Talk:USB flash drive/Archive 3

Over 29 titles redirected here
17-March-2008: The following redirect to "USB flash drive":
 * USB flash drives (plural), pen drive, pen drives, pendrive, pendrives, thumb drive, thumb drives, thumbdrive, thumbdrives, jump drives (not "jump drive"), JumpDrive, key drive, key drives, keydrive, keydrives, DiskOnKey, pocket drive, USB flash memory, USB key, Flashdisk, USB stick, USB memory stick, USB flash memory drive, USB hard drive, flash key, memory key, Clip drive, disk on key, usb memory card, usb memory, etc.

All those name variations were created due to the importance of the article, as some consider USB flash drives as making floppies obsolete. Plus, all those names have thousands or millions of search-engine hits, with "pocket drive" having 182,000 Google hits, "clip drive" having 36,500 hits. A less common subject would not warrant 29 aliases. -Wikid77 (talk) 06:07, 17 March 2008 (UTC)
 * looks good but please do not add alternative names to the main article. As you list above there are at least a dozen common alternatives and at least 30 other terms that are used and if one is added then proponents of the others want them listed to. Rather than ended up with a giant list (which is what we had at one point See here ) - SimonLyall (talk) 04:53, 21 March 2008 (UTC)

I don't see a problem with listing at least the more common alternative names, or indeed all of them. It seems perverse to say that because there are many synonyms then none of them should be listed - which seems tantamount to suggesting that the alternative names aren't used. Surely the more alternative names there are, the more important it is to list them! (So that readers understand that terms they may be familiar with, e.g. pen drive, refer to the same thing as a USB flash drive.) How about just having a heading 'Alternative names' and list them there?

Another name incidentally which used to be used is 'data stick' - I've added a redirect from it. Ben Finn (talk) 16:52, 25 November 2008 (UTC)


 * Please read the old version of the article here, we ended up with a huge number of names, many of which were very localised and some were fake ( "dick drive", etc ). We will need multiple references for each of them which in turn will clutter up the article. Other articles do the same like Automobile, it has over a dozen terms redirected to it ( like car and Motor Car ) but doesn't list the hundreds of others terms for car on the page - SimonLyall (talk) 19:21, 25 November 2008 (UTC)


 * Yes but the difference with "USB Flash Drive" as opposed to cars, is that everyone knows the term car and hardly anyone uses any other term, but with flash drives there are a LOT of terms that are in common use as alternatives. It might behoove us to list say 3 or 4 of the most common alternatives. Floodo1 (talk) 04:14, 14 February 2009 (UTC)


 * No. "Auto" is frequently used in certain localities.  "Motor" is also frequently used at least here in the UK.  There may be other similarly regional terms that slip my mind at this point.  There are many problems with long lists of terms making them impossible to verify and maintain:
 * is it notable enough? There was someone a while back wanting to use the term "fishstick" on the basis that a single teacher used it.
 * is it a brand name? Is a vacuum cleaner a Hoover when it isn't made by Hoover?  Is a washing machine made by Hoover still a Hoover?
 * is it accurate? Some terms are not.  USB hard drive generally refers to a mechanical rather than flash storage device.  These are not memory cards of any description.
 * is it ambiguous? USB key can refer to an authentication device.  Flash disk is more general than USB flash drive - I have some PCMCIA flash disks right here.
 * In summary, a long list of alternative is a very bad idea. The fact that someone somewhere uses a particular nomenclature does not make that usage notable or even correct.  We are already bending over backwards with all those redirects.  If we considered everyone's own pet name here as well it would end up that you wouldn't see the wood for the trees. CrispMuncher (talk) 21:21, 14 February 2009 (UTC)
 * Agreed, a long list is a bad idea. Which is why I suggested just putting 3 or 4 of the most popular alternatives. Who knows how someone got to the USB Flash Drive page, maybe they do it WITHOUT using a redirect and maybe it will help them to see the most popular other names...."thumbdrive" is definatley heavily used around here (to cite one). In any case there seems to be a consensus that a long list is a bad idea :) Floodo1 (talk) 01:06, 15 February 2009 (UTC)
 * The problem is that if we include 3-4 terms then people will start adding just one more that everybody *I* know uses and we will then have to have a criteria as to which ones get included and where the cutoff is and so on. I'm not sure how listing alternatives names helps somebody who has found the page anyway, if they type in "jumpdrive" and get here we still have the information they want. - SimonLyall (talk) 01:49, 15 February 2009 (UTC)


