Wikipedia:Reference desk/Archives/Computing/2010 October 11

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.

October 11

MacBook trackpad problem

I have a 2007-era MacBook. If I press down with my palm to right of the trackpad (not touching the trackpad at all, but completely to the right of it), it will consistently register a click. To the left, no. Why is this happening, and can it be fixed? --140.232.178.118 (talk) 00:27, 11 October 2010 (UTC)

This is, I suspect, a mild design flaw. I have a 2006-7 era MacBook Pro that will occasionally move the mouse with pressure on the left palm rest. you might try turning on the 'ignore accidental trackpad input' option in the preferences. --Ludwigs2 01:12, 11 October 2010 (UTC)

I found the source of my problem: a swollen battery. When I remove the battery, the trackpad button works nice and crisp with no hypersensitivity. My 3-year Applecare has expired, so I'll have to see whether they're willing to replace it. --140.232.179.169 (talk) 01:57, 11 October 2010 (UTC)

Ah, interesting. FYI, you can get decent replacement batteries on-line for about 2/3 the cost of a direct-from-apple replacement, assuming applecare has expired. --Ludwigs2 17:15, 11 October 2010 (UTC)

Quick question - Nvidia cards

 Done I'd say

Hello, everyone. Quick question - Which card would be better for general work and watching movies (and having nice desktop effects): Nvidia 9400GT or 8600GT, upgrading from (no laughing, please!) an 8400GS. The bus is PCI-E. All three have 512 megs. The system is a P4 3,06 with 2 Gigs of RAM running Linux. Because of space constraints in the box I don't have much to choose from. Prices for both of the new cards are similar. Thanks :) --Ouro (blah blah) 08:36, 11 October 2010 (UTC)

The 9400GT is the fastest. Are you sure about the 512 MB? Compare [1] and [2] and you'll see there's more graphics memory, faster graphics memory, and a faster GPU.--Best Dog Ever (talk) 09:28, 11 October 2010 (UTC)

Yes, I'm sure about the 512 MB. So You'd go with the 9400 despite it being in the entry-level range and the 8600 in the performance range as shown here? I do not do any graphics-intensive games, but I want the user interface, working on large documents in OOo and any possible graphics manipulation (simple photo trimming and editing mostly) plus of course films to be swift. --Ouro (blah blah) 09:57, 11 October 2010 (UTC)

The 9400GT is perfectly fine for your requirements (you would even be able to play most pre-2010 commercial games on that in lower resolutions). The later series is usually better when it comes to Nvidia due to tech changes in a rapidly changing sector of computing...Sandman30s (talk) 11:28, 11 October 2010 (UTC)

Thanks loads! I don't play games, I prefer to spend my free time away from the machine whenever I can. Cheers! --Ouro (blah blah) 12:09, 11 October 2010 (UTC)

I strongly disagree with the claim "The later series is usually better when it comes to Nvidia due to tech changes in a rapidly changing sector of computing". There are very many cases when a 'newer' card is a lot worse then an older card, whether Nvidia or ATI. This is even more acute if you're comparing the GeForce 9 series vs the later GeForce 8 series (8800 GT, 8600 GT etc i.e. not the original 8800 GTS/G80) since AFAIK there weren't any major feature changes between the two. You'd need to look at the benchmarks to be sure but from an educated guess and a quick search the 8600GT is in fact better then the 9400GT in nearly every way (the only areas I can think of where the 9400GT is likely to come out on top is power consumption and therefore heat, the 9400GT also has PCI-express 2.0 support but it seems a moot point since neither card is every likely to get close to saturating a PCI-express 1.0 link). Nil Einne (talk) 18:17, 11 October 2010 (UTC)

You can't compare the bottom of the range of 9000 series to top of range 8000 series. The top of the range of a new series is always better than the top of range of an old (9800GT vs 8800GT). However, comparing the 9600GT to top of range 8800GT, I think the 9600GT has the edge (I don't want to quote websites here, I've had both cards and the 9600GT just felt better on my computer). This comparison applies to every other Nvidia series change since about the 5000 series. Of course it is a generalization and there are exceptions, but clock speed is not everything... new series of cards always include new features and pipeline optimizations and because of hardware support for new features, takes away the need for that extra clock speed used for DirectX software calculations. Sandman30s (talk) 06:45, 12 October 2010 (UTC)

Who's comparing the bottom of range to the top of range? Also to be frank claims like 'just felt better on my computer' are silly particularly without specifying under what conditions and what applications this 'feel good' occured (double blind? games?). This may not be the science desk, but it is still a reference desk. For some actual references [3]. (I'm not saying this is definite, a single review never is.)

