Wikipedia:Reference desk/Archives/Computing/2011 June 16

Welcome to the Wikipedia Computing Reference Desk Archives
The page you are currently viewing is an archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

June 16

Ubuntu Desktop 10.04 LTS - Separate Partition Sizes

Good Morning, Everyone!

  I would like to place each of the following ten eight directories on its own partition, and I need to know the required size (assuming a base installation) for each partition (except of course the /home directory partition). I've looked around on help.ubuntu.com, but haven't been able find any information on this.

• /boot
• / - the system root is to contain all the directories (bin, dev, etc, lib, media, mnt, proc, sbin, sys & temp) which are not to be placed on their own partitions.
• /home
• /opt
• /root
• /srv
• /usr
• /var
• /lib
• /sbin
• /temp

  Thanks as always. Rocketshiporion♫ 03:36, 16 June 2011 (UTC)

Although I can't actually answer your question (the easiest way would just be to do an install with whatever options you want and just see for yourself), but are you actually putting the partitions on different physical disks? If not then I don't really see any advantage to making them all separate partitions, it's just going to trip you up later when one of them hits capaciy and you have unused space elsewhere on the same disk. I would just use the entire disk keeping their default partitions as you can always split them up to different physical disks later if need be.  ZX81  ^talk 04:33, 16 June 2011 (UTC)

I agree with ZX81. I use three partitions: an extremely minuscule /boot partition, and two "major" partitions: / and /home, of which /home is about three to four times the size of /. JIP | Talk 18:12, 16 June 2011 (UTC)

/lib and /sbin often contain vital files for a lot of programs your system will need to actually mount those partitions (eg, my '/bin/mount' uses libraries from '/lib'. Separating those may make your system unbootable, unless you go through a lot of trouble setting up an initrd to handle it. It's more common to split off just /boot and /home. /usr, /opt and /var should also be designed to be easily separated (see the file hierarchy standard). I'd recommend JIP's partition setup too, and would only split off /opt or /var if I had reasons to expect to be storing a lot of data there. Generally, I'd say allocating 20GB for everything but /home should suffice for a Linux installation. Unilynx (talk) 22:26, 16 June 2011 (UTC)

I intend to put each partition on a different iSCSI volume, and there's no problem if any partition hits capacity, as the iSCSI volumes can be easily resized. However, as separating /lib and /sbin might make the computer unbootable (now that would be a problem), I now think I'll seperate only /boot, /home, /opt, /root, /srv, /usr and /var. Rocketshiporion♫ 03:46, 17 June 2011 (UTC)

You might also want to look at Logical Volume Manager (Linux). Just sayin'. -- 88.67.157.239 (talk) 19:29, 19 June 2011 (UTC)

Nope. Won't be needing Logical Volume Manager (Linux). The iSCSI volumes are on a SAN head, and they can easily be resized. Rocketshiporion♫ 10:16, 20 June 2011 (UTC)

Acquisition of home LED and LCD monitors, navigating marketing jargon

I did check the archives. I'm having problems distinguishing the computer monitor marketing terms LCD and LED. I'm assuming that LED in this case means an LCD monitor with LED backlighting? Additionally, given the high degrees of marketing, it becomes difficult to make a rational comparison and evaluation of cost/benefit when I'm looking to run something around dual 1920x1200 with pivot. Web searching results in marketing guff after marketing guff sites. Attempts to make a rational choice in person have been restricted by product variety (pivot to portrait is an uncommon feature) and sales pressure. My market region is Australia, so there are limitations in online purchasing from non-Australian suppliers, but generally no major product differences between markets. I am aware that I can't afford the time to immerse into specialist communities, so I'm looking for high order help with differentiating and evaluating products. My primary purpose is writing, and reading high resolution photographs of aged documents in an academic context. Fifelfoo (talk) 03:40, 16 June 2011 (UTC)

That's correct; a monitor or flat-screen television marketed as "LED" just means "LCD with LEDs for backlighting". Our article is the rather unfortunately-titled LED-backlit LCD television. As for how to buy, sorry, I've never found a great source. I use newegg.com's user reviews for a lot of electronics, but for TVs, I haven't found any substitute for looking at one at the store. Comet Tuttle (talk) 06:51, 16 June 2011 (UTC)

Most web sites have a "compare" option that puts all the stats up side-by-side for a few items (5 is a typical limit). There are lots of stats to compare, though, so here are a few that I think matter most:

1) Brightness. This is particularly important if you're in a room that's not easy to darken. (I just bought black-out drapes for my room, so that's another way to go.)