 * Seems like list of synonyms is more apropos of Wiktionary, USB flash drive.Wikipedia is not a dictionary. Until have a WP:RS that indicates what the 3-4 most common terms in worldwide English usage are, not much point in discussing inclusion here. Zodon (talk) 07:46, 15 February 2009 (UTC)

What does "speed 120X" mean?
I think the article needs a more complete explanation about the speeds of current drives. Also in the sentence "The highest current overall file transfer speeds are about 10-60 MB/s", 10-60 MB/s is a too wide range... --Lefter 12:28, 21 March 2008 (UTC)


 * The problem is that the claimed maximum speed of USB 2.0 is 54 mbits/sec, but it's pratically impossible to cap that out. It's more realistic to to see 10-15. 66.215.20.28 (talk) 23:20, 28 May 2008 (UTC)

Thanks! But I meant the max. speed of the flash chips itself not the USB 2.0 port. The article states that with most drives the speed of the flash chip is much lower than the max. speed of USB 2.0. Also I have met on the Internet specifications like "speed 120X" but I cannot find anywhere what speed 1X equals. Is it Floppy, USB, CD-ROM, or DVD speed? I think it would be nice if someone adds this to the article. --Lefter 12:33, 3 June 2008 (UTC) —Preceding unsigned comment added by Lefter (talk • contribs)


 * It's based on the CD-ROM speed rating, where 1X = 150 kB/s. -- Coneslayer (talk) 13:00, 3 June 2008 (UTC)

New Picture
Hi, I made a photo of my own Toshiba TransMemory Flash drive and feel like adding it on here to give the photos a consistent quality to them. I'll replace the the current photo of the black Flash Drive next to a ruler, which is a bit unaesthetic and rustic in my opinion, (and the first picture already has a ruler for comparison) with my own. Tell me what you think, if you want to replace it, feel free to bring it up here and go ahead. Thanks

Gamer112 (talk) 18:54, 21 April 2008 (UTC)

JFFS/YAFFS
Spoon recently added a line stating that JFFS2 and YAFFS2 are useful on a UFD. However, my understanding is that these systems are mainly useful on flash memory that can be directly accessed (e.g. SmartMedia through a raw reader), by assisting with wear levelling, cacheing, etc. However, a UFD abstracts the memory structure, and its controller provides similar intermediate features, such that these filesystems shouldn't make a significant difference.

Unless anyone objects (and has persuasive evidence that JFFS2 or YAFFS2 provide advantages over other filesystems on a UFD), I'm going to remove the sentence. EJSawyer (talk) 07:00, 9 May 2008 (UTC)

Running web applications
I added the content below about running web applications, which I think is useful and significant information (and referenced), but is was removed by this edit, with an irrelevant edit summary about ¨running mail servers¨. Rather then a preemptive revert, I think this is worthy of discussion here. If mail servers are worth mentioning also, then they can be too. Any comment from other editors on this? Peter Campbell 13:18, 8 June 2008 (UTC)
 * The way I see it is that Flash drives are several gigabytes these days and most computers will boot off them. So you can pretty easily create a flash drive that any machine will boot that will do anything a normal computer will. Thus I don't really think there is a need for individual sections on Web server/mail server/Video game simulator/media server/tourist kiosk/inventory controller/tax filer/voting machine "on a bootable USB flash drive". Perhaps a series of examples - SimonLyall (talk) 20:21, 8 June 2008 (UTC)
 * Yes, that is the way you see it, but many are unaware of the versatility of flash drives and what you can actually run on them, so I think this content adds value and depth to the article. By all means expand the list if other compelling applications and examples are around.  I don't think many would be too interesed in running a mail server, but quite a few people may be interested in having a local blog or wiki for testing or personal use.  Peter Campbell 13:45, 9 June 2008 (UTC)
 * I have re-added the content below as it is relevant to the article and I think of interest to many. Peter Campbell 13:58, 19 June 2008 (UTC)



