Wikipedia:Reference desk/Archives/Computing/2013 April 4

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.

April 4

wireless pointing stick keyboard?

Does anybody make a wireless keyboard with a pointing stick (you know, those little widgets between the g and h and b keys to do mousy things like many laptops have?) I know there are a few wired pointing stick keyboards still made, but i might as well go for the moon, right? My fingers are too big for touchpads and use of a normal mouse always makes me wish i had three hands. Gzuckier (talk) 15:54, 4 April 2013 (UTC)

I'm pretty sure that the pointing sick is exclusive for laptops only because some people prefer the pointing stick over the mousepads. So yeah I'm sorry no I really doubt that they make them. One other thing, why do you not like normal ordinary mouses?RunningUranium (talk) 00:23, 5 April 2013 (UTC)

I found this, it's not exactly waht you want, but it could solve your problem. I hope it helps. — Preceding unsigned comment added by RunningUranium (talk • contribs) 09:03, 5 April 2013 (UTC)

On my IBM ThinkPad, it's called a TrackPoint, so you might also use that term when searching. StuRat (talk) 20:14, 6 April 2013 (UTC)

"automated response might cause WWIII"

A topic such as "automated response might cause WWIII" seems highly appropriate right now when one of China / Japan / US / North Korea / etc might trigger our next World War unintentionally. To the history of the triggers that caused wars, we now must add: computer causes. — Preceding unsigned comment added by 149.135.147.4 (talk) 20:36, 4 April 2013 (UTC)

The Reference Desk is intended for questions that can be answered with factual information. It is not a discussion forum. What you have written is not a question. Looie496 (talk) 20:45, 4 April 2013 (UTC)

It's not a new concern; people have been worried about it since the 1960s at least. In the United States, the fear of automated or accidental nuclear war has led to non-automated procedures. During the Cold War there were a number of close calls attributable to computer or detector errors (e.g. 1983 Soviet nuclear false alarm incident and several in the US in 1979-1980). Currently, accidental nuclear war just redirects to nuclear warfare, but arguably you could make a full article out of it, as it is a rich topic of both real history and popular culture. --Mr.98 (talk) 00:13, 5 April 2013 (UTC)

Windoes 8 CHKDSK

How can you run Chkdsk in Windows 8? When I try to do it at the command line, it says that I have to be in "elevated mode", what ever that is. Bubba73 ^{You talkin' to me?}

How do you run Command Prompt? If you click on an item to do so, try right-clicking said item and clicking "Run as Administrator" instead. If you don't, open up "cmd" in C:\Windows\System32 (at least there's where it is in Windows 7) by doing the same thing (right-click "Run as Administrator"). I'm not sure how 8's Start Menu operates (not sure how different from 7's), but you could also try typing "cmd" into the Start Menu's search box, right-clicking the "cmd.exe" (or "Command Prompt") that appears, then clicking "Run as Administrator". -- 143.85.199.242 (talk) 21:20, 4 April 2013 (UTC)

That worked, thanks

Resolved

. Bubba73 ^{You talkin' to me?} 21:38, 4 April 2013 (UTC)

Why didn't they just say "run as administrator" rather than the obscure "elevated mode"? Bubba73 ^{You talkin' to me?} 23:22, 5 April 2013 (UTC)

CHKDSK isn't really needed on modern hard drives. The S.M.A.R.T. system is pretty much standard and there are plenty of tools to query it. Used that quite a bit in my Dell tech support days back around 2002-ish, still seems to be the go-to. — The Hand That Feeds You:^Bite 20:02, 6 April 2013 (UTC)

Carbonite tech support is claiming that their software isn't working right on my computer because my HD is having way too many errors. I want to know if that is true. Do the SMART programs just monitor it, or does the system know to fix errors, etc? Bubba73 ^{You talkin' to me?} 21:50, 6 April 2013 (UTC)

