Computing desk
< August 25	<< Jul \| August \| Sep >>	August 27 >

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.

August 26

Putty recycle bin hack

A student in one of my classes asked if there is such a thing that allows files deleted during an SSH session in Putty to be automatically copied to the Windows recycle bin. He (and another student) claimed that they were told that this exists. I cannot see how it would work. It would be a rather annoying hack to program. You'd have to first recognize that you connected to a Unix/Linux system. Then, you'd have to detect the use of "rm filename" typed in. When you see it, you must suppress the enter key and launch an SFTP session to copy the file to the recycle bin. Then, allow the file to removed. But, if someone decided to go through all that trouble, I'd like to let the students know where they can download this hack. Has anyone head of it? -- kainaw ™ 02:44, 26 August 2009 (UTC)[reply]

I've used PuTTY for a while, and never heard of such an extension or tool. However, there are a variety of ways to make a unix system use a "safer" delete. One way is to alias "rm" to some other command (e.g. by editing the .bashrc file):

alias rm='\mv \!^ .trash/'

Assuming that you have created a .trash/ directory. Then, you could map that to a shared network drive, (or something). (Conceivably, your students have been loosely describing an alias mapping rm to scp or something - but I don't know how they're scp`ing directly to a Windows recycle bin - one could imagine a convoluted setup, running openssh on windows and using SSH keys to avoid typing a password for each transfer... can you say "unmaintainable"?). Personally, I do not think it is good practice to alias the basic shell commands (especially "volatile" commands like mv and rm), because it gets the user into a false sense of comfort. Then, when they migrate to an unconfigured server, their safe "rm" is no longer safe - and they have not developed the discipline to use the shell properly. Nimur (talk) 04:33, 26 August 2009 (UTC)[reply]

X11 on MacBook

I'm trying to install OpenOffice on a MacBook running OS X 10.4. The latest version that I can find that will run on it is version 2.4 but that requires X11. And according to Apple's site, I can't download a version of X11 for 10.4. The install discs (with X11 on them) are literally 1000 miles away. What I'm trying to do is make it so that my father can open MS Word .doc files on his PowerPC MacBook without having to buy (or steal) MS Word. And getting him to install X11 once he goes home and then walking him through the install process for OOo over the phone is out of the question. I'd like to find a version of OOo Aqua that doesn't need X11 but they don't seem to have one for US English. Does anyone have any suggestions as far as either dealing with the root issue of opening Word docs or the issue of installing X11? Thanks, Dismas|^(talk) 03:52, 26 August 2009 (UTC)[reply]

Root issue: how about AbiWord? X11: Only thing I can think of, barring borrowing disks, is possibly downloading an ISO from some torrent site. Technically not legal, though ethically I think one could live with it (why X11 is only available on the disk, I have no clue and have always thought silly). --68.50.54.144 (talk) 04:04, 26 August 2009 (UTC)[reply]

Maybe this works? Not sure though.. —Preceding unsigned comment added by Sealedinskin (talk • contribs) 07:10, 26 August 2009 (UTC)[reply]

Get it from here: http://cjmart.in/2006/12/01/install-x11-on-tiger-without-the-tiger-install-dvd/ --Spoon! (talk) 07:59, 26 August 2009 (UTC)[reply]

Thanks guys! I downloaded that guy's copy of X11 and then went through the Apple Software Updater to update it. Opened up OOo and it worked like a charm! Thanks! (I might have gone with AbiWord but I already am familiar with OOo, so tech support for my dad over the phone will be easier with something that I already know) Dismas|^(talk) 17:54, 26 August 2009 (UTC)[reply]

This is probably too late but why don't you try NeoOffice instead? It's OS X-native. --antilived^{T | C | G} 20:14, 26 August 2009 (UTC)[reply]

It is too late. He went back home this morning. Thanks though. Had never heard of that. Dismas|^(talk) 22:04, 27 August 2009 (UTC)[reply]

about router

when a ip datagram is transfered from one interface to another one on the router,what changes would happen to the MAC address sections of the datagram?please good guys tell your opinions. —Preceding unsigned comment added by Newton2009 (talk • contribs) 06:40, 26 August 2009 (UTC)[reply]