Running web applications
A web server such as XAMPP can be loaded onto a USB drive and used to run Apache HTTP Server, MySQL database, and interpreters for scripts written in the PHP and Perl programming languages. Such a platform can then run web applications loaded onto the same USB drive such as:
 * Mediawiki - the software that powers Wikipedia
 * Drupal - an open source content management system
 * I reverted this addition, per SimonLyall's argument. And regarding the statement "quite a few people may be interested in having a local blog or wiki for testing or personal use" I would add that even more people who want to learn about USB are not at all interested in this. Yes, I have a local lamp installation on my computer and it is simply installed on the hard disk. I think web developers (the kind that installs PHP scripts) represent less than 1 in 1000 of the people who own a computer, and of those only a minute fraction would be interested in running it from a USB stick. Moreover, Wikipedia is not a how-to manual. If running software from USB is to be discussed, the logical thing would be to discuss (for example) how more and more people are running software from USB while switching from one physical computer to the other, rather than carrying a laptop around. This should then be properly referenced. Han-Kwang (t) 17:10, 19 June 2008 (UTC)
 * Agree, running server software on a desktop by booting to USB is probably not a common use ( virtualisation is probably a more common way of doing this). I also think the text that was added was to much about giving examples of applications rather than the concept. Mentioning XAMPP, Drupal and Mediawiki doesn't really enhance the text. - SimonLyall (talk) 11:15, 20 June 2008 (UTC)


 * Why revert without discussion first? Many people I have spoken to are fascinated that you can run these type of applications on a USB key.  Just because "1 in 1000" (your guestimate) might actually do it does not detract from the interest in this.  Dicussing running other software across multiple machines is fine - but this has nothing to do with the content you just removed.  If it is worth adding, then add it too.  This is not a valid reason for removing content.
 * This area seems to be covered in the Live USB article. I don't think anything beyond a link to that is needed and we already have that. - SimonLyall (talk) 11:26, 21 June 2008 (UTC)


 * The Live USB link is good, but it covers a separate topic. Running a bootable operating system from a USB drive is different from running the AMP stack and web server applications - for these you don't need to boot from the USB drive. You just start Apache and MySOL then access the web appications via a URL.  To restate - many people find this topic of great interest and are unaware of this capability - which is why I think it should at least be mentioned in this article.  Peter Campbell 11:01, 22 June 2008 (UTC)


 * Ah, I see now. So you are just running an executable file sitting in a folder and the folder happens to be a USB flash drive. See my comment further down - SimonLyall (talk) 11:18, 23 June 2008 (UTC)


 * It is a lot more than "just running and exectable file". The XAMPP or WOS exe runs a program which then starts the web server and MySQL, both running from the USB drive.  A web browser can then access to web application (e.g. Mediawiki, Drupal, Joomla etc) via URL.  I think from the apparent confusion about his functionality in this discussion it is well worth mentioning in the article, and other relevant articles such List of portable software Peter Campbell 04:56, 25 June 2008 (UTC)
 * The problem is it isn't flash drive specific. The software could be easily downloaded as a zip or something and then unpacked and run. Sure it's cute software but it is not tied to USB flash drives. Lets put it this way, you could add the same blurb about it to the articles about CDROMs, external Hard drives, Network Mounts, the local hard drive of the computer, etc. - SimonLyall (talk) 07:51, 25 June 2008 (UTC)


 * This is not a problem, it is in fact a feature of portable software. I really don't understand your issue; software can often be "downloaded and run". If other people want to add similar content to the other hardware/media articles then they are free to do so; but we are talking about USB drives here. Running significant web applications such as Mediawiki (and a web server environment) is cute and quite significant, so I think it is notable and therefore worth mentioning. I think readers should be made aware of this possibility in the article, although no details are required. 06:21, 27 June 2008 (UTC)


 * I don't think you are going to change the opinions and create a consensus by repeating your arguments over and over. Han-Kwang (t) 08:10, 27 June 2008 (UTC)