SMART just reports information logged by the drive controller. If the "Reallocated Sectors Count" (SMART parameter 0x05) is more than a few, and particularly if it's increasing, the disk is failing and you should get a new one. If the disk is new then it should be under the manufacturer's warranty and you should start an RMA replacement dialog with them (they'll surely require that you download their own disk health check program, which will recover the same SMART data). -- Finlay McWalterჷTalk 22:07, 6 April 2013 (UTC)

Thanks, I'm running HD Tune on it right now. Bubba73 ^{You talkin' to me?} 22:44, 6 April 2013 (UTC)

"Elevated mode" is related to User Account Control. "Run as Administrator" sounds like it's for running the program as a different user (namely Administrator), but I think it actually just runs it elevated if the current user is an administrator. If you are an administrator, chkdsk is telling you the truth, while Explorer is kind of bending the truth. If you aren't, it's the other way around. -- BenRG 05:09, 7 April 2013 (UTC)

Partitioning a 3TB drive

Hi! I bought a 3TB HDD that contains 3 1-TB platters. Is there a performance advantage in partitioning the drive into 3 1-TB parts (or subsections thereof, e.g. 500 GB, 500GB; 1TB; 250GB, 750GB), such that no partition spans across more than one platter? Is partitioning software conscious of the drive's platter composition, so that if I specify to GParted to create 3TB partitions, it doesn't push a few sectors over to another platter because of adding some metadata? I'd ideally like to use 3 partitions: 1 exclusively for Windows, 1 for Ubuntu, and 1 to share between them, but I probably wouldn't make them each exactly 1TB large unless there was some benefit to that partitioning scheme. Thank you for any info or advice :) --el Aprel (^facta-_facienda) 21:34, 4 April 2013 (UTC)

There is a performance disadvantage in having a block or sequence of blocks on only one platter, and drives will typically organise things so that a single block (or logically adjacent LBAs) is distributed across a cylinder (meaning it's on all platters) because that means each head (one per surface) can read the data concurrently with its brothers. In practice you have no control and little knowledge of how a disk controller organises itself. Makers of server-grade storage devices, which use higher performance drives like Cheetah, may be able to extract from their disk-maker partners information (and, crucially, guarantees that information will remain valid) about disk organisation, but for the rest of us don't know, and really don't need to. -- Finlay McWalterჷTalk 21:49, 4 April 2013 (UTC)

Thank you for your response, Finlay. Is my understanding of your comments correct, that for a consumer-grade HDD without knowledge of the fine-grained internal details, there is no performance-preferable partitioning scheme, so one might as well partition the drive however one wants, and the internal controller of the HDD will optimize for performance as best as it can?--el Aprel (^facta-_facienda) 22:28, 4 April 2013 (UTC)

As long as you are using an up-to-date version of gparted, and use it as described here, it won't let you organise the partitions in a wrong way. -- Finlay McWalterჷTalk 22:57, 4 April 2013 (UTC)

For sequential reads, the outer edge of a platter is the fastest part of the disk, while the inner edge is the slowest. Some operating systems will reorganize the disk contents to put frequently-accessed data in the faster areas, and you can force this by partitioning and putting data in the "fast" partition. --Carnildo (talk) 02:42, 5 April 2013 (UTC)

To amplify what Carnildo is saying, with most consumer level drives, the fastest transfer rate and fastest seeks times are at the beginning of the disk and slowest at the end with a gradual slow down over the disk, regardless of how many platters are in use. See for example [1] [2]. However to amplify what Finlay McWalter is saying, the actual translation between performance for a certain purpose and raw performance metrics is quite complicated, see for example [3]. Even so, it's normally better to keep the most important (frequently accessed) stuff at the beginning of the disk. Nil Einne (talk) 03:25, 5 April 2013 (UTC)

More relevant to Windows(-specific filesystems), yes? ¦ Reisio (talk) 05:50, 5 April 2013 (UTC)