2) Size. Although bigger is only better to a point.

3) Connectors. It has to have connectors which are compatible with your PC.

There are many other things that are probably not important, in your case:

A) High refresh rate (over 60 Hz). That's useful mainly for video.

B) Low response time. Again, only important for video.

C) High contrast ratio. Again, mainly useful for movies, although it might have some impact on still pictures, too. One confusing thing here is that they sometimes give a "dynamic" contrast ratio, which just means it gets darker when they turn off the back-light. That's cheating, to my way of thinking.

D) LED back-light. I don't think this makes much difference. A true LED monitor (which does exist, for large industrial signs) would be much brighter, but the cost would be absurd, and I'm not sure if they could even scale that down to a reasonable size for home use. So, yes, "LED" is just mostly marketing jargon.

E) TV tuner.

Also, I'd guess that the "pivot to portrait mode" feature will severely limit your choices and up the price, so are you sure you really need that feature ? StuRat (talk) 07:39, 16 June 2011 (UTC)

Yes, I'm extremely certain regarding pivot. 95%+ of archival documents are in portrait mode. Extended screen reading via scrolling single pages in landscape severely degrades my user experience. Most website "comparison" charts provide meagre, market driven comparisons (though, I will consider newegg as a resource for information given the high praise I've heard elsewhere). I don't see the advantage in size without a corresponding resolution change: bigger 1900x1200 pixels aren't better. Thanks for the grounding assistance. Fifelfoo (talk) 07:52, 16 June 2011 (UTC)

There's an optimal screen size for any resolution. Too small, and your eye can't distinguish between the tiny pixels, so adjacent pixels seem to blur together. Too large, and you see each pixel as a big block, and you have to move your head to read the whole screen. These effects are dependent on your visual acuity an viewing distance, as well as screen size and resolution.

Regarding marketing, I certainly wouldn't trust any comparison offered by an individual manufacturer, as they will find a way to make their product win. However, why would a comparison provided by a retailer carrying multiple product lines be biased ? Yes, they want to sell you something, but they make money regardless of the brand you choose, so they should provide you a fair comparison. (Yes, they might theoretically want to steer you toward higher profit items for them, but biased comparisons are likely to piss off both their customers and other suppliers, so the risk isn't worth it.) StuRat (talk) 06:39, 17 June 2011 (UTC)

Re D: Actually an OLED display would by most definitions be considered a true LED monitor and the problem here is producing the panels in large sizes (particularly at resonable cost) not small sizes (where they are starting to become common e.g. even in some cheap mobile phones and other portable devices). The current best appears to be a 15" inch Nil Einne (talk) 11:23, 16 June 2011 (UTC)

Regarding the "pivot to portrait mode", couldn't you just use a regular screen, permanently rotated 90° (with the image rotated 90° to match at the graphics driver) ? That is, do you need the ability to switch that monitor back to horizontal ? With a dual monitor set-up, perhaps one could be permanently in portrait mode and the other permanently in landscape mode. I bet this approach would save you lots of money. StuRat (talk) 01:43, 19 June 2011 (UTC)

Characters added to text file

How many times would the letter ÿ have to be added to a text file in order for the size to increase from 781,403 bytes to 239,667,608 bytes? CambridgeBayWeather (talk) 04:04, 16 June 2011 (UTC)

The answer would depend on how the file is encoded. It couldn't be a simple ASCII file, but that leaves several possibilities, which use different numbers of bytes. Our article on character encoding may provide some useful explanation. Looie496 (talk) 04:17, 16 June 2011 (UTC)

Yes, you will have to specify the encoding of the file to know for sure, but ÿ will likely map to either 1 (e.g. CP1252) or 2 (e.g. the Unicode charsets, even UTF-8) bytes (see here). Therefore, you the answer is either 238,886,205, or half that (if it's mapped to two bytes): 119,443,102.--el Aprel (^facta-_facienda) 04:25, 16 June 2011 (UTC)