 * (unindent) I read things on Wikipedia on a daily basis that I was unaware of before and that I find of great interest. That doesn't mean that every article needs to discuss everything I might find interesting. If XAMPP satisfies the requirements for notability, you could start a new article about it. If it is not notable, then it is also questionable whether it deserves a separate section in another article to which it is only (IMO) loosely related. Han-Kwang (t) 12:01, 22 June 2008 (UTC)


 * XAMPP already has an article, so it meets notability criteria. It is provided as an example only, there are other AMP stacks such as WOS. A Personal wiki article exists too, which is an example of wikis that can be run on USB drives - which is not currently apparent in this article. The last edit you reverted was not a separate section, it was a paragraph addition to the existing Application Carriers section. The topic meets notability and (with interwik links) criteria, and is referenced too. Why should you be the arbiter of what information is provided? Wikipedia is for capturing and storing information, not censoring it. Peter Campbell 00:15, 23 June 2008 (UTC)


 * Let's go for the simple solution which would seem to cover everything everyone seems to want - why not linking to List of portable software with a suitable comment?
 * That would give a resonably comprehensive listing of the types of software that can be used, without being overly specific, or duplicating information? Nuwewsco (talk) 07:33, 23 June 2008 (UTC)


 * Peter, let's not make it personal, OK? If there is a consensus for inclusion, I will (grudgingly) accept it, but so far that doesn't seem to be the case. I like Nuwesco's suggestion. The length of List of portable software also shows why it is not a good idea to try to include all kinds of USB-storage-capable programs in this article. Han-Kwang (t) 08:17, 23 June 2008 (UTC)


 * Good suggestion Nuwewsco, I have added the content in question to the List of portable software. Nothing personal Han-Kwang - I do suggest discussing this type of content prior to reverting it.  Peter Campbell 04:12, 17 September 2008 (UTC)

Testing for consensus on inclusion of web app content
Testing for consensus on the inclusion of the above content, to broaden the poll from 1 for and 2 against: Can more editors indicate whether they support or oppose including this content in the article. Peter Campbell 11:10, 21 June 2008 (UTC)
 * oppose - It's just an executable ( and some other fles) that is sitting in a folder. The folder can be on a USB flash drive, CD, network mount, floppy, normal hard drive etc. Not USB flash drive specific enough for more than a pointer. Perhaps something like " Many types of portable software can be stored, distributed and run from USB Flash Drives ", maybe something about autorun - SimonLyall (talk) 11:18, 23 June 2008 (UTC)
 * The ability to run web apps is NOT on par with the ability to boot a live operating system, which itself is able to run web apps. This is akin to including in an article about cars an in-depth commentary about types of car racing. Just because it is something that can be done with flash drives does not neccesitate in-depth discussion thereof. A brief mention of the possibility should be sufficient for this article, which is how all of the other uses have been handled. Floodo1 (talk) 22:16, 14 July 2008 (UTC)


 * support - this content expands the information provided in the "Application carrier" section which already mentions U3 apps but nothing else. The ability to run web apps is on par with the ability to boot a live operating system,. It also provides a means for testing and evaluating this type of software without impacting the host computer.  If more interesting applications are found they can be added too.  Peter Campbell 23:26, 23 June 2008 (UTC)

Fragmentation
We need to careful with these assertions that fragmentation is of no consequence on a flash drive - it isn't true. Contiguous files on a flash drive can be transferred in bulk with a limited number of commands. Fragmented files on the other hand require extra commands for each non-contiguous section which need issuing repeatedly as each previous request is completed. The protocol overhead imposed is nothing like the penalty imposed by a seek on a real disk but to say that there is no penalty is overstating things. CrispMuncher (talk) 21:56, 28 August 2008 (UTC)
 * Moreover, from what I understand from reading Flash memory, erase-write operations on NAND flash need to be done on a per-block basis, blocks being much larger than typical 1 or 4 kB allocation units of the file system, which could be a pretty hefty overhead on a fragmented file. Han-Kwang (t) 21:57, 29 August 2008 (UTC)
 * Interesting, and counter to my understanding. I've certainly seen it said often enough that defragmenting flash drives has no advantages, and does have disadvantages; doesn't necessarily mean it's correct though. As it stands the article would still be incorrect if things are as you say, as it says elsewhere that defragmentation is undesirable. I did add a reference to support this when I reverted your edit, but am happy to be corrected. Whatever the consensus is, we need a reference (at the moment the only reference is against defragging); and it would be useful to have some numerical idea of the speed hit of fragmented USB drives, and some idea of the reduction in life due to many writes. You'll infer that I'm dubious that defragmentation is signifcantly advantageous, but I'm always willing to learn. There was an unquestionably wrong statement that electromagnetic fields affect hard drives which I've corrected again. Thanks for your input. Pol098 (talk) 23:23, 29 August 2008 (UTC)


