Wikipedia:Reference desk/Archives/Computing/2007 November 26

= November 26 =

A problem with the Windows Sound Recorder (on Windows Vista)...
I have Windows Vista on my laptop, and I like the new Sound Recorder. But now, there's a problem. It doesn't pick up a sound when it's recording. It used to pick up a sound, but now it doesn't. How can I make the Sound Recorder pick up a sound in recording again? —Preceding unsigned comment added by Sirdrink13309622 (talk • contribs) 04:46, 26 November 2007 (UTC)


 * It's probably to do with the 'mixer' settings on your PC. Go to the Control Panel then Sounds & Audio (or whatever it's called under Vista), click on the Audio tab then click on Volume under Recording Properties.  You should see what looks like a mixing desk with sliders for the various inputs you have, ie master volume, microphone, line in etc.  If you can't see them all, you will need to go to Properties and tick all the boxes, this will show them all then just make sure whatever input you are trying to record from is not set to mute and has the volume slider set accordingly (ie not at the bottom of the slider).


 * If that doesn't work, it might be something to do with the settings in the Sound Recorder itself but I am not familiar with it so cannot comment on that. GaryReggae (talk) 12:55, 26 November 2007 (UTC)

Windows XP: File | Save As
Windows XP: File | Save As

In the dialog box of any program in Windows XP when you select: File | Save As

There is an option in the "View Menu" to select "Thumbnails"

Is there any registry tweak to make "Thumbnails" the default choice?

If so, what is the tweak?

Also, is there any way to change the "default size" and "default location" of the dialog box?

Thanks for any info...

Kd6iex (talk) 07:16, 26 November 2007 (UTC)

programming
name and discribe four programming languages

jack —Preceding unsigned comment added by 212.49.85.162 (talk) 09:35, 26 November 2007 (UTC)


 * Welcome to the Wikipedia Reference Desk. Your question appears to be a homework question. I apologize if this is a misevaluation, but it is our policy here to not do people's homework for them, but to merely aid them in doing it themselves. Letting someone else do your homework does not help you learn how to solve such problems. Please attempt to solve the problem yourself first. If you need help with a specific part of your homework, feel free to tell us where you are stuck and ask for help. If you need help grasping the concept of a problem, by all means let us know. Thank you. See further our articles on Programming Language, List of programming languages, and Comparison of programming languages. antilivedT 09:53, 26 November 2007 (UTC)


 * Four languages:
 * SNOBOL - renowned for its powerful string processing capabilities
 * PL/I - has advanced features including recursion and structured programming
 * Simula - model the world as objects!
 * Befunge - it's Turing-complete with a syntax perfect for those with good visual thinking skills
 * Enjoy! --Sean 14:15, 26 November 2007 (UTC)


 * You are soooo evil...but you forgot the all-important Whitespace language - which is ecologically sound because program listings save on printer ink and allow for easy paper recycling! :-)


 * Dear OP:


 * If you wish to remain sane, please ignore Sean and don't even THINK of looking up Whitespace or Befunge (nor FALSE, nor Brainfsck, nor LOLCODE, nor INTERCAL, nor Malbolge, nor Shakespeare) because your brain will be permenantly warped and you'll probably get a zero on your homework!


 * For a more reasonable set of choices to write about try: BASIC] (A simple teaching language), C++ (A mainstream programming language upon which many others such as JAVA are based - nearly every modern video game is written in C++), Assembly language (what the computer itself speaks) and PHP (a server-side web programming language - the language that Wikipedia is written in).


 * If you really want to find something a bit different, some other significant languages in the history of programming are: Prolog (a really 'different' logic-based language), Fortran (the first true high level language), ALGOL (the precursor of C, C++, Pascal, Java, PHP, and almost all other 'modern' languages - but hardly every used these days), JavaScript (a client-side web programming language), Forth (a stack-based language), PostScript (a programming language that's primarily used for talking to printers(!)), Smalltalk (an early object-oriented language - a great influence on C++ and hence on many other modern languages), LISP (a popular language for AI applications), COBOL (the language of choice for banking and accountancy), Ada (what you get if you let the military arm of your government design your programming language (by committee) and then pass laws to force you to use it!)
 * I'm sure others can provide other significant choices - these happen to be mine. SteveBaker (talk) 15:49, 26 November 2007 (UTC)


