While Windows 95 may have reduced the GDI resources problem of Windows 3.x,
The fix was minimalistic - especially since that it was kept in with no changes for at least 5 years (to Windows ME).
At the time, it seemed to be an improvement as it looked like you'd never hit the limit. Now, try reading Slashdot when you suddenly get moderator priviliges - with most browsers (including IE and Firefox), you'll hit the limit and lose track of whatever you were trying to read.
And that's why I switched to XP - 16-bit limitations are never helpful to developers or power users.
WinXP still has GDI limits - thankfully, problems that arise are mostly contained to one application.
Knowing that, does anyone remember their IQ test(s)?
Not really - I was 6 when I took it. After that, they administer some other form of tests (e.g. academic capability tests at Grade 3, 6, 9, etc.)
However, I was told that I got one specific question of note wrong: Which smelled worse: 6-month old fish (in fridge), Spagetti Sauce, and two other things. I circled the second answer - which obviously caused me to be marked down on English ability when it was really testing personal taste. (At the time, I didn't like Spagetti.)
I was also told that I got a 300-IQ in the spatial relationships subset. If IQ tests are prone to spikes like that, then perhaps it's not a good test.
It's not as bad as ambigious questions. For example, "1, 2, 4... what comes next?". The "correct" answer is 8, but using a mathematical formula to determine the polynominal for a sequence brings up 7. Unless the answers are guarenteed to have one correct answer (and can be proven as such), all these questions do is consider how similar you think to the test writer(s).
Microsoft was still using FAT16 at this time, which meant that the cluster size was 32kB if you had a "large" disk (600 MB) and nearly half your disk space was wasted if you had lots of small files. It wasn't until Windows 98 that Microsoft advertised the revolutionary new filesystem (FAT32) that would reclaim that space for you.
Actually, that feature appeared in Windows 95B.
Of course, it's generally better to split your HD into seperate partitions when cluster size was increasing. While you could use basic utilities when you want to reinstall, there are specialized programs that could non-destructivly repartition your hard drive to reclaim the wasted space. (For example, Partition Magic.)
It was also being recommended by professionals - if your operating system was becoming overloaded with DLL Hell, you can wipe c:\ and reinstall Windows. (If you have a re-imager, then that might be a problem unless it only restrictes itself to formatting C:\.)
...Windows 95, despite all it's many flaws, was a lot of fun. It was stupidity to use it in a situation requiring stability, but as a gaming platform and all around PC OS, it was great to have at the time.
That's W98/ME.
I've used W95 in a gaming situation and encountered my first problem with what should be a "stable" game - X-Wing Collector's edition. In particular, the video card driver (or some other driver) would not function consistantly, with the application simply closing to the desktop with *NO* error message (just a standard beep). This problem, as you know, is the same class of problem that plagued every other application since the inception of DirectX - Crash To Desktop with no message (not even an "Illegal operation"). As a side note, the game in question isn't well programmed either - it *REQUIRES* a sound card and joystick, even though the original edition did not (and it was designed to be playable without either).
My computer was a simple Pentium 133MHz, with an Avance Logic 2301 video card (even though the boot-up/chipset said 2302). It was powerful enough to play Quake and other leading edge games at the time. While it did have Windows 95, I made a significant attempt to disable it from auto-launching from startup, and make it return to Dos when it was finished (in the same way that Windows 3.11 functioned.) Dos games could easily get Sound and VESA video - Windows, however, was slightly less stable because of a bit more abstraction.
It took until DX5, WHQL drivers and W95B for things to even become close to stable. Even then, I've seen in-game access violations blamed on improper or faulty video card drivers (when it's a fault with the game code trying to access memory it shouldn't) with the same CTD symptom. The only thing that is worse is a "Send to Erik" button for a bug tracking system that doesn't work.
75% of Americans thought that "God helps those who help themselves." was a teaching from the bible - look as hard as you like, it isn't there
Of course, one of the Bible stories says that if you don't at least put your money in the bank, you'll be punished. Nowadays, you'll lose it all to "service charges" if you do so.
Even so, it is indirectly mentioned in the bible - Jesus was dared by the devil to jump from the roof of the Jerusulum temple. It indicates that one should not foolishly place their faith in God helping them whenever.
of course - these seem to be scaled down versions of Computer Stupidities. One of the entries is accidently dropping a camera - not very spectacular.
It took him a week to notice someone had "punched" the lcd on his laptop?
Actually, the nephew confessed a week later. If you punch it lightly enough, you could potentially damage internal components (by vibrations) without damaging the LCD itself.
This is hard to belive, but does seem plausable enough to punish the nephew. The actual problem could be different (such as random HD data corruption - which I know exists in rare conditions.)
Windows loyalists, did you ever get Outlook Express to understand the concept of multiple email accounts and different stationary/spam filters for each?
No. I used the Mozilla mail reader instead of Outlook.
Figure out how to stop Windows from dumping icons all over the screen?
Do you mean by the icons created by programs duuring install? That's a problem with developers being too stupid to write a proper install script. This can happen in any operating system, even in a Mac or under Linux.
