Wikipedia:Reference desk/Archives/Computing/2011 February 28

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.

February 28

RAID 1+1 Array

  I now want a RAID 1+1 array; is there any Windows-compatible hardware or software which implements RAID 1+1? I prefer hardware-based RAID to software-based functionality. Rocketshiporion♫ 00:50, 28 February 2011 (UTC)

How is a RAID 1+1 array different from a RAID 1 array? Just set up a all your disks in a single RAID 1 array and you get exactly the same functionality. Also, I can't point you to any hardware solutions although I would be surprised if none exists. Taemyr (talk) 15:12, 2 March 2011 (UTC)

Therein lies the difficulty - all the hardware implementations of RAID 1 of which I'm aware only support two-mirror sets, and I want a four-mirror array. Rocketshiporion♫ 08:16, 3 March 2011 (UTC)

Copy and run a file into C:\ drive and all folders, sub folders, subsub folders.etc

I urgently need to find a way to mass attempt to copy a file from a USB into every folder, subfolder and subsubfolder in C:\ and if it works attempt to run the .exe file. It will be even better if a log can be produced showing where it was sucessful or not in copying and/or running a .exe. Can anyone figure out how to put such an instruction into command prompt or powershell? General Rommel (talk) 01:21, 28 February 2011 (UTC)

Do you mean that you want to copy tha same executable into the root of every directory and all sub-level directories on C: (which I'm presuming is your Boot Drive), and then execute all the copies of the executable? Rocketshiporion♫ 04:27, 28 February 2011 (UTC)

And if that is what you mean, I'm curious to know why anyone would want to do such a thing? Other than as a rather arbitrary test of disk drive reliability, it wouldn't do anything that you couldn't achieve in other ways. AndyTheGrump (talk) 04:50, 28 February 2011 (UTC)

Rocketshiporion, that is exactly what I'd like to do, and to AndyTheGrump, I just want to test something about my computer. General Rommel (talk) 08:11, 28 February 2011 (UTC)

(Assumption 1: you're on windows). Assuming access to suitable tools (e.g. excel, openoffice), it's fairly easy to take a recursive listing of a folder structure from a command prompt (e.g. dir /s /a:d > filename.txt), load into excel and parse out distinct folders using (in column B) a formula such as =IF(MID(A1,2,1)="D",RIGHT(A1,(LEN(A1)-17)),""). You can then use these as the basis of a batchfile, by putting together a repeated list of copy xyz.exe c:\dir\dir\ strings by formula, notably using syntax such as +A1&A2&A3, where A1-A3 contain the sub-elements of the command. It's a job of a couple of minutes. --Tagishsimon (talk) 14:32, 28 February 2011 (UTC)

I have a not so high understanding of Excel and unfortunatley I don't know what you are talking about. When I looked up command dir, it dos not seem to help me do the job. AS well as that I can't run scripts or batch files either. So des anyone have an idea? General Rommel (talk) 23:16, 28 February 2011 (UTC)

I can't believe that you really want to copy an executable file into every folder of your hard disk. That's nuts. Perhaps just you want to run it once per hard disk folder with that folder as the current directory? If so, you can try this:

       cd /d c:\
       for /d /r %i in (.) do ((cd /d "%i" && u:\foo.exe) || (echo %i >> u:\failed.txt))

where u:\foo.exe is the program, and u:\failed.txt is a log of directories where the execution failed. This could be adapted to actually copy the executable into each directory, but I'm not going to tell you how unless you tell me why you need to do it (to satisfy my curiosity, if nothing else—I expect some reward for the time I volunteer here). If you just want to know which folders you have write/execute permissions in, you can use AccessEnum. -- BenRG (talk) 05:00, 1 March 2011 (UTC)

Well I'll tell you. I want to put a executable on a locked down DET laptop given to students in year 9. All I want to do is run Google Chrome (they only let you use IE for a web browser, which I find insulting), so that is why. So now will you please help? I'll attempt to ask someone to make an virtual medal for you (now where am I'm going to get that?) if you want recognition. General Rommel (talk) 08:53, 1 March 2011 (UTC)