 * Oh, I don't know -- the assignment asked for four programming languages, described, and it didn't say they had to be, like, good languages or anything. So using INTERCAL, Brainfuck, Befunge, and LOLCODE (hey! that's a new one on me) could lead to a very interesting answer, and if the professor was so clueless or closed-minded as to grade the OP with a zero for that answer, well, I think the OP needs a new professor. —Steve Summit (talk) 02:38, 27 November 2007 (UTC)


 * Come on - let him impress his teacher with a list like:
 * C++: Why let the programming language keep you from making terrible memory management mistakes?
 * Perl: Others shouldn't be able to read your code when you are done writing it.
 * MUMPS: You shouldn't be able to read your own code either.
 * Lisp: Because all the other languages do it wrong.
 * Only four? I could go on and on and on...  Of course, I'm ignoring the esoteric programming languages. --  k a i n a w &trade; 16:13, 26 November 2007 (UTC)
 * Here, let me correct that for you:
 * C++: Is like an extremely mean attack dog. If it's on your side, it's awesome - but you've got to let it know who's boss or it'll turn on you...and you really wouldn't enjoy that.
 * Perl: The world's first write-only language. The syntax looks like transmission line noise.
 * MUMPS: The only language where the time between writing your code and being unable to read it is actually a negative number. If MUMPS were only used to write unimportant applications, this might be forgivable. Sadly, almost the only users of MUMPS are in the healthcare industry (Eeeeek!).
 * Lisp: To paraphrase the Slashdot meme:
 * Buy shares in a tiny struggling company whose only product is replacement ')' keys for PC keyboards,
 * Encourage everyone to use Lisp.
 * Profit!
 * SteveBaker (talk) 04:27, 27 November 2007 (UTC)
 * SteveBaker (talk) 04:27, 27 November 2007 (UTC)


 * My take:
 * C++: For when you actually need to control what's going on, but don't use unless necessary
 * Perl: Because UNIX fogies spent too much time writing books about it 20 years ago to realize how badly it needs to be annihilated and how useless it is as a good scripting language
 * MUMPS: Meh, doesn't look as bad as Whitespace :)
 * Lisp: Valuable only because it influenced Ruby. Ugly, inefficient. LOGO is kind of nifty though.
 * -- ⁪ffroth 04:12, 28 November 2007 (UTC)
 * You don't think Lisp was valuable?!? Just about every AI program in existance was written in Lisp until relatively recently!  Even today, Lisp is exceedingly cool and hasn't really been superceded.  The only serious problem it has (which I alluded to in my reply) is that its syntax is rather bracket-laden - you can easily come to the end of a chunk of code and find that it ends with 20 or more close-parenthesis characters!  Ruby and LOGO can't come close to Lisp's importance as a language! SteveBaker (talk) 14:25, 28 November 2007 (UTC)
 * So it was historically important; that doesn't mean that the language isn't slow and unweildly for all but a few specialized applications -- ⁪ffroth 03:25, 29 November 2007 (UTC)
 * Your characterisation of Lisp as "slow and inefficient" reveals your lack of insight. There is no reason inherent to Lisp why it should be any slower or less efficient than, say, Python. Such matters are a concern of the implementation. Marnanel (talk) 04:06, 29 November 2007 (UTC)
 * A language can be built for speed independent of the implementation -- ⁪ffroth 01:35, 30 November 2007 (UTC)
 * More to the point - "slow and inefficient" relative to what? Since there are no other languages that do what Lisp does - there is nothing to compare it against.  There are Lisp compilers that'll reduce your code to machine-code - but because Lisp programs can be self-modifying (intentionally so because it's for AI and you'd like for programs to be able to learn and extend themselves) - there is a need to carry the Lisp compiler around inside every executable.  So generally it's better that Lisp is interpreted...but you definitely have a choice. SteveBaker (talk) 21:50, 29 November 2007 (UTC)
 * A better point, but my point is that what it uniquely can do isn't very friendly to modern processor architecture and digital logic, so it's slow. A lot of C statements practically boil down to a few processor instructions, while Lisp isn't nearly as friendly.. that's what I mean by slow -- ⁪ffroth 01:35, 30 November 2007 (UTC)


 * Every time a discussion arises about programming language quirks, I feel obliged to point out the greatest programming language ever concieved by man, Prolog. I had been programming for nearly 15 years in various procedural, fuctional and object-oriented languages, but I have never been as fascinated or had as much fun doing stuff in Prolog. It's something every programmer should treat themselves. 213.112.18.120 (talk) 01:47, 27 November 2007 (UTC)


