Those are a lot of good points. Ok... Idea number two: Replace the "undeliverable" message with one telling the spammers that they've won a "free boat"... NByz 20:24, 25 January 2007 (UTC)[reply]

The main point I was trying to make above is that you cannot send a message to the spammers, they do not provide a valid return path for responses, anything you send will usually end up going to another victim. In other words, sending any type of response back to the from: address in SPAM is not a good idea! Droud 23:57, 25 January 2007 (UTC)[reply]

Droud's point is a good one, and worth repeating. Never set up automated bounce messages for spam: if you do, you are contributing to the spam problem. On some days, more than half of the spam I see is not original spam, but rather bounce messages to spam that was forged by spammers as if to have come from me. (What's worse, these bounce messages are trickier to filter, since I don't want to train my spam filters to reject bounce messages in general, because of course I still want to see bounce messages from messages I did send.) —

talk) 13:14, 26 January 2007 (UTC)[reply

]

• Should be free and light-weight
• Should work with pre-recorded files, and on selections (not necc. the whole file)
• Should be able to make lots of different voices
• Should have male-to-female voice conversion
• Should be easy to use (Easy to learn would be nice too, but isn't a requirement)
• Should support either Windows or x86 Linux —The preceding
  unsigned comment was added by Kjoonlee (talk • contribs) 03:42, 24 January 2007 (UTC).[reply
  ]

Thank you. :) --Kjoonlee 03:38, 24 January 2007 (UTC)[reply]

I don't think what you seek exists due to the male-to-female conversion (which is a highly complex problem). Bolster those programming skills and start an OSS project to write it! On a more serious note, have you considered text to speech software instead of stalking reading stories yourself? If you're set on the creepy changed voices, try out funny voice. Droud 04:07, 24 January 2007 (UTC)[reply]

It could be done, using a voice recognition system, then feeding that output into a text to voice converter with the proper voice selected. However, you would get errors in both steps, making the output likely full of errors. StuRat 07:50, 24 January 2007 (UTC)[reply]

I just saw a commercial on Cartoon Network for some some new toy. Bionics, I think - wasn't paying attention. But, one of the toys in the commercial is a voice-changer that changes your voice into the voices of one of the many characters from the cartoon associated with the toys. Just thought I'd mention it because it is related to this question. Of note - the kid in the commercial selected different voices with the writband selector and they all sounded exactly the same to me. I haven't seen the cartoon. Perhaps all the characters in the cartoon sound exactly the same. --

(talk) 09:59, 24 January 2007 (UTC)[reply

]

It probably just does some standard manipulation to your voice, like the pitch shift used to create "chipmunk voices". If the same effect was used on the actor's voice to create the cartoon character's voice, and if your voice is somewhat close to the actor's, you may end up with a result somewhat close to the cartoon character's. StuRat 22:05, 24 January 2007 (UTC)[reply]

Well, the important thing is that my accent has to be still there in the final recording. (This will be a labour of love on my part, and I hope it helps my girlfriend get better at English.) If I were recording Alice in Wonderland, for example, I'd be using my own voice for everything Lewis Carroll says. I only want the Mad Hatter and the March Hare to sound different, and I don't want Alice to sound masculine. --Kjoonlee 13:50, 24 January 2007 (UTC)[reply]

Can't you just shift the pitch of your voice yourself ? StuRat 22:07, 24 January 2007 (UTC)[reply]

Or use Audacity to change it for you... and record it.... and add some other fancy effects and stuff... --antilived^{T | C | G} 05:03, 25 January 2007 (UTC)[reply]

This is what audiobook readers do normally. It's really not that difficult to do. 203.49.220.190

A commercial product called AV Voice Changer looks excellent, but it is priced accordingly! (cleaned off-thread post) Droud 00:01, 26 January 2007 (UTC)[reply]

I was trying to burn a free game for one of my friends who has dial up. THe copy would not burn so: I tried to take out the cd. It doesn't open. any suggestions on how to get it out? —The preceding

Most CD drives have a small pinhole under the drive door. Get a paperclip, straigten it out and gently push it into that hole. That'll trigger the drive to open the door. Alternately, reboot your computer. When it's starting up, but before the main OS loads, press the Eject button and the drive should open. If that doesn't work, tell us more about what OS you're running and what kind of CD drive you've got. -- Kesh 04:50, 24 January 2007 (UTC)[reply]

If it's a Macintosh system, shut down and as soon as you turn it back on hold down the mouse button. Almost all Macintoshes will understand that as a command to eject all disks.