Besides, I store most of the stuff on the desktop anyway.
Get rid of the MS Network and Network Neighborhood because you never use it?
Sure, if you want to disable ability to connect to the Internet. You can remove the Icon in Windows XP with no problem, as it has the ability to hide these critical icons.
Get Windows to recognise your new graphics card without a struggle?
Yes - it uses a default VESA interface since Windows XP - it also has some stock Nvidia drivers, and the WHQL drivers work with minimal problem.
Stop Real Player from firing up at startup and immediately demanding your whole machine's resources?
Yes - by not installing that POS in the first place.
Completely remove software you uninstall, without having to go in and manually delete folders?
Yes - by using third-party uninstall scripts such as CleanSweep 95 (now no longer available.)
Confess to you where it hides ALL the cookies and let you clean them ALL out?
Under IE: Tools -> Internet Options -> Delete Cookies. Under Mozilla: A similar procedure.
Not show the taskbar?
Why on *EARTH* would you want to do that? It's essential if you want to do any serious multitasking. It's so critical that I have to have it twice as large as it would normally be.
But if you want, you can either kill explorer.exe from the task manager (although you need to find a way to make sure it stays down - WinXP auto-reloads it), change the default Window shell to something other than explorer.exe, or do some other interesting trick.
If you don't like the taskbar taking up 20 or so pixels, then resize it so that it takes up a minimal amount of screen real-estate. Alternativly, just enable the auto-hide option available in the start menu properties.
g. you must hit delete, and be told that removing this program may affect your system state.
Actually point g. is interesting. Windows will warn you, GNU/Linux will not warn you, yet none of them does the obvious - start a deinstallation program.
Starting a deinstallation program should only be necessairy if the program needs to store information outside of its main program directory (such as additional information in a search path), and the directory used to store user-information (e.g. a sub-directory just off of the user's home directory.)
Most applications are unintrusive and will have no problem being forcefully removed. The only reason a problem arises is if the installation script refuses to install because the program is already installed (as detected by a sentry written somewhere on the disk.)
The exception is with system-level applications (such as virus scanners, which aren't an issue), those that make arbitrary changes (e.g. Install StarForce drivers, which have no independant installation script), and most forms of spyware. Usually, these are the exception rather than the rule.
Either you are trolling, or you have tried soemthing like Red Hat 6 and nothing since. Every single problem and "hurdle" you come up with are things that have been dealt dealt with long ago.
That is true for the literal cases provided, but is there a music player that can understand all the different types of music that Winamp does? As soon as you look deeper, you'll understand better.
In particular, I have a bunch of music files - some in WMA (unencrytped, hopefully), some in the format WinAmp uses for ripping output (probably M4A), some MIDI files, some Impulse Tracker files, and a variety of other formats that I didn't know existed until now. Any player worth it's salt can handle MP3 or OGG - the real issue is with handling all the different music files.
This is not counting the fact that the last version of mPlayer that I used rendered the console that was used execute "startx" unusable: some library keeps writing stuff to standard whatever. Obviously, a misconfigured default install for that application.
it's easy to connect to a Linux remotely while not the case with Windows.
The steps for enabling Remote desktop are listed here. While you won't be able to connect when another user is on the machine, there is at least one third-party patch that allows otherwise. You still need to grab a copy of the client if you do not have it already.
After following these steps, it's no more difficult than getting regular Windows networking to work. (Which it doesn't - WinME doesn't fully understand WinXP's Network security system, by asking for a password for $IPC without giving the option to try a different username. This, however, is another story.)
and put the crappy OEM reinstall CD in
Never do that. Those OEM CDs are known to wipe the entire system to factory configuration.
As one alternative, find a friend who's willing to copy the Windows XP CD (without the CD-key) - this saves a lot more trouble and time, as well as gives the ability to reinstall without wiping.
If you consider this a breach of the license (Microsoft does, but basic morality and/or default copyright law may be different), you can also simply compress the initial installation to a CD immediatly after acquiring the system - either using burning software or a backup program. (Latter is preferred with this route.) In this case, it's your choice whether to wax the operating system on a reinstallation.
I can probably do whatever I want to do, it's just that some things take time.
That's the problem with Linux - figuring out how to do stuff generally takes time.
While the GUI section has been gradually improving, there's still some major documentation issues (e.g. documentation for some packages fail to describe what should be essential information.) There's also the case of certain "core" applications being incapable of interpreting non-native CR-LFs, which has been known to cause compilation to fail in some applications if they were initially pulled from CVS through WinCVS.
While I do know where to find the documentation (man pages, apropos, the info command, and the HOWTOs), this is barely cutting it as these are four independant help sections that I have to check for general information about the system. Microsoft has the advantage here, as there is only one place you need to check to learn operating system basics (or API, etc.) and can maintain it because they have a dedicated team focused on writing documentation.
If it weren't for that, I would have switched completely over (aside from platform specific things, such as becoming a Minesweeper Consultant and Solitaire Expert). There's more than enough coders going around - what's really needed is writers that get documentation out to minimize time wasted figuring out stuff by glaring at potentially messy source code.
