Wikipedia:Reference desk/Archives/Computing/2013 July 27

= July 27 =

Quote characters in Word vs. Wikipedia
I feel that I should know this by now, but I don’t.

For the double quote character, Word uses an oblique and slightly concave symbol (e.g. the “real” story). That’s fine for work that is never used outside that context. But the problem comes when I prepare some text in a Word document before moving/copying it to a Wikipedia page. It’s not a big deal for simple text, but when it’s for table markup, what I need is the vertical character, e.g.


 * style="background:blue"

That simpy does not work when it’s:


 * style=”background:blue”.

Trying to change the character in Word never seems to work. I can cut and paste to my heart’s content, but Word still changes " to “ or ”. I have to resort to manual fixes within Wikipedia itself, and that can be extraordinarily tedious when a table has many entries.

The same issue arises with the single quote/apostrophe character (‘ vs '), but, while that rarely affects tables, I’d just like to have my text the way my aesthetic sensibilities dictate and not be governed by Word’s default protocol.

I’m sure there’s a simple solution to this, one that I should have discovered years ago. Any clues, folks? --  Jack of Oz   [pleasantries]  00:43, 27 July 2013 (UTC)


 * Turn off "smart quotes" in the autocorrect preferences. 00:53, 27 July 2013 (UTC) — Preceding unsigned comment added by Bobmath (talk • contribs)


 * (edit conflict)Methods for this seem to be different depending on what version of Word you have: here's the tutorial for Word 2003, also Word 2007. --Yellow1996 (talk) 01:01, 27 July 2013 (UTC)


 * The feature you describe is called "smart quotes" or "curly quotes"; disable them as described in this post. But really your problem is a symptom of a deeper one - Word isn't a good choice for producing wikitext, as it is apt to arse about with your input in an attempt to be "helpful". You'd be better off with a decent modern plain text editor (once that supports unicode) like Notepad++. 146.90.36.204 (talk) 00:59, 27 July 2013 (UTC)


 * Ah ha! That works.  Thanks for the quick responses.  --   Jack of Oz   [pleasantries]  01:22, 27 July 2013 (UTC)

Security training for the office
Hi! I am looking for ways to protect my small internet business against scams, etc. So I'd like to train employees (typical office workers) not to install random programs, to use safe passwords and to keep them secret, not to reply to spam, not to pay random invoices sent to them, how to discover spam, etc ... basic stuff. I've been looking for some comprehensive lists on what to include in such a training, but did not find anything. All I found is websites explaining how to secure web servers which is not what I need here. Any ideas on what to search for? bamse (talk) 00:49, 27 July 2013 (UTC)


 * As far as recognizing when legit invoices are submitted, one common system is to give each company you deal with a billing number, and only consider invoices submitted with a proper billing number. Of course, it's always possible that one of those companies may either intentionally or unintentionally submit an improper bill, or that somebody else will get hold of their billing number, but the idea is to reduce the volume of possible scams so that the remaining bills can be examined more carefully.  And, if you keep getting attempted scams using a particular billing number, then you should have a talk with that company, at the very least, and change their billing number.  If it happens again, sever your relationship with them. StuRat (talk) 00:58, 27 July 2013 (UTC)


 * I'd also encourage you to have some computers with no internet access, where you keep critical data, like your customer database and financial records. StuRat (talk) 01:00, 27 July 2013 (UTC)


 * External hard drives can be useful as well if you are trying to protect important data (encryption helps too.) --Yellow1996 (talk) 01:09, 27 July 2013 (UTC)
 * I would look for anything about Social engineering (security), Internet fraud, scams etc.. See for example the external links in the two articles. Ssscienccce (talk) 16:00, 27 July 2013 (UTC)


 * Training the employees won't be enough. You'll need a billing and audit system. Ideally, payments will have to go through a clearing person, that actually pays the bill. Others would only be able to indicate that a bill has to be paid or would get limited funds from this person for a business trip or such stuff. You could also have an office manager who pays minor stuff like pens, paper a new printer, and a manager who deals with the bigger stuff. Otherwise, your employees would be able to rip you off too.
 * Regarding securing PCs: simply have a similar system, where a senior employee manages the computer as admin and does not let anyone install anything.
 * Simply limit the points where something could go wrong. OsmanRF34 (talk)

agree with osman, that should work just fine
 * Iskánder Vigoa Pérez (talk) 12:47, 28 July 2013 (UTC)

Laptop battery
Hello. I just got a new laptop, and I've been told by some people that removing the battery when a power outlet is available would increase battery life because leaving it in would equate to constant low-level depletion followed by recharging (I understand that constant depletion and recharging wears down battery life). Is this true, or do laptops have a function where they stop drawing from the battery once it's fully charged and plugged in? And if it does make a difference, is this difference significant?  Bramble claw  x  02:47, 27 July 2013 (UTC)