So basically you found that you couldn't run your .exe in a particular folder so you want to see if it can run from some other folder on the C drive? Odds are if they locked down the laptop sufficiently to block the .exe from running on C:\ and a random selection of other folders, then it probably won't run anywhere. That being said there are possibly other means of accomplishing what you want (run a .exe file not approved by whoever set up security on your laptop). For example, have you tried just running the .exe off the USB drive? That often works for me. Also, it would be helpful if you tried to not give the impression of taking offense at the responses here, you may not realize it but your request is extremely bizare and many people here (computer professionals or otherwise) would never have a need to do what you're doing. Also, you were given a solution above (using batch files and excel would do what you want to do) and turning it down without providing more information about what your goal is, isn't very helpful. The reality is that your method of getting chrome to work probably will be a lot of effort for little gain. Good luck! Chris M. (talk) 20:17, 1 March 2011 (UTC)

I should have mentioned, that you cannot run .exes from a USB drive as well, and there are loopholes because there was one (until group policy took care of that). I'm sorry if I sound that I just shrugged away from the first response, but as I said already, unfortunatley I don't know what it means, and I can't run batch files. That is why I would like a better answer please. Thank you for noticing that though. Perhaps I sould say thanks for a good response Ben! General Rommel (talk) 22:47, 1 March 2011 (UTC)

@General Rommel - If you'd like to give anyone a virtual medal, refer to Wikipedia's barnstars. Rocketshiporion♫ 07:45, 2 March 2011 (UTC)

I will if BenRG wants a reward for helping people at the reference desk. Once I figured how to make oneGeneral Rommel (talk) 09:17, 2 March 2011 (UTC)

A quick search suggests these laptops tend to be locked down fairly well and many people have tried different ways to get round the restrictions [1] (see comments) [2] but when uncovered these tend to be locked down too [3] [4] and also tampering is grounds for confiscation as well as what sounds like an unpleasent visit from various people in your school and involved in the program and in some cases possibly including punishments like detention or even suspension. In other words, it may be prudent to consider the wisdom of breaking the security restrictions of something it sounds like you don't own particularly if (as it sounds like from here) you don't really know what you're doing and you're posting publicly about it. Nil Einne (talk) 15:41, 2 March 2011 (UTC)

I know the risks of asking here, but I'm am quite sure that a officer will not mind. I am aware of the risks. As I hae said above, all I want to do is ise Google Chrome. General Rommel (talk) 01:12, 3 March 2011 (UTC)

Try putting the executable in a .zip archive. Use Windows Explorer to browse the zip file and run the .exe. decltype (talk) 08:33, 3 March 2011 (UTC)

I was AFK for a few days, but I'll follow up for what it's worth. I don't want a barnstar or any other reward; I was just in a silly mood because the question was silly (or seemed silly, at any rate). Having heard the explanation I understand what led you to ask the question, but it is really unlikely that this global-copy attack will do anything useful (as other people have always said). I promised to give you this:

       cd /d c:\
       for /d /r %i in (.) do (copy /y u:\foo.exe "%i" && "%i\foo.exe" && (echo %i >> u:\succeeded.txt))

but I recommend strongly against actually trying it. -- BenRG (talk) 02:29, 4 March 2011 (UTC)

I tried it, and unless it was typed wrongly (I didn't understand anything when I typed 'help for') it failed. Thanks anyway Ben!!!!!! General Rommel (talk) 03:17, 4 March 2011 (UTC)

Garage Band: recording with real instrument, converting to software instrument

I've been trying to figure out how to do this. I'll record myself playing guitar with Garage Band '08. What I want to do is take the notes I have just played and convert them into notes for a software instrument. Is this possible? I can copy and paste the guitar information in the top track section, but when I look in the bottom track editor section, the information doesn't paste into the software instrument track. When I try playing the software instrument track, it doesn't do anything. Is there any way to do what I'm trying to do or does it just not work like that? Thanks 129.3.151.117 (talk) 03:15, 28 February 2011 (UTC)

