Wikipedia:Reference desk/Archives/Computing/2010 January 9

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.

January 9

genuine software assurance

The most popular operating system manufacture uses generic disks with meaningless Proof of License, Certificate of Authenticity, a hologramed disk and a product code, which are meaningless to the buyer as far as determining whether of not the software is genuine prior to purchase. To make things worse the same manufacturer also has a volume licensing system that does not require end user activation or verification. Why not a copy protection method that can be bought and sold by individuals in which the software is on a disk of special media like movies are on DVD which can not be copied and therefore do not even need a product key much less activation or verification? Wake up most popular operating system manufacturer before its too late. 71.100.3.13 (talk) 00:03, 9 January 2010 (UTC)

The reference desk is for asking concrete factual questions, not starting debates or for complaining about things you don't like. Please take up your complaints with the vendor you dealt with, with ebay, or with Microsoft. -- Finlay McWalter • Talk 00:06, 9 January 2010 (UTC)

I thought the W and its RD were to obtain truthful and accurate information for everyone, which is my purpose. What is yours, to pick a fight in defense of your employer? 71.100.4.171 (talk) 12:16, 9 January 2010 (UTC)

Please see WP:NPA. I'm not aware that anyone on the RD works for Microsoft, and it's not relevant anyway Nil Einne (talk) 12:29, 9 January 2010 (UTC)

And adding to the above, movie DVDs CAN be copied. See our article DeCSS. ZX81 ^talk 00:23, 9 January 2010 (UTC)

Thanks. 71.100.4.171 (talk) 12:16, 9 January 2010 (UTC)

In fact, DVDs are exceptionally easy to copy. Nil Einne (talk) 12:29, 9 January 2010 (UTC)

In order for your computer to install the software it has to be able to read the software on the disk. If it can read the software then it can write the software to a new disk.

True. I am aware of the Analog hole so the media would have to be able to defeat the analog type hole copy process. In other words the copy process would have to do something to the contents that could be detected. 71.100.4.171 (talk) 12:16, 9 January 2010 (UTC)

Even if you used some sort of disk that could be read by standard hardware, but you needed special hardware to write it, this would still not stop counterfeiters. Counterfeiters are often large scale operations that would have no trouble buying whatever special machines they needed to copy the disks.

Short version : No one knows how to make a disk that can't be copied with the right equipment. Most experts agree that this isn't even possible. APL (talk) 02:58, 9 January 2010 (UTC)

Too late for what? --Mr.98 (talk) 04:15, 9 January 2010 (UTC)

China has already defeated Microsoft Genuine Advantage. 71.100.4.171 (talk) 12:16, 9 January 2010 (UTC)

assemble-able cell phones

Why there are no options yet to buy parts of a cell phone, and assemble it the way we like. It would be much more interesting if we could build cell phones like our desktop computer. --V4vijayakumar (talk) 08:00, 9 January 2010 (UTC)

Depending on exactly what you mean by "parts", I can only suggest that cell phones are so cheap, that making them from a kit would not be economically viable. To package a kit, PCB, components and instructions etc, for personal assembly, and make a profit the supplier would probably have to charge a lot more than a cheaper ready-built mobile phone. Secondly, as the parts are mostly surface mount for minimum size (among other reasons) and meant for machine assembly, construction could be very challenging. It is certainly possible, but likely not profitable. --220.101.28.25 (talk) 09:01, 9 January 2010 (UTC)

I repaired cell phones for Sprint in the early part of the 2000s. They just had one PCB inside. All the chips were (and are) soldered onto the board. The CPUs in smart phones nowadays are ARM processors from companies like Samsung. Only the most diehard computer enthusiasts actually solder chips onto PCBs nowadays in any device. Instead, they just insert new parts (like memory, CPUs, PCBs, etc.) into the computer. You can't just insert parts onto the system board of a cell phone, though. I imagine you could put a newer, faster CPU into an older iPhone or Blackberry, but the risk of damaging the board from ESD or the soldering iron should discourage you from trying. You can buy replacement logic boards for cell phones, but not individual chips. It's remotely like building your own laptop. Some people actually build their own laptops, but it's a pain because everything is miniaturized. So, there isn't a market for parts for doing things like that, because no one wants to.--Drknkn (talk) 09:40, 9 January 2010 (UTC)