 * There are two separate issues there. One is whether flash drives are affected by fragmentation, the other is whether defragmentation is worth doing.  Since most defragmentation utilities are made assuming a hard disk medium, they do not tend to take into account the characteristics of a flash drive.  If one was going to defragment a flash drive, one would want to write each block at most once; whereas with a hard disk that is not an important consideration, so hard disk defragmenters are not suitable to defragmenting a flash drive.  Whether defragmentation is worth doing is a cost/benefit tradeoff; whether fragmentation has an effect is just asking if it has a cost. Zodon (talk) 05:45, 30 August 2008 (UTC)


 * Well, I did some experiments and put the results online on http://www.lagom.nl/misc/flash_fragmentation.html . Maybe this is suitable as a reference? I won't put original research on Wikipedia, but I can conduct it outside Wikipedia and let someone else use it as a verifiable source :-) Han-Kwang (t) 10:17, 30 August 2008 (UTC)

Very interesting reference. Hard information on this topic is hard to find. Another much-touted point is wear; what seems to be missing is empirical information of the lifetime of real flash memory devices in terms of, say, FAT32 file operations. Han-Kwang's reference says "A standard hard disk defragmentation tool might not be the ideal tool to defragment a flash drive, since it is not aware of the allocation block size inside the flash drive. This may slow down the defragmentation procedure". This implies that a HD defragmenter will do a good job, but might take a long time. Possibly frequent defragmentation in place this way, requiring huge numbers of writes, would significantly shorten the effective life of a device (it might wear out before becoming obsolete); perhaps the best way (perfect defragmentation with minimal wear) to defragment flash devices is by backing up (to a hard drive) and restoring on a file-by-file, not sector, basis? In the light of what's been said here, the "File systems" section could do with amending. Pol098 (talk) 11:54, 30 August 2008 (UTC)

A quick back-of-the-envelope calculation based on the reference quoted above (write 15MB with 4000 accesses takes 492"; write 486MB with 1 access, 291") could be interpreted as a write speed of 1.67MB/s and a "write access time" of 120ms. Read speed 7.4MB/s, "read access time" 0.6ms. This assumes the large file is written as a single block; a block size of, say, 128kB (which gives the most sensible answers) would give 9.3MB/s and 63ms "write access time". FWIW (not too much). Pol098 (talk) 12:59, 30 August 2008 (UTC)


 * Yes, my mistake: 120 ms rather than 8 ms. But I'll do some more tests, it seems that the numbers are not consistent when I don't fill the whole device; maybe wear-leveling is playing a role as well. Han-Kwang (t) 17:36, 31 August 2008 (UTC)

Comparison with other portable storage
Hang on - this should be an article in it's own right ("Comparison of portable storage", or similar), not tucked away here! Nuwewsco (talk) 18:21, 31 August 2008 (UTC)

Permanent erasure?
When I erase a file from my flash drive, I get the message: "Do you want to permanently erase this file?" If no trace is left on the flash drive, then it would be the preferred medium for privacy--either my privacy or some terrorist's privacy. Is there, in fact, no way to recover an erased file from a flash drive? Dynzmoar (talk) 13:03, 4 September 2008 (UTC)

There are many levels of answering your question. On the simple level, the question would be, are the bits still there? With most computer tech, the answer would be YES, the bits are still there after "deleting" a file. The only way a user can be quite confident that the bits are gone is to delete the file and then fill up the device with other data. (Even then, assume that file name debris may persist in the directories.)