Detection of the notes you play, and conversion to a software instrument format (e.g., MIDI), is an incredibly difficult challenge. Converting recorded-audio to MIDI has been discussed many times on the desk (see this archive-search), and the answer is generally summarized as "currently impossible" or at least "extremely difficult," due to inherent theoretical challenges in the detection of the notes. In this discussion from November 2010, I described some research-grade software technology that can convert MP3 to MIDI, and some of the reasons why this research is still far from being a "one-click conversion" software tool. What you could do is obtain an electronic guitar with a MIDI interface: Guitar synthesizer or "MIDI guitar" and play through that. And, with professional-grade software (and a high-end, hybrid synthesizer/guitar), you can record both the waveform of the music as well as the MIDI note meta-data. As far as I know, this is not something GarageBand can do - but it can probably take in the MIDI track only, which may be suitable for your needs. Nimur (talk) 18:37, 28 February 2011 (UTC)

MacBook battery bar problem

I have a first-generation MacBook running Snow Leopard, and I've recently started having trouble with the battery indicator. If I unplug the computer, it still shows the charging icon, right through the whole cycle, at which point it will die with no low battery warning. Sometimes I can prod it into showing an unplugged icon by repeatedly unplugging and replugging, but it's inconsistent and laggy - sometimes showing an unplugged icon while the machine is plugged in. How do I fix this? --140.232.184.35 (talk) 06:49, 28 February 2011 (UTC)

Odds are it is a problem with the battery — see if you can't find another battery and see it does the same behavior. --Mr.98 (talk) 16:17, 28 February 2011 (UTC)

I just replaced the battery a few months ago (because the original one was bulging), and the new battery has been working fine. So no, I'm not going to waste money replacing a perfectly functional battery. How do I fix the indicator bar? --140.232.177.147 (talk) 18:54, 28 February 2011 (UTC)

First, reset the System Management Controller.[5] If that doesn't work, then calibrate the battery.[6] ---— Gadget850 (Ed) ^talk 19:00, 28 February 2011 (UTC)

Actually, I'd reverse that - I had this problem when I first installed a new apple battery in the same kind of machine, and calibration fixed the problem. I think Apple introduced this weirdness in the last update (somehow it's recalling information from the last battery and getting confused about the new battery?) --Ludwigs2 19:12, 28 February 2011 (UTC)

Free PDF reader - not Adobe

I do not like bloatware, so I want to replace the Adobe reader with some other. The Adobe reader is about 45MB, and I think I recall reading that it is several times that when loaded into the computer. What would people recommend from personal experience please? I use Windows XP. I have already looked at various Wikipedia articles, but nothing evaluates them rather than just describing them. 92.15.3.182 (talk) 11:58, 28 February 2011 (UTC)

I've used Sumatra PDF and it's very small, under 2MB. It handles most .pdfs fine but can occasionally have trouble with complex .pdfs 82.43.92.41 (talk) 12:35, 28 February 2011 (UTC)

If you haven't used Adobe Reader for a while, you might want to give it another try. It sure has been bloated, slow, and difficult to use in the past (version 7 is the worst in my opinion). However, the more recent releases of Adobe Reader (including the current version, Adobe Reader X) is actually faster and easier to use than its predecessors. And I tried it on the very same hardware I once used to run Adobe Reader 7. There is an Adobe Reader alternative called Foxit Reader, but I like Adobe Reader better. 118.96.165.14 (talk) 12:47, 28 February 2011 (UTC)

I use Okular. I hear that it has a Windows version. -- kainaw™ 13:21, 28 February 2011 (UTC)

everyone uses Foxit. You won't be disappointed. Ignore the respondents above. 109.128.222.233 (talk) 14:39, 28 February 2011 (UTC)

I see three problems in as many sentences here:

• 1) Not everybody uses Foxit - because I (and evidently some of the above respondents) do not.
• 2) "You won't be disappointed" is an absolute.
• 3) This is a reference desk. We do not advise people to ignore other people's answers, especially when they offer relevent advice. --KägeTorä - (影虎) (TALK) 16:31, 28 February 2011 (UTC)

Okay, but he should still just install foxit. 109.128.222.233 (talk) 16:47, 28 February 2011 (UTC)

For me, Foxit Reader does not support form filling. In both Windows and Linux, it does not support catalog indexing and has a tendency to mangle copied text. So, why install it? I will certainly be disappointed. Please make some attempt to comprehend that this is a reference desk. If you want to state absolute facts, they need to be true. If they are questionable, you need to provide references to back them. -- kainaw™ 17:09, 28 February 2011 (UTC)

I seem to remember IrfanView as being able to view .pdf documents. The only disagreeable (to me) feature of this (free) software was its ugly icon. It uses Ghostscript to render .pdfs, but as for its accuracy I cannot say as I have not used it recently. --KägeTorä - (影虎) (TALK) 17:28, 28 February 2011 (UTC)

Evince is available for Windows as well. For the technically-inclined, it does not use ghostscript (so certain subtle PDF-rendering artifacts are avoided). Here's a comparison of Evince vs. Adobe Reader features. Nimur (talk) 18:30, 28 February 2011 (UTC)

I just use preview - does everything adobe reader does, and more, and is generally snappier to use. or if you want something more sophisticated, get Skim. --Ludwigs2 19:15, 28 February 2011 (UTC) wrong platform, sorry --Ludwigs2 19:16, 28 February 2011 (UTC)

The icon is a red cartoon cat which is squashed because it has been run over. Since I like Irfanview, I am going to claim that this is a lovely icon, and is also a very funny joke, unfortunately accessible only to Bosnians. 213.122.11.147 (talk) 23:08, 28 February 2011 (UTC)

It's icon is roadkill? I knew it was a contorted red cat, but I did not know that........ --KägeTorä - (影虎) (TALK) 00:56, 1 March 2011 (UTC)

Google Chrome has its own PDF reader. Downloading Chrome just for this is probably overkill, but if you have it already, or are otherwise looking to use a different browser to the troubled IE, this might be an option for you. I've found Chrome's PDF viewer to be fast and reliable. -- Finlay McWalter ☻ Talk 19:46, 28 February 2011 (UTC)

This advice is orthogonal to everything above, but my advice is to permanently abandon your dislike of bloatware, since your objection to it is based on either RAM or disk space. Here in 2011, RAM is cheaper than it has ever been in history, hard disk space even more so, and who cares if software takes up more RAM than "it ought to", since today's operating systems use a page file and will immediately take care of the "problem" if you somehow fill up your computer's RAM with such software. I had the same attitude fifteen years ago, but these days I see this posture as being analogous to complaining that 1.44MB floppy disk drives are needlessly slow. Comet Tuttle (talk) 20:31, 28 February 2011 (UTC)

If someone was just running Adobe Reader and nothing else then it shouldn't be a problem on a modern computer. But considering most programs try to take ~300MB each, even running just a few programs at the same time can cause a 2GB system to be used up quickly. Upgrading RAM isn't always an option especially on netbooks or laptops, and why upgrade more RAM so that badly written / bloated programs can run when the same task can be done by a small / well written program? 82.43.92.41 (talk) 21:29, 28 February 2011 (UTC)

Sorry, I think you missed my comment about the page file. That's a key component of why oversized software isn't a real problem anymore. Comet Tuttle (talk) 21:47, 28 February 2011 (UTC)

No. From page file: "The backing store for a virtual memory operating system is typically many orders of magnitude slower than RAM. Additionally, using mechanical storage devices introduces delay, several milliseconds for a harddisk. Therefore it is desirable to reduce or eliminate swapping, where practical." Anyone who has exhausted RAM and then needs to thrash around by repeatedly querying the swap space can attest that it is a frustrating experience, since the computer's performance becomes obscenely slower. Titoxd^{(?!? - cool stuff)} 22:08, 28 February 2011 (UTC)