If it's genuinely a router (unfortunately people seem sometimes to call switches routers, which they entirely aren't) then it'll send a new ethernet frame (with its own and the destination router's MAC addresses) with the received IP packet encapsulated within it. You shouldn't look upon it as being the same packet with its MACs rewritten. -- Finlay McWalter • Talk 08:14, 26 August 2009 (UTC)[reply]

Mobile phone internet protection

Do mobile phones have anti spam, anti virus and defence against hackers from the internet? Simply south (talk) 09:40, 26 August 2009 (UTC)[reply]

yes,it does.There are some mobile security software in the market.If you're intrested in it,search "mobile security software" for it.—Preceding unsigned comment added by [[User:{{{1}}}|{{{1}}}]] ([[User talk:{{{1}}}|talk]] • [[Special:Contributions/{{{1}}}|contribs]])

Thanks, however that seems to imply that whenever you get a mobile with the internet, there is no protection originally in the phone. Simply south (talk) 18:32, 26 August 2009 (UTC)[reply]

Fan controller schematic: comments needed

I've designed a schematic a personal desktop cooling system, which is essentially a temperature sensor, a microcontroller, 4 old computer case fans and a fear-instilling, Doctor Octopus-like suspension, powered by an ancient ATX supply. Being a complete beginner when it comes to hobbyist electronics (I've put together a few designs I found online, but this is my first time actually building something I designed from the ground up all by my lonesome), I was hoping some of the more experienced people here were willing to comment on, and generally audit, my schematic -- I want to be really sure everything is correct before I order any parts. The design can be found at [1] (warning: large image!).

The function of the microcontroller in this design is to read in the temperature from the LM335 through its ADC and output a PWM signal to the fans. Multiple ADC inputs are used in case I someday decide to use multiple temperature sensors; each ADC input has a corresponding PWM output, so each fan can have a separate sensor. The design can also easily be modified to support up to 3 additional fans using the 6 microcontroller pins that currently have pulldown resistors.

Thanks in advance! --Link ^(t•c•m) 10:29, 26 August 2009 (UTC)[reply]

I'm by no means an expert but I am a bit puzzled by your design: why do you need R9-R12? Why not just replace the two resistors with a 1k and save some parts and soldering? I probably wouldn't bother with calibrating the LM335 unless you actually have the equipment for it (it's not like you require much precision anyway), and you should include connectors for in-place programming (it'll make debugging a lot easier). Unused pins don't really need to be pulled-down (I don't know about the ATTiny26 but most new Atmels have internal programmable pull-up resistors), and you should leave the unused ADCs unconnected, with a wirepad for future expansion (currently it's almost impossible to connect extra sensors). And one last point, you do know that Molex gives out 5V too right? You might be able to do away with the 7805 too. --antilived^{T | C | G} 11:13, 26 August 2009 (UTC)[reply]

Thanks for the reply! The R9-R12 are needed for the manual operation and tuning modes (which I have forgotten to post about); by setting S1-S4 to the constant 5V rail, the fans run independently of the µc and their speed can be adjusted down by means of the aforementioned pots. In microcontroller-based mode, it will still be possible to decrease the base speed of the fans using the exact same pots; e.g. if the fans just produce too much blast. I definitely agree on the ISP and extra connectors (I'll get cracking!). Regarding the 5V line on the molex: that is already connected, but from what I've heard, ATX supply voltages are quite unstable, so I decided it wouldn't be a bad idea to add a regulator on that line to stabilise it for the µc and temperature controller. --Link ^(t•c•m) 16:03, 26 August 2009 (UTC)[reply]

The manual mode with adjustment will, however, defeat the purpose of PWM (efficiency) though, since you're just using the Darlingtons in active mode. If I were doing this I'd connect the pots to the left-over ADC pins and do the manual override in software instead. And I just noticed you've connected the 5V line to the input of the 7805. That will give you minimal filtering as the 7805 requires about a volt (at this load) of voltage drop for reliable filtering. If you want a clean power supply you need to connect the 7805 to the 12V line, or otherwise remove the 7805. --antilived^{T | C | G} 20:03, 26 August 2009 (UTC)[reply]