And both answers assume the text file is not compressed. Indeed, if it was compressed, then, depending on the compression method, you might never get to that size by adding the same character. StuRat (talk) 07:50, 16 June 2011 (UTC)

Thanks all. I'll check it again at work and see. CambridgeBayWeather (talk) 14:39, 16 June 2011 (UTC)

You'd have to eventually. By the time that you've added $256^{239,667,608}$ of them, you'll have created more distinct files than than can be represented in 239,667,608 bytes. But an uncompressed file that large is firmly situated in the realm of the unpractical. Paul (Stansifer) 14:58, 16 June 2011 (UTC)

Theoretically, yes, but no file system would actually support such a large file. StuRat (talk) 01:50, 19 June 2011 (UTC)

I checked again, it is Windows-1252 and it adds 1 byte for every instance of the letter. It turns out that it is caused by the modem not hanging up correctly. The icon vanishes from the tray so it appears that all is well. CambridgeBayWeather (talk) 23:05, 16 June 2011 (UTC)

Problem with WLAN under Ubuntu

I am able to make a connection, but it is extremely slow. The problem is only under Ubuntu, if I try a different OS, the connection is OK. I tried disconnecting IPv6, Global DNS servers, ioctl vs. external, uninstalled WICD and re-installed Network Manager, installed WICD again, but nothing worked. This Ubuntu connection worked fine in the past, but I made some PPP connection, started torrents and also a connection status report program. One of these screw it up, but I dont know where the problem can be...

You'll want to ask this question at the computing desk. Try WP:RD/C. --Jayron32 01:27, 16 June 2011 (UTC)

@ Jayron32 - this is the Computing Reference Desk. Rocketshiporion♫ 06:19, 19 June 2011 (UTC)

If an IP is shared between multiple people in reality, could they be accused of sockpuppetry?

Read above headline. --Jeff (talk) (contribs) (email) 08:25, 16 June 2011 (UTC)

No, I don't think so. Sock puppetry is the use of multiple accounts by one editor (for the purposes of deception), not the use of one account by multiple editors. There is a clause in Wikipedia:Username policy which says "Sharing an account – or the password to an account – with others is not permitted, and doing so will result in the account being blocked" - but this clearly only applies to registered accounts. Gandalf61 (talk) 08:37, 16 June 2011 (UTC)

Jeffwang is asking about shared IP addresses, not shared accounts. If several people in a household (or office) had Wikipedia accounts then they would have the same IP address, and this would be indistinguishable from a single person using multiple accounts, so it could look like sockpuppetry, but per WP:AGF there would need to be evidence of misuse of the accounts for an accusation. See also Wikipedia:Signs_of_sock_puppetry. AndrewWTaylor (talk) 09:34, 16 June 2011 (UTC)

There's no policy or rule against it, but in practice you are running a much higher risk of being banned because of collateral damage as checkusers often just ban everyone on a shared ip if the user agents match, which happens if you share a computer with someone else who uses Wikipedia or edit from public computers with have a set pre-installed browser on them. 82.43.90.27 (talk) 10:28, 16 June 2011 (UTC)

list of -fu to practice to become an elite hacker?

Hi,

I've just started out on the path to becoming an elite ninja hacker. I have installed several varieties of Linux and BSD in some virtual machines. Next, I learned how to get vi to save, to quit, and to type much of what I write into it in the wrong mode. I likewise learned how to start and exit emacs. I know C++ and Perl from before I entered on the path towards becoming an elite ninja hacker: my quesiton is, could you suggest an elite ninja hacker training program of techniques and technologies I should practice? Obviously bash helps, but what else! Thanks for a detailed regimen you could suggest. — Preceding unsigned comment added by 188.29.180.253 (talk) 11:24, 16 June 2011 (UTC)

Sounds like you mean hacker as in taking over machines rather than hacker as meaning a good programmer. Might I suggest from the films I've seen that you need four screens, a very fast computer with a GPU supercomputer for password cracking and access to a DS5 comms link plus a steel door and an underground passageway for escape. I believe the main thing you really need though is the mindset where if you see an ad selling worms for fishing via mail you think how you could send worms to somebody else via it. Dmcq (talk) 12:20, 16 June 2011 (UTC)