If you take a look at some of the things being done with Quake 1, you will find that they are quite impressive - particularly when you bare in mind that they are basing it on 10 year old software.
When I was working with Quake-C, I was bumping into limitatons. There were a limit of 600 entities on a map, which made it very dificult to put a waypoint system in some of the larger maps.
There was also a limit on the number of instructions allowed to be executed before Quake kills the game ("Runaway loop error"), which was low enough to kill what would normally be considered acceptable on a now high end system. Both of these limits get tripped when you try to stress the game (e.g. porting an AI player over to the larger Team Fortress game.)
Source ports fixed these problems, but not soon enough to do any good.
The precedent has already been set countless times over the last 20 years: Computers are like physical property - if you enter without permission you are breaking and entering no matter if the key is in the door, under the carpet or dangling from a piece of string.
Let me complete your analogy: I rent at an apartment room. Instead of using the officially sanctioned apartment key, I choose to enter by picking the lock with a BIC pen (or climb through the window, etc.) While you may be arrested for break and enter in this case, you will never be charged as you are authorised to be in the room. (You could be liable for another offence - but break and enter is not one of them.)
As for the laptops, what exactly was hacked, and why should it be considered unauthorised? As far as I'm hearing, the greatest known penetration is the local account, where the password was printed on the back of the iPaq. Considering this to be hacking is pointless, as the laptop can be opened (forcefully if necessairy), hard drive extracted and modified, and returned with modified data. Even when tha HD is modified, it's still no permanent damage since it gots reimaged anyway.
Besides, after I modify the laptop, I would probably claim it to be lost. It's criminal theft/fraud, but is still not hacking.
The 13 students charged violated that policy, said Kutztown Police Chief Theodore Cole, insisting the school district had exhausted all options short of expulsion before seeking the charges. Cole said, however, that there is no evidence the students attacked or disabled the school's computer network, altered grades or did anything else that could be deemed malicious.
Either this means there at at least one genius prodidy in the group of 13, or there is no hacking. The former is unlikely, as that prodigy would have zero use for a school. The latter is much more believable at this time. However, until there is credible evidence (or at least circumstantial) saying that there the network is hacked, the school doesn't have any legitimate case.
"What benefit would we get from NOT GPLing this code".
That's basically a limited analysis. What you forget is that any piece of code can immediatly improved on and suddenly generate a small profit stream. As soon as it is released, the cahnges are easily made withoit help from the company, this sealing off that source of potential revenue.
While there are ways around this problem, they generally require planning beforehand (e.g. source release scheduling).
Besides - a lot of people haven't read the GPL fully enougd te understand it's ramifications. There's bound to be at least one newbie lawyer or administrator that thinks nothing can't be dane against code obfuscators.
But truthfully, the competitor will instead say "I have no need for that/already got one."
Do you know that for sure?
R&D costs cannot be recouped as easily. Any competitor that needs to cut back will explore existing software before attempting to write their own.
As a variation on a theme: "Give a man a fish, you feed him for a day; teach a man to fish, you created competition."
2) Trade secret. (Coke formula for example.) Companies like to keep secrets cause that way they look smarter than I are. Again, bullcrap. Pepsi has NO interest in stealing Coca Col'as formula.
Of course Pepsi isn't interested in Cake's formula. However, this does not mention other companies that would just love to get a boost in sales with something like counterfeit Coka Cola.
There's plenty of arguments for GPLling code - the trick is finding some that work best for the business. Anything else is BuckShot reasoning.
IMO, it is vital to make homework not feel like homework in order to get children interested in their schooling again and combat their growing apathy.
No, it's necessairy to make homework be appropriate to the level of development, while preventing it from being a tedious, menial exercise.
For me, math homework failed both requirements. The teachers usually assigned easy problems that you had to do each night - this took a lot of paper and time, and ultimatly gave no benefit for those who already learned how to do arithmetic.
Similar problems are occurring with teaching history - while the topic is important, the focus of the education system attempts to pound dates info the minds of students, while completely ignoring the more appropriate or serious aspects (e.g. The date of DDay might not be ultra-important, but the story about German Generals in-fighting over resources is more critical to the purpose of teaching history in the first place.)
How many times have you missed the doorbell, telephone or significant other trying to get ahold of you in the middle of a Doom/Marathon/Unreal fragfest?
Actually, I missed the doorbell/phone because it is hard to hear at the distance involved. Even when I'm not playing games, I can only hear a faint noise that usually gets dismissed as a standard background distraction.
Even when I do hear the phone, I have to get up from my computer (which isn't a trivial action in some cases) and run to the phone or door. If I do get the phone, it's usually just before the answering machine kicks in.
Naturally, things that are barely audible can't be noticed when there is a much stronger or important stimulus occurring. If some information canèt be processed, you can be sure that faint details are going to be lost. The same occurrs with memory - if you do a meanial task, you can be sure it won't be remembered when there is something more interesting.
Get an old PC(C64, Apple 2xx, IBM PCjr, etc) with basic built in,
No - there's almost no way to store what you want to work with. While it is true that you could use Floppy disks, don't count on being able to find or keep them for a long time. In addition, most old PCs have minimal similarities to modern operating systems.