By parts I mean, screen, camera, processing unit (processor, ram/rom), external storage, bluetooth ear piece, physical keyboard, etc. If there were some open standards then these all could fit together, like legos, and made look like today's regular cell phone. I hope, it will happen one day. --V4vijayakumar (talk) 10:44, 9 January 2010 (UTC)

The most important reasons imho are price and the fact most people buy cell phones for calling purposes only, laptops dominate when people need mobile computing power. I'm no engineer but I know cell phones are designed for battery life & size and standards like that would make development hard in these areas. You could always get a chip from TI, hook it to a display and a keyboard and call it a "phone", but as a private person you'd have hard time getting licenses so that you could run something else than linux on it. --91.145.88.180 (talk) 11:09, 9 January 2010 (UTC)

fork() system call

I have read all about fork() system call - its functions and its working - and how its used to create a copy of the calling process. But what I am unable to understand is what is the purpose of fork()? When the child process is exactly same as the parent process, just executing at a different location - then what's the use of executing it? Komal.s110 (talk) 08:31, 9 January 2010 (UTC)

"different location" that is the point of fork. you get "different location", entire new address space, another 4 GB memory to play with. --V4vijayakumar (talk) 09:15, 9 January 2010 (UTC)

fork (system call) sums it up, especially the sample code. --91.145.88.180 (talk) 10:36, 9 January 2010 (UTC)

Fork is usually used with exec (operating system) call to replace child process with different code. Lukipuk (talk) 16:58, 9 January 2010 (UTC)

The key is that parent and child know who is who (via the return value from the fork() call), and can act differently. In particular, the child often starts doing something completely different (including loading a new program into the process and executing it). But even if both execute the same code, you now have two threads of executions and they can do different things. One typical pattern for servers is that the parent only accepts network connections and forks of one child for each connection. The child then talks with the client over the new connection and terminates itself when the connection is broken. --Stephan Schulz (talk) 20:40, 9 January 2010 (UTC)

You could be forgiven for asking "isn't fork() a weird way of starting a new process?", and on the face of it you'd have a point. The first time you read code with a fork() you have to step back a bit to realise that one branch of an if runs in one process and the other branch in another, a paradigm you almost never see elsewhere in C programs. But fork() has some nice features which can make for simpler programming; chiefly because the child inherits so much from its parent. Firstly it gets a copy (not a shared view) of the parent's data segment, so all variables that the parent has set up are copied to the child. Moreover the child inherits the file descriptors of the parent, and any mmaped memory blocks. All of this makes setting up the child easy and makes getting the parent and its new child talking together straightforward. Perhaps a more obvious API would be just to say "make a new process", which is essentially what Windows' createProcess() call does - children on Windows don't inherit so much (not the copied data, not the mapped memory, but they do optionally get the file descriptors). So on Windows (or a comparative API) you have to package up stuff you're sending to the child and push it through the various parameters that the child gets. The Windows way is arguably cleaner, but also more labourous. And properly handing how fork() works (particularly in respect of file descriptors) can require labour too. The cynic might observe that when Thompson, Richie et al. were grinding Unix out in the tiny environment of the PDP, a fork model made for a very simple kernel implementation and some very succinct applications. BTW, Linux has a rather interesting extension to fork(), clone() - clone allows more control over what the child does (and doesn't) inherit. You can implement fork() using clone() (but I think Linux has a fork() kernel entry, at least for historical reasons) and Linux also uses clone() to implement threads. -- Finlay McWalter • Talk 21:22, 9 January 2010 (UTC)

auto sorting incoming email

Is there a way in Outlook Express for an incoming email to automatically create a new folder in the name of the sender or the subject line if one does not already exist and then send the incoming email there? If not what email program can do this even if it Linux based? 71.100.4.145 (talk) 11:55, 9 January 2010 (UTC)