 * Yeah - I know exactly what you mean! I have a VERY clear memory of (somewhat) learning Prolog.  I'd seen an article (I think it was in Byte magazine) about how you could implement the three laws of robotics in Prolog.  So I bought the book - found a place where I could run programs and I started reading.  I was getting more and more excited about the purity and clarity (and oddness) of the language - right up until the third paragraph of page 65 (Why do I remember that from 20 years ago?  I have no clue.)...that was where they started to talk about 'cut'.  Cut is the rough equivelent of a GOTO statement in Prolog...except that this one is absolutely needed in order to make your programs actually run in less than the life of the universe!  By the bottom of page 67, I gave up on Prolog. &lt;sigh&gt; SteveBaker (talk) 04:27, 27 November 2007 (UTC)

Web --> unix terminal interface
Hi,

I've got a server. I've also got web access at work. I want to be able to open a command shell on my linux server from work, but can't SSH to it. So I'm looking for something in CGI or javascript or similar that can put a terminal in a web page. Any suggestions?

--Psud (talk) 11:50, 26 November 2007 (UTC)


 * Try webmin. Morana (talk) 12:45, 26 November 2007 (UTC)


 * You don't say why you can't use SSH, but using a Java applet-based SSH client might be an option. No installation on the client is necessary, assuming it can run Java applets.  There are several implementations, but I've used Mindterm with success.  --Sean 13:46, 26 November 2007 (UTC)


 * One presumes that our OP's employer has blocked outgoing SSH ports to prevent employees doing exactly this. While you certainly can put together something to get shell access via web pages, it's awfully risky.  You need secure password-locked access and you need to not be sending the password in plaintext.  These are the kinds of concern that SSH is designed to protect against.  I'm not saying it's impossible - just that you have to exercise large amounts of paranoia.  Also, if your employer doesn't want you to have shell access to your server during your time at work, bypassing that is exactly the kind of thing that could get you into a lot of trouble.  If you believe you have a genuine business need to do this then perhaps your IT department could be asked to unlock outgoing SSH access. SteveBaker (talk) 15:04, 26 November 2007 (UTC)

Megaupload safe?
Is the site "Megaupload" safe to use, that is, it won't give you spam or a virus or something if you use it? Thanks in advance. External link: 67.42.180.114 (talk) 13:24, 26 November 2007 (UTC)
 * You can safely upload and download things that others have stored there. I've never had a computer virus myself, so I don't know about any popups or whatever giving you one.  --Sean 14:07, 26 November 2007 (UTC)
 * Thank you. I have Adblock Plus, so I'm safe from popups. 67.42.180.114 (talk) 14:42, 26 November 2007 (UTC)
 * I don't know what kind of things people upload and download there. If it's text files and pretty pictures, fine.  But if it's programs (applications, screensavers, web browser toolbars, etc.), then you're at rather dire risk of all sorts of mayhem if you download, install, and run those programs.  (Remember, installing and running code on your computer is like having sex -- don't do it unless you have good reason to be sure of the health and trustworthiness of your partner.) —Steve Summit (talk) 16:15, 26 November 2007 (UTC)


 * Except sex is a lot more fun, on the whole. --24.147.86.187 (talk) 16:22, 26 November 2007 (UTC)
 * It's just a normal downloading experience, just most likely with larger files. But I agree that the popups may give you problems - not that I've encountered any.  x42bn6 Talk Mess  17:38, 26 November 2007 (UTC)
 * Megaupload and rapidshare— it depends on what files you're downloading, and that depends on who uploaded them. If somebody randomly posted a rapidshare link on 4chan it would be wise to wait for several testimonies if you really want to download the file (unless of course you have Mac OS X). Also, most files on there I believe have been .rar-ed or .zip-ed, so there's no way to tell what's in them aside from the size and title until you have downloaded it. If you upload a file to megaupload and want to download it, or have your friend download it, it's safe. The unsafe comes when you randomly download other people's files. Mac Davis (talk) 04:51, 27 November 2007 (UTC)
 * Steve Jobs isn't going to break you out of prision if you unintentionally downloaded 100MB of child porn -- ⁪ffroth 03:58, 28 November 2007 (UTC)
 * I don't see why it matters much if things have been RARed, or ZIPed unless of course you have a heavily data limited or slow connection. If you're expecting to download images or a movie and find there is "Kournikova SEX VIDEO.mpg.exe" inside the RAR/ZIP/7zip/whatever then delete it. If it has what you expect, then why does it matter it was compressed? There is a slight risk the image or movie may try to take advantage of some security loophole particularly if you use an unupdated Windows system and Windows Media Player but this doesn't change when the file is compressed. The only additional problem is the slight risk of a RAR/ZIP/7zip security loophole. Of course it's more annoying to have wasted your time but anyway... Nil Einne 08:12, 1 December 2007 (UTC)
 * I wouldn't recommend megaupload. I prefer http://mihd.net or http://drop.io F (talk) 23:46, 27 November 2007 (UTC)