I agree clock speed is not everything, but what 'new features and pipeline optimisations' are you referring to? As I've already said, comparing the later 8 series to the 9 series, there aren't really any significant new features. I hope you're aware that the 8800GT uses the same G92 based chip that is used certain models of the 9800X line (well basically all use the G92 but there were two versions of it a 65nm and 55nm).

Presuming it's clear from all benchmarks that the 8800GT is better or the same as the 9600GT, are you still going to insist that the 9600GT 'feels better'? (Of course you haven't specified the clock speeds and memory sizes of both cards but I'm presuming spec.) If so perhaps you could explain what causes this 'feel better' behaviour.

P.S. I hope you understand why I refer specifically to the later 8 series. It's something that I hope doesn't need to be explained to someone offering help on comparisons of various Nvidia cards. I also hope you understand the big difference between the move from say the 7 line to the 8 line and the 8 line to the 9 line.

P.P.S. I would point out this is getting a bit off topic anyway but I dislike claims likely to be misleading. The OP's question was about the 9400GT vs the 8600GT and from what I've seen so far there's no question in terms of performance, a 512MB 8600GT is better if it's running at the original 8600GT spec speeds. (I emphasise spec because it wasn't really established whether the 8600GT 512MB was at normal spec speeds, as I pointed out the 512MB version often uses slower memory). For example, something potentially relevant to the OP [4]. Nil Einne (talk) 11:34, 12 October 2010 (UTC)

I don't have the time nor the energy to hunt for references and data sheets and comparison charts and debate this any further. I agree that it wasn't very encyclopaedic of me to say 'it felt better' but sometimes first-hand experience with things beats a page full of numbers. You win. Thank you for your feedback. Sandman30s (talk) 17:59, 12 October 2010 (UTC)

The NVIDIA page I linked to says that the 8600 GT actually has 256 MB of memory. Like I said, it also has a slower GPU and the memory operates at a slower speed. Every part that impacts performance is slower in the 8600 than the 9400.

Nevertheless, if you won't be playing games, the 8600 GT will work fine.--Best Dog Ever (talk) 18:10, 11 October 2010 (UTC)

512MB 8600GTs definitely exist (relying on official specs to know what exists for graphics cards is usually a rather bad idea). Also what on earth are you talking about? The page you linked to lists the memory speed of the 8600 GT as 700mhz and the 9400GT as 400mhz. (Both have a 128 bit bus.) How is the slower? The core clock is marginally slower at 540mhz vs 550mhz and there's a bigger difference in the shader clock at 1140mhz vs 1400 mhz but the 8600GT has 32 stream processors while the 9400GT has 16 stream processors so the 8600 mhz is still likely to come out on top. All these specs are somewhat moot though since the OP should be looking at the cards they're intending to buy, not generic Nvidia specs. Particularly for the 512mb 8600GT since higher memory cards often have slower clocks. Having said that, it's not clear to me why the OP wants to upgrade at all. I admit I don't know much about the Linux desktop but is it really that demanding that the 8400GS isn't more then enough? Nil Einne (talk) 18:25, 11 October 2010 (UTC)

My mistake. Yes, it looks like the 8600 GT is faster. I don't know about Linux desktop effects, but I tested Vista Aero a while ago on a desktop with a GeForce 7300 GT, and it ran great.--Best Dog Ever (talk) 18:37, 11 October 2010 (UTC)

Wow, I didn't think... I must have looked at everything too superficially earlier on. Nil, especially Your input is appreciated. Truth be told, I'm just having concerns whether my graphics sometimes tend to be choppy because the card is inadequate or because the nvidia drivers for Linux are not perfect yet (might be both). It might be that this improves in the futue (Fedora 14 is coming out in less than a month officially, the test release is already there). I guess what the 8400 GS has to offer should be more than enough to smoothly operate compiz and these certain effects I have on, but I'm just having doubts as to whether I need more power in there or not. Thanks for Your input, (and I can't think of another word to put here other than) friends. Good night. --Ouro (blah blah) 21:21, 11 October 2010 (UTC)