At the paranoid level, the question is, if the bits have been overwritten, can the data be recovered? This is mostly a theoretical question. It would usually take very special skills and equipment, costing thousands or millions of dollars, to try to get back the information. Yet this can be done, at least with hard drives. This is why there is special software to overwrite the data on hard drives over and over and over... Similar issues may apply to flash drive tech. (Ghosts of previous bits.) But if people who have these kinds of resources are after your data, there are easier ways for them to steal it.

If you need to be damn sure information is erased, physically destroy the medium. If you want practical deletion and re-use, overwrite the bits a number of times with other mixed data. (The only way to be sure information stays private is to never type it on any keyboard.) -69.87.204.110 (talk) 11:55, 16 September 2008 (UTC)


 * In view of wear leveling and possible re-mappings of data blocks by the drive controller, how do you erase/overwrite the bits? Seems like it may require different techniques (or at least different software) than those used on a hard disk. Zodon (talk) 20:23, 16 September 2008 (UTC)
 * The same applies to a lesser extent to the bad sectro remapping on hard drives. The blunt answer is if you need to protect data against casual snooping by script kiddies etc then a simple overwrite of the sectors you can access is probablly sufficiant, if you need to protect your information from those with more resources then physical destruction is probablly the safest option. Plugwash (talk) 00:07, 6 October 2008 (UTC)

Macintosh
This article needs much more info about file transfer between computers on flash drives. It is great to say that "FAT or FAT 32 file system" is universal. But that is just beginning. What other formats are actually used, on Linux and Mac, and what are the reasons they should or shouldn't be used, for various purposes? Particularly, why would anyone with a Mac use non-FAT on a flash drive, what advantages might there be? -69.87.204.110 (talk) 11:55, 16 September 2008 (UTC)


 * No it doesn't. Choice of filesystem has nothing to do with USB flash drives.  IMHO this page already has enough stuff on it that is of questionable relevance to the nominal subject of this article., and the fact that flash drives (may) have filesystems on them does not make filesystem selection relevant.    We already have a page on this—Comparison of file systems.  Admitted, that page could probably do with more descriptive text rather than simply being a collection of tables, but that is not reason to include what that article should be here. CrispMuncher (talk) 20:49, 16 September 2008 (UTC)

Not obsolete
"Obsolete devices

Audio tape cassettes are no longer used for data storage. High-capacity floppy discs (e.g. Imation SuperDisk), and other forms of drives with removable magnetic media such as the Iomega Zip and Jaz drives are now obsolete and no longer an option."

I'm assuming 'audio tapes' is refering to Compact Cassettes. I hope it doesn't refer to Digital Audio Tapes. The Iomega Zip drive is not obsolete and is still avaliable, and therefore still an option. http://store.iomega.com/section?p=4760&secid=39520

S. Coates

87.112.78.14 (talk) 18:10, 24 October 2008 (UTC)

Agreed. I suggest you all go look up the definition of "Obsolete" before you post. While it is true that audio tapes are no longer used for data storage (unless you're still using an obsolete computer), floppy disks still have their uses. In fact, a lot of motherboards still come with on on-board port for a floppy dive. Why? Allow me to quote http://en.wikipedia.org/wiki/Floppy_disk.



"It should be noted that Windows XP still requires the use of floppy drives to install third-party RAID, SATA and AHCI hard drives, unless the install cd were modified to include these drivers with programs made to customize a Windows XP install cd, such as nLite"



Yeah. That's right. If you want to install Windows XP onto a SATA drive and you haven't rolled the SATA drivers into your XP installation media, you need a floppy drive and a floppy disk to add the SATA drivers to the installation. See http://en.wikipedia.org/wiki/F6_disk for more information. A USB flash drive is not supported for this purpose and there no way to make XP support it for that purpose. Same goes for optical drives. Ask any CompTIA A+ certified technician, or better yet go look it up in any CompTIA A+ reference manual or exam study guide.

Now, considering that Windows XP is hardly "obsolete" I'd say that this is a pretty good reason to keep floppy drives around for a while, wouldn't you?Allthenamesarealreadytaken (talk) 03:19, 27 October 2008 (UTC)


 * Or you set the SATA adapter into non-AHCI mode (in BIOS Setup) and then plain Windows XP can be installed. ;) --Xerces8 (talk) 11:03, 22 December 2008 (UTC)