From what I've been able to tell from similar research, You're mostly going to need an awesome soundtrack and a hot babe. APL (talk) 01:30, 17 June 2011 (UTC)

I recommend to work through W. Richard Stevens' Advanced Programming in the Unix Environment (how is that a red link?) and Lions' Commentary on UNIX 6th Edition, with Source Code. Skip The Art of Unix Programming, Raymond was in poser mode when he wrote it. The UNIX-HATERS Handbook and the Jargon File are ok (although you might want to get an earlier version, the last ones are somewhat polluted by wannabe political correctness). --Stephan Schulz (talk) 12:31, 16 June 2011 (UTC)

A book on computing theory and/or algorithm design would be good too. (For example The Art of Computer Programming) So many people think that this stuff is "obvious" or something they can figure out on their own and keep making the same stupid mistakes. APL (talk) 01:36, 17 June 2011 (UTC)

Pick a specific task and learn how to do it. Otherwise, you'll just be a script kiddie who downloads work that others did and pretends to know what he is doing. Those who do the work actually know about computers. For example, can you exploit a buffer overflow error in IE9 if you don't know how to read through the machine code and see how it is handling buffers? There is a reason that a lot of the real "hackers" are graduate students - it takes a lot of dedication to learning about the hardware to get good at hacking the software. -- kainaw™ 12:37, 16 June 2011 (UTC)

If your intent is to become a black-hat hacker, you are going to need a very strong sense of paranoia. If "188.29.180.253" is your real IP address then you have already failed the first test.

If your intent is to become a computer guru that those in the know will call a "hacker" in admiration for your ability to bend (your own) computer to your will, then you've made a good start, but you've got a long way to go. You need to cultivate a strong sense of curiosity and obsessiveness that causes you to pick a task that you don't know how to do, and learn everything you can about every detail of the tools you might use to complete that task, and then stay up long hours of the night to finish the task, post it somewhere, and pick a new task that you don't know how to do.

That's how I'd approach it anyway, there are multiple paths up the mountain, but I don't think you'll find an easier one. APL (talk) 01:30, 17 June 2011 (UTC)

This depends solely on what you'd like to do...and the amount of time (and money) you're willing to put in. There is a difference between knowing how to start emacs (or load a vhd) and knowing how to locate and reliably exploit a dangling pointer. If you're up for it, you should learn assembly with Assembly Language Step-by-Step: Programming with Linux ISBN 0470497025...though you may not get far. Smallman12q (talk) 15:51, 17 June 2011 (UTC)

I agree with APL in that if you're intending to become a black hat hacker, you would need an irrational sense of paranoia. But before you embark on this path, be warned that black-hat hacking is a serious crime in most jurisdictions. Your IP address, 188.29.180.253, geolocates to London. All I need to do to find you (i.e. if I had the legal authority to do so) is to check with your ISP which customer was using that IP address at 11:24am on 16-June-2011 (UTC), and I would have your billing information. The police in most countries do have that legal authority, and would swoop down on you in next to no time. Rocketshiporion♫ 04:55, 20 June 2011 (UTC)

parody video of startup (full of buzzwords) on techcrunch?

someone told me about a great video on techcrunch (must have been recently) parodying a startup, i.e. the buzzwords, and I would like to see this video, hwoever I'm having trouble finding it. does my description ring a bell for anyone? — Preceding unsigned comment added by 188.29.180.253 (talk) 15:10, 16 June 2011 (UTC)

Remote Desktop from Linux to Windows?

I use Linux at home but Windows at work. Is it possible to open a Remote Desktop session from Linux to Windows, assuming both computers are on the same VPN? JIP | Talk 18:14, 16 June 2011 (UTC)

Yes, rdesktop supports the protocol (RDP) that Microsoft Remote Desktop uses. There are other alternatives as well. Nimur (talk) 18:19, 16 June 2011 (UTC)

Also FreeRDP. ¦ Reisio (talk) 18:10, 17 June 2011 (UTC)

Does a device like this exist?