No offense, but when was the last time you sat there, arms crossed, and waited for your PC to stop thrashing the hard disk? Swapping out the 45MB that the original poster is complaining about takes about 0.5 of one second these days. Of course I remember the bad old days when thrashing made it horrifying to run into low-RAM situations — I owned a Macintosh with no hard disk, you see — and those days are just a bad memory now, unless you're unfortunate enough to have to use old hardware with not much RAM and with a nearly full hard disk. As stated below, the original poster has taken a principled position, rather than a practical one. Comet Tuttle (talk) 23:39, 28 February 2011 (UTC)

It's happening right now on my computer, even just maximizing windows of already running programs takes 5 - 10 seconds, opening a new tab in Firefox takes ages, basically doing anything takes a long time. I would not want to lump another ~60mb of Adobe Reader on top. Process explorer reports 1.3GB of RAM in use, with 2.5GB of pagefile. Windows 7, 2GB of installed RAM. I really appreciate programs which don't use loads of RAM for no reason. 82.43.92.41 (talk) 23:56, 28 February 2011 (UTC)

Happens to me all the time. On pretty high end machines too.

I've noticed that there are two schools of computer use. Some people are single-taskers : they do one thing at a time, they rarely use tabbed browsing, and when they are done with a task they close all related programs and start 'fresh' on their next task. If they need to reference a web page for their work, more often than not they'll print it out.

I am not one of those people. At any given time I've probably got a two dozen firefox tabs open in at least a couple of windows. At least one development IDE, Thunderbird, a word processor, a calculator, and a bunch of file browsers.

If you're from the first group, I'd agree that "software bloat" isn't much of an issue, but if you're in the second group your choices are to change your mindset (Which is an admission of tool failure) or to seak out programs that start quickly and then can sit idle using a minimum of CPU power and RAM.APL (talk) 15:33, 3 March 2011 (UTC)

Paging is much slower than that in practice because the reads/writes are generally not sequential. If, say, you're reading non-sequential 64K blocks and it takes 8ms per block, that's about 8 MB per second. Physical RAM is always in use, even when the OS reports it as free; if nothing else, it's caching file data that isn't currently mapped into the address space of any user process. Any time an application needs more RAM, something is discarded to make room for it, at the cost of extra disk accesses later if that data is ever needed again.

The best argument for speed optimization is that there are far more users than programmers. If it takes a man-week to implement an optimization that saves the average user one second per day, and there are 1,000 users and the optimization remains in the code base for one year, then it was worth it. Actually, though, the most efficient way to save users' time is probably to improve the UI or documentation. And a botched attempt at optimization may cost users time instead, if it introduces a new bug. -- BenRG (talk) 04:24, 1 March 2011 (UTC)

Yes - this is a principled position rather than a practical one. If everybody writes programs which are ten times bigger than they need to be, well, however cheap hard drives and memory are, an order of magnitude difference is still likely to cost people a bit of money on hardware, and the same difference in the opposite direction means paupers can use old and cheap machines, and makes interesting design variations possible. So taking a moderate, reasonable amount of care to reduce bloat is just what software designers ought to do. It's not vital, but it's pleasant, and so using needlessly huge programs feels slightly unpleasant - not because they won't work, but because pandering to the ethos of bloat is very slightly immoral. Having said that, in many cases the impression of bloat may be an illusion, and the hugeness of a program may in fact be because it's very useful and does all kinds of things you'll eventually discover you need. Besides, excessive optimization is a bad thing. 213.122.11.147 (talk) 22:40, 28 February 2011 (UTC)