Not really. There may be other reasons for wanting multiple partitions in non Windows systems and with some systems it may be difficult to ensure stuff within a partition remains at the beginning of the partition because of different allocations schemes but the actual raw performance advantage doesn't change. While as I mentioned, the raw performance advantage doesn't always translate into better performance in a certain app (for such purposes you need to benchmark whatever app you're using), generally speaking it's more likely to help then hurt presuming the choice is arbitrary. There are also benchmarks on *nix systems if anyone chooses to disbelieve. (I actually wonder whether the raw metrics more closely correlate to app performance in *nix systems because it's likely the various optimisations and caching algorithms are tailored to Windows systems but I suspect it's probably not that simply since some of the workloads are not that dissimilar across operating systems.) But the fact that most systems on average benefit from improved transfer speeds and particularly improved seek times shouldn't exactly be surprising. Now as I hinted earlier, if you are forcing abnormal behaviour or there are other factors at play, then you have to consider whether it's really a good idea. But on the other hand, if you have a data partition and a system partition/s and you're trying to decide whether to put the system partition/s at the beginning before the data partition or at the end after the data partition then yes, generally speaking it will be better to put the system partition/s first to keep the frequently accessed/important stuff at the beginning (this may also reduce the risk of boot issues in some cases but I wouldn't consider that an important factor). In a case like this, where there is Windows, Ubuntu and data, the logical choice would be for the Windows and Ubuntu partitions to come first. The question of which one of Windows or Ubuntu will come first will generally depend on how much you expect to use each one. Depending on your plans, you could also consider a more complicated partitioning scheme but in such cases it's probably better to ask in a more specialised forum (there are complicating issues here since of course if you spread out the important data over more of the disk you increase the seek times). P.S. I wasn't able to find any benchmarks online in my quick search earlier but I currently own 2 Seagate 3TB, 1TB/platter drives. These vary from 197MB/s at the beginning to around 90MB/s at the end for reading with very similar performance at writing. This benchmark was done in Windows but with an unpartition disk using HD Tune and were repeated a few times including some times without much usage of the computer. They do start slow for some reason at about 175MB/s with a fairly linear rise until they reach the 197MB/s speed after a few GB. Seek times are not that dissimilar from for other HDs. The sort of benchmark raw metrics are fairly similar to pretty much every example I've seen for a consumer HD (only the rise at the beginning is a bit unusual). Nil Einne (talk) 07:23, 5 April 2013 (UTC)

Meant more in regard to file systems; how the more common Windows and Unix file systems tend to order things differently (some starting at the beginning of a disk[?], some not). Afraid I don’t know much about it. ¦ Reisio (talk) 09:14, 5 April 2013 (UTC)

According to this article, Linux spreads out files over the entire partition, leaving lots of space in between for file expansion, whereas old Windows FSs like FAT wrote files adjacent to one another, and still NTFS doesn't spread files apart adequately, leading to fragmentation. However, from the information above, it follows that NTFS and adjacently-ordered FSs would see an average-performance benefit over ext* on a not-so-full partition.--el Aprel (^facta-_facienda) 22:57, 5 April 2013 (UTC)

Assuming fragmentation isn't an issue.--el Aprel (^facta-_facienda) 23:06, 5 April 2013 (UTC)

Is maximizing usage of the outer edges of platters really as simple as partitioning the drive with gparted and then using the partition(s) closer to the front for the most I/O activity (e.g. OS, system)? Maybe I'm missing something (seems too easy). For one, I know whenever installing a Linux distro, the default/suggested partition scheme puts the main partition first and the swap following it. Wouldn't you want the swap first for faster I/O, especially on memory-parched systems? (Maybe this is a separate question, but I'm still trying to wrap my head around the [platter edges]-partitioning concept.)--el Aprel (^facta-_facienda) 22:57, 5 April 2013 (UTC)

The outer edges are only faster for sequential access. Swapping often involves extensive random access, in which case no part of the disk is substantially faster than any other part. --Carnildo (talk) 00:07, 6 April 2013 (UTC)

