Wikipedia:Reference desk/Archives/Computing/2008 September 6

src: upload.wikimedia.org

September 6

Video Wikipedia:Reference desk/Archives/Computing/2008 September 6

OS Kernels

How can I repair the damaged or lost kernels of Win XP SP2? --Omidinist (talk) 03:51, 6 September 2008 (UTC)

Hi Omi, can you give us the exact error message you are getting? --mboverload@ 03:52, 6 September 2008 (UTC)

CD doesn't play and this message appears: 'The application failed to initialize properly (c0000006). Click on OK to termiante the application.' --Omidinist (talk) 12:01, 6 September 2008 (UTC)

What makes you believe that is a kernel problem? If you had a damaged or lost Windows Kernel, the chances of you seeing anything past boot is very slim. - Jimmi Hugh (talk) 15:06, 6 September 2008 (UTC)

I downloaded SP3. Then I was forced to delete it because it caused some unwanted effects, like damaging my ADSL connection. I think it may have removed or damaged some kernel. I am not an expert though. Thanks for any help.--Omidinist (talk) 16:15, 6 September 2008 (UTC)

This forum post says it might be a hardware issue. Xenon54 17:00, 6 September 2008 (UTC)

Omni, there is no need to "talk technical" with us. =) Talking about kernels without knowing exactly what it is will confuse people about your question. Also, I can assure you that your ADSL connection is not physically damaged. =) --mboverload@ 19:25, 6 September 2008 (UTC)

Maps Wikipedia:Reference desk/Archives/Computing/2008 September 6

iTunes problem

hi, ok:

i have itunes on an external HD along with all my songs, however, from what i can tell, all the 'info' about itunes (eg libraries, play count etc info) was stored on my computer (windows XP) so when i 're-set' my computer after a virus problem this was killed. Thus when i open up itunes now, from my external HD, its as if ive just got it. This is of-course a big problem for me. However, because i kinda foresaw an event like this i regularly saved a 'backup' version of the 'previous iTunes library' file on my ext HD. Is there any way that i could load that 'previous iTunes library' up, bareing in mind all the song file paths are the same...? thanks, --81.76.41.196 (talk) 13:52, 6 September 2008 (UTC)

Well, not sure, but I think you could just look for the iTunes library file on your internal HDD, then delete it, then copy-paste the "previous itunes library" file into where the old one was. flaminglawyer^c 18:52, 6 September 2008 (UTC)

I am not sure with Windows but that is what I would do on a Mac. Kushal (talk) 16:52, 7 September 2008 (UTC)

If you look through the options there is an "Import Library" option in iTunes. Use this and point it at the backed up library. This will probably solve your issue TheGreatZorko (talk) 08:18, 8 September 2008 (UTC)

If any of these solutions worked for you, please let us know, as I am heading down the same path very soon and would like to know the best solution. cheers. 10draftsdeep (talk) 18:12, 9 September 2008 (UTC)

src: super-memory.com

Firmware engineering.

I have a friend who is a firmware engineer - he's looking for work right now. Is there an umbrella organisation for firmware engineering? He's looking for a list of companies that engage in firmware engineering in Texas (and maybe elsewhere) to get an idea of where to send his resume to.

SteveBaker (talk) 15:01, 6 September 2008 (UTC)

Steve, send me a Wiki email. I know of a company that is deep into firmware programming, not sure if they need anyone at the moment though. --mboverload@ 01:48, 7 September 2008 (UTC)

src: upload.wikimedia.org

Typing in Devnagari Alphabet

I have to type many pages of text in Devnagari font.But I am familiar with typing in QWERTY keyboard only i.e. Latin Alphabet.Is there any software(freeware) that transliterates the text typed in roman to devnagari? Thank you 202.70.74.161 (talk) 16:50, 6 September 2008 (UTC)