Get a copy of QBasic, QuickBasic, GW-Basic, or FreeBasic. While still marginally related to current computers, it's much better than just using a random old PC. As a last resort, you can use VB - but it is not recommended as it places a distance between coding and the user.
And that not all - as the child grows older, there on how could other applications could possibly do things quickly. At that time, grab a C/C++ compiler (such as MinGW or Cygwin at the least), a Java compiler, or a C# compiler. Give yourself a crash course in that language, and show that the new system does things much more effeciently overall. Make sure you collect offline references as well - either a programming book (excluding "* for dummies" and "Learn * in * [units of time]"), an API reference or list of valid functions, and/or a tutorial on the programming language.
Remember - a programming language is useless unless your offspring has easy access to it. The BASIC interpreters in the old computers accomplished this with minimal problem. However, as older computers relied on floppies, the more advanced Assemblers and Compilers tended to be *much* harder to use (and therefore fail to attract young learners.)
Tell them that any preferential treatment of a particular class or category of users (only users who own a copy of IE) is considered discrimination.
That's just like saying that most private companies that release a Macintosh software product (e.g. MacFixer, a computer diagnostic program for the Macintosh) would be discriminating against other users that use things like an Amiga. It's just as absurd.
The trick is to minimize browser-specific development by working for the standards first. When problems arise with IE, then you can do changes from the base package to get stuff working.
BTW, you should say IANAL in the same way that I do.
I'd need a really good argument to not use Java. A decent Java developer is harder to find than a C++ or even a C# guy. Tools like Eclipse are free and there is a lot of free online documentation.
Generally, Java picked up an early rep of being "slow" (which is no longer as true with JIT), and that repelled some people. Also, Java can be slightly harder to learn from C++ as OOR is now considered mandatory in that language.
If you want a reason to use C#, then you might as well take a peek at the language to see how it works. This isn't really a reason for C# - just a reason to understand a programming language.
And it's easier to build a quick GUI in Java than C++.
Not really - it's about the same in either language, provided that you know how to program for the GUI.
The only reason it's "harder" to build a GUI in C++ is because you have to write for some random GUI system that you have to learn about (because the base language didn't include any portable GUI function calls, leaving the work to 3rd-party libraries.)
anti-cheating/hacking program that was born from the cesspit that Counterstrike became IIRC, and is now the industry standard for preventing cheating and hacking in online first person shooters.
It's also reported to be ineffective.
The initial version of PB, which appeared on CS servers, picked up a reputation for kicking players for setting cvars to non-default values, even when they wouldn't be in use normally. (e.g. saying which yaw a third-person camera should point at, even when the cvar for the camera was turned off.)
My last problem with PB - I've seen it attempt to kick me 120 times in a row because of "No packet flow". Not suprising, since I was the listen server. (It updated successfully, but that error message shouldn't have appeared in the first place.) It's also capalbe of launching easily, but requires shutting down the game in order to diable PB.
I guess that's why the other major games (such as Half-Life and UT series) have stuck with their own custom anti-cheat protection. While they aren't necessairly more effective, they both feel more trustworthy because they don't have a rash of false positives (or at least no recent false-positives).
None of their other source releases (every game dating back to Wolfenstein 3D) were tampered with in any way before releasing them
Even if that's true, it doesn't mean that the source code is in an instantly usable condition. Wolf 3D requires some obscure assembler product and would not compile out of the box for most users. (However, it's trivial to replicate a Wolf3D engine - no need to optimize it to Wolf3D's levels, since computers have enough processing power to do things quickly enough.)
Of course, the source did change between the latest version and the source release. Doom, as you mentioned, only compiled under Linux. There was a problem with the demos from Doom 1.9 losing sync in the source code release - this was traced to a change made between 1.666 and 1.9 which was somehow rolled back. I'm not sure if this has been solved, but it would definatly take a while to fix.
This is followed by Quake 1 source release. As you know, the latest Dos version is 1.08 - the source code release was 1.09. (While WinQuake was 1.09, I'm sure there's some minor, trivial changes that were made.) The soruce package also came with QuakeWorld, version 2.30 - later than the final binary release.
All in all, the source code isn't guarenteed to be "true and pure". But that doesn't matter at this stage, since those engines have evolved much since then.
2: That doesn't apply so much to games, the distributed development model is less effective since you need a coherent structure.
In most of the Quake engine-level mods I've seen, a coherent structure is formed within individual dev-teams. These dev-teams then proceeded to combine information and produce a Quake Standards Group to ensure that the Quake enginges remain as fully compatable as possible, by showing which builtin functions should be asigned to a given Quake-C builtin call, and so on.
3: For the amazing stuff people have done with the Quake 2 engine Nexuiz,
Nexuiz is based around Quake 1, Normal version. You can tell by the fact that it uses Quake-C, as well as the lack of predictive netcode (e.g. you have to wait until the server acknowledges weapon change before it is recognized on the HUD.)