Tools --> Message rules --> Mail.--Drknkn (talk) 12:01, 9 January 2010 (UTC)

This allows manual creation of a rule for a particular email not a system rule to do this automatically for every email. 71.100.160.154 (talk) 12:31, 10 January 2010 (UTC)

How do older computers compare to more modern machines in terms of power usage?

Generally speaking I am concerned with how much older computers (anything from Pentium III-IV level PCs down back to very old machines, including early 8-bit home computers)tend to compare to more modern machines in terms of power usage. Obviously there will have in recent times been some drive for efficiency and the addition of power-saving features, but I am guessing possibly some factors like the need for increased cooling as more a higher component density causes chips to become hotter, for example.

(I am already aware of thing such as flat-screen monitors using less energy than CRT ones.)

Mostly any such information is hard to come by.

Can anyone give me any hints? —Preceding unsigned comment added by 82.17.144.241 (talk) 12:48, 9 January 2010 (UTC)

Most modern PCs are rather good at conserving power when usage is low. I'm lazy to dig up some specific refs but I suggest you check out SilentPCReview (including forums) & perhaps Anandtech and look in the history of the RD archives since I'm pretty sure I've given some decent links before albeit not directly related to comparisons of old computers. Particularly if you use IGP, you could easily have a comp using under 100W under most normal usage for the average user i.e. a low load. Using energy efficient processors or those designed for laptops helps and also lower power chipsets. And also choosing a good well sized PSU that meets the 80 Plus standard [1] (which is a relatively recent invention). This will beat the shit out of a P4 in power usage and would probably beat many P3s as well. Of course, if you really want low usage, you should choose something like a netbook or other miniITX platform (you can get these which aren't in netbooks) or perhaps even something like a ALIX [2] or Soekris [3] which use about 5W which I expect would beat even many 8bit computers. Nil Einne (talk) 20:10, 9 January 2010 (UTC)

Some quick results A pg3 & A pg4; B pg3, B pg4 & B pg5; C pg4; D pg3, D pg4, D pg5 & D pg6. Compare this to some older results including P4s. Finding P3 figures may be harder and you may have to look in other places or ask in the forums but I suspect you can find something. You can of course get some idea by PSU sizing but many people like to highly oversize their PSU so I'm not sure how good an indicator that is. Nil Einne (talk) 20:27, 9 January 2010 (UTC)

eBay and PayPal buyer protection policy limits

A case study has revealed that both eBay and PayPal have a 45 day limit on buyer protection against counterfeit software although both claim that they do not endorse sale of counterfeit software. This seems contradictory to me since counterfeit software can be detected for an indefinite amount of time. IMHO fraud and counterfeit should not be included in the 45 day limit. Is there a good reason why it should and is there a place in either article that is appropriate to publish these facts? 71.100.4.171 (talk) 13:00, 9 January 2010 (UTC)

We are not eBay. Complaining to us will not help you in any way.

We also are not likely to even agree with you enough to make you feel better. APL (talk) 17:53, 9 January 2010 (UTC)

LOL... but the Wikipedia does have articles on both organizations of which the accuracy either is or is not subject to increasing so long as its editors are committed to finding and reporting the truth. 71.100.4.171 (talk) 18:58, 9 January 2010 (UTC)