The main reason I personally chose to do this with PWM and a microcontroller is that it appears to be extremely difficult to scale the LM335 output precisely with analogue components. None of the cheap µcs I can get my hands on have multi-channel DACs, so I figured I'd use PWM. It's not so much about efficiency for me as it is about getting it done. It's more of a learning and hobby project anyway. Also, if I did the manual override in software, I'd lose the ability to use the full-manual mode without using more potentiometers. As for the 7805, good call. I reckon it's best to hook it up to the 12V line, because I don't want to brick a microcontroller due to a spike on the 5V rail. Thanks! --Link ^(t•c•m) 21:01, 26 August 2009 (UTC)[reply]

With a bit of software you can maybe designate different ranges of the potentiometer to different settings, like if the value from ADC is >240 (8-bit ADC) then it's in full auto mode, otherwise set the speed according to the potentiometer's setting (full manual mode)? --antilived^{T | C | G} 07:14, 27 August 2009 (UTC)[reply]

How self-sufficient is the Internet?

That is, if everyone in the world disappeared, how long would it be before the internet stopped working? Obviously, it wouldn't be updated, but would simply viewing a page be possible after a day? A month? A year? Thelb ⁴ 13:39, 26 August 2009 (UTC)[reply]

It would probably depend on one important thing - electricity. All webservers, switches, routers, fibre repeaters, modems, telephone exchanges and computers are powered by electricity from some source. The main datacentres and exchanges will have some for of UPS backup which in some cases will last them for a day or two (depending on whether it's battery backup or generator backup) but apart from that, when the power station stops, it all stops. Nanonic (talk) 14:30, 26 August 2009 (UTC)[reply]

Part of the answer is: How long would electrical power plants continue to run, unmanned and unmaintained? Tempshill (talk) 15:23, 26 August 2009 (UTC)[reply]

Another way to view the question: How long will the information remain intact? Assume all humans disappear at once. Then, a thousand years later, some aliens come and turn the electricity back on. Will they be able to use the Internet to gather information about our culture? Please don't get into a discussion about what they will think of us when they see half the Internet contains porn. I'm just slightly rephrasing the question in such a way that "stopped working" could mean "is not able to function even if another form of intelligence were to get it operating again". -- kainaw ™ 16:02, 26 August 2009 (UTC)[reply]

A thousand years should be fine for some storage, but degredation will no doubt be existent. — neuro^(talk) 16:03, 26 August 2009 (UTC)[reply]

Half? You my friend are an optimist... or pessimist.

This is an active area of research. It's not only the technological failure you need to worry about - even if the technology infrastructure persists, the connectivity may fail catastrophically. Without a central governing authority, many researchers theorize that the internet will cease to be economically viable, and may schism into disconnected sub-networks. (What is astonishing is that this has not happened yet, despite 30+ years of evolution. Amazingly, the "internet" continues to be economically viable; to have universal connectivity to a single, globally-addressable network; and to be unambiguously the internet, as opposed to many competing large networks that are not mutually accessible, as we see with mobile telephony. (It is theorized that the status-quo is only stable due to government mandate - i.e., phones operated by one brand can successfully dial and call phone-numbers hosted by competitors - as this represents an economic loss for all involved parties). A "stable system", allowed to relax along capitalist profit-motivation lines, is theorized to result in in-network-only phone calls (and similarly, in-network-only IP addresses for the computer networks. Evolution of the Internet AS-Level Ecosystem (Srinivas Shakkottai et al.); Internet Interconnection and the Off-Net-Cost Pricing Principle (Laffont et al) , and A Market Model for Networked Systems (by me) ^{[original research?]}. Nimur (talk) 23:10, 26 August 2009 (UTC)[reply]