Though I do think there are probably real performance problems when you start talking about a bloated OS running many bloated programs at once. I'm somewhat dismayed to find that 3 GhZ of processing power and 4 GB of RAM doesn't add up to much anymore, when you are layering that over Windows 7 with MS Office and a bunch of bloaty other things on there. I'm not convinced that I'm really getting a lot more of import done with that processing power than I would have been getting done with my old Pentium III running at 550 MhZ and with 256 GB of RAM. Yes, things are "prettier" and I know that takes up a LOT of processing power, and websites have gotten very complicated and all that jazz, but there's part of me that sort of expected that I'd really be seeing a truly better overall experience by now, above and beyond the specialized benchmarks. --Mr.98 (talk) 18:58, 1 March 2011 (UTC)

Thanks, I have installed Sumatra PDF which is now about 3MB, and it seems to work very well. The other free PDF readers suggested were several times larger. I had replaced OpenOffice by Abiword, which is about 20MB compared to the formers 400MB. I am now looking for a suitible spreadsheet. The old Visicalc, a version of which still works in a DOS window, is only a 19 kilobyte (repeat kilobyte) zip file. 92.24.191.30 (talk) 12:58, 1 March 2011 (UTC)

If you are looking for a simple office suite with very small footprint, have you ever considered "cloud office suites" that are pretty popular nowadays? They are accessible anywhere (from any Internet-enabled computer with a modern web browser), require no installation and updating, use very little disk space and memory (that can be easily reclaimed after they have been "exited"), load pretty quickly (as fast as your Internet connection), more secure (primarily because they are web pages running in a web browser —most of the processing is done in their vendors' servers— and they are regularly updated by their vendors as new security problems are discovered), and … they are free. 118.96.158.165 (talk) 02:42, 2 March 2011 (UTC)

If you are interested in cloud office suites, here are some links to the most popular ones: Google Docs (article), Docs.com (article). 118.96.158.165 (talk) 02:51, 2 March 2011 (UTC)

Sorry, I am not, but thanks. I have since found a 490KB spreadsheet that appears to do a good job http://www.321download.com/LastFreeware/page14.html#Spread32 (last freeware version S32118.zip) and a less than 2MB wordprocessor http://members.multimania.nl/textshield/ Other small applications are available at http://tinyapps.org/ http://www.portablefreeware.com or even http://www.softpedia.com 92.15.15.174 (talk) 20:23, 3 March 2011 (UTC)

Wikimedia Commons

About 12-14 hours ago I wanted to upload a few pictures to Commons using Internet Explorer. The first one or two went as usual. Then after entering all of the correct information and clicking on the upload button it started returning a IE unable to (do not remember exact) page. I would use the back up button and I would get returned to an unfilled picture upload information page. I would then re-enter the information exactly the same and then click on the upload button and it would work correctly uploading the picture. This happened about 2-3 times and then I started getting a pop up small window saying Internet explorer stopped working. The program tried to recover the tab and after a few tries I got a page saying it was unrecoverable. I went to a couple of other non Wiki websites and they worked fine, I then tried Commons again with the same results. I restarted my computer with the same situation (all other sites working including Wikipedia just not Commons), gave up thinking I would retry latter. It is now latter after my computer has been shut down for the night and again the same situation. I cannot get onto the Commons main page or any of the Commons pages I have bookmarked, I also tried to get there by going through Google with no success. Always the same small pop up window and the attempt to recover the tab. Anybody else with this problem? any thoughts?
--RifeIdeas Talk 14:16, 28 February 2011 (UTC)

The Commons site now seems to be working, I just uploaded one picture. We will see if it continues.
--RifeIdeas Talk 14:59, 28 February 2011 (UTC)

I uploaded several (7+) pictures and took a break. Just came back and am having the same problem of not being able to get onto any page within WikiMedia Commons. Again any thoughts? ? ? ? ?
--RifeIdeas Talk 23:23, 28 February 2011 (UTC)