This works fine for hardcore gamers who are going to play every ouce of the game anyway, but for casual players it sucks. They bought a racing game so they could drive a Porsche through downtown Tokyo, only to discover they need to drive a Tercel for 49 races in order to unlock the Porsche, and that they have to make it around Denver in 14 seconds before they're allowed to race in Tokyo.
And that's why there should always be a way to forcefully unlock unlockables. I generally tire of having to do a procedure over and over again just to regain full access to a game - especially when I'm forced to reinstall.
Most decent games do this, even if it means hunting through config files.
Commandos would be a perfect example of unlockables gone wrong - to unlock the final mission, you had to do everything quickly and with minimal injuries. The only problem is that the game now runs too quickly on modern computers - and there is no way of telling if you are exceeding the time limit until after the mission.
Episode 3, Mission 6 of Doom had a secret exit that required you to use self-damaging rocket propulsion to get to.
The rocket jump is unnecessairy - you can simply use the switch through the wall.
Of course, Doom was one of the first to intorduce straferunning to a large scale - the bane of map designers as it made shortcuts across what should be considered impossible routes. It was so popular that the developers of Rainbow Six included it in their first of the series.
Warcraft3, everything costs 1 gold for the computer.
No, it only costs one gold for the faction. And even then, it's only used to help campaign AI player (which is not capable of expanding beyond it's initial borders, among having other limitations.)
If this were true with a standard skirmish AI, then it wouldn't be possible to choke the AI from well-needed resources.
We've only just moved into the 3d era of video games recently, and we're not close to standardizing the 3d game development process, so we're even further away from writing effective AI. Galactic Civilizations, a strategy game, has an acceptable AI (which runs as crippled by default - it's that strong.)
Likewise, Operation Flashpoint has an acceptable unit AI. While the player generally can defeat AI players, this is generally caused by the limited vision of the enemies.
Both of these games have an AI player that is trusteh to be effective enough. The actuol problem with AI development is that some developers insert a lobotomized version rather than doing proper development.
I know this is a bit offtopic, but why is there not as much focus on making a decent AI for things other than games?
The main reasons is because regular AI programming has higher standards - and generally requires a Masters to even consider doing anything serious. Placing a focus on a more advanced AI system world therefore need more Masters, which in turn requires a focus placed on education.
Game AI programming is much easier, since you just have to write an AI that always makes the best actions (e.g. drone toward player, shoot with perfect accuracy, etc.) and cripple it from there. This applies for almost any game, although exceptions do exist.
Basically, game AI programming involves not having AI players do ultimatly stupid moves or actions (e.g.g throw away a pawn for no good reason.)
Slashdot Mirror
The fix was minimalistic - especially since that it was kept in with no changes for at least 5 years (to Windows ME).
At the time, it seemed to be an improvement as it looked like you'd never hit the limit. Now, try reading Slashdot when you suddenly get moderator priviliges - with most browsers (including IE and Firefox), you'll hit the limit and lose track of whatever you were trying to read.
And that's why I switched to XP - 16-bit limitations are never helpful to developers or power users.
WinXP still has GDI limits - thankfully, problems that arise are mostly contained to one application.
Not really - I was 6 when I took it. After that, they administer some other form of tests (e.g. academic capability tests at Grade 3, 6, 9, etc.)
However, I was told that I got one specific question of note wrong: Which smelled worse: 6-month old fish (in fridge), Spagetti Sauce, and two other things. I circled the second answer - which obviously caused me to be marked down on English ability when it was really testing personal taste. (At the time, I didn't like Spagetti.)
I was also told that I got a 300-IQ in the spatial relationships subset. If IQ tests are prone to spikes like that, then perhaps it's not a good test.
It's not as bad as ambigious questions. For example, "1, 2, 4... what comes next?". The "correct" answer is 8, but using a mathematical formula to determine the polynominal for a sequence brings up 7. Unless the answers are guarenteed to have one correct answer (and can be proven as such), all these questions do is consider how similar you think to the test writer(s).
Actually, that feature appeared in Windows 95B.
Of course, it's generally better to split your HD into seperate partitions when cluster size was increasing. While you could use basic utilities when you want to reinstall, there are specialized programs that could non-destructivly repartition your hard drive to reclaim the wasted space. (For example, Partition Magic.)
It was also being recommended by professionals - if your operating system was becoming overloaded with DLL Hell, you can wipe c:\ and reinstall Windows. (If you have a re-imager, then that might be a problem unless it only restrictes itself to formatting C:\.)
That's W98/ME.
I've used W95 in a gaming situation and encountered my first problem with what should be a "stable" game - X-Wing Collector's edition. In particular, the video card driver (or some other driver) would not function consistantly, with the application simply closing to the desktop with *NO* error message (just a standard beep). This problem, as you know, is the same class of problem that plagued every other application since the inception of DirectX - Crash To Desktop with no message (not even an "Illegal operation"). As a side note, the game in question isn't well programmed either - it *REQUIRES* a sound card and joystick, even though the original edition did not (and it was designed to be playable without either).