Is there a small, handheld device thats main purpose is to display picture files? It would add to its usfullness if sound commands were an option. — Preceding unsigned comment added by 109.74.50.52 (talk) 19:18, 16 June 2011 (UTC)

I think digital photo frame fits the bill. 75.155.136.49 (talk) 20:22, 16 June 2011 (UTC)

Exactly. --Ouro (blah blah) 07:40, 17 June 2011 (UTC)

Delete duplicate files

Resolved

On my Windows 7 computer, in several folders I somehow got hundreds of duplicate files, e.g. filename (1).ext for every filename.ext. There are hundreds of these. Is there an easy way to delete them? (Doing them individually takes a lot of time and I can't figure out any way to sort them so I can mark a block.) Bubba73 ^{You talkin' to me?} 19:46, 16 June 2011 (UTC)

Could you search for "(" or "(1)" and delete all the resulting files? AndrewWTaylor (talk) 20:37, 16 June 2011 (UTC)

Searching for "(1)" comes up with a large number of files that don't have "(1)" in them, so it doesn't work. Bubba73 ^{You talkin' to me?} 02:38, 17 June 2011 (UTC)

You can easily delete all files that have (1) in their names from the command prompt. Hold the Shift key and right click on the folder with the offending files, then choose "Open command window here". In the command window, type del "* (1).*" to delete all files with " (1)." in their names. (The * is a wildcard character.) This will delete the files permanently, not move them to the Recycle Bin. If you're not entirely sure you want that to happen, you could instead type mkdir deleteme & move "* (1).*" deleteme. That will move the files into a subfolder named "deleteme", which you can then delete in whole or in part from Explorer. -- BenRG (talk) 03:14, 17 June 2011 (UTC)

That works perfectly! (I created the new folder with right-click, new folder instead of mkdir.) The command line wins again! Thank you very much! Bubba73 ^{You talkin' to me?} 03:36, 17 June 2011 (UTC)

Command line forever! There are things that are easy with the command line that you can't even come close to doing in the Windows GUI. Bubba73 ^{You talkin' to me?} 05:05, 17 June 2011 (UTC)

Overwriting CD-Rs to destroy data

CD-R#Security_risk says that it's possible to write over CD-Rs to destroy the data they contain, but the article seems to suggest that this functionality needs to be implemented in the hardware. Is that true? Why can't some software tell any CD-R burner to write 1s to all the blocks on the CD? Or, take the image of the CD, reverse all the 0s and 1s, and then write that image on top so that the 1-blocks aren't burned twice. Is this possible? Thanks.--el Aprel (^facta-_facienda) 23:44, 16 June 2011 (UTC)

Well it actually says "overwriting" (in quotes), because it's not really overwriting it, but literally destroying the disc (which has the same result that the data has been overwritten/removed). CD-R's simply can't be overwritten with new data, you'd need a CD-RW for that and the only way to destroy the data is to the destory the CD using whatever method you prefer :)  ZX81  ^talk 00:10, 17 June 2011 (UTC)

Well, he's not talking about adding new useful data to the disk. He's talking about taking the portions of the disk that weren't written to (the "zeros") and writing them so that they become ones. Now there are no zeros and nobody can tell which ones are the original ones.

Without knowing too much about it, I strongly suspect that the firmware on standard hardware isn't capable of doing that.

The hardware is designed with the one goal of usefully writing data to the disk. There are all sorts of markers and stuff that the drive will look for to get itself lined up properly. If it can't find an "empty" section of disk, it probably won't write. This would be an important safety feature. APL (talk) 01:18, 17 June 2011 (UTC)

Wouldn't it be simpler either just to break the CD into a few pieces with your hands, or to use a shredder (I know paper shredders which can also destroy credit/debit cards and have special-size openings for discs)? This, I suppose, quite effectively destroys the data, wherby in case of filling up the empty space on a CD with ones the process would practically leave the actual data untouched, in which case a skilled enough computer expert could probably still retrieve it. Am I right? --Ouro (blah blah) 07:39, 17 June 2011 (UTC)