68.39.174.238 21:37, 26 January 2007 (UTC)[reply

]

(Note: I'm not the same guy from the question above, it's just coincidental.)202.10.86.63 07:59, 24 January 2007 (UTC)[reply]

What type of computer do u have, as different computers use different ways to fix this problem? --|K.Z|Z.K| Do not vandalize... 08:27, 24 January 2007 (UTC)[reply]

You'll probably need to replace your CD-Drive. It's fairly easy, and CD drives are fairly cheap (DVD-RWs are down to about

£20 now), but obviously you need to get into the case. The case screws are a problem. Try reading this post on a forum. --h2g2bob 14:49, 24 January 2007 (UTC)[reply

]

Well, your drive is getting power if it will open and close the tray. Sounds like it's gone bad, or the disc has. Try another disc and if that doesn't work either, it's a bum drive. As to the screw, it sounds like you need a Torx or Hex key wrench. -- Kesh 01:16, 25 January 2007 (UTC)[reply]

I have a picture that I'd like to try to upload, but I'd like to crop it first. However, the only image-editing software I have is Paint, which will make the picture into a Paint file. Can anyone recommend some image-editing software that is:

1. Free
2. Downloadable
3. Relatively small (I don't want to download more than I have to)
4. Fairly simple (All I want to do is crop, and I don't want to have to read a huge readme to find out how)
5. Will leave the picture in a file format that can be uploaded to commons.

Thanks very much for any help :-) Skittle 10:40, 24 January 2007 (UTC)[reply]

I'll assume you mean Windows Paint (because Mac/Linux/etc users usually start their questions "I have a Mac and..." :-) In Paint, select "File -> Save As...", there is a dropdown menu of file types. Weregerbil 11:09, 24 January 2007 (UTC)[reply]

Thanks, I'll have a go. I assumed Paint, being MSPaint, only came on Windows :-) Shows my ignorance. Skittle 11:27, 24 January 2007 (UTC)[reply]

Argh! I hate Paint. Is there a way to crop a big picture from the left effectively? Cropping from the right I can do, but I assume I have to select the image I want to keep if I want to crop from the left. However, the image I want doesn't fit on the screen and there's no 'zoom out' option. Skittle 11:33, 24 January 2007 (UTC)[reply]

Crop the right and bottom sides, rotate by 180 degrees, crop the left and top sides (that are shown to the right and bottom respectively), rotate by 180 degrees... TERdON 18:06, 24 January 2007 (UTC)[reply]

I can crop it appropriately in Word, but then I can't save it as an image file. And when I try to copy-paste back to paint it all goes horribly wrong. Does anyone know either how to crop this properly in Paint, or where I can get some software as described above? Skittle 12:05, 24 January 2007 (UTC)[reply]

For a Paint quickie: click the rectangle tool, select the area you want, drag the area with the mouse to top left, then resize the image area by dragging the bottom right corner. The trick is to guess that you can drag a selected area. Or after selecting the area copy&paste it to a new document. For other programs there is

List of raster graphics editors and Comparison of raster graphics editors, can't say which ones are good though. Weregerbil 12:28, 24 January 2007 (UTC)[reply

]

The problem with that is that I can't select all of the image I want, because it doesn't fit on the screen. Oh wait, sudden thought, I can move it in pieces if I'm very careful... I may be back sobbing in a moment... Skittle 12:32, 24 January 2007 (UTC)[reply]

Okay, moving it in pieces seems to have worked. I've lost a few slices of image a few pixels tall, but it's not noticable. Thanks for the ideas people! Skittle 12:46, 24 January 2007 (UTC)[reply]

There used to be a great program for this from Microsoft called PhotoEd. I believe it was a hidden program on an old version of Office. Does the program still exist? Can you get it? I don't know because I haven't had a Windows machine since 1998 and I use Gimp for all my graphics needs. --

(talk) 13:07, 24 January 2007 (UTC)[reply

]

If you still want a decent free image editor, give Paint.NET a shot. Matt Eason ^{(Talk • Contribs)} 13:42, 24 January 2007 (UTC)[reply]

NNNNNOOOOOOOOOOOOOOOOOOO..... Don't use an image editing program to crop files, unless you're cropping lossless bitmap images. Use jpegcrop instead. --Kjoonlee 13:54, 24 January 2007 (UTC)[reply]

How to use jpegcrop:

1. Run jpegcrop
2. Open JPG file
3. Select portion of image by dragging
   • Drag edges of selection to fine-tune results
     • Clicking on the "Block Grid" toolbar button helps
4. Right-click selection, Save Frame As