I would personally investigate more about what's causing any issues you encounter before bothering to upgrade. In particular choppy graphics could be other things like the processor not keeping up for example. (I'm not very good at remote diagonostics and don't know Linux so can't really offer much more help.) Nil Einne (talk) 11:34, 12 October 2010 (UTC)

HTML5 canvas typography

I want to print x² on a HTML5 canvas graph.

context.fillText('x<sup>2</sup>', 100, 100);

It doesn't work.

Certainly I can manually arrange the positions of each character so it shows up like x² on Canvas.

I can also put rasterized images on canvas.

Is there a better solution?

Is there a way to create more complex things such as ²
₁H
? -- Toytoy (talk) 09:06, 11 October 2010 (UTC)

Perhaps overlay a normal markup layer over the canvas with z-index? --Sean 17:55, 12 October 2010 (UTC)

Can you use unicode context.fillText('x²', 100, 100); Graeme Bartlett (talk) 09:41, 13 October 2010 (UTC)

²₁H with subscript 1 in unicode, but I don't know how to make the two characters combine ̩²H is looking closer with the combining diacritical mark with superscript 2. Graeme Bartlett (talk) 09:55, 13 October 2010 (UTC)

QList from Qt and C++ operator

Resolved

 – Aha moment right after I post this... antilived^{T | C | G} 12:03, 11 October 2010 (UTC)

I'm trying to use the indexOf() method from the QList class in Qt4, containing a custom class called "Resident". I've implemented the == operator in the Resident class as below:

bool Resident::operator == (Resident& resident) {
	return this->name == resident.getName();
}

But it gives the error no match for ‘operator==’ in ‘n->QList<T>::Node::t [with T = Resident]() == t’ at compile time if I try to use the indexOf() method:

QList<Resident> users;
int id = users.indexOf(someUser);

It works if I do it manually:

int id = 0;
QList<Resident>::iterator i;
for (i = users.begin(); (i != users.end()) && !(*i == someUser); i++) {
	id++;
}

But I would prefer getting indexOf() to work. How do I implement the == operator properly so that indexOf() would be happy? Thanks for your help. --antilived^{T | C | G} 11:58, 11 October 2010 (UTC)

And a few seconds later I've figured it out - the parameter of the == operator cannot be passed by reference (no & allowed). I've removed the & and it's all fine now. --antilived^{T | C | G} 12:03, 11 October 2010 (UTC)

The equality operator should accept constant objects both on the left-hand side (the hidden this parameter) and on the right-hand side (the explicit parameter). Thus your operator should be

        bool Resident::operator == (Resident const& resident) const { ... }

or possibly

        bool Resident::operator == (Resident resident) const { ... }

but probably the former, since the latter will cause the RHS argument to be copied.

The fact that your modification worked is something of an accident. Probably QList<T>::indexOf is implemented more or less the same way as your explicit search, with the test being (*i == argValue) and argValue having the type T const&. If the test had been written (argValue == *i) instead, your modified version would have failed as well. -- BenRG (talk) 21:50, 11 October 2010 (UTC)

Windows product activation of XP VirtualBox virtual machine

I've installed Windows XP home edition as a VirtualBox virtual machine (PUEL license) under ubuntu 10.04, using an OEM CD, legit, not previously used. Everything seems to be running smoothly, and I will need to do a Windows Product Activation within a couple of weeks.

• First question: Is there any reason to expect problems with the windows product activation?
• Second question: If product activation is successful, will I be able to move the virtual machine to a different physical machine, maintaining the validity of the product activation? Note: I'm not asking whether moving the machine is acceptable according to Microsoft's EULA, but whether it will technically work, i.e. whether the virtual machine will detect that it has been moved.