NFS CARBON problem
I had just bought this need for speed carbon collector's edition dvd but it does not work,it displays the nfs carbon picture and all of a sudden theres a black screen and then it returns to normal,i tried everything except for Ram the game requires 512 ram but i have 504 mb of ram but in the easy info it gives ok.my system's info are: Operating System       Microsoft Windows XP Professional Service Pack 2 (Build 2600) 504 mb ram System Memory 502.2 Processor Speed 3192.2 DirectX Version (Get Updates)    9.0c Display Device (Get Drivers) Intel(R) 82945G Express Chipset Family Display Memory 128.0 Display Driver 6.14.10.4497 Sound Device  Realtek HD Audio output Sound Driver 5.10.00.5324 —Preceding unsigned comment added by 202.70.64.41 (talk) 17:18, 26 November 2007 (UTC)
 * You don't have a graphics card. There's your problem. - Woo ty   [ Woot? ]  [Spam! Spam! Wonderful spam! ] 17:31, 26 November 2007 (UTC)
 * This is the graphics card (or onboard chip, more likely): "Intel(R) 82945G Express Chipset Family". --Sean 18:01, 26 November 2007 (UTC)
 * I know - the emphasis was on discrete 'graphics card'. Onboard graphics isn't going to cut it. - Woo ty   [ Woot? ]  [Spam! Spam! Wonderful spam! ] 20:15, 26 November 2007 (UTC)
 * Especially with only 8MiB of VRAM. --antilivedT 00:12, 27 November 2007 (UTC)


 * Probably the deeper problem is that the Intel graphics chips use memory out of main CPU RAM for texture maps and such. Since you already don't have enough RAM to run the game - when the Intel chip steals maybe half of it...you DEFINITELY don't have enough.  Time to upgrade! SteveBaker (talk) 04:01, 27 November 2007 (UTC)


 * And I'm pretty sure Carbon requires some sort of hardware pixel and vertex shading, which I'm not even sure the very newest intel onboard has. They claim to have this feature but it is done on the CPU so it is very very slow. If you want to play this (and more modern games (Your PCs manual should tell you this)) I would suggest buying a new graphics card (If you only have PCI then you are out of luck, if you have AGP I suggest an x1600, if you have PCI-Express then I recommend a GeForce 7600 of some sort. With this I would buy some more ram to bring you up to 1gb. I am guessing you have 1 512mb stick in there at the moment and you can get another very cheaply. Use the memory checker at www.crucial.com to see what sort of RAM you have and what you can upgrade to.TheGreatZorko (talk) 10:07, 29 November 2007 (UTC)

Searching google with 51 words,How?

 * I want to search following 51 words in google at one go how do I do it ?Mahitgar (talk) 17:33, 26 November 2007 (UTC)
 * Terms I want to serch at one go at google:विकिपेडीआ OR विकिपेडिआ OR विकिपेडीया OR विकिपेडिया OR विकीपेडीआ OR विकीपेडिआ OR विकीपेडीया OR वीकीपेडिया OR वीकीपेडीआ OR वीकीपेडिआ OR वीकीपेडीया OR वीकीपेडिया OR वीकीपेडिया OR वीकिपेडीआ OR वीकिपेडिआ OR वीकिपेडीया OR वीकिपेडिया OR विकिपीडीआ OR विकिपीडिआ OR विकिपीडीया OR विकिपीडिया OR विकीपीडीआ OR विकीपीडिआ OR विकीपीडीया OR वीकीपीडिया OR वीकीपीडीआ OR वीकीपीडिआ OR वीकीपीडीया OR वीकीपीडिया OR वीकीपीडिया OR वीकिपीडीआ OR वीकिपीडिआ OR वीकिपीडीया OR वीकिपीडिया OR विकिपिडीआ OR विकिपिडिआ OR विकिपिडीया OR विकिपिडिया OR विकीपिडीआ OR विकीपिडिआ OR विकीपिडीया OR वीकीपिडिया OR वीकीपिडीआ OR वीकीपिडिआ OR वीकीपिडीया OR वीकीपिडिया OR वीकीपिडिया OR वीकिपिडीआ OR वीकिपिडिआ OR वीकिपिडीया OR वीकिपिडिया