Win7 Laptop, Wrong Language

Resolved

My Power Settings window now has one power plan setting option in the wrong language. The rest are in English. Can anyone think why? KägeTorä - (影虎) (TALK) 22:51, 4 April 2013 (UTC)

Yes. Somebody screwed up. The person who set up the language support for that application accidentally put the Turkish version into the file of English translations. Either that or the application text was originally written in Turkish, and no English translation was provided for those items. Looie496 (talk) 02:42, 5 April 2013 (UTC)

So, basically, you are saying it is 'supposed' to be like this, as in, it was like that all along? Either nobody has noticed it, or Microsoft or HP (whichever is responsible) has not bothered to fix it? KägeTorä - (影虎) (TALK) 09:59, 5 April 2013 (UTC)

It would be HP, and of course it isn't supposed to be like that, but it's easy for things like that to slip by -- especially if you are, for example, using British English rather than the default form of English. Translation of computer messages is a mass-production job, and it's easy for errors to creep in. Regarding whether it has been like that all along, I haven't the slightest idea how old your computer is. Looie496 (talk) 15:39, 5 April 2013 (UTC)

Sorry, I will clarify (my dialect can be confusing): 'supposed to be like that' = 'was like that when I bought it (as I have done nothing to change these settings since)'. 'Like that all along' = ibid. :) Anyway, cheers. Interestingly, only one of my Win7 Home Premium laptops has this problem. My other one is Dell. KägeTorä - (影虎) (TALK) 16:04, 5 April 2013 (UTC)

Since the top line in your picture says "HP recommended", that power-saving utility is definitely a specific HP product. It may well have been programmed by an outsourced contractor in Turkey -- there is really no way to know. Looie496 (talk) 16:21, 5 April 2013 (UTC)

Actually, the power plan page is part of Windows 7, but the manufacturer can add settings to this page. Power plans: frequently asked questions says Windows provides plans called Balanced, Power saver, and High performance, and the computer manufacturer might provide additional power plans. This FAQ page also says you can create your own plans, see Change, create, or delete a power plan (scheme). User-created plans also appear on this page, but when I played around with it, the user-created plan only had a name and I couldn't find a way to add a second description line. Maybe it's available in a registry setting or I just didn't find the setting.

So based on this information, the Turkish text might have been created by the manufacturer or maybe by a user or a program with appropriate access to create a power plan. Make sure the Turkish plan isn't the active one, then across from the Turkish text, try clicking on the blue text to "Change plan settings", then see if there's an option to delete the plan. You can't delete Windows-provided plans or the active plan, so if it's not active and is deletable then it might be a user-created plan that someone or some program created. --Bavi H (talk) 03:21, 6 April 2013 (UTC)

Your picture says the text is Turkish, so I guess you already tried translating it? Google Translate says it's "Gaming Mode Power Plan". Perhaps a game you installed by a Turkish developer also installed this power plan? --Bavi H (talk) 03:54, 6 April 2013 (UTC)

In fact, I had been using the Gaming Mode power plan a while back - I remember now that it was English (I didn't bother translating the Turkish one, so I didn't know), so, yes, it looks like it has been replaced by something. Interestingly, my computer has sounded recently like it is overclocking. This is how I came to look at the power plans in the first place, and am thinking this is very suspect. I cannot believe it is just a case of having got past the proofreaders, because I do that for a living. I will have to look through my utilities (the ones which have power-saving options) and then my games. I was looking forward to a nice weekend off...... Anyway, thanks! KägeTorä - (影虎) (TALK) 08:12, 6 April 2013 (UTC)

Just as I suspected - it was installed by IObit - I use their Advanced System Care, and there used to be a separate program called Game Booster (directly translated into Turkish: Oyun Güçlendirici. Now it's included in the package when you get ASC. Problem sorted. Thanks. KägeTorä - (影虎) (TALK) 08:18, 6 April 2013 (UTC)