Thanks, 95.34.148.81 (talk) 12:46, 11 October 2010 (UTC)

At my workplace, I seem to either create a new Windows 7 (VirtualBox) VM or copy an existing one to another computer at least once a week. We have a volume license, so activation is not a problem. Assuming XP works about the same, then 1) you will not have any problems with activation, but your license will be considered used, and 2) if you move it to another computer, you will need to re-activate it with a valid license; unless you have a volume license, then do not expect to be able to use the same license key again. 124.214.131.55 (talk) 14:18, 11 October 2010 (UTC)

But to what extent does the windows software detect the hardware that the virtual machine is running in? For example, the Mac address of the network card that the virtual machine sees is not the Mac address of the physical card, and will not change when the machine is moved. When the hardware is emulated, how can Windows detect that the physical machine has changed? --95.34.148.81 (talk) 14:37, 11 October 2010 (UTC)

I'd recommend seeing our article Windows Product Activation as it lists the 8 categories of hardware that it checks/compares. OEM licences are sold for the original machine only (unlike retail which can be transferred). If you activate it on the Virtual machine then that would be using up your licence and it should be expected that trying to reactivate on a different machine would fail (since the hardware would be completely different). It doesn't matter that the original machine is virtual/isn't physical, it's still a machine so that's what would count. I'd recommend simply not activating it until it's installed on the proper physical machine to avoid breaking the licence agreement as well as activation woes you're sure to face.  ZX81  ^talk 15:00, 11 October 2010 (UTC)

Just adding to my previous post as I think I may have misunderstood your question. When you say moving to a "different physical machine" do you mean that on this new machine the Windows XP installation will still be virtualised? If so then you won't have any problems at all. You can literally install VirtualBox on the new machine, copy the virtual hard disk files across/setup the configuration as before and it'll continue to work with no idea that the outside hardware has changed and as such the activation won't be challenged either (assuming you set it up with the same memory/optical drive configuration etc).  ZX81  ^talk 15:04, 11 October 2010 (UTC)

Yes, exactly. I mean copying the .vdi file which implements the virtual machine, along with the stuff in ~/.VirtualBox, to a machine on which Ubuntu with virtualbox is installed beforehand. When I do this with the non-activated machine, it behaves exactly similarly, the activation countdown gives the same number of days left until activation etc. 95.34.148.81 (talk) 15:30, 11 October 2010 (UTC)

It will probably see the actual CPU's model and stepping and (if present and enabled) the processor serial number. The other hardware is emulated and shouldn't change. If you're moving between systems with identical CPUs and there's no PSN, then even that shouldn't change. Note that it is possible to detect that you're running within a particular VM product and take additional steps, but I've never heard of Microsoft doing that and I'm pretty sure that I would have heard about it if they did. -- BenRG (talk) 19:14, 12 October 2010 (UTC)

Thanks, everyone! --95.34.148.81 (talk) 11:57, 13 October 2010 (UTC)

I checked it out with this utility program (note: compiled program, use it only if you trust the website, or on "discardable" virtual machines). It confirms that BenRG is right, the real processor's details are reported. 95.34.148.81 (talk) 10:56, 14 October 2010 (UTC)

SVG indefinite repeat / radial gradient

I have this bit of code I've been working on, its basically got several ellipse shapes overlaying each other and they are creating a circular rainbow. These however switch colours as time goes on. Basically I want them to change colour and back again but then continue doing this indefinately. The problem I have is: putting repeatCount="indefinite" into my code just repeats the colour changing back and not the colour switching first. How do I put this action in correctly?

Below is an example of one of the circles, it switches from red to violet and back again and I would like this to be continuous

<ellipse id="ellipse7" fill="#EE82EE" cx="400" cy="600" rx="150" ry="150">
 <animateColor id="ellipse7" attributeName="fill"
 from="#FF0000" to="#EE82EE" 
  begin="0s"
  dur="10s"/>
  <animateColor id="ellipse7" attributeName="fill"
  from="#EE82EE" to="#FF0000" 
  begin="12s"
  fill="freeze"
  dur="10s"/>
</ellipse>

Also one more question, can this effect be applied to ellipses with a radial gradient? Thanks in advance195.49.180.89 (talk) 15:03, 11 October 2010 (UTC)

As for the first question, you can merge the two animateColor elements to one, and put repeatCount="indefinite" to that:

