
Perl version numbers used to be written in decimal (for a long time the current version was 5.00503), but have moved to a more conventional numbering scheme where that would be written as 5.5.3.
As for the frequency of updates: the line I've heard is that the dev kernels (2.odd.x) are alpha versions for developer use, and the "stable" kernels (2.even.x) are beta versions for public testing. If you want something that's "release quality", pick a reliable and non-bleeding-edge distribution and use their release kernels instead (I believe Debian stable currently runs heavily security-patched versions of either 2.4.17, 2.4.18 or 2.2.20, depending on CPU architecture - any of these are fine, as long as they support your hardware).
The Computer Lab at Cambridge University (your second link) is housed in the William Gates building, but apparently Linux is the environment of choice there.
(Many mathematicians at Cambridge certainly use Linux - Red Hat is usually the distro chosen, I think.)
What some people are suggesting is that people using illegal copies of Windows should be allowed to install security patches, at least the important ones, in order to reduce the damage done when a worm starts spreading - if illegal copies can't be patched, every illegal copy is an extra carrier for worms. The only way MS pay for that is in extra bandwidth for the Windows Update servers, which I suspect would be a pretty small cost (particularly if the next big worm DoSs Microsoft yet again, in which case having more updates downloaded would probably be a net saving).
The other side of the argument is that Microsoft should have no obligation to support illegal copies, and indeed should reduce the functionality of illegal copies in order to encourage people to buy a copy instead; this is the philosophy MS currently follow, to some extent, by having Windows Update and service packs not install on copies with a bad CD-key.
The problem with using patches as an area of reduced functionality is that most people don't particularly care about the security of their computer at the best of times, so it's not a big deterrent to illegal copying; at the same time, illegal copies getting worms and such affects everyone on the Internet, whether they're illegal Windows users, legit Windows users, or not even using Windows.
(There's also the argument that Microsoft have tacitly encouraged illegal copies in the past in order to get more market share, which I think might be what you're referring to, but the above applies whether you believe this or not.)
Microsoft should set the updates to automatically remove the operating system from anyone who is not a legit user
False positives under MS's current policy are merely an annoyance, but if they followed your policy and their warez-detection algorithm got any false positives whatsoever, it'd wipe the OS of a legit user - I for one wouldn't appreciate that. Microsoft have, um, a bit of a reputation problem as it is:-) and I can't imagine it'd get any better if it became public knowledge that their security updates sometimes deleted the operating system.
I can't imagine it would kill that many warezed copies either (once word got around), it'd just encourage anyone with an illegal copy not to install patches, and since that has a negative effect on the rest of the Internet, it'd be irresponsible.
*** now talking on #hypothetical-warez-channel - Topic: Get your XP isos here! <w4r3z-k1dd1e> don't install yesterday's critical update whatever you do, I got burned by it <@l33t_d00d> how's that? <w4r3z-k1dd1e> it deleted my OS! <w4r3z-k1dd1e> had to reinstall it <@l33t_d00d> lol, didn't you know? <@l33t_d00d> some of the patches do stuff like that <@l33t_d00d> safest way is to skip them all <w4r3z-k1dd1e> doesn't that make your pc not secure? <@l33t_d00d> heh, whatever <@l33t_d00d> that's what *they* tell you <w4r3z-k1dd1e> ah, k *** l33t_d00d has changed topic to "Remember kids, patches are for the weak"
Is that really what you want the warez kiddies to be thinking, and if so, would your answer change when the next Code Red/Nimda/Slammer/Sasser/... turns up?
In Deus Ex, particularly if you're running out of health, incapacitating guards with the riot prod and pepper spray is often a very effective method of getting past (a couple of hits from the riot prod will knock most things unconcious, or if you just need to get past, one prod hit or catching them with pepper spray will stun them for long enough). The tranquilizer darts fired from the mini-crossbow are also very useful, or there's always the tear-gas grenades. With EMP grenades too, you don't even need to destroy the guard robots.
Indeed, one of the NPCs will criticise you and give you less equipment if you kill too many opponents in the first level (this was spoiled slightly by counting knockouts as kills, thus penalizing everyone except stealth players; the first patch changed this so non-lethal weapons were advantageous too, I think).
It's a pity you can't just knock out guard dogs when they try to eat you, but the developers didn't bother giving them separate "unconscious" and "dead" states, so you have to kill those if you can't avoid them (if you use a "non-lethal" weapon, they die anyway).
- a decent editor to supplement whatever's in the base system these days (used to be joe, but if I reinstalled now I'd pick Vim) - sudo - aptitude (dselect is useful, but unpleasant) - less, if it's not already there - ssh (both client and server) - links and/or lynx for documentation - CVS (I keep meaning to switch my config-file repositories to something better though - I've tried Subversion and GNU Arch, and I intend to try out Darcs next) - Mutt
and if this machine needs a GUI:
- XFree86 - xterm (specifically uxterm, my systems all run in UTF-8) - KDE 3, or XFCE 4 if KDE's too big to be usable, or (Open|Flux|Black)box if I'm getting desparate - Firefox
I know what you mean. Idea: assuming you don't have 10GB of *new* music and a lot of that is older CDs, Ebay them or something - obtain a legal copy cheaply, and don't give the cartel your money.
You could also downloadsomemusiclegally, of course, or buy from non-cartel musicians (I have CDs by two of the artists I linked, and a third one on order).
A while ago I saw something about a Deus Ex film being under consideration - if done right, that could work.
Unfortunately it would probably get overly Hollywoodized (good guys made 100% good, bad guys made 100% evil, JC Denton made into a generic big-gun-wielding action hero) but if instead it was scripted/filmed in a way that reflected the style of the game, it'd be impressive.
(Of course, the other issue is that its anti-authoritarian plot and characters' musings on the nature of freedom might not be too popular politically at the moment...)
it's pretty tough to take linear algebra notes on a computer
Tried LaTeX? If you mean the same thing by "linear algebra" as I do (vectors, vector spaces, linear maps, matrices) it shouldn't be too bad if your course is mostly abstract - abstract algebra with just Latin+Greek letters, superscripts and subscripts is trivial to do in LaTeX, although writing out vectors and matrices explicitly *is* a bit of a hassle. A vector:
(mathematical correctness not guaranteed, it's the wrong time of night to be trying to remember the Normal distribution, but that might be right if I'm lucky;-)
I've (just about) successfully taken Representation Theory notes (involving integrals and inner products) on a laptop, in lectures given by a fast lecturer (there wasn't much time to waste when handwriting, so I'm surprised I managed to keep up at all). I gave up after 3 lectures or so, since writing LaTeX took up valuable brain power which I needed to use on understanding the lectures (also, diagrams required switching to my usual pen+paper and scanning it later).
Make your mod and all necessary files available in a safe place on the internet that isn't file planet.
I'd love to (and I agree Vileplanet is undesirable), but paying excessive bandwidth costs for people to download doesn't appeal to me (and yes, I do pay for my web hosting, but that's orders of magnitude cheaper).
I used to do Unreal Tournament mods: nothing big like ChaosUT, just some fixes and addons related to custom model support. I was also the official source for a couple of people's custom models, since I was the last to work on them and I had a more stable web server than the authors did. The two models and my Advanced Model Support mod were about 1MB each, the Bonus Pack 4 fix was only a couple of hundred K.
Despite their smallish size, I had to move all the large downloads onto Vileplanet, because they were sucking too much bandwidth; at the point my then-web host started limiting and charging for bandwidth I'd been using a consistent 2.5GB/month for several months, and I hadn't even made a release recently or anything; that was just normal traffic. At my then-current host's prices, bandwidth charges would have been something like 10 times the cost of my basic hosting account, and I'm sure I'd have spiked over that level considerably when I made a new release.
I hope you can see that if you're someone like ChaosUT, hosting mods that are, say, 30 megabytes, getting as many downloads as I did (a couple of thousand a month), while hosting your own downloads, would be rather painful, and actually being popular (I was, after all, in a pretty niche part of the mod scene) would make your bandwidth prices go up even further.
It's interesting to read the EULAs for games with a thriving mod scene, particularly those where the developers actively encourage mods. In many cases, popular mods are technically in breach of the EULA; I assume the problem is that, say, Epic or Valve encourage modding, but Infogrames/Atari or Sierra are responsible for setting the EULA.
(Of course, many modders, particularly those who do maps, models etc. alone rather than being part of a major mod/TC team, don't understand copyright and are happy to release mods without actually giving anyone permission to distribute the mod...)
I don't even think IPSec allows for you to communicate with machines on the same LAN on the same Subnet.
I don't know about the Windows implementation, but KAME (the *BSD IPSec stack, also used in Mac OS X, Linux 2.6 and Debian's patched Linux 2.4) looks as though it will do that fine.
Set up a policy for all traffic from anywhere to your Windows box, and vice versa, to have mandatory encryption in tunnel mode.
You will then need to to set up more specific policies for UDP port 500 (isakmp), and for protocols 50 (esp) and 51 (ah), to avoid trying to apply IPSec to them, since they're what IPSec itself uses (if you don't de-restrict these, you have a chicken and egg problem). You may also want to allow non-IPSec'ed DNS, or ssh, or whatever
Totally untested configuration (you may need to reverse the order of the lines):
#!/usr/bin/setkey -f # This config is for the restricted box # On the gateway, exchange the "in" and "out" keywords flush; spdflush;
# IPSec gateway is 192.168.0.1 # Restricted box is 192.168.0.2
# ISAKMP over UDP spdadd 192.168.0.1[500] 192.168.0.2[500] udp -P in none; spdadd 192.168.0.2[500] 192.168.0.1[500] udp -P out none;
# Encrypted IPSec data spdadd 192.168.0.1 192.168.0.2 esp -P in none; spdadd 192.168.0.2 192.168.0.1 esp -P out none;
# "Signed" IPSec data spdadd 192.168.0.1 192.168.0.2 ah -P in none; spdadd 192.168.0.2 192.168.0.1 ah -P out none;
# Everything else spdadd 192.168.0.2 0.0.0.0/0 any -P out ipsec esp/tunnel/192.168.0.2-192.168.0.1/require; spdad d 0.0.0.0/0 192.168.0.2 any -P in ipsec esp/tunnel/192.168.0.1-192.168.0.2/require;
Viruses that claim to be from your email provider are even sillier in countries whose registrars don't sell second-level domains (like the UK and Australia - there are a few second-level domains like co.uk, org.uk, net.uk, com.au etc., and you can buy third level domains like pseudorandom.co.uk).
Because my email is at pseudorandom.co.uk, I get occasional mails from "Co.Uk administrator " telling me my account is going to be suspended. Hmm, why don't I believe that? It's like getting an email from postmaster@com...
In a typical GUI environment, there tends to be something of a duality. Either you muddle through with whatever features a programmer has deigned to give you, or you make the effort to fire up your heavyweight Real Programming Environment (Visual Something-or-other, Delphi, the Java IDE du jour, Emacs...) and construct a proper program; there's a split between being an ordinary user, and being a programmer.
I agree that there's not a great difference between writing a snippet of shell script and writing a snippet of Perl or Python or something (or VBscript or Javascript via WSH if you're in Windows), and Windows with WSH scripting has the potential to be as expressive as Unix with shell scripting; however, you do have to make the conscious decision "this would be easier as a program, I will write a program now."
Shell script in a CLI does have the advantage that you're already in the shell, and you're already used to the shell from using it interactively, so there isn't the conceptual leap; in interactive use you're an ordinary-user-but-acting-a-bit-like-a-programmer, and the switch to programmer-but-also-an-ordinary-user mode isn't as large. (Sorry, that's a bit cumbersome, hopefully you can see what I mean though...)
You can switch from single commands to pretty complicated write-only one-liners within a few lines, without having to think "at this point it would be easier to stop using the usual interface and write a program instead". When you eventually do decide a program is required (which is at a higher level of complexity than it would be in a GUI) you can just cut and paste your previous one-liner, indent it nicely and you have the beginnings of a proper program already.
The other advantage of the CLI in Unix specifically is that a *lot* of stuff is scriptable, since it's designed around command line arguments and pipes. If you want to, say, make thumbnails of images, in Unix you already have ImageMagick at your disposal and ready to be scripted, but in other environments you have to hope you can find a "remote-controllable" app. Mac OS (by which I mean the proprietary GUI bit, not the Darwin Unix bit) actually has this advantage to some extent *without* being a CLI, thanks to many apps accepting "remote control" via Applescript; however, it still has a conceptual leap between "click on things" and "write an Applescript" (if you can record keyboard/mouse "macros", I suppose that's an intermediate step - sorry, I haven't played with Mac OS much).
So... it's the same interface as my mobile phone (a Nokia bought about 18 months ago, but older Nokias had basically the same interface for a few years), but applied to a multimedia asset player? I fail to see how that's an innovation worth protecting.
I would have thought the record companies' worry should be the other way round. If the line drawn by the law between legal format-shifting/fair use and illegal duplication is fairly close to people's view of the line between ethical and unethical, I would have thought people would tend to obey the law, but if format-shifting seems perfectly reasonable but is in fact illegal, then customers who copy their CDs to other formats are breaking the law anyway, so might well reason that they have nothing to lose by copying their friends' CDs too ("might as well be hung for a sheep as for a lamb").
That's a really bad solution for dialup users, who rely on their cache being large enough to hold a decent volume of common images (for instance, to make Slashdot non-painful on dialup, you want the top-left logo and the comment icons to be in cache).
I seem to remember that the first UT aimbot was "Funbot", written by Darkbyte[S&D] after discussing the possibility of UT cheats with players who insisted UT was cheat-proof. It was a proof of concept, but was released after he foolishly sent a copy to someone who claimed to be an anti-cheat developer. Was that you?
(UT players might also know Darkbyte as the co-author, with Dr.Sin, of Client-Side Hack Protection.)
I haven't used Funbot, but I was sent a copy by someone who'd collaborated with the CSHP developers, when he was checking over one of my mods for possible attacks. The decompiled source is surprisingly simple, just a hook into a loophole in the UT "sandbox", and a bit of vector maths to do the aiming.
I'm surprised nobody ever made a native-code UT aimbot: as far as I know, the aimbot/CSHP arms race was entirely in Unrealscript, although admittedly I never dug into the cheat world particularly far.
OK, the obligatory rant:
Cheating in online games sucks the fun out of the game. You're amusing yourself and spoiling the game for everyone else on the server; I hope it's obvious how childishly destructive that is.
I use a modified version of mserv, a weighted-random online jukebox, with several separate userIDs; when away from the computer with my music on it, I use a MP3 CD player with some loosely themed CD-RWs.
According mserv's top 20 lists (of songs most likely to be played), it looks like I've biased the "personalities" as follows:
work: top-40-ish rock, trip-hop
games: metal, punk, heavier end of rock, techno
code (concentrated coding/deep hack mode): mostly 90s trance, some (particularly electronic) rock
smcv (general all-purpose setup): any of the above
I see absolutely no reason for people to be listening to music while in any sort of educational institution.
I'm a maths student; like many students here, I do a lot of work on problem sheets (i.e. homework) between or after lectures, in the university maths department canteen/common room. I find wearing earphones and listening to music helps me to concentrate on what I'm doing by masking out random noises, the conversation on the next table, or whatever; also, even if I'm not making progress, I find music makes it less frustrating/boring to work on one question for a long time.
If I'm with friends I usually leave the headphones off, unless I want to concentrate (e.g. imminent deadline:-), in which case the headphones not only give me music to listen to, but act as a mild "do not disturb" sign (I suppose I mean "don't talk to me just for the sake of it, but if you want to ask me something specific, go ahead"). So, yes, I'm being slightly antisocial, but the alternative is hiding in the library or my room and being completely antisocial.
Yes, it's important to have enough money. That's not the same as saying worth is measured in money. Is it worthwhile earning a huge salary, if you're too busy to spend it and too stressed to enjoy it?
However, I start to wonder where your benefit is. You are - out of principle - not making any money out of this, because it is open-source and you and your buddies insist that it must be absolutely free. So you are putting all of that time and energy into this project for what? Fame? To found a career? Come on.
For fun? For interest? To prove to yourself that you can? For the feeling that you've contributed something useful to the world?
Now if most linux distros agree to stick to a common Filesystem Hierarchy system (http://www.pathname.com/fhs/) then you can use all sorts of packages together.
I could use Fedora packages, apt packages, debian packages, gentoo build scripts and all sorts of stuff and pluss get support for closed sourced software easily in any distro of my choosing.
It's not that simple. Shared library versioning, in particular, is a pain.
Debian (my distro of choice) devotes a lot of maintainer time to managing tricky transitions between library versions - if an app directly or indirectly links to incompatible libraries (perhaps you're using Mozilla linked against one major version of libpng, and also linked against Gtk which is linked against a different major version of libpng), it will break.
Debian's complex dependency system copes, but only because Debian packages are consistent and more-or-less come from a single source. If Red Hat's libfoo package isn't 100% compatible with Debian's libfoo package, mixing Red Hat and Debian packages which depend on libfoo will probably result in subtle breakage.
Incidentally, Debian does have numbered releases, every time a new stable release is made ("woody" is Debian 3.0); it also has release revisions (woody is currently at version 3.0r2, which a Windows admin might compare to "Debian 3.0 service pack 2"). It's just testing and unstable that have more-than-daily updates.
Softlink (aka symlink): Has it's own file, but any access gets routed to the real file. This is what most links I deal with are. If you delete the real file, the link breaks, but stays there.
Hardlink: Doesn't have own file. Behaves exactly like the original in all ways, but has a different name. If you delete a hardlink or the original file, nothing happens until *all* instances are gone. Hardlinks must be on the same partition.
FYI, as far as I know, symlinks are just files containing a relative filename, with a special flag in the filesystem (at the same low level as "this is a directory") which marks it as a symlink rather than an ordinary file. The size reported for symlinks under Linux/ext2 is certainly the number of bytes in the filename it points to.
A shortcut is basically a file containing a filename (and other details like the icon), but without the same low-level support in the kernel (open()ing a shortcut gives you its data, not that of the referenced file).
A Mac OS alias (what you get if you run "ln -s" on a HFS+ partition) is what a Windows shortcut should have been (it has more data than just the filename, like the Windows shortcut, but is transparent to apps using the Unix API, like a Unix symlink).
The version I've always heard is that hundreds of years ago the only way to be educated was a private tutor. When they were introduced, "public schools" (schools where pupils' parents pay fees) were called that to differentiate between private tuition and a public school.
The terminology is a bit unfortunate, now that private tuition doesn't happen and state schools are more public than "public schools", but that's how the English language works;-)
Schools entirely paid for by taxes are "state schools" (as in "separation of Church and State", not as in "Washington state").
Indeed. I've heard it said that stable is for people who need stability, unstable is for people who can cope with instability, and testing is for people who're not paying attention.
Slashdot Mirror
Is it just me or is the pace of kernel releases accelerating dramatically. Just a quick look at the number, 2.6.6 seems to indicate that we are now 3/5 through the lifespan of the 2.6 kernel.
.17 or .18).
Version numbers are not usually decimal, they're usually a sequence of arbitrary dot-separated integers - don't think of decimal fractions, think of books with chapter 1, section 1.2, subsection 1.2.34, paragraph 1.2.34.5 (this scheme is common in maths and computer science textbooks).
Linux 2.4 is up to 2.4.26, 2.2 is up to 2.2.20ish, and I'm sure 2.6 will go well into the 20s too. (Updates are also more frequent in the early stages of a release kernel - I seem to remember 2.4 settled down to infrequent, minor updates somewhere around
Perl version numbers used to be written in decimal (for a long time the current version was 5.00503), but have moved to a more conventional numbering scheme where that would be written as 5.5.3.
As for the frequency of updates: the line I've heard is that the dev kernels (2.odd.x) are alpha versions for developer use, and the "stable" kernels (2.even.x) are beta versions for public testing. If you want something that's "release quality", pick a reliable and non-bleeding-edge distribution and use their release kernels instead (I believe Debian stable currently runs heavily security-patched versions of either 2.4.17, 2.4.18 or 2.2.20, depending on CPU architecture - any of these are fine, as long as they support your hardware).
The Computer Lab at Cambridge University (your second link) is housed in the William Gates building, but apparently Linux is the environment of choice there.
(Many mathematicians at Cambridge certainly use Linux - Red Hat is usually the distro chosen, I think.)
What some people are suggesting is that people using illegal copies of Windows should be allowed to install security patches, at least the important ones, in order to reduce the damage done when a worm starts spreading - if illegal copies can't be patched, every illegal copy is an extra carrier for worms. The only way MS pay for that is in extra bandwidth for the Windows Update servers, which I suspect would be a pretty small cost (particularly if the next big worm DoSs Microsoft yet again, in which case having more updates downloaded would probably be a net saving).
:-) and I can't imagine it'd get any better if it became public knowledge that their security updates sometimes deleted the operating system.
The other side of the argument is that Microsoft should have no obligation to support illegal copies, and indeed should reduce the functionality of illegal copies in order to encourage people to buy a copy instead; this is the philosophy MS currently follow, to some extent, by having Windows Update and service packs not install on copies with a bad CD-key.
The problem with using patches as an area of reduced functionality is that most people don't particularly care about the security of their computer at the best of times, so it's not a big deterrent to illegal copying; at the same time, illegal copies getting worms and such affects everyone on the Internet, whether they're illegal Windows users, legit Windows users, or not even using Windows.
(There's also the argument that Microsoft have tacitly encouraged illegal copies in the past in order to get more market share, which I think might be what you're referring to, but the above applies whether you believe this or not.)
Microsoft should set the updates to automatically remove the operating system from anyone who is not a legit user
False positives under MS's current policy are merely an annoyance, but if they followed your policy and their warez-detection algorithm got any false positives whatsoever, it'd wipe the OS of a legit user - I for one wouldn't appreciate that. Microsoft have, um, a bit of a reputation problem as it is
I can't imagine it would kill that many warezed copies either (once word got around), it'd just encourage anyone with an illegal copy not to install patches, and since that has a negative effect on the rest of the Internet, it'd be irresponsible.
*** now talking on #hypothetical-warez-channel - Topic: Get your XP isos here!
<w4r3z-k1dd1e> don't install yesterday's critical update whatever you do, I got burned by it
<@l33t_d00d> how's that?
<w4r3z-k1dd1e> it deleted my OS!
<w4r3z-k1dd1e> had to reinstall it
<@l33t_d00d> lol, didn't you know?
<@l33t_d00d> some of the patches do stuff like that
<@l33t_d00d> safest way is to skip them all
<w4r3z-k1dd1e> doesn't that make your pc not secure?
<@l33t_d00d> heh, whatever
<@l33t_d00d> that's what *they* tell you
<w4r3z-k1dd1e> ah, k
*** l33t_d00d has changed topic to "Remember kids, patches are for the weak"
Is that really what you want the warez kiddies to be thinking, and if so, would your answer change when the next Code Red/Nimda/Slammer/Sasser/... turns up?
In Deus Ex, particularly if you're running out of health, incapacitating guards with the riot prod and pepper spray is often a very effective method of getting past (a couple of hits from the riot prod will knock most things unconcious, or if you just need to get past, one prod hit or catching them with pepper spray will stun them for long enough). The tranquilizer darts fired from the mini-crossbow are also very useful, or there's always the tear-gas grenades. With EMP grenades too, you don't even need to destroy the guard robots.
Indeed, one of the NPCs will criticise you and give you less equipment if you kill too many opponents in the first level (this was spoiled slightly by counting knockouts as kills, thus penalizing everyone except stealth players; the first patch changed this so non-lethal weapons were advantageous too, I think).
It's a pity you can't just knock out guard dogs when they try to eat you, but the developers didn't bother giving them separate "unconscious" and "dead" states, so you have to kill those if you can't avoid them (if you use a "non-lethal" weapon, they die anyway).
I nearly agree:
:-)
- a decent editor to supplement whatever's in the base system these days (used to be joe, but if I reinstalled now I'd pick Vim)
- sudo
- aptitude (dselect is useful, but unpleasant)
- less, if it's not already there
- ssh (both client and server)
- links and/or lynx for documentation
- CVS (I keep meaning to switch my config-file repositories to something better though - I've tried Subversion and GNU Arch, and I intend to try out Darcs next)
- Mutt
and if this machine needs a GUI:
- XFree86
- xterm (specifically uxterm, my systems all run in UTF-8)
- KDE 3, or XFCE 4 if KDE's too big to be usable, or (Open|Flux|Black)box if I'm getting desparate
- Firefox
(OK, so my list is 10 +/- 2 entries
I know what you mean. Idea: assuming you don't have 10GB of *new* music and a lot of that is older CDs, Ebay them or something - obtain a legal copy cheaply, and don't give the cartel your money.
You could also download some music legally, of course, or buy from non-cartel musicians (I have CDs by two of the artists I linked, and a third one on order).
A while ago I saw something about a Deus Ex film being under consideration - if done right, that could work.
Unfortunately it would probably get overly Hollywoodized (good guys made 100% good, bad guys made 100% evil, JC Denton made into a generic big-gun-wielding action hero) but if instead it was scripted/filmed in a way that reflected the style of the game, it'd be impressive.
(Of course, the other issue is that its anti-authoritarian plot and characters' musings on the nature of freedom might not be too popular politically at the moment...)
it's pretty tough to take linear algebra notes on a computer
d }x = 1
;-)
Tried LaTeX? If you mean the same thing by "linear algebra" as I do (vectors, vector spaces, linear maps, matrices) it shouldn't be too bad if your course is mostly abstract - abstract algebra with just Latin+Greek letters, superscripts and subscripts is trivial to do in LaTeX, although writing out vectors and matrices explicitly *is* a bit of a hassle. A vector:
\[
v = \left(\begin{array}{c}
1 \\
2 \\
3
\end{array}\right)
\]
A matrix:
\[
M = \left(\begin{array}{ccc}
\sin\theta & \cos\theta & 0 \\
-\cos\theta & \sin\theta & 0 \\
0 & 0 & 1
\end{array}\right)
\]
Or if applied maths is more your thing, an integral:
\[
\int_{-\infty}^{\infty} \frac{1}{\sqrt{2\pi\sigma^2}} \mathrm{e}^{-\frac{(x-\mu)^2}{2\sigma^2}}\mathrm{
\]
(mathematical correctness not guaranteed, it's the wrong time of night to be trying to remember the Normal distribution, but that might be right if I'm lucky
I've (just about) successfully taken Representation Theory notes (involving integrals and inner products) on a laptop, in lectures given by a fast lecturer (there wasn't much time to waste when handwriting, so I'm surprised I managed to keep up at all). I gave up after 3 lectures or so, since writing LaTeX took up valuable brain power which I needed to use on understanding the lectures (also, diagrams required switching to my usual pen+paper and scanning it later).
Make your mod and all necessary files available in a safe place on the internet that isn't file planet.
I'd love to (and I agree Vileplanet is undesirable), but paying excessive bandwidth costs for people to download doesn't appeal to me (and yes, I do pay for my web hosting, but that's orders of magnitude cheaper).
I used to do Unreal Tournament mods: nothing big like ChaosUT, just some fixes and addons related to custom model support. I was also the official source for a couple of people's custom models, since I was the last to work on them and I had a more stable web server than the authors did. The two models and my Advanced Model Support mod were about 1MB each, the Bonus Pack 4 fix was only a couple of hundred K.
Despite their smallish size, I had to move all the large downloads onto Vileplanet, because they were sucking too much bandwidth; at the point my then-web host started limiting and charging for bandwidth I'd been using a consistent 2.5GB/month for several months, and I hadn't even made a release recently or anything; that was just normal traffic. At my then-current host's prices, bandwidth charges would have been something like 10 times the cost of my basic hosting account, and I'm sure I'd have spiked over that level considerably when I made a new release.
I hope you can see that if you're someone like ChaosUT, hosting mods that are, say, 30 megabytes, getting as many downloads as I did (a couple of thousand a month), while hosting your own downloads, would be rather painful, and actually being popular (I was, after all, in a pretty niche part of the mod scene) would make your bandwidth prices go up even further.
It's interesting to read the EULAs for games with a thriving mod scene, particularly those where the developers actively encourage mods. In many cases, popular mods are technically in breach of the EULA; I assume the problem is that, say, Epic or Valve encourage modding, but Infogrames/Atari or Sierra are responsible for setting the EULA.
(Of course, many modders, particularly those who do maps, models etc. alone rather than being part of a major mod/TC team, don't understand copyright and are happy to release mods without actually giving anyone permission to distribute the mod...)
I don't even think IPSec allows for you to communicate with machines on the same LAN on the same Subnet.
d d 0.0.0.0/0 192.168.0.2 any -P in ipsec esp/tunnel/192.168.0.1-192.168.0.2/require;
I don't know about the Windows implementation, but KAME (the *BSD IPSec stack, also used in Mac OS X, Linux 2.6 and Debian's patched Linux 2.4) looks as though it will do that fine.
Set up a policy for all traffic from anywhere to your Windows box, and vice versa, to have mandatory encryption in tunnel mode.
You will then need to to set up more specific policies for UDP port 500 (isakmp), and for protocols 50 (esp) and 51 (ah), to avoid trying to apply IPSec to them, since they're what IPSec itself uses (if you don't de-restrict these, you have a chicken and egg problem). You may also want to allow non-IPSec'ed DNS, or ssh, or whatever
Totally untested configuration (you may need to reverse the order of the lines):
#!/usr/bin/setkey -f
# This config is for the restricted box
# On the gateway, exchange the "in" and "out" keywords
flush;
spdflush;
# IPSec gateway is 192.168.0.1
# Restricted box is 192.168.0.2
# ISAKMP over UDP
spdadd 192.168.0.1[500] 192.168.0.2[500] udp -P in none;
spdadd 192.168.0.2[500] 192.168.0.1[500] udp -P out none;
# Encrypted IPSec data
spdadd 192.168.0.1 192.168.0.2 esp -P in none;
spdadd 192.168.0.2 192.168.0.1 esp -P out none;
# "Signed" IPSec data
spdadd 192.168.0.1 192.168.0.2 ah -P in none;
spdadd 192.168.0.2 192.168.0.1 ah -P out none;
# Everything else
spdadd 192.168.0.2 0.0.0.0/0 any -P out ipsec esp/tunnel/192.168.0.2-192.168.0.1/require;
spda
Viruses that claim to be from your email provider are even sillier in countries whose registrars don't sell second-level domains (like the UK and Australia - there are a few second-level domains like co.uk, org.uk, net.uk, com.au etc., and you can buy third level domains like pseudorandom.co.uk).
Because my email is at pseudorandom.co.uk, I get occasional mails from "Co.Uk administrator " telling me my account is going to be suspended. Hmm, why don't I believe that? It's like getting an email from postmaster@com...
OK, how about a different perspective:
In a typical GUI environment, there tends to be something of a duality. Either you muddle through with whatever features a programmer has deigned to give you, or you make the effort to fire up your heavyweight Real Programming Environment (Visual Something-or-other, Delphi, the Java IDE du jour, Emacs...) and construct a proper program; there's a split between being an ordinary user, and being a programmer.
I agree that there's not a great difference between writing a snippet of shell script and writing a snippet of Perl or Python or something (or VBscript or Javascript via WSH if you're in Windows), and Windows with WSH scripting has the potential to be as expressive as Unix with shell scripting; however, you do have to make the conscious decision "this would be easier as a program, I will write a program now."
Shell script in a CLI does have the advantage that you're already in the shell, and you're already used to the shell from using it interactively, so there isn't the conceptual leap; in interactive use you're an ordinary-user-but-acting-a-bit-like-a-programmer, and the switch to programmer-but-also-an-ordinary-user mode isn't as large. (Sorry, that's a bit cumbersome, hopefully you can see what I mean though...)
You can switch from single commands to pretty complicated write-only one-liners within a few lines, without having to think "at this point it would be easier to stop using the usual interface and write a program instead". When you eventually do decide a program is required (which is at a higher level of complexity than it would be in a GUI) you can just cut and paste your previous one-liner, indent it nicely and you have the beginnings of a proper program already.
The other advantage of the CLI in Unix specifically is that a *lot* of stuff is scriptable, since it's designed around command line arguments and pipes. If you want to, say, make thumbnails of images, in Unix you already have ImageMagick at your disposal and ready to be scripted, but in other environments you have to hope you can find a "remote-controllable" app. Mac OS (by which I mean the proprietary GUI bit, not the Darwin Unix bit) actually has this advantage to some extent *without* being a CLI, thanks to many apps accepting "remote control" via Applescript; however, it still has a conceptual leap between "click on things" and "write an Applescript" (if you can record keyboard/mouse "macros", I suppose that's an intermediate step - sorry, I haven't played with Mac OS much).
So... it's the same interface as my mobile phone (a Nokia bought about 18 months ago, but older Nokias had basically the same interface for a few years), but applied to a multimedia asset player? I fail to see how that's an innovation worth protecting.
I would have thought the record companies' worry should be the other way round. If the line drawn by the law between legal format-shifting/fair use and illegal duplication is fairly close to people's view of the line between ethical and unethical, I would have thought people would tend to obey the law, but if format-shifting seems perfectly reasonable but is in fact illegal, then customers who copy their CDs to other formats are breaking the law anyway, so might well reason that they have nothing to lose by copying their friends' CDs too ("might as well be hung for a sheep as for a lamb").
That's a really bad solution for dialup users, who rely on their cache being large enough to hold a decent volume of common images (for instance, to make Slashdot non-painful on dialup, you want the top-left logo and the comment icons to be in cache).
Non-rant:
I seem to remember that the first UT aimbot was "Funbot", written by Darkbyte[S&D] after discussing the possibility of UT cheats with players who insisted UT was cheat-proof. It was a proof of concept, but was released after he foolishly sent a copy to someone who claimed to be an anti-cheat developer. Was that you?
(UT players might also know Darkbyte as the co-author, with Dr.Sin, of Client-Side Hack Protection.)
I haven't used Funbot, but I was sent a copy by someone who'd collaborated with the CSHP developers, when he was checking over one of my mods for possible attacks. The decompiled source is surprisingly simple, just a hook into a loophole in the UT "sandbox", and a bit of vector maths to do the aiming.
I'm surprised nobody ever made a native-code UT aimbot: as far as I know, the aimbot/CSHP arms race was entirely in Unrealscript, although admittedly I never dug into the cheat world particularly far.
OK, the obligatory rant:
Cheating in online games sucks the fun out of the game. You're amusing yourself and spoiling the game for everyone else on the server; I hope it's obvious how childishly destructive that is.
I use a modified version of mserv, a weighted-random online jukebox, with several separate userIDs; when away from the computer with my music on it, I use a MP3 CD player with some loosely themed CD-RWs.
According mserv's top 20 lists (of songs most likely to be played), it looks like I've biased the "personalities" as follows:
work: top-40-ish rock, trip-hop
games: metal, punk, heavier end of rock, techno
code (concentrated coding/deep hack mode): mostly 90s trance, some (particularly electronic) rock
smcv (general all-purpose setup): any of the above
I see absolutely no reason for people to be listening to music while in any sort of educational institution.
:-), in which case the headphones not only give me music to listen to, but act as a mild "do not disturb" sign (I suppose I mean "don't talk to me just for the sake of it, but if you want to ask me something specific, go ahead"). So, yes, I'm being slightly antisocial, but the alternative is hiding in the library or my room and being completely antisocial.
I'm a maths student; like many students here, I do a lot of work on problem sheets (i.e. homework) between or after lectures, in the university maths department canteen/common room. I find wearing earphones and listening to music helps me to concentrate on what I'm doing by masking out random noises, the conversation on the next table, or whatever; also, even if I'm not making progress, I find music makes it less frustrating/boring to work on one question for a long time.
If I'm with friends I usually leave the headphones off, unless I want to concentrate (e.g. imminent deadline
Yes, it's important to have enough money. That's not the same as saying worth is measured in money. Is it worthwhile earning a huge salary, if you're too busy to spend it and too stressed to enjoy it?
However, I start to wonder where your benefit is. You are - out of principle - not making any money out of this, because it is open-source and you and your buddies insist that it must be absolutely free. So you are putting all of that time and energy into this project for what? Fame? To found a career? Come on.
For fun? For interest? To prove to yourself that you can? For the feeling that you've contributed something useful to the world?
Now if most linux distros agree to stick to a common Filesystem Hierarchy system (http://www.pathname.com/fhs/) then you can use all sorts of packages together.
I could use Fedora packages, apt packages, debian packages, gentoo build scripts and all sorts of stuff and pluss get support for closed sourced software easily in any distro of my choosing.
It's not that simple. Shared library versioning, in particular, is a pain.
Debian (my distro of choice) devotes a lot of maintainer time to managing tricky transitions between library versions - if an app directly or indirectly links to incompatible libraries (perhaps you're using Mozilla linked against one major version of libpng, and also linked against Gtk which is linked against a different major version of libpng), it will break.
Debian's complex dependency system copes, but only because Debian packages are consistent and more-or-less come from a single source. If Red Hat's libfoo package isn't 100% compatible with Debian's libfoo package, mixing Red Hat and Debian packages which depend on libfoo will probably result in subtle breakage.
Incidentally, Debian does have numbered releases, every time a new stable release is made ("woody" is Debian 3.0); it also has release revisions (woody is currently at version 3.0r2, which a Windows admin might compare to "Debian 3.0 service pack 2"). It's just testing and unstable that have more-than-daily updates.
Softlink (aka symlink): Has it's own file, but any access gets routed to the real file. This is what most links I deal with are. If you delete the real file, the link breaks, but stays there.
Hardlink: Doesn't have own file. Behaves exactly like the original in all ways, but has a different name. If you delete a hardlink or the original file, nothing happens until *all* instances are gone. Hardlinks must be on the same partition.
FYI, as far as I know, symlinks are just files containing a relative filename, with a special flag in the filesystem (at the same low level as "this is a directory") which marks it as a symlink rather than an ordinary file. The size reported for symlinks under Linux/ext2 is certainly the number of bytes in the filename it points to.
A shortcut is basically a file containing a filename (and other details like the icon), but without the same low-level support in the kernel (open()ing a shortcut gives you its data, not that of the referenced file).
A Mac OS alias (what you get if you run "ln -s" on a HFS+ partition) is what a Windows shortcut should have been (it has more data than just the filename, like the Windows shortcut, but is transparent to apps using the Unix API, like a Unix symlink).
The version I've always heard is that hundreds of years ago the only way to be educated was a private tutor. When they were introduced, "public schools" (schools where pupils' parents pay fees) were called that to differentiate between private tuition and a public school.
;-)
The terminology is a bit unfortunate, now that private tuition doesn't happen and state schools are more public than "public schools", but that's how the English language works
Schools entirely paid for by taxes are "state schools" (as in "separation of Church and State", not as in "Washington state").
Indeed. I've heard it said that stable is for people who need stability, unstable is for people who can cope with instability, and testing is for people who're not paying attention.