Zap it in a microwave oven for one or two seconds (certainly no longer) - a very effective destroyer of CDs in my opinion. Astronaut (talk) 13:00, 17 June 2011 (UTC)

If you happen to have a spare microwave sure [1]. If not, perhaps not a great idea particularly if you do it regularly and it's a shared microwave (even if you don't mind potentially damaging the microwave and the fumes other people might not be so happy). Also [2] suggests (and from the images I've seen I would have to concur) microwaving might not be that effective some parts seem to remain relatively okay (although the FAQ also mentions recovering data from a single wipe of HDs, something most evidence suggests isn't possible). Most professional recovery services don't seem to mention what level of damage they can recover from optical discs but you could try asking them if they can recover any data from microwaves discs if you're really interested. If you're truly paranoid some of the suggestions in that FAQ or [3]. P.S. Just don't use this product [4]. Well it may work well, I don't know but the stuff in the FAQ is mostly bullshit so I wouldn't trust it. Nil Einne (talk) 14:18, 17 June 2011 (UTC)

The FAQ doesn't seem that bad to me, and the product seems probably okay. But I would trust one of LiteOn's disc-destroying drives ($25 shipped from Newegg) more than this product ($19 shipped from the maker). -- BenRG (talk) 18:26, 17 June 2011 (UTC)

Sorry I closed the window. It wasn't the FAQ but [5] I was thinking of which discusses how "Most drives spin at over 24,000 RPMs". It also says "Even the most advanced algorithms can only correct 2 bit errors per 32 bit codeword" which I'm resonably sure is incorrect since I just read about this, it's only partially describing the situation in audio CDs and is completely missing the extra error correction layer in CD-ROMs.

Finally there's "Unlike an analog signal (an old record player, for example) the codewords and length (n,k) of 0's and 1's are fixed and finite. Therefore, the small sections destroyed by the Optical-Strip cannot be simply ommitted without affecting the rest of the data in a binary string. In other words, breaking up the key sections on a track destroys the rest of the track because of digital processing. Each Optical Strip that is applied breaks every data track twice, and even the smallest data files take up many tracks"

Which if I understand what they're saying correctly is misleading. Sure you might not be able to record undamaged data but if you have a large text file you could potentially recover parts of the text file which may not be useless. I say potentially because I don't deny if you use their product the disc will almost definitely become unreadable in a drive because amongst other things it looks like you'll destroy the TOC and leadin completely and also the drive would likely keep losing tracking. And I have no idea whether data recovery companies will bother with a disc damaged in that way.

Nil Einne (talk) 23:28, 17 June 2011 (UTC)

That article may be referring to LiteOn's SmartErase [6] [7] [8] which I think just writes over with random data. I suspect that is likely to work just as well if not better as trying to 'reverse' the original writing. Given the complex nature of the CD/DVD encoding system combined with the error correction I'm not sure whether a reverse of the original data will even produce the desired output. As APL said, most drives are not designed to be able to write to sectors that have already been written to. Nil Einne (talk) 15:55, 17 June 2011 (UTC)

The data, after error correction and LFSR scrambling and more error correction and subchannels and interleaving, is eight-to-fourteen modulated, and the output of that is what actually goes on the disc. The ones are encoded as pit-land or land-pit transitions, and the zeros as a lack of transition. So flipping all of the bits wouldn't do it; you would have to add or remove an odd number of 1 bits early in the writing process, and leave everything else alone. But I doubt that the accuracy of the writing process is good enough to obliterate the data this way (it would certainly make it unreadable to an ordinary drive, but not under a microscope). And this is all academic because any drive that will let you write to an already-written CD-R will presumably have a special disc-destroying mode that doesn't go through the usual encoding process. -- BenRG (talk) 18:26, 17 June 2011 (UTC)

Though one must be wary of toxic fumes, simply burning the disc (i.e. with a fire) will have the same dye-destroying effects as overwriting it with a CD-R and be a lot more thorough. For the truly paranoid, who want no readable remnants after you are done, then I would suggest that this is the way to go. Dragons flight (talk) 20:21, 17 June 2011 (UTC)

Thank you all for your interesting responses.--el Aprel (^facta-_facienda) 19:34, 18 June 2011 (UTC)