Computing desk
< June 18	<< May \| June \| Jul >>	June 20 >

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 19[edit]

IPCONFIG Question[edit]

I am using Windows 7. If I open a command prompt (so-called DOS box), and issue the ipconfig command, it displays my IP address as 10.22.163.105. It doesn't display a 71.*.*.* IP address. However, if I go to my Wikipedia talk page, log out, and post a test edit to my talk page, it attributes the edit to 71.191.8.90. 10.22.163.105 is a non-routable address. Is there a way to display the IP address that the world sees (short of editing Wikipedia logged out, which is usually deprecated anyway)? Robert McClenon (talk) 01:09, 19 June 2014 (UTC)[reply]

The public IP address doesn't belong to the Windows machine but to some router between it and the public Internet, so you have to connect to that router or to some host on the other side to find out what it is. There are Web sites that will echo your public IP back to you, such as whatismyip.com (not an endorsement). You could also edit a Wikipedia page logged out, insert ~~~, press "Show preview" to see the IP, and then exit without saving the change. -- BenRG (talk) 04:01, 19 June 2014 (UTC)[reply]

Thank you. 10.22.163.105 is, as you say, the non-routable address assigned to my computer by my Linksys router. IPCONFIG shows that the gateway address is 10.22.163.240, which is the non-routable address of the router itself. I can log on to the router. Is there a way that I can get the router to show me the address assigned to me by Verizon? Robert McClenon (talk) 20:11, 19 June 2014 (UTC)[reply]

It will be shown somewhere in the web interface (exactly where depends on the router). If it's running Linux and you can get a shell prompt, you can run ifconfig, which is similar to Windows' ipconfig. -- BenRG (talk) 01:55, 20 June 2014 (UTC)[reply]

@Robert McClenon: See network address translation.--Jasper Deng (talk) 20:22, 20 June 2014 (UTC)[reply]

Percent encoding[edit]

So I'm playing around with percent encoding, and I find that https://en.wikipedia.org/wiki/Guaran%C3%AB produces a page Guaranë. However, strip everything but the last character, resulting in https://en.wikipedia.org/wiki/%AB, and you get «, a typographic figure known as a guillemet. Why doesn't %AB produce the same character both times? Nyttend (talk) 03:52, 19 June 2014 (UTC)[reply]

URLs are supposed to be UTF-8 encoded, and C3 AB is the UTF-8 encoding of U+00EB, "Latin small letter e with diaeresis". AB by itself is not valid UTF-8, and some software somewhere tried to guess what it was supposed to mean. It guessed Latin-1 (or more likely Windows-1252), in which AB stands for U+00AB, "Left-pointing double angle quotation mark". -- BenRG (talk) 04:08, 19 June 2014 (UTC)[reply]

128 GB USB 3.0 flash drives[edit]

Can anyone figure out why the SanDisk 128 GB USB 3.0 flash drive listed at the top of this page is 3x more expensive than the other 128 GB USB 3.0 flash drives? Based by the number of and substance of reviews, it apparently it's apparently a very popular drive . I don't understand why anyone would pay $143 for a 128GB drive when they can pay ~$50. What am I missing? A Quest For Knowledge (talk) 13:32, 19 June 2014 (UTC)[reply]

Because (like other high-end flash drives) it scores several times the read and write performance of cheaper drives in benchmarks. Sandisk and other large flash manufacturers make a range of flash products with a radically different range of speeds, and naturally charge much more for the faster ones. It's my understanding that the high-speed products feature multiple parallel blocks with parallel write and erase circuitry (essentially multiple devices on the same silicon), whereas the cheaper bulk ones have only a single channel. This makes most sense for uses where speed is vital, like SD cards in cameras; I'm not sure there are so many use cases where higher performance is worth paying lots more for on USB. -- Finlay McWalterᚠTalk 13:50, 19 June 2014 (UTC)[reply]

By way of example, Sandisk's offerings for SD cards is here, where you can compare the performance of their most expensive kinds to the cheapest ones (well, they don't even list speeds for the cheap "standard" ones). See also Secure Digital#Speed class rating. -- Finlay McWalterᚠTalk 13:57, 19 June 2014 (UTC)[reply]

@Finlay McWalter: Thanks, that explains it. I'm planning to use the drive so I can play a copy of my entire music collection in my car. My 'master' music collection is kept on a network drive on my home network. My plan is copy the music from the network drive to the USB drive. Read speed shouldn't be an issue when playing a song. Writing might be an issue when backing up my music collection to the USB drive, but as long as I can find a utility that only copies the delta's, it should be quick enough after the initial backup. A Quest For Knowledge (talk) 14:36, 19 June 2014 (UTC)[reply]

rsync, or Robocopy on Windows. -- Finlay McWalterᚠTalk 14:46, 19 June 2014 (UTC)[reply]

@Finlay McWalter: Do you have any opinion on RichCopy? A Quest For Knowledge (talk) 16:05, 19 June 2014 (UTC)[reply]

No, I don't know anything about it. I use rsync exclusively. -- Finlay McWalterᚠTalk 16:23, 19 June 2014 (UTC)[reply]

Cheap flash memory uses as many as 4 bits per cell (16 discrete voltage levels), impairing write durability and reliability in general.[1] If you've got another copy of the data though, maybe this is tolerable. Good memory these days uses 2 bits per cell. 1 bit per cell (SLC, single level cell) is almost extinct. 173.228.123.145 (talk) 15:21, 19 June 2014 (UTC)[reply]

The price differential is a possible indication these could be among the cheap fake capacity flash drives coming out of China. Such drives have apparent capacities 32GB up to 1TB (sometimes more), but are actually low capacity drives (typically 4, 8, or 16 GB) that have been hacked so that they appear to have a much higher inflated capacity. Files will write without apparent errors, will look okay in folders, but any data written past the true capacity will be quietly lost and irretrievably corrupt when read back. Most of the fakes involve an assortment of generic and novelty styles, but convincing counterfeits of brand name drives such as Kingston, ADATA, SanDisk, and Toshiba also exist. Flash memory cards are also involved. If you purchase one, I recommend that you run a thorough test of the true capacity before trusting any valuable data to it. Fill it with data and then verify that you can actually read that data back intact by opening/viewing/playing the files, especially those at the end. The h2testw tool created by a German tech magazine "c't" (published by heise.de) is one of the more recommended tools for testing for fake capacity flash.