Here's an online one: http://207.172.130.5:8080/cgi-bin/webitrans.pl . If you're really going to be doing a lot, it might be worth just changing your keyboard layout and learning how to type Devanagari directly. --Sean 17:41, 6 September 2008 (UTC)

src: upload.wikimedia.org

RAZR's water-resistance

I am confused. Someone I know's razr (the phone) got dropped in a swimming pool. It was only in there for like a second, then it was retrieved and dried with a large fan. For about a week, its buttons didn't work, then they all started working again, except for the 1 button. I, just recently, "accidentally" sent my RAZR through a washing machine in my pocket. I forgot to move my clothes from the washer to the dryer, but when I did, about 5 hours later, I realized that my phone was in there, so I took it out and dried it with the same large fan. My phone worked, and still does work, perfectly (all the buttons), after only an hour of fanning. Can someone explain that to me? 75.66.48.112 (talk) 18:46, 6 September 2008 (UTC)

Assuming that the device doesn't short-circuit and fry itself when it gets wet, the next biggest issue is impurities in the water mucking up the works when the water evaporates. Depending on what exactly is in the water, you might be dealing with a conductive residue or an insulating one. Each causes its own set of troubles for the electronics. Buttons are often a first casualty because of the fact that they often require exposed conductors to work properly.

Probably what happened is that the second dunking dissolved some thin film (of pool chemicals?) that was left by the pool water. Of course, when the tap water from the washing machine dried it probably left behind it's own residue of salt, fluoride, and assorted other minerals, which may cause long-term problems for your device. (Corrosion is a big issue here.)

If you feel comfortable taking the device apart, it would probably last longer if you took it apart and gently wiped down each of the pieces with the purest alcohol you can get. (Sometimes it's sold in really expensive tiny bottles as "electronics cleaner" but I just use the drugstore varieties without any problem.)

As an aside, the battery may have been shorted temporarily while the device was dunked. This could reduce the battery's lifespan.

If you don't feel comfortable taking the device apart, perhaps it's not worth worrying about. Most people get new phones every year or two anyway. APL (talk) 19:10, 6 September 2008 (UTC)

FYI: A good, cheap source of nearly water-free alcohol is gas line dryer ("Drygas", etc.). You can usually get it in isopropal or methanol variations, but for working on electronics, I'd say you want the isopropyl variant.

Atlant (talk) 16:36, 13 September 2008 (UTC)

src: i2.wp.com

Rollover images

I can't figure out how to easily make a rollover image thing. By that, I mean showing image1, but when you put your mouse over image1, image2 replaces image1. Can someone show me an easy bit of code to do this? No Flash, I don't have a flash maker. flaminglawyer^c 18:49, 6 September 2008 (UTC)

Here's a super-basic cascading style sheets and HTML rollover:

  <html>    <head>      <style type="text/css">        a#foo       {display:block; width:100px; height:50px; background:url("a.jpg"); }        a#foo:hover {background:url("b.jpg");}      </style>    </head>    <body>      <a id="foo"></a>    </body>  </html>  

where a.jpg and b.jpg are both 100x50 pixels. -- Finlay McWalter | Talk 19:10, 6 September 2008 (UTC)

Without CSS... <img src='a.jpg' onmouseover="this.src='b.jpg'" onmouseout="this.src='a.jpg'" /< -- kainaw(TM) 21:25, 6 September 2008 (UTC)

If you are going to do a Javascript solution (like Kainaw suggests) you should probably pre-load the roll-over images so that they don't have a significant delay in loading. Google "javascript image preload" for about a million examples.

And it's worth noting that the CSS option is relatively limited in how IE supports it (only supports the :hover pseudoclass with A tags, nothing else). --98.217.8.46 (talk) 21:55, 6 September 2008 (UTC)