Another type of item where fraud could be detected later would be stolen goods of any sort. So, the "good reason" is a good reason for eBay, not for you. eBay (PayPal is a subsidiary) does not want unlimited liability as a result of its buyer protection program. I have no inside information, but I would guess that the reason for the 45-day limit is that if eBay set, say, a 1-year limit during which a buyer could get a refund for counterfeit software, the seller would often disappear from eBay's systems before the buyers could file claims. Then eBay would be refunding your money instead of the seller. eBay would lose money that way, so they try to minimize that. Legally, I don't think eBay has any requirement (at least in the US) of even having a buyer protection insurance plan in the first place. (PayPal didn't, a long time ago; it is an innovation.) Technically, eBay is not involved in the transaction, supposedly, and it's a private transaction between the buyer and seller. As far as inserting information into the Wikipedia articles about how you got defrauded, WP:OR doesn't belong there. If you have WP:RELIABLE sources that have notable facts that aren't in the articles, feel free to add those facts (with inline citations of course). Comet Tuttle (talk) 19:24, 9 January 2010 (UTC)

The majority of companies and individuals are out there looking to make a profit. This is the central guiding principal of their thinking. I was there myself at one time. Now I look at transactions from a customer service point of view, i.e. what will make the customer happy, not just reasonably satisfied. The way I got here was from watching the results of unhappy customers take their business elsewhere and yes it can even happen to a business that has a major section of the business locked in. I learned that no matter how big and broad you are no one can afford in the end to let any customer go away unhappy. All to often do businesses loose sight of the long term in exchange for the short term and go back on their claims or make exceptions which nullify their claims. Consumers may be slow and companies may be able to ride for a while on the momentum of past happiness but eventually they will catch on and sour. If you want verifiable proof for an insert into any article then check the records of companies who have eventually succumbed to their greed. In this case, however, it is a matter of verifiable fact that fraud and counterfeit are permitted by eBay and PayPal if not discovered within 45 days. Proof of the intent to defraud other customers can be found in the fact that eBay will not allow feedback after 45 days, which might warn others. If the Wikipedia does not want to include this fact then the Wikipedia is not better. Of course though in the case of software all eBay or PayPal need do is to hold the funds until the manufacture approves the transaction such as Microsoft saying the software passed its Windows Genuine Advantage screening on the buyer's computer. 71.100.160.154 (talk) 22:48, 9 January 2010 (UTC)

45 days is a month and a half which to me seems quite a reasonable time to work out if the product is functional or not. If the bidder believe that it'll take them longer than 45 days to verify the product then they probably shouldn't bid on it. Ebay's rules about the 45 days are quite clearly visible (it took me less than 30 seconds to find that without needing to sign in or create an account). ZX81 ^talk 23:54, 9 January 2010 (UTC)

So its the customers fault if he is still going by the original rules that got eBay up and running and it is okay now for eBay and PayPal to claim they do not support fraud and counterfeiting but only for the first 45 days. If a customer buys an operating system in Oct. or Nov. for a Christmas or Hanukkah present then and the receiver does not open it until Dec 25th then 45 days is simply not enough. I think you have missed the moral and ethical point entirely here since discovery of counterfeit money has no clock from its creation to the time of discovery anywhere in the world. I am reminded of the complaints we heard about such doing prior to WWII by the German people. EBay and PayPal are now supportive of criminal activity and my dollars are best spent elsewhere to prevent myself from being cheated. 71.100.160.154 (talk) 12:38, 10 January 2010 (UTC)

Okay then. Spend them elsewhere. Quit posting here, go buy a real copy of windows from a reputable vendor, or sue the person who sold that to you. Don't troll here. 99.241.68.194 (talk) 18:06, 10 January 2010 (UTC)

Microsoft does not provide buyers with a list of "reputable" vendors nor a means for the user to verify the authenticity of its software nor its record of ownership prior to any sale. If the Wikipedia article does not include these facts then the Wikipedia article is inaccurate. --71.100.15.157 (talk) 20:15, 13 January 2010 (UTC)

No original research. Please feel free to cite published criticism. For more, please see WP:NOR. 69.150.163.1 (talk) 19:11, 11 January 2010 (UTC)

The Wikipedia has an article on street light interference which references articles published elsewhere rather than because it has anything to do with reality likewise placing the validity of the Wikipedia criteria for discussion or publication in serious doubt. --71.100.15.157 (talk) 20:15, 13 January 2010 (UTC)

In-Page Links

So, I am developing a site and I wanted to know how to add an in-page link as wikipedia does on everyone of their pages. For example, you can go to a page on wikipedia, then under "Contents [hide]" you can click one of the links to jump to a certain point on the page within that page.