--Kjoonlee 14:01, 24 January 2007 (UTC)[reply]

For image cropping and other rudimentary operations, you don't need (and sometimes don't even want) a full-blown image-manipulation program. In particular, many image-viewing programs will not only let you crop, shrink, and rotate the image being viewed, they'll also let you save the modified version. Two such viewers I'm familiar with that can do this are

talk) 13:56, 24 January 2007 (UTC)[reply

]

Low tech solution: You can use the grey square at the far bottom right to crop the picture from bottom and right. Then use tools, flip or (tools, rotate) to flip (or rotate) the image so the top-left corner of the image is in the bottom-right corner of the screen. Then crop the bottom-right again. Finally, use the same tools, whatever as before to flip (or rotate) the image the right way up again. Harder to explain than do, honest! --h2g2bob 14:24, 24 January 2007 (UTC)[reply]

You should check out GIMP for Windows, GIMP is an open source PhotoShop like image editor for linux, and windows now too :D Aetherfukz 16:13, 24 January 2007 (UTC)[reply]

Use the program Irfanview . It is very simple, and user friendly. --Codell ^{[ Talk]} 03:11, 2 February 2007 (UTC) Thanks everyone! In the end I used jpegcrop, but I shall remember the other tips (particularly flipping the overall image in Paint) for other occasions. Thanks again! Skittle 16:25, 24 January 2007 (UTC)[reply]

It takes GNU sort (from

Not sure what fancy shmancy algorithm they use but sorts are computationally expensive. IIRC a bubble sort takes at least n squared loops where n is the number of items --froth^T 17:22, 24 January 2007 (UTC)[reply]

I believe that GNU sort uses some optimizations that, when presented with a certain type of data set. You might want to consult the full documentation for it to be certain (eg, the INFO docs.) --Mdwyer 22:50, 24 January 2007 (UTC)[reply]

Try setting the environment variable LANG to C, with the line

export LANG=C

These days, a number of the GNU core utilities use, by default, your native internationalized locale, which valiantly tries to do all sorts of culturally appropriate, but vastly more computationally expensive, Unicodey conversions and comparisons. Setting LANG to C forces it to use the old, unsophisticated, fast, native charset operations. I've had the slowness problem most acutely with the grep family of utilities, but it's perfectly imaginable that sort is similarly afflicted. —

talk) 05:13, 25 January 2007 (UTC)[reply

]

That said, 0.8 seconds is pretty fast for 50000 entries. Sorting is computationally expensive (unless you're sorting numbers, in which case it's quite fast at O(n)), even optimized quicksort (which is what I assume they are using) takes O(n log n). Oskar 09:26, 25 January 2007 (UTC)[reply]

Well, once upon a time, 0.8 seconds might have been "pretty fast", but today it's dog-slow. Me, I'd expect something more like grendelkhan's second result, and in fact, on my computer I can confirm it exactly:

head -50000 /usr/dict/words | shuffle | time sort > /dev/null

0.26 real 0.08 user 0.01 sys

—

talk) 01:56, 26 January 2007 (UTC)[reply

]

Aha! It was the LANG settings. After setting LANG=C (it was en_US.UTF-8), it sorts those 50000 strings in 0.08 seconds. Spiffy! grendel|khan 15:20, 25 January 2007 (UTC)[reply]

Of course, this solution is only helpful if you actually are sorting content that's using the traditional one byte per character. If there might really be UTF-8 in there, which is an increasingly likely possibility these days, and you want it to sort correctly, you're stuck with the slow sort. --Anonymous, January 26, 2007, 01:00 (UTC).

Good point, although of course this depends on your definition of "sort correctly". If you want è, é, é, and ë to sort anywhere near e, and not between ç and ì, then yes, you've little choice but to use the slower, localeized sort. But if you want to sort by Unicode code point, it turns out that a naïve byte-oriented sort on UTF-8 gets you that. (I suppose I should cite a source for this. The Unicode Standard 5.0, §2.5, p. 35: "A binary sort of UTF-8 strings gives the same ordering as a binary sort of Unicode code points.") —

talk) 02:06, 26 January 2007 (UTC)[reply

]

You would either need another Program that reads a file your PHP script generated (probably a plaintext file) and prints it out. Or you could control your printer itself with PHP Printer functions. Aetherfukz 16:17, 24 January 2007 (UTC)[reply]