 * Just type them in the google search box and click the google search button. Google automatically applies an "or" between the words, but pushes pages with the most words to the top of the results list. --  k a i n a w &trade; 17:55, 26 November 2007 (UTC)
 * Are you sure about this? I'm pretty sure they've always been ANDed by default.  --Sean 17:57, 26 November 2007 (UTC)
 * It's worth checking the Google Cheat Sheet for this. OR will work here, although you'll be able to search for part of those words, since Google limits the number of words to something like seven (if I recall correctly).  --  JSBillings  18:05, 26 November 2007 (UTC)
 * Wait, it looks like google has raised that limit to 32, although I can't find any definitive limit on Google's site. -- JSBillings  18:08, 26 November 2007 (UTC)


 * As for the and/or usage in Google. A search will return documents with all words (and) if they are available.  Then, it will return documents with some of the words (and/or mix) if they are available.  Then, it will return documents with at least one of the words (or) if they are available.  That is why I stated that the OR wasn't required since it will be implied if no documents contain all of the words.  I did not know Google has a word limit.  I assumed the character limit of the query string wouldn't be an issue with only 51 words. --  k a i n a w &trade; 18:47, 26 November 2007 (UTC)
 * URLs can only be 256 characters long, so if the words are just 5 letters long then the search query alone will exceed the limit -- ⁪ffroth 01:49, 27 November 2007 (UTC)
 * WHAT?!? Where did you get that idea from?  This Wikipedia URL is 280 characters:  SteveBaker (talk) 03:26, 27 November 2007 (UTC)


 * The shortest URL limitation I've ever seen was in an early version of Lynx at 1024 characters. Currently, IE5 is the only semi-popular web browser that I know of with a low limit (2048 characters if I remember correctly).  All other web browsers have much larger limitations. --  k a i n a w &trade; 13:08, 27 November 2007 (UTC)


 * It does sound kind of short doesn't it? I remembered that little mis-fact from the terrible reference HTML 4 for Dummies which I read 7 years ago.. maybe IE5 had that limit? -- ⁪ffroth 02:09, 28 November 2007 (UTC)


 * Thanks to all for inputs and enthusiastic support and very educating info I am recieving here,ofcourse further discussion and inputs is most welcome and I expect this discussion to continue a little further and it is quite interesting to follow guidance from all of you.Mahitgar (talk) 13:30, 27 November 2007 (UTC)

32-bit vs 64-bit time
32-bit time will cause the year 2038 problem in January 2038, when the number of seconds since January 1, 1970 causes signed 32-bit integer math to wrap around. Does 64-bit time use the same signedness, granularity and epoch? From the Sun article, the Sun will become a red giant in about 5e9 - 6e9 years. Ignoring the comparably negligible difference between January 1, 1970 and now (November 26, 2007), this will mean that it will happen at roughly (very roughly) at time 0x0038CC1780063000 to time 0x0044281C333AA000. Both of these fit nicely not only to signed 64-bit integers, but also to signed 56-bit integers, which means that by clever programming we could save 12.5% of computer memory used to store the time. J I P | Talk 17:56, 26 November 2007 (UTC)


 * Okay, that last thing you said is just stupid. It's a fine example of premature optimization, "the root of all evil". On most systems it wouldn't even make any difference in memory usage, because 64-bit integers are aligned to a multiple of two bytes or greater. —Keenan Pepper 18:08, 26 November 2007 (UTC)
 * I know. I forgot to add a smiley. J I P  | Talk 18:33, 26 November 2007 (UTC)


 * There is no standard definition of 64-bit time the way there is for the Unix "time_t time(time_t*)" call. The latest Single Unix Specification addresses it only in the marginal notes.  In all likelihood, at some point in the next few years, they'll say that implementations are allowed to do a 64-bit transition with time in the same way they did with calls like lseek.  That means they'll probably just make time_t a 64-bit signed integer, maybe with some transition functions like time64, and leave everything else the same.  --Sean 18:22, 26 November 2007 (UTC)


 * There are already plenty of computers where  is 64 bits (retaining, of course, the conventional interpretation as seconds since 1970).  Debian Linux running on my Alpha does it that way, for instance.  (On that machine, regular old   is 64 bits, too, which I personally believe is as it should be on a 64-bit machine.)  And most modern processors and compilers have good support for 64-bit arithmetic even if they're not wholly 64-bit architectures, so I'm confident we can expect to see   on more and more implementations switch over to 64 bits over the next several years, and if that were all it took to solve the "Y2.038K problem", we'd be home free.