So, here is a page from my site, http://thirteenpercent.webs.com/cellularmembrane.htm, and for example I want to be able to click on "Phospholipids" and then it should jump to that point on the page (if you scroll down you'll see it) where it begins to talk about Phospholipids.

Thanks, 74.184.100.154 (talk) 18:26, 9 January 2010 (UTC)

With a #, like http://en.wikipedia.org/wiki/Phospholipids#Simulations -- Finlay McWalter • Talk 18:31, 9 January 2010 (UTC)

And to make that target you have an HTML A tag with a NAME field - see http://www.htmlcodetutorial.com/linking/_A_NAME.html -- Finlay McWalter • Talk 18:35, 9 January 2010 (UTC)

Thank you! 74.184.100.154 (talk) 19:17, 9 January 2010 (UTC)

Note that you can make a link reference itself. That has the effect of placing the link at the top of the screen, quite good for section alignment. for example <a name="fig1" href="#fig1"> Figure 1</a> -- SGBailey (talk) 20:31, 9 January 2010 (UTC)

Excel Circular reference

I want to have a controlled circular refrence in Excel (I use Excel 2003). I have Q2 "=RAND()" ; P23 = "0" or "1" ; P2 "=IF($P$23=1,P2,Q2) to form a latched random value, otherwise everytime I do an edit the random value changes. I know I could copy the random value from Q2 and paste special | values it into P2, but there are reasons why I prefer not to do this. The code "works" fine, except that every time I change P23 (Latch the value control) from 0 to 1 it pops up a message box warning of a circular reference. Is there a way I can persuade it that I'm happy and for it to not bother with the message box? -- SGBailey (talk) 18:35, 9 January 2010 (UTC)

It appears that although the latch of the random value works, it stops other parts of the spreadsheet evaluating. So I probably can't do this. However the question stands, albeit modified: how can I tell excel that this circular reference is ok and it should evaluate it and carry on? -- SGBailey (talk) 20:27, 9 January 2010 (UTC)

Excel will allow circular references as long as you have switched on iterative calculation (I think it used to be called "goal seeking" in earlier versions of Excel). In iterative calculation mode, Excel will repeatedly re-evaluate all cells in a worksheet up to, say, 100 times unless it reaches a point where no cell changes by more than a given amount (say 0.001) before this. In this mode, circular references are an intentional part of an iterative model, and are not a problem.

However, if all you want to do is to stop RAND changing the contents of cell Q2 every time you make an edit, you can just switch from automatic calculation mode to manual calculation mode. In manual calculation mode, Excel only re-evaluates the worksheet when you explicitly tell it to - F9 is the shotcut key for "calculate now", although I think there is a menu option for this as well. Gandalf61 (talk) 13:15, 10 January 2010 (UTC)

Thanks -- SGBailey (talk) 17:16, 10 January 2010 (UTC)

Unusual behaviour by antivirus program

I've just done a full anti-virus scan by the free version of Superantispyware. My OS is XP. During the scan the program reported cscript.exe as a worm. The path was c:/windows/system32/cscript.exe SuperAntiSpyware recommended that I move it to the virus vault, but cscript.exe was in use. It would not delete either. It would aparantly rename though, by changing the end to vir. When the scan was complete I was surprised that SAS report that nothing had been found, despite the earlier alert. On a whim I checked for updates for SAS and one or two were loaded. That is unusual because I thought I had checked for an update just before scanning. I looked manually in windows/system32 for cscript and could not see it. I did a quick scan with MalwareBytes and nothing was found. Do I have anything to worry about? 78.147.235.55 (talk) 21:07, 9 January 2010 (UTC)

Cscript is not a worm. It's unusual for it to be in use, though. In any case, if you delete it, Windows will probably recreate it instananeously. Make sure Windows put it back. Next time, be sure to Google such files before allowing SAS to mess with your system files like that.