 * I would expect anything as expensive as a laptop to have a smart recharging system which avoids overcharging. Only when you get to the cheapest items, like walkie-talkies, do I suspect that they cut corners and rapidly destroy the battery to lower the price a dollar or two. StuRat (talk) 09:44, 27 July 2013 (UTC)


 * A decent laptop battery should work well for at least two years, and tolerably well for at least one more, even if left in the computer, even if used regularly. After that, a new one is not very expensive - it's US$/EUR 129 for the MacBook Pro, which probably is on the high side, as notebooks go. That's about 10 cents a day, or less than a typical Starbucks' Latte per month. I suspect that the frequent removal and reseating of the battery, and leaving a large weak spot in the bottom of your notebook in between, result in a much higher expected damage, even if the process would make a significant difference in battery lifetime. --Stephan Schulz (talk) 13:31, 27 July 2013 (UTC)


 * I differ on $129 not being expensive. Considering that that probably exceeds the residual value of the laptop at that time, it would make more sense to buy a new laptop rather than buying a replacement battery that expensive.  The laptop can continue to be used when plugged in only, at that point. StuRat (talk) 23:43, 27 July 2013 (UTC)
 * That depends on the laptop, of course. Sure, for a $299 netbook, that is expensive. For a $3000 MacBook Pro, there is a significantly higher residual value. --Stephan Schulz (talk) 21:40, 28 July 2013 (UTC)


 * Hmmm... I've been using my Toshiba Satellite since 2008. I always have it plugged in whenever I use it, and the battery seems to be just fine (though then again, I wouldn't know unless I tried it without the plugin!) I read an article once that said this is a safe practice for laptops, and it's worked out for me so far! Also, Li-ion batteries should never be kept at 0% for too long. --Yellow1996 (talk) 19:24, 27 July 2013 (UTC)


 * Lithium-ion batteries will degrade with time, even if you don't use them. Discharged or fully charged batteries will degrade faster, optimum charge seems to be 40%. But the main reason for removing the battery would be the temperature. From the battery life section: Loss rates vary by temperature: 6% loss at 0 °C (32 °F), 20% at 25 °C (77 °F), and 35% at 40 °C (104 °F).  (that's loss of capacity, in one year) Ssscienccce (talk) 16:10, 27 July 2013 (UTC)


 * It would also depend on whether your laptop is really used as a desktop PC: if you only need battery power on rare occasions, you're really wasting an unused battery. In that case I would consider storing it (properly wrapped) in the refrigerator. But that's because I like the idea of having an old spare laptop for situations in which I wouldn't risk taking my good laptop. Can't do that without a battery, and not worth the expense of a new one, so... Ssscienccce (talk) 18:53, 27 July 2013 (UTC)


 * Don't get obsessed about making your battery last longer. Is your time input worth 6% more battery life? Go with Stephan's suggestion and plan to replace your battery or get a complete new laptop some time in the future. Do a write-off plan for it. OsmanRF34 (talk) 17:35, 27 July 2013 (UTC)


 * One thing - I don't recommend getting one of those cheap no-name laptop batteries. I bought one that had twice as many batteries as the standard one, but on a full charge it would last only about 20 minutes.  Just my experience, your mileage may very.  Bubba73 You talkin' to me? 00:08, 28 July 2013 (UTC)


 * You mean twice as many cells per battery ? That wouldn't be a good thing, putting it outside of the acceptable specs and possibly damaging the laptop. StuRat (talk) 00:30, 28 July 2013 (UTC)


 * Yes, twice as many cells per battery. The place I bought it from said that it was a replacement for the original and had twice the life (because of twice the batteries).  But besides that laptop, I've had similarly bad results with no-name camera batteries.  Bubba73 You talkin' to me? 00:47, 28 July 2013 (UTC)

A good Hard-drive recommendation?
I have a Standard Western-Digital Hard-drive (WDC WD500AAKX-001CA0). some Folders are loaded slow with it, for example a folder with Hundreds of photos \ videos. everything else in the computer seems to work fine. the slow loading of some folders (also in scrolling) is the only problem. Are there any Hard-drives today that could load such folders easily and without any sensible waiting time? THX Ben-Natan (talk) 12:48, 27 July 2013 (UTC)


 * According to Western Digital's information on that drive, it's a SATA-3 7200 rpm desktop-grade drive from c.2011. That's a perfectly decent drive that should be performing well. Assuming S.M.A.R.T. isn't showing any errors, that it's formatted to a modern filesystem like NTFS, and it somehow hasn't become fragmented (which is pretty rare with NTFS anyway) then the fault isn't in the drive, but rather the OS, its drivers or services, or some other ancillary software service like Explorer's thumbnail system. -- Finlay McWalterჷTalk 13:09, 27 July 2013 (UTC)