My computer was a simple Pentium 133MHz, with an Avance Logic 2301 video card (even though the boot-up/chipset said 2302). It was powerful enough to play Quake and other leading edge games at the time. While it did have Windows 95, I made a significant attempt to disable it from auto-launching from startup, and make it return to Dos when it was finished (in the same way that Windows 3.11 functioned.) Dos games could easily get Sound and VESA video - Windows, however, was slightly less stable because of a bit more abstraction.
It took until DX5, WHQL drivers and W95B for things to even become close to stable. Even then, I've seen in-game access violations blamed on improper or faulty video card drivers (when it's a fault with the game code trying to access memory it shouldn't) with the same CTD symptom. The only thing that is worse is a "Send to Erik" button for a bug tracking system that doesn't work.
Of course, one of the Bible stories says that if you don't at least put your money in the bank, you'll be punished. Nowadays, you'll lose it all to "service charges" if you do so.
Even so, it is indirectly mentioned in the bible - Jesus was dared by the devil to jump from the roof of the Jerusulum temple. It indicates that one should not foolishly place their faith in God helping them whenever.
of course - these seem to be scaled down versions of Computer Stupidities. One of the entries is accidently dropping a camera - not very spectacular.
Actually, the nephew confessed a week later. If you punch it lightly enough, you could potentially damage internal components (by vibrations) without damaging the LCD itself.
This is hard to belive, but does seem plausable enough to punish the nephew. The actual problem could be different (such as random HD data corruption - which I know exists in rare conditions.)
Do you mean by the icons created by programs duuring install? That's a problem with developers being too stupid to write a proper install script. This can happen in any operating system, even in a Mac or under Linux.
Besides, I store most of the stuff on the desktop anyway.
Sure, if you want to disable ability to connect to the Internet. You can remove the Icon in Windows XP with no problem, as it has the ability to hide these critical icons.
Yes - it uses a default VESA interface since Windows XP - it also has some stock Nvidia drivers, and the WHQL drivers work with minimal problem.
Yes - by not installing that POS in the first place.
Yes - by using third-party uninstall scripts such as CleanSweep 95 (now no longer available.)
Under IE: Tools -> Internet Options -> Delete Cookies.
Under Mozilla: A similar procedure.
Why on *EARTH* would you want to do that? It's essential if you want to do any serious multitasking. It's so critical that I have to have it twice as large as it would normally be.
But if you want, you can either kill explorer.exe from the task manager (although you need to find a way to make sure it stays down - WinXP auto-reloads it), change the default Window shell to something other than explorer.exe, or do some other interesting trick.
If you don't like the taskbar taking up 20 or so pixels, then resize it so that it takes up a minimal amount of screen real-estate. Alternativly, just enable the auto-hide option available in the start menu properties.
Starting a deinstallation program should only be necessairy if the program needs to store information outside of its main program directory (such as additional information in a search path), and the directory used to store user-information (e.g. a sub-directory just off of the user's home directory.)
Most applications are unintrusive and will have no problem being forcefully removed. The only reason a problem arises is if the installation script refuses to install because the program is already installed (as detected by a sentry written somewhere on the disk.)
The exception is with system-level applications (such as virus scanners, which aren't an issue), those that make arbitrary changes (e.g. Install StarForce drivers, which have no independant installation script), and most forms of spyware. Usually, these are the exception rather than the rule.
That is true for the literal cases provided, but is there a music player that can understand all the different types of music that Winamp does? As soon as you look deeper, you'll understand better.
In particular, I have a bunch of music files - some in WMA (unencrytped, hopefully), some in the format WinAmp uses for ripping output (probably M4A), some MIDI files, some Impulse Tracker files, and a variety of other formats that I didn't know existed until now. Any player worth it's salt can handle MP3 or OGG - the real issue is with handling all the different music files.
This is not counting the fact that the last version of mPlayer that I used rendered the console that was used execute "startx" unusable: some library keeps writing stuff to standard whatever. Obviously, a misconfigured default install for that application.
The steps for enabling Remote desktop are listed here. While you won't be able to connect when another user is on the machine, there is at least one third-party patch that allows otherwise. You still need to grab a copy of the client if you do not have it already.
After following these steps, it's no more difficult than getting regular Windows networking to work. (Which it doesn't - WinME doesn't fully understand WinXP's Network security system, by asking for a password for $IPC without giving the option to try a different username. This, however, is another story.)
Never do that. Those OEM CDs are known to wipe the entire system to factory configuration.
As one alternative, find a friend who's willing to copy the Windows XP CD (without the CD-key) - this saves a lot more trouble and time, as well as gives the ability to reinstall without wiping.
If you consider this a breach of the license (Microsoft does, but basic morality and/or default copyright law may be different), you can also simply compress the initial installation to a CD immediatly after acquiring the system - either using burning software or a backup program. (Latter is preferred with this route.) In this case, it's your choice whether to wax the operating system on a reinstallation.
That's the problem with Linux - figuring out how to do stuff generally takes time.
While the GUI section has been gradually improving, there's still some major documentation issues (e.g. documentation for some packages fail to describe what should be essential information.) There's also the case of certain "core" applications being incapable of interpreting non-native CR-LFs, which has been known to cause compilation to fail in some applications if they were initially pulled from CVS through WinCVS.
While I do know where to find the documentation (man pages, apropos, the info command, and the HOWTOs), this is barely cutting it as these are four independant help sections that I have to check for general information about the system. Microsoft has the advantage here, as there is only one place you need to check to learn operating system basics (or API, etc.) and can maintain it because they have a dedicated team focused on writing documentation.
If it weren't for that, I would have switched completely over (aside from platform specific things, such as becoming a Minesweeper Consultant and Solitaire Expert). There's more than enough coders going around - what's really needed is writers that get documentation out to minimize time wasted figuring out stuff by glaring at potentially messy source code.
Let me complete your analogy: I rent at an apartment room. Instead of using the officially sanctioned apartment key, I choose to enter by picking the lock with a BIC pen (or climb through the window, etc.) While you may be arrested for break and enter in this case, you will never be charged as you are authorised to be in the room. (You could be liable for another offence - but break and enter is not one of them.)
As for the laptops, what exactly was hacked, and why should it be considered unauthorised? As far as I'm hearing, the greatest known penetration is the local account, where the password was printed on the back of the iPaq. Considering this to be hacking is pointless, as the laptop can be opened (forcefully if necessairy), hard drive extracted and modified, and returned with modified data. Even when tha HD is modified, it's still no permanent damage since it gots reimaged anyway.
Besides, after I modify the laptop, I would probably claim it to be lost. It's criminal theft/fraud, but is still not hacking.
As for accessing school networks, it's unlikly: http://apnews.excite.com/article/20050809/D8BSI8I
Either this means there at at least one genius prodidy in the group of 13, or there is no hacking. The former is unlikely, as that prodigy would have zero use for a school. The latter is much more believable at this time. However, until there is credible evidence (or at least circumstantial) saying that there the network is hacked, the school doesn't have any legitimate case.
That's basically a limited analysis. What you forget is that any piece of code can immediatly improved on and suddenly generate a small profit stream. As soon as it is released, the cahnges are easily made withoit help from the company, this sealing off that source of potential revenue.
While there are ways around this problem, they generally require planning beforehand (e.g. source release scheduling).
Besides - a lot of people haven't read the GPL fully enougd te understand it's ramifications. There's bound to be at least one newbie lawyer or administrator that thinks nothing can't be dane against code obfuscators.
Do you know that for sure?
R&D costs cannot be recouped as easily. Any competitor that needs to cut back will explore existing software before attempting to write their own.
As a variation on a theme: "Give a man a fish, you feed him for a day; teach a man to fish, you created competition."
Of course Pepsi isn't interested in Cake's formula. However, this does not mention other companies that would just love to get a boost in sales with something like counterfeit Coka Cola.
There's plenty of arguments for GPLling code - the trick is finding some that work best for the business. Anything else is BuckShot reasoning.
No, it's necessairy to make homework be appropriate to the level of development, while preventing it from being a tedious, menial exercise.
For me, math homework failed both requirements. The teachers usually assigned easy problems that you had to do each night - this took a lot of paper and time, and ultimatly gave no benefit for those who already learned how to do arithmetic.
Similar problems are occurring with teaching history - while the topic is important, the focus of the education system attempts to pound dates info the minds of students, while completely ignoring the more appropriate or serious aspects (e.g. The date of DDay might not be ultra-important, but the story about German Generals in-fighting over resources is more critical to the purpose of teaching history in the first place.)
Actually, I missed the doorbell/phone because it is hard to hear at the distance involved. Even when I'm not playing games, I can only hear a faint noise that usually gets dismissed as a standard background distraction.
Even when I do hear the phone, I have to get up from my computer (which isn't a trivial action in some cases) and run to the phone or door. If I do get the phone, it's usually just before the answering machine kicks in.
Naturally, things that are barely audible can't be noticed when there is a much stronger or important stimulus occurring. If some information canèt be processed, you can be sure that faint details are going to be lost. The same occurrs with memory - if you do a meanial task, you can be sure it won't be remembered when there is something more interesting.
Get a copy of QBasic, QuickBasic, GW-Basic, or FreeBasic. While still marginally related to current computers, it's much better than just using a random old PC. As a last resort, you can use VB - but it is not recommended as it places a distance between coding and the user.
And that not all - as the child grows older, there on how could other applications could possibly do things quickly. At that time, grab a C/C++ compiler (such as MinGW or Cygwin at the least), a Java compiler, or a C# compiler. Give yourself a crash course in that language, and show that the new system does things much more effeciently overall. Make sure you collect offline references as well - either a programming book (excluding "* for dummies" and "Learn * in * [units of time]"), an API reference or list of valid functions, and/or a tutorial on the programming language.
Remember - a programming language is useless unless your offspring has easy access to it. The BASIC interpreters in the old computers accomplished this with minimal problem. However, as older computers relied on floppies, the more advanced Assemblers and Compilers tended to be *much* harder to use (and therefore fail to attract young learners.)