I would guess in-network-only calls would rapidly result in a monopoly by one provider. That would be the only way everyone could call everyone. People may be willing to install multiple free IM clients, but far fewer people would be willing to buy and pay line rental on multiple phones (some already do for various reasons, of course, but I doubt many people would). --Tango (talk) 23:32, 26 August 2009 (UTC)[reply]

I've taken the liberty of correcting some ambiguous grammar in my earlier post. I should reiterate that our current best economic models of network connectivity are certainly incomplete, as they predict an outcome which we are not observing (yet). So, at best we can say that we do not really understand how "the internet" is managing to scale so well; or why its architecture is so resilient to major changes in hardware, software, audience, and usage patterns. It's very hard to predict how the network as a whole would respond to a catastrophic "wipeout" event - but I think a safe assumption is that certain high-redundancy systems would persist long after any humans were around to maintain them. Those systems, however, would fail to be "connected" via a global internet. Nimur (talk) 05:23, 27 August 2009 (UTC)[reply]

For the electricity question The Straight Dope estimated that if humans all disapeared, the electrical grid (in North America) would collapse in under 24 hours leaving only a few isolated islands of working electricity. (Their analysis is more detailed than that, and a fun read. I recommend clicking.) APL (talk) 02:47, 27 August 2009 (UTC)[reply]

Question about SD card moved from WP:MCQ

I have a sd media card that some how got encripted is there a device i can purchas to be able to uncript those files —Preceding unsigned comment added by 71.136.34.82 (talk) 03:18, 26 August 2009 (UTC)[reply]

Could you explain how exactly you know the data on the card is encrypted? Tell us what error message you are getting, and how it is you believe the files became encrypted. Tempshill (talk) 15:22, 26 August 2009 (UTC)[reply]

If the device is encrypted, you'll need to know the key (pass) to decrypt it. It cannot be decrypted without it, unless you brute force, which I will not go into. — neuro^(talk) 15:41, 26 August 2009 (UTC)[reply]

If the files are properly encrypted, there is no feasible way (no device, no software) which can unencrypt them without the password. You can guess the password, e.g. by brute force method, as Neurolysis has hinted at. Nimur (talk) 23:15, 26 August 2009 (UTC)[reply]

Maybe you mean the files have got DRM (I.E. Content Protection for Recordable Media), and you aren't allowed to access them? Unfortunately I'm not aware of a reliable way to break that (and it may be illegal in some jurisdictions). I don't think it's that hard, it's just that few devices use them so it hasn't been well studied. It's possible you'd need a specially designed card reader or at least one with updatable firmware but this is just a random guess. I know some phones (MicroSD obviously) have bugs where they ignore DRM requirements under certain circumstances, e.g. when you aren't supposed to be able to copy the file although that's clearly only when you can already access the file. Then again if you can't access the files with any device you may be SOL anyway Nil Einne (talk) 19:28, 27 August 2009 (UTC)[reply]

Simple HTML Question: Forcing Side Scroll

I have the "shifting content" problem with a website i am making. I looked online for a way to force scroll bars to always be on, and i got this:

html { height: 100%; margin-bottom: 1px; }

From the following sites:

http://hicksdesign.co.uk/journal/forcing-scrollbars-now-even-better http://www.communitymx.com/content/article.cfm?cid=528a0

Neither of these sites are clear as to where this code goes. Is it javascript or straight html? between what tags should it go? Can someone please help me? Thanks!

71.214.145.141 (talk) 17:47, 26 August 2009 (UTC)[reply]

It's neither, it's CSS. It either goes in a special section of an HTML page or in a dedicated .css stylesheet document that's included. This tutorial should get you started. -- Finlay McWalter • Talk 17:51, 26 August 2009 (UTC)[reply]

Perfect, i put the code after the <style type = "text/css"> tag, it works fine! thanks!

71.214.145.141 (talk) 17:57, 26 August 2009 (UTC)[reply]

I'd recommend html { overflow-y: scroll; } though - clearer, and doesn't cause a 1-pixel scrolling scrollbar. Unilynx (talk) 22:00, 27 August 2009 (UTC)[reply]

Resolved