 * I agree - I don't think the HD is the problem. I have some folders that are slow for some reason - when I go into them, the bar has to go across the top, which takes quite a few seconds.  It is reindexing or something.  OTOH, you could get a Solid-state drive, but I don't know if that would help much with your problem.  Bubba73 You talkin' to me? 13:57, 27 July 2013 (UTC)


 * Look at the OS and installed shell extensions - it seems some rogue process is scanning through all your images / videos. (Oops, Finlay beat me to it.)
 * Don't trust M$ on NTFS - it gets fragmented just as easily as FAT.
 * And maybe you should get MyDefrag, a highly customizable free-of-charge defragger which actually places the folders close together to minimize access times - its Data Disk (Weekly) script will probably work just fine in your case. There is an "analyze only" script too, which lets you look at the actual fragmentation, too. You can even "point" at a cluster and it'll tell you what file it belongs to and how badly fragmented it is.
 * P.S. "weekly" is just a name. I'd run it no more than quarterly or monthly - unless I move huge amounts of data. - ¡Ouch! (hurt me / more pain) 16:00, 27 July 2013 (UTC)
 * However, NTFS limits the performance hit due to fragmentation somewhat. The allocation map is closer to the data, so the seeks are faster than they used to be on FAT partitions. Much of the fragmentation potential is not in the file system but in the way the operating system handles it. If it writes a file into the first free cluster, fragmentation is quite severe. If it tries to find a contiguous space, many small files will never fragment.

Hook...thingies...in keyboard keys
Underneath the larger keys (space, return etc) on my keyboard (and indeed, every keyboard I've removed the keys of), there is a small metal bar that hooks into place, as if to secure the key. But the keys are already secured in place, the same way every small key is - by the...pop-in thingy that...pops in. What's the point of the little metal bar? 94.5.20.77 (talk) 15:58, 27 July 2013 (UTC)
 * I mean, I just got done cleaning my keyboard, and I had about six bars missing, judging by the number of spaces for them on the keyboard base. But I didn't even use all of the bars I did have, since they don't seem to do anything and it's just more fiddly work. 94.5.20.77 (talk) 16:00, 27 July 2013 (UTC)


 * Wide keys, especially the spacebar, have a bar underneath them that distributes the load horizontally, so pushing the key anywhere on its length pushes down the switch in the centre. If you were to remove that bar, the spacebar would work okay if you pressed it in the middle, but if you pressed it on the side, instead of working normally, it would sag on that side, possibly not register at all, and wouldn't bounce back up in an acceptable way. You probably have one of these under at least right-ctrl and possibly left-ctrl, tab, and enter. -- Finlay McWalterჷTalk 16:28, 27 July 2013 (UTC)
 * Interesting. I have one large key that sticks (the + key on the numpad) if I push it far from the centre. I'll stick the bar underneath it and see if it helps. Thanks. 94.5.20.77 (talk) 17:10, 27 July 2013 (UTC)

Changing RAM/FSB clock of ADM box

 * Sorry, AMD box, not ADM.

Hi everyone.

I have an AMD running at 1066/533 MHz RAM/bus speed. What are the tolerances for changing the clock? I've read that too bad a deviation can even corrupt SATA or SCSI data transfer, so that would be pretty dire.

Can I fix the clock at 1000/500 without problems, or would that already be dangerous? (I only know of the 10% tolerances on voltage, not on any timing.)
 * 217.255.179.48 (talk) 16:55, 27 July 2013 (UTC)
 * Underclocking should make a system more stable, so I don't see a reason why not. I never had a problem when I lowered the clock speed. I assume you only change clock frequency, not multipliers or memory timing. Ssscienccce (talk) 19:28, 27 July 2013 (UTC)

Yes Ssscienccce, from a thermal point of view, it should be more stable. I'm asking because I have high 70s on CPU usage (when encoding video - and low 80s once in a blue moon), so, obviously, the last 20% of CPU power are "wasted" because the HDD can't transfer more data no matter how fast the CPU processed it. But I read some years ago that some systems which were overclocked and thermally stable lost some bits on I/O transfers. I'd guess that some bits got lost between ticks of the clock of the receiving end (the HDD or its controller in that case), and that caused corruption - is that correct?

Now if I underclock a system, is there a risk that one bit could be read twice?

Oh yes, I'm leaving memory timing as it is (RAS, CAS, etc.) - but if I could, I would keep RAM and FSB where they are and lower the CPU speed from 2333 to 2132, a straight multiple of both 533 and 1066. As I see it, the excess 201 MHz mostly go to waste anyway, and are there (as I understand it) only to avoid losing the MHz arms race, no matter how inconsequential that would be with respect to actual performance.