 * Unfortunately, that's not all it will take to solve the Y2.038K problem. The big bugaboo, I fear, will be the same as it was for Y2K: on-disk records.  There are lots of systems that store binary data on disk, and lots of them store timestamps as four-byte, 32-bit  's.  It's likely to be exactly as hard to handle all of those as it was for on-disk two-digit years, and I'm afraid our grandchildren are going to be just as scornful of the shortsighted old Unix programmers of yore as we all were of the shortsighted COBOL programmers who were behind the Y2K problem.  (But of course that's not fair, either, because there were programs written in virtually every language that were bitten by Y2K, not just the COBOL ones.) —Steve Summit (talk) 02:17, 27 November 2007 (UTC)


 * The 2038 problem isn't all THAT serious - nothing like on the scale of Y2K - only people who used signed integers to manipulate absolute dates are in trouble - and we've known that was a bad idea for a very long time. Those who went with unsigned numbers are safe until 2076 or so.  Since the storage problem is VERY unlikely to be affected by signed/unsigned issues, and VERY few people will have packed dates into 31 bits, we only need to fix code in order to survive to 2076...it's exceedingly unlikely that any large database files on disk will need work before then.


 * Incidentally, it's not just DEC Alpha's that are using 64 bit time_t's. Both my AMD64 PC SuSE 10.2 Linux system and my I32 PC SuSE 9.3 Linux machine have time_t being a typedef of "__kernel_time_t" which in turn is typedef'ed as "long".  However, sizeof(long) on the I32 machine is 4 - so  time_t is a signed 32 bit number (Aaarrgghh! Big problem in 2038!) - but the AMD64 machine reports sizeof(long) to be 8.  AMD64 machines are very common these days - so an awful lot of people are already running 64 bit time values and if they have screwed up and put 31 or 32 bit times on disk, they already know it!


 * As an aside: You might also be tremendously excited to see this photo I took off my Linux machine's screen on the occasion of the first UNIX billenium: http://www.sjbaker.org/gallery/main/billenium.png ...it's running a program that reports the return from 'time' whenever the value changes. My family sat around with puzzled looks on their faces and Tux toys in their laps counting down the seconds.  OK, I can tell that you aren't tremendously excited (neither were they)...but still...cool, or what?! :-)  SteveBaker (talk) 03:17, 27 November 2007 (UTC)


 * You were lucky that the program missed 999999998 and not 1000000000; that would have been much less photogenic! --Tardis (talk) 15:44, 27 November 2007 (UTC)


 * Okay, my own random comments:
 * Keenan Pepper: You're right that using 56 bits to store an "extended" time_t is a ridiculous optimization, although I once got code published in a book (C Unleashed, by Heathfield et al.) showing how to do exactly that. It was tactical: my thinking was that there are some programmers (and managers) out there who would think that 64 bits is "clearly overkill" and would reject any suggestion that they use 64 bits for their own on-disk time records.  But it's easy to see that 32 bits are too few -- plenty of us will be alive in 2038.  (I'm enough of a Unix nurd that I hope to be.)  So 56 bits was a good compromise.  (Or maybe I used 48, I forget.)
 * SteveBaker: Who says using signed integers to manipulate absolute dates is a bad idea? What if we want to represent our birthdays?  (And if nobody does, why are you worried that your I32 machine will have big problems in 2038? :-) )
 * Anybody who liked Other Steve's countdown video will love Image:Year 2038 problem.gif.
 * —Steve Summit (talk) 23:44, 27 November 2007 (UTC)
 * Oh yes, it's clearly overkill to use a single more byte. Sheesh we only have 12 billion to work with here on the server! Why not just add like 1 or 2 more bits? Surely moore's law doesn't apply to the growth of seconds. -- ⁪ffroth 02:02, 28 November 2007 (UTC)