At the time I write this, the SanDisk drive you mention no longer appears. However, it's worth noting that several unusually low priced drives at the top of the list are sold by a third party rather than NewEgg, and the styles shown are among the common generics often involved in the fake capacity scam. Such fakes are common on eBay and many Chinese wholesale sites but also appear in other marketplaces such as Amazon and now apparently NewEgg. This scam has been ongoing for many years and shows little sign of letting up. Some resellers might not be aware that they are selling fakes, having themselves been scammed by the wholesalers. Google "SOS Fake Flash" for more info. 24.254.222.118 (talk) 17:45, 21 June 2014 (UTC)[reply]

Correction: When I first clicked the OP's link above, the results were sorted by lowest price and the top two items were generic drives offered for $13.99 and $14.99. Those are the ones I suspect to be fakes. At the other end of the spectrum, I agree with Finlay McWalter's comments. Speed, reliability (including error correction), encryption, and other features affect price. 24.254.222.118 (talk) 20:39, 21 June 2014 (UTC)[reply]

IP address with new format[edit]

Thread retitled from "IP Address".

I'm a recent changes patroller, and I keep noticing a different kind of IP address sometimes, they look like this: 2001:5C0:1000:A:0:0:0:A67. What does this kind of address mean or what is it used for, and how is this different from a "standard" IP, like 106.414.1.1? Ben^Yes? 21:04, 19 June 2014 (UTC)[reply]

The former is IPv6, the latter is IPv4. This question was asked by an anon user about their own IP address a few days ago. CS Miller (talk) 21:11, 19 June 2014 (UTC)[reply]

OK, yeah I've heard of them, I just don't know what they were. Thanks! Ben^Yes? 21:17, 19 June 2014 (UTC)[reply]

For an explanation of why IPv6 is being implemented, see IPv4 address exhaustion. There are "only" 2**32 or 4 "binary billion" IPv4 addresses available, some of which have special purposes. There are 2**128 IPv6 addresses available, and 2**128 is BIG. Robert McClenon (talk) 21:21, 19 June 2014 (UTC)[reply]

(ec)That is an IPv6 address. The "normal" IP protocol is more than 30 years old and uses 32 bit addresses, normally written as 4 8-bit-bytes in decimal notation, separated by dots. IPv6 is the latest standard, and uses 128 bit addresses, written as 8 16-bit numbers in hexadecimal notation, separated by colons. One reason for IPv6 is that we have been running out of IPv4 addresses for a while now. --Stephan Schulz (talk) 21:17, 19 June 2014 (UTC)[reply]

I've actually never noticed IPv6 addresses until just recently, and at the time I thought it was just a funky username. :/ Kurtis ^(talk) 01:13, 20 June 2014 (UTC)[reply]

So did at least one Admin, who blocked it — they thought it was a gibberish username (a blockable offense). Facepalm ensued. - ¡Ouch! (^{hurt me} / _{more pain}) 06:44, 23 June 2014 (UTC)[reply]

IPv6 is more than 15 years old itself... -- BenRG (talk) 01:56, 20 June 2014 (UTC)[reply]

IPv6 was enabled for all of Wikimedia's properties two years ago. At that time, if I remember rightly, we were seeing around 1% of IPv6 visitors. I wrote a little IRC bot today to monitor the en.wikipedia recent-changes feed, and of the last 1000 ip edits, 21 came from IPv6 addresses (but bear in mind that those numbers are so low that a single person editing or not could easily double or half the ratio). So its use here appears to have roughly doubled; I'd expect that it'll continue to increase over the next few years, as ISPs roll out IPv6 capable routers to households. -- Finlay McWalterᚠTalk 18:30, 20 June 2014 (UTC)[reply]

Google currently gives ~3.25-3.30%, with a weekly peak of ~3.7-3.8% on Saturdays, at their statistics page. To me at least, it seems to be resembling logistic growth if you look at the weekly peaks; it's pretty much exponential for the past 3 years and the present (more evidence: this and this are both consistent with logistic growth).--Jasper Deng (talk) 05:04, 21 June 2014 (UTC)[reply]

@Finlay McWalter: Also, I'm not sure I'm allowed to do a simple z-test here, but doing so with your result (2.1%) gives a (95%) confidence interval of 1.2% to 3.0%, so I'd probably say 3% is a reasonable estimate of current IPv6 usage given Google's statistics and yours. Then again, what we have here is original research, so we'll just have to wait for how it plays out. Hopefully all admins will know enough about it by the time it becomes as common to block IPv6 addresses as IPv4.--Jasper Deng (talk) 05:15, 21 June 2014 (UTC)[reply]

I ran a longer run overnight, with about 10 times the sample size, and it came out to 3.2% IPv6. -- Finlay McWalterᚠTalk 11:32, 21 June 2014 (UTC)[reply]

I am revising the section heading from IP Address to IP address with new format, to distinguish this section from #IP Address (under "June 17") on this page. This is in harmony with WP:TPOC, point 12 (Section headings). The new heading facilitates recognition of the topic in links and watchlists and tables of contents.

—Wavelength (talk) 19:02, 20 June 2014 (UTC)[reply]