If you have a viral infection, your computer will usually be very slow and certain programs will not work. If that's not happening right now, then the chances of you having an infection are reduced significantly, although not completely.--Drknkn (talk) 21:22, 9 January 2010 (UTC)

I've realised that Avast! found the suspect item while I was doing a scan with SAS - hence the confusion. It has now found it again. 78.147.235.55 (talk) 22:11, 9 January 2010 (UTC)

desktop computer slowdown and performance

suppose i am opening 8 website tabs in my browser at the same time. all of these are now loading simultaneously. while this is happening i click on one of the tabs loading, but it takes up to 10 seconds for the computer to respond to my click and switch to the tab. from what i gather, the computer cannot process my request to switch to a certain loading tab because all of the processing power is being used to load the web pages. this makes the computer unresponsive to the users request. wouldn't it be possible to make the computer always respond to the users actions? so when the user clicks on the tab while so much is going on, the computer stops processing the webpages for a split second, switches to the tab in real time and then continues loading the pages. meaning the computer would put priority on what the user wants rather than what the computer is currently working on. is this strictly a software limitation, or are computers simply not fast enough? is an intel core 2 duo not 'fast enough' to process a tab switch while loading several web pages?

another thing, i find that with any computer, be it a windows 95 desktop to a brand new mac, there is a delay caused when inserting an optical disk. so if i am browsing the web and at the same time insert a disc into the optical drive, the computer kind of lags for a few seconds, even if the user is not trying to directly view the disc's contents. why does it lag? can't the computer handle the optical disc drive as a lower priority and accept the users scrolling in the browser window as the most important?

are there articles explaining these supposed software flaws/limitations? anyone have info on similar types of computer phenomena, where computers are slow and unresponsive no matter how fast they make them? Bonusbox (talk) 23:00, 9 January 2010 (UTC)

I think the lag from the CD-ROM drive is more of a hardware issue. Anytime you press a key on your keyboard or move your mouse or insert a CD, an IRQ interrupts the CPU (your Core 2 Duo), causing it to stop what it's doing to pay attention to the IRQ. CD-ROMs send many IRQs. So, if your CD-ROM drive goes bad and refuses to turn off, then get ready for Hell on Earth.

I think you're over-emphasizing your CPU, also. There is also memory (more = better), memory speed (higher = better), hard drive speed (higher = better), hard drive fragmentation (so defragment your drive every once in a while), and much more. As for the browser, yes, Windows can set priorities for different tasks. If you open the Task Manager and right-click on a process, you can set the priority manually to give it a higher or lower priority over other processes. But, if Windows gives a higher priority to your web browser automatically, then certain more important programs will have to wait in line. Further, have you ever had a web page go into an endless loop? What would happen if the tab hosting that page had a high priority? Would your computer stop responding completely? Possibly.--Drknkn (talk) 23:58, 9 January 2010 (UTC)

(edit conflict) :How much memory do you have? From what you've written I'd hazard a guess that the slowdown is actually being caused by insufficient memory rather than processor problems. I quite regularly open more tabs than that without any problems, but I have 6Gb of RAM. Regarding the second part of your question and the optical disk, I think the problem isn't so much software, but hardware. If the optical drives uses IDE/PATA (all older drives do and even some newer ones) then they would suffer from the problem where PATA can only perform one command at a time and the computer MUST wait until that's finished before sending another. Because the disc spinup takes a second or two this is the delay you get. There's more information on this in the "Serialized, overlapped, and queued operations" section in the above link. As far as I know newer SATA optical drives don't suffer the same problem. ZX81 ^talk 00:05, 10 January 2010 (UTC)