There was a problem for users trying to open Commons pages in Internet Explorer, reported here at WP:VPT and at Commons:Village_pump#Commons_not_working_with_IE.2C_again. It was caused by Javascript and is apparently resolved now. If you're still having problems, the old versions of Commons files may be stored on your computer - try clearing your cache to see if this fixes it. --Kateshortforbob _talk 10:50, 1 March 2011 (UTC)

like a live powerpoint, but a video

is there any way I can make a video on my computer of me speaking into the mic while flipping through powerpoint slides, this is called a screencast I just realized, so I have my answer: yes, as well as the search term to use. Thank you. 109.128.222.233 (talk) 21:36, 28 February 2011 (UTC)

Camtasia will do the job for you. --KägeTorä - (影虎) (TALK) 21:38, 28 February 2011 (UTC)

Or CamStudio for something free. ¦ Reisio (talk) 22:17, 28 February 2011 (UTC)

Motorola Droid Series Bootloaders

Can someone please explain to me what people mean by locked bootloaders? I know what a bootloader is, but can't one just flash a new bootloader onto a Droid 2 or Atrix or whatever that doesn't check the digital signature of the kernel? As I understand, a bootloader is not hardwired. --Melab±1 ☎ 22:24, 28 February 2011 (UTC)

Not necessarily. I'm not sure what method they employed, but they could have protected the memory (which could potentially be reflashed if you removed the chip and had the tools, depending on if the pins required are accessible or not), or it could have been manufactured with a ROM bootloader, which is unchangeable. Basically, yes, it can be hardwired. --Wirbelwindヴィルヴェルヴィント (talk) 23:10, 28 February 2011 (UTC)

Almost all consumer electronics devices these days boot from a flash memory part. Many such devices (nicknamed "bootsectored flash") feature write-protectable sectors. These were designed to give devices a basic loader that couldn't be corrupted even if software running the flash-write algorithm on the same part went crazy and tried to wipe the whole part. Typically such a basic loader executes in place (after the most minimal hardware setup); it checksums the next phase of the loader and only executes it if it's valid. This initial loader is written onto the flash part in one of two ways:

• In the factory of the flash manufacturer themselves, en-masse. Flash manufacturers will (for an additional cost) perform this for large orders, before the flash devices are put on the tape (the plastic roll that holds them until they're fitted onto the final PCB)
• In the phone (etc.) manufacturer's factory, where a production-line appliance connects down over the chip (e.g. with pogo-pins or a zebra connector) and performs the programming operation with the flash device mounted in the PCB.

Either way it requires either grounding a given physical pin or applying an external voltage to it. The whole idea is that this physical connection isn't normally available in the field, and the protected sectors can't be erased by software.

This process was invented purely for reliability; you don't want a bad update wrecking a whole field population of units. But it can be adapted to enforce the manufacturer's desire to lock-out third-party or altered firmware. Replace that checksum with a robust public-key cryptographic check and the system can refuse to run any firmware not signed by the manufacturer.

The obvious hack would be for someone to open up the device (e.g. phone) and (with skill and dexterity, as the pins are very small) apply the same external voltage as the manufacturing appliance did - that would unprotect the bootsector segments, allowing an unsigned loader to be written. To prevent that, a manufacturer may choose to glob epoxy over the part (making access to the pin very difficult), or to snip off the relevant pin (which logically will be held in the no-op position by an internal pull-up or pull-down resistor).

When manufacturers really want to keep out custom firmware, they have to do more. The bootsector validation code in modern game consoles is written into the chipset (so you can't just swap in a fresh boot flash); that's more difficult still to circumvent. -- Finlay McWalter ☻ Talk 01:25, 1 March 2011 (UTC)

Good grief. That was a really detailed expansion of what I was saying. One point I would like to add that you didn't expand on, is that the chip likely will still need to be removed in able to access the pins required for reflashing. For example, the all so common TI OMAP's pins are not accessible without removing, as the pins are not exposed at all on the circuit board. --Wirbelwindヴィルヴェルヴィント (talk) 17:27, 1 March 2011 (UTC)

Mac OS X Interface Builder

I have a mac running Mac OS 10.5, and I use interface builder a lot for developing software. There's a problem I have though. Whenever I put things in tab views, there are no resize handles and no visible alignment guides. It's like the tab view is obscuring those or something. Is there any way to fix this? Hope someone can help, and thanks in advance. --Thekmc (talk) 22:35, 28 February 2011 (UTC)

This is a followup to the above question

So, I can make an Avi file while I'm talking and advance the slides as I'm doing my presentation. Great. Is this the right way to do what I'm doing, or am I repeating built-in Powerpoint functionality??