Thank you all. Your advice has inspired a masterpiece (in training) (the home sign has a rollover). flaminglawyer^c 23:34, 6 September 2008 (UTC)

src: upload.wikimedia.org

Interview coding tests

Reading the thread above about coding samples at job interviews leads me to ask a question about "Interview coding tests". I'm an Electronics engineer and have interviewed more folk than I have been interviewed. When interviewing, I am usually seeking another electronics engineer, often with a firmware/software skill as part of the job. Generally I ask the candidate to write some code to the following spec: I often have a handful of resistors and need to know their value when connected in parallel. Please write a program to do that. It never ceases to amaze me how many "Senior engineers" with "years of experience" in electronics and C have no idea how to approach this, don't know the equation for parallel resistors, have no idea of how to input values, do conditional tests, loop code, initialise variables or anything. The good ones do it just like that. One chap was good and honest, saying he wasn't into software but would do it with excel and proceded to explain that idea well. Some folk get very hung up on what language to write in (my answer - any at your choice, you just have to explain what it does even if it is a made up language). Some folk insist on doing it with a fixed number of resistors. Most make it far more complex than it needs to be. Anyway, what interview coding questions have you come across? (And if anyone wants to, how would you answer my question...)? -- SGBailey (talk) 20:01, 6 September 2008 (UTC)

Remember, SGBailey, we aren't forums - in my honest opinion, here isn't the right place to ask "what interview coding questions have you come across" - because it is so open ended. It's like asking "what is your favourite colour," it has a different answer for everyone who comes along. Your second question (about total value of parallel resistors) would be OK, but it seems like you already know what it is (if you ask it at interviews.89.240.241.186 (talk) 20:37, 6 September 2008 (UTC)

I rarely ask people to write code; I find you get a lot more info (and a lot more done in the interview time) if you ask them to read code. If you really look at what most engineers spend their time really doing, it's as much reading, understanding, debugging, fixing, and extending others' code than writing their own on a tabula rasa. So one test I set a few years ago had a small C program and asked them to say why the function didn't seem to return what its comment said it should (it was something like returning a pointer to an automatic or something like that) and determine if a given Java program was threadsafe or not (various versions either deadlocked or concurrently accessed memory without protection); I figure that's a much better test of C or Java skill than "write a C program that reverses a string" or whatever. For algorithms and datastructures I'd just talk to the person - having them actually code it isn't a good use of precious interview time. For anything other than short-term-single-task contractors I think tests that drill down to their specific knowledge of some arcane bit of the class library is pointless (anyone who understands how the memory interface to a memory-mapped modem works can pick up how ftell works by themselves, anyone who understands how a dynamic VM rewrites code to maximise cache locality can read the manual for the collections package themselves). The one thing I've utterly been unable to do (even after a 3 day "interviewing and hiring" class) is Find-The-Jerk; otherwise competent people fail big (and can take whole projects down with them) because of their deep personality defects, defects that they all seem to be able to conceal through the interview and initial hiring phase. -- Finlay McWalter | Talk 20:48, 6 September 2008 (UTC)

And I'm going to confess that I nicked several of the C examples from old ads in Dr Dobbs Journal that advertised a fancy lint by showing some bits of C code that had problems that weren't obvious but that their lint could find. The threadsafe examples were paraphrased from Bil Lewis' pthreads book. And a bunch of the rest were things that had come up over the last couple of years, reduced to the simplest sensible case. -- Finlay McWalter | Talk 21:00, 6 September 2008 (UTC)

It's interesting to hear you say that about the "find the jerk" issue, since I've experienced it myself (to say the least). I'm trying to change jobs now, and near the top of my resume I say "I'm funny, friendly, have broad interests, and am easy to get along with", which is mostly true. --Sean 22:30, 6 September 2008 (UTC)

The interviewing class I mentioned (too long ago for me to remember enough about it) centred around positing scenarios and asking how people would react. It was partially to get out of the rut of the interviewee just describing what they'd done, and the interviewer describing what the project is. It was also supposed to find people would would be a poor fit for the team. But I (and other technical interviewers on the course) found it didn't work terribly well for engineers. Partly because unfunny narrowminds with flat effect can still be valuable engineers. But a major problem (and potentially a significant legal hazard to any interviewer) is that many of the interviewees were non-native speakers and often from cultural backgrounds where job interviews were very formal things - so some candidates were clearly very technically strong but it was difficult to establish a good rapport. It would be foolish, unfair, and quite probably illegal to reject a candidate on that basis. -- Finlay McWalter | Talk 22:51, 6 September 2008 (UTC)