But there's no option to do that in the BIOS, and I downloaded two programs which don't even have the option to lower any speed settings. So i've basically given up on downloadable clock adjustment tools. Too soon?
 * 217.255.143.116 (talk) 06:55, 29 July 2013 (UTC) :(
 * I'm not sure you have a problem with your HDD transfer rate, reading and writing usually happens in blocks and is quite fast, so unless you are encoding more than 10-30MB per second, HD access shouldn't be a major factor. Most cpu's today have multiple cores, if that's the case (if you're talking total CPU usage) 70 to 80% wouldn't be bad. Not all applications can use more than 1 core, those that can aren't always optimized for it, and optimization has it's limits. For an application to use all cores for 100% is rare, because threads often need the results of others and will wait until that data becomes available.
 * Nowadays it's not so easy to identify bottlenecks, for example, on my Athlon II X4, when I increased the base clock from 200 to 250 but kept all other frequencies like northbridge, memory and CPU the same by changing the multipliers, benchmarks showed no change at all, but video encoding was much faster (from 121 sec to 89 sec for a testfile). Why, I don't know.
 * I've heard of data corruption when the SATA controller used the frequency of the FSB/northbridge/whatever it's equivalent is called now. Only affected specific boards and chipsets I think, and there were ways around it. But in general, lowering clock speeds is safe.
 * Most recent CPUs lower the clock rate automatically when idle, for individual cores, so depending on the specific processor, lowering the speed may have very little effect if energy consumption is your main concern. Ssscienccce (talk) 23:36, 29 July 2013 (UTC)

So, all frequencies (CPU, RAM, FSB) were the same, and still you got faster performance? That sounds indeed funny, even more so since the gains are more than the anticipated 25%.

Did you adjust the frequencies while in Windows? The gains look what I get when I re-encode the same file with different settings, because after the first run through a "small" file is in the cache. And only then do the readings go near 100% on the MOST loaded individual core.

Does anybody know how long it takes a core to power down? Is it like the HDD / VGA sleep intervals, or on the order of seconds?
 * 217.255.170.192 (talk) 06:00, 30 July 2013 (UTC)
 * I believe it takes more like milliseconds (if not microseconds). --Tardis (talk) 01:02, 31 July 2013 (UTC)

Wow. I had no eye, dear. So I gather that it doesn't matter much if the CPU runs for, say, 80 milliseconds at 2.3GHz, then "sleeps" for 20ms to let the HDD catch up, or 88ms at 2.1GHz, then sleep for 12ms, becuase the total clocks under load are still the same. Looks like I don't touch the timings then.

Thanks, Ssscienccce and Tardis. 217.255.177.143 (talk) 11:10, 31 July 2013 (UTC)

Simple image viewer for Linux with mapping function
I would like to find a Linux program with the functionality of the old "kuickshow" KDE app. Specifically, I would like to be able to browse directories of photos, in a manner similar to Nautilus, except that there should be a convenient way to translate geotags in the photo into a position on a map.

I know that digiKam has this functionality to some extent, but digiKam is a pain &mdash; it won't display your photos (or at least I haven't figured out how) unless you put it in one of its private databases, called an "album". I haven't figured out all the ramifications of that and don't really want to; all the organization I want is already contained in the directory tree structure.

Anyone have any leads on this? Kuickshow was just perfect (in addition it would also let you edit the "Comment" field in the EXIF data, which is a lovely feature), but unfortunately it used an obsolete imlib (or however you spell that) and there doesn't seem to be a replacement compatible with up-to-date Ubuntu. --Trovatore (talk) 20:04, 27 July 2013 (UTC)

Nvidia video card problem
Hi,

My friend has a problem with his computer.

His screen sometimes work and sometimes doesn't.

(When the screen 'doesn't work') The screen is black, but its lamp blinks.

More hardware and software info:

Windows 7 64bit Home Edition

The CPU is Intel

Screen LG ips235

Did anyone hear about similar problem?

Thanks.Exx8 (talk) 23:05, 27 July 2013 (UTC)


 * Have you checked for a loose connection of the cable to the monitor or to the PC ? Have you tried a different monitor and cable ?  You should take those steps before assuming it's the graphics card.  From what you describe it sounds like an intermittent electrical connection to the monitor, to me, so also check that the power cord to the monitor isn't loose.  There are two different types of power cord plugs, in the US at least, which are close enough to fit in the other slot, but never fit very well, so this might be the problem if the power cord doesn't fit snugly. StuRat (talk) 23:15, 27 July 2013 (UTC)


 * I also think it's probably a loose connection. The cords (at least the ones I have) can be tricky to plug in (on both ends); and to get a snug fit, the help of a screwdriver is often needed. Check to see if either end is loose. --Yellow1996 (talk) 17:46, 28 July 2013 (UTC)