I want to know what is System 360 Operator. Please explain it with examples and pictures. Thanks Tariq Mehmood possible phone # removed —The preceding

Probably a person who operated an IBM System/360. Perhaps you meant "System/360 operator panel"? You can use Google image search to find plenty of pictures of that. -- mattb @ 2007-01-24T18:15Z

I'm attempting to transwiki Weapons in Dead Rising and Weapons and items from The Legend of Zelda series to Strategywiki before they are deleted. But when I attempt to import them at strategywiki, it says they are too large. I'm a sysop there obviously, and I use export here (with history, to satisfy GNUFDL), save the file, then import there and get that error. Is it a setting I'll have to get the owner of SW to change? Are all edits necessary for GNUFDL? Is there another way around this? I'd appreciate any help possible.

Weapons in Dead Rising —The preceding

Yes you're required to provide access to the history. A lot of mirrors just link to the wp article and say "look here for the original history" --froth^T 19:20, 24 January 2007 (UTC)[reply]

If you have access to the server, it may be a problem in php.ini which restricts uploading of large files. Edit php.ini and change the maximum upload size value (can't remember official name but the file should be well commented) and it will probably work. If you don't have access to the server, contact the server admins (that may well be one of the bureaucrats). --h2g2bob 22:48, 24 January 2007 (UTC)[reply]

I STRONGLY suggest that you save a copy of these articles yourself right away because these damn deletionests delete these aritlces very fast and without warning. Thats how i got my fav articles deleted; I've been working on them for so long. I was gonna transwiki it or make a wiki myself but I was too late and the next time i checked it was too late and they alradly deleted them. 12 pieces of valueable work we have been working on suddenly lost :( --Taida 02:58, 25 January 2007 (UTC)[reply]

That suddenly? Unless they were speedied (and in that case there is at least partial responsibility for the creator to read the CSD), they would have sat on AFD for at least 3-4 days.--inks^T 05:53, 25 January 2007 (UTC)[reply]

In any case if you want the text of a deleted article, if it isn't anything objectionable you can just ask an admin. They can view deleted articles. --24.147.86.187 15:12, 25 January 2007 (UTC)[reply]

Not only that, we can also temporarily undelete them and move to a subpage of your user page for you to save a copy (this would be the ideal way if you want to use Special:Export on them). The place to request that is at

WP:DRV. --cesarb 18:03, 25 January 2007 (UTC)[reply

]

I have a copy of both articles, it must be an upload size limit, I know the server admin, I'll let him know, thanks.

Chris M. 15:19, 29 January 2007 (UTC)[reply

]

What are the differences in these two methods of storing uncompressed audio? Are there advantages to either method (or any other), particularly in quality? Thanks. Someoneinmyheadbutit'snotme 23:30, 23 January 2007 (UTC)[reply]

I'm guessing this is homework (see the top of the page), so I'll just link to

PCM, and ask you to think about the pixelation-type effects at low amplitudes. If this isn't homework, feel free to leave a shocked post at my slander :D --h2g2bob 22:42, 24 January 2007 (UTC)[reply

]

I am shocked. No, seriously, this is for personal interest. I've been recording recently and am a bit of an audiophile. Someoneinmyheadbutit'snotme 20:15, 25 January 2007 (UTC)[reply]

Computer audio typically uses a 16 bit sample (65,000 possible values) and max 24 bit samples (16.7 million possible values), whereas floating point values are stored in 32 bits (4 billion possible values). The human ear can just barely discern the difference between 16 and 24 bit, so floating point is not particularly beneficial, and takes more storage space for the same audio. If you're looking for high quality recordings, sampling frequency is vastly more important than the format of each sample (once you get to 16 bit samples, that is!). Droud 00:15, 26 January 2007 (UTC)[reply]

Well, I'm aware of all that, but Audacity gives the option of both 32-bit floating point and 32-bit PCM, which begs my question. Someoneinmyheadbutit'snotme 02:20, 26 January 2007 (UTC)[reply]

If you're going to store raw audio, 16 bit fixed-point PCM should be more than enough if the audio has been downsampled and/or dithered properly.