Also, there are memory mapped modems??? --Sean 22:31, 6 September 2008 (UTC)

(me not expressing self well; instant interview FAIL) In embedded systems it's quite common to have IO devices like modem chips, sound chips, network control chips wired to their own chip select and thus their control and data registers appear in the cpu's address space like magic bits of volatile memory. It's mostly like Intel's IN and OUT stuff, and entirely unlike a memory mapped file (and so "memory mapped" is perhaps a poor choice of words). -- Finlay McWalter | Talk 22:41, 6 September 2008 (UTC)

I've been through a lot of interviews recently. I'm a Senior games programmer with a speciality in graphics. The procedure may vary in other disciplines. But in 99% of cases, it goes like this:

• You contact them via email, attach a resume.
• Response (within a few days) is a request to do a "phone interview".
• The phone interview is sometimes a programming test - sometimes a general chat about what you did - what they do - why you want to work there, etc.
• If the phone interview didn't involve a programming test, then mostly they'll email you a written programming test.
• Then (if you didn't screw up) you get into a face-to-face interview - and in almost every case they want you to talk to the people you'll be working with - briefly go over incentives, vacation, health care - then goodbye - we'll be in touch. Often they'll want you to write a simple function on the white-board. A binary search or the 'atoi()' function for a "known-good" string or detect whether a given string is a palindrome...something of about that complexity. But doing it "live" in front of a whole room of critics can be painful if you get stressed out easily - you're generally being nudged into making 'off-by-one' errors or failing to detect the special cases (What happens if you ask whether an empty string is a palindrome?). There may be a discussion of the efficiency and legibility of your solution.

The programming tests are all pretty similar - whether done on the phone or in writing. Generally they start with some really simple logic puzzles ("You've got a 5 gallon bucket and a 3 gallon bucket and and unlimited water supply - measure out 4 gallons of water"), then there is a trivial C programming bug to track down (99% of the time the answer is "you returned a pointer to deallocated memory" or "you ran off the end of the array" - but there are variations on the theme). Next comes something ikky to do with C++ derived classes with virtual/non-virtual functions that somehow get messed up in a non-obvious way. After these "bug finding" things we get some short essay questions - usually stuff about linear algebra or how 'casts' work or what an abstract data class is. Then they generally wind up with a programming problem with some 'meat' to it - that takes some subtlety to complete and maybe needs 40 lines of code if you "get it". But it's pretty variable. SteveBaker (talk) 23:56, 6 September 2008 (UTC)

src: super-memory.com

Dropdown menu coding

Can someone tell me an easy bit of code to get a dropdown menu effect? Not like a filling in forms dropdown; but like on ebay, the homepage, when you hold your mouse over the "Categories" tab, a thing comes down with a bunch of different links. Something like that.

When I look at examples of it on sites that use this, the only thing I can find is a class="dropdown", or something like that. And little old me with close to 0 html experience doesn't know what that means. flaminglawyer^c 23:40, 6 September 2008 (UTC)

This is a nice article on drop downs using CSS/HTML only (OK you need a bit of javascript for it to work in IE). It might be a bit too complex/technical for you right now but it's far more elegant than the pure javascript way. --antilived^{T | C | G} 00:23, 7 September 2008 (UTC)

MaxDesign has a number of tutorials for CSS-based lists. Under "Listmatic," you'll find examples like this dynamic vertical list and this dynamic horizontal list. Examples and code provided. --- OtherDave (talk) 01:21, 7 September 2008 (UTC)

The horizontal one doesn't have an onmouseout event and so the drop down menu stays drop down even if I move the mouse somewhere else. Hardly a good example (and even less so for something to be learnt from). --antilived^{T | C | G} 07:47, 7 September 2008 (UTC)

Sorry to have taken your valuable time. Good thing you weren't the OP, huh? --- OtherDave (talk) 19:32, 7 September 2008 (UTC)

Source of the article : Wikipedia