How shoudl I get the video OR powerpoint-with-built-in-voice-over-and-slide-advances (if that exists) to my cliwent??? In the former case, shoudl I just e-mail him the whole AVI? Or, is there some site I can put it up on, like Youtube, only it is private for just one person? 109.128.222.233 (talk) 22:56, 28 February 2011 (UTC)

In PowerPoint 2007, I see that under the "Slide Show" tab, there's a button that says "Record Narration".

Otherwise, an AVI or other movie file is going to be too large to e-mail, and YouTube videos are not private; so you could use a file transfer service like DropSend, or, if you have your own website, create a private, password-protected directory for your client on it, upload your AVI to that directory, and e-mail (or tell them via phone, or whatever) your client the directory URL and password. Comet Tuttle (talk) 23:23, 28 February 2011 (UTC)

Though I've never tried, I think you can do this in Google docs. Google of course own youtube, and googling "google docs video" (this is all a bit incestuous, isn't it?) turned up a result titled "Play your video files in Google Docs". These uploaded documents can be made private to a limited audience (which is the default setting). 213.122.11.147 (talk) 23:27, 28 February 2011 (UTC)

Javascript - what does constructor prototype equality signify?

I'm in the process of analysing a piece of javascript malware I found when idly browsing. I've successfully deobfuscated it, bar a few lines at the top that I don't understand. They're below (don't worry, these lines themselves are harmless):

String.constructor.prototype.mama='BAR';
rNEW=Object.constructor.prototype.mama;

if('FOO'+rNEW=='FOO'+'BAR'){ 
    doc=document
};
// otherwise doc is uninitialised and subsequent calls to it will fail

At the end, if things are working well (for the malware author) doc will be the DOM document object, which he later manipulates to his wicked advantage. But I don't understand the purpose of the check he makes. What he's essentially doing, to my mind, is checking whether String.constructor.prototype == Object.constructor.prototype. In my testing, this evaluates to true in Rhino, node.js, and Firebug; and indeed in all three String.constructor == Object.constructor is true too. So I guess he's checking to see if the environment is conducive to his manipulations.

If that assumption is correct, presumably there is some javascript environment in which this isn't true, in which case he doesn't set doc and the subsequent code (his attack) will fail.

Does this check make any sense to anyone? In what environment would it fail? If he's just checking to see if he's in a DOM environment, wouldn't typeof(document)!='undefined' work just as well? Or is he checking to see if some framework or tool has hooked one of those constructors (but if so, why)?

I don't, incidentally, think this is an artefact of his automated obfuscator: it does dumber string-substitution and string-splitting things. -- Finlay McWalter ☻ Talk 23:48, 28 February 2011 (UTC)

I'm no JavaScript expert, but it seems like he's creating a new variable in the Object Constructor Prototype. To do this, a security-manager must not be preventing write-access to the object constructor prototype. Your "equivalence-check" (String.constructor.prototype == Object.constructor.prototype) only checks for equality - his checks for equality and write-access (to add a new variable "mama"). This tutorial shows how you can override the String constructor-prototype. The malware author probably wants to append a function to the String object constructor, so that any new string in your javascript environment will be over-rided and will contain his variables and functions as a payload. I don't know if that method actually circumvents any type of security-policy or access controls on any reasonable JavaScript engine; but if it did, it could open the door for XSS or other attacks. It seems like it's his intention to append functions - not variables - to each string instance - probably for the purposes of snooping; and his check is determining whether he can modify the String constructor. Nimur (talk) 21:19, 1 March 2011 (UTC)