If you're going to do post-processing later on, do use 24bit fixed-point PCM, 24bit floating-point PCM, or something better. --Kjoonlee 13:43, 26 January 2007 (UTC)[reply]

Oh, and also, please consider using lossless audio compression to save space. I personally recommend FLAC and WavPack. --Kjoonlee 13:43, 26 January 2007 (UTC)[reply]

People can't hear the difference between 24 bit and 32 bit, so it's just a matter of how much space you want to waste. FLAC is an excellent way to save about half the space a raw PCM file consumes, but is slower to use due to the compression and decompression steps. Most sampling hardware (sound cards) can't actually record better than 24 bits, while CDs (16 bit) and DVDs (24 bit) don't store better, so storing that sound as 32 bit PCM would be useless. Droud 13:59, 26 January 2007 (UTC)[reply]

I disagree with Droud about sampling rate. 44.1 kHz should be more than enough fidelity for end-results. Also, 24bit floating-point PCM would be the same size as 24bit fixed-point PCM. --Kjoonlee 13:48, 26 January 2007 (UTC)[reply]

24 bit is 16.7M possible values whichever way you look at it, encoding to floating point just slows down processing since floating point values are scaled (the value is stored as a number and an exponent, not a straight number). As far as 44.1kHz being an ample sampling frequency - people hear frequencies, not samples. A higher sampling rate more accurately describes frequencies than does a higher sample accuracy. In any case, anything over 16 bit, 44.1kHz is overkill, unless you're an audiophile like the questioner. Droud 13:59, 26 January 2007 (UTC)[reply]

Regarding sampling rate, see

Hearing (sense)#Hearing in humans and Nyquist rate and you'll see why 44.1 kHz is considered sufficient for the human range of audio perception (the reason 44.1 kHz was chosen exactly is a little more complicated, but not worth getting into here). There are some folks who believe that interference and harmonics of higher frequencies that we cannot directly perceive can produce audible results once rendered back into sound, but I've never seen a particularly credible study or quantification of this. -- mattb

@ 2007-01-26T14:09Z

Please accept my apologies Someoneinmyheadbutit'snotme, there are quite a few homework questions get submitted and I was in a bad mood :( I doubt there's much I can add to the above - there's not much advantage to using floating point to be honest --h2g2bob 15:58, 26 January 2007 (UTC)[reply]

All right, I already know all of these things. All I'm asking is: Is there a difference between 32-bit floating point audio and 32-bit PCM audio at the same sampling rate? Someoneinmyheadbutit'snotme 20:12, 26 January 2007 (UTC)[reply]

Floating point audio is slower to mix, transform, and play than its integer (PCM) counterpart due to differences in how floating point and integer calculations are performed on a microprocessor. If you're using hardware instead of software, it really doesn't matter at all, except that PCM enjoys much wider support than floating point. Droud 03:19, 27 January 2007 (UTC)[reply]

Thanks. Someoneinmyheadbutit'snotme 18:15, 27 January 2007 (UTC)[reply]

I transferred from my i.s.p.(claranet) to broadband with talktalk. I contacted claranet regarding the transfer and they inquired if i would like to keep my old e-mail address whilst i still had a few months to run with claranet, i seemed quite happy with the setup but just recently i struggle to receive many e-mails from talktalk awindow appears that an error has occurred whilst downloading e-mails from claranet, this appears to block my e-mails from talktalk. is there a way i can romove the claranet download or will i have to contact claranet to finally disconnect me from theresite —The preceding

Are you using a webmail interface? Outlook? Outlook Express? Eudora? Checked your POP settings, access, and passwords? It would be much easier and faster (IMHO) for you to call your ISPs tech support line.--inks^T 21:35, 24 January 2007 (UTC)[reply]

Grampus4 says he/she's using Outlook express in the title. I have no idea about Outlook, as I've never used it. To edit the accounts you have listed, look for an item called "Accounts", probably in the tools menu. I'd suggest checking the help files (really, they do contain some help). Also, can you post the actual error message you get, as this will help people see what's wrong? --h2g2bob 22:24, 24 January 2007 (UTC)[reply]

Ah yes, so s/he does, my bad. :) Having re-read that a few times, it sounds like there may be 2 OE accounts, one of which is failing to authenticate/connect/'work' for some reason. But I'm sure that if the settings for a particular account are correct, it should work regardless of what happens with other accounts.--inks^T 23:23, 24 January 2007 (UTC)[reply]

In using Excel for spreadsheets I have forgotten how to total a column or row--JmHoran 21:41, 24 January 2007 (UTC)[reply]

In the cell you want the total to appear in type "=SUM(A1:A7)" without quotes, and substitute A1:A7 for the range of cells that you want to add.--inks^T 22:05, 24 January 2007 (UTC)[reply]

Or go to the empty cell below or to the right of the line you want to total and click the sigma symbol to the top right of the standard toolbar. Anchoress 22:47, 24 January 2007 (UTC)[reply]