That's just like saying that most private companies that release a Macintosh software product (e.g. MacFixer, a computer diagnostic program for the Macintosh) would be discriminating against other users that use things like an Amiga. It's just as absurd.
The trick is to minimize browser-specific development by working for the standards first. When problems arise with IE, then you can do changes from the base package to get stuff working.
BTW, you should say IANAL in the same way that I do.
Generally, Java picked up an early rep of being "slow" (which is no longer as true with JIT), and that repelled some people. Also, Java can be slightly harder to learn from C++ as OOR is now considered mandatory in that language.
If you want a reason to use C#, then you might as well take a peek at the language to see how it works. This isn't really a reason for C# - just a reason to understand a programming language.
Not really - it's about the same in either language, provided that you know how to program for the GUI.
The only reason it's "harder" to build a GUI in C++ is because you have to write for some random GUI system that you have to learn about (because the base language didn't include any portable GUI function calls, leaving the work to 3rd-party libraries.)
It's also reported to be ineffective.
The initial version of PB, which appeared on CS servers, picked up a reputation for kicking players for setting cvars to non-default values, even when they wouldn't be in use normally. (e.g. saying which yaw a third-person camera should point at, even when the cvar for the camera was turned off.)
My last problem with PB - I've seen it attempt to kick me 120 times in a row because of "No packet flow". Not suprising, since I was the listen server. (It updated successfully, but that error message shouldn't have appeared in the first place.) It's also capalbe of launching easily, but requires shutting down the game in order to diable PB.
I guess that's why the other major games (such as Half-Life and UT series) have stuck with their own custom anti-cheat protection. While they aren't necessairly more effective, they both feel more trustworthy because they don't have a rash of false positives (or at least no recent false-positives).
Even if that's true, it doesn't mean that the source code is in an instantly usable condition. Wolf 3D requires some obscure assembler product and would not compile out of the box for most users. (However, it's trivial to replicate a Wolf3D engine - no need to optimize it to Wolf3D's levels, since computers have enough processing power to do things quickly enough.)
Of course, the source did change between the latest version and the source release. Doom, as you mentioned, only compiled under Linux. There was a problem with the demos from Doom 1.9 losing sync in the source code release - this was traced to a change made between 1.666 and 1.9 which was somehow rolled back. I'm not sure if this has been solved, but it would definatly take a while to fix.
This is followed by Quake 1 source release. As you know, the latest Dos version is 1.08 - the source code release was 1.09. (While WinQuake was 1.09, I'm sure there's some minor, trivial changes that were made.) The soruce package also came with QuakeWorld, version 2.30 - later than the final binary release.
All in all, the source code isn't guarenteed to be "true and pure". But that doesn't matter at this stage, since those engines have evolved much since then.
In most of the Quake engine-level mods I've seen, a coherent structure is formed within individual dev-teams. These dev-teams then proceeded to combine information and produce a Quake Standards Group to ensure that the Quake enginges remain as fully compatable as possible, by showing which builtin functions should be asigned to a given Quake-C builtin call, and so on.
Nexuiz is based around Quake 1, Normal version. You can tell by the fact that it uses Quake-C, as well as the lack of predictive netcode (e.g. you have to wait until the server acknowledges weapon change before it is recognized on the HUD.)
And that's why there should always be a way to forcefully unlock unlockables. I generally tire of having to do a procedure over and over again just to regain full access to a game - especially when I'm forced to reinstall.
Most decent games do this, even if it means hunting through config files.
Commandos would be a perfect example of unlockables gone wrong - to unlock the final mission, you had to do everything quickly and with minimal injuries. The only problem is that the game now runs too quickly on modern computers - and there is no way of telling if you are exceeding the time limit until after the mission.
The rocket jump is unnecessairy - you can simply use the switch through the wall.
Of course, Doom was one of the first to intorduce straferunning to a large scale - the bane of map designers as it made shortcuts across what should be considered impossible routes. It was so popular that the developers of Rainbow Six included it in their first of the series.
No, it only costs one gold for the faction. And even then, it's only used to help campaign AI player (which is not capable of expanding beyond it's initial borders, among having other limitations.)
If this were true with a standard skirmish AI, then it wouldn't be possible to choke the AI from well-needed resources.
We've only just moved into the 3d era of video games recently, and we're not close to standardizing the 3d game development process, so we're even further away from writing effective AI.
Galactic Civilizations, a strategy game, has an acceptable AI (which runs as crippled by default - it's that strong.)
Likewise, Operation Flashpoint has an acceptable unit AI. While the player generally can defeat AI players, this is generally caused by the limited vision of the enemies.
Both of these games have an AI player that is trusteh to be effective enough. The actuol problem with AI development is that some developers insert a lobotomized version rather than doing proper development.
Game AI programming is much easier, since you just have to write an AI that always makes the best actions (e.g. drone toward player, shoot with perfect accuracy, etc.) and cripple it from there. This applies for almost any game, although exceptions do exist.
Basically, game AI programming involves not having AI players do ultimatly stupid moves or actions (e.g.g throw away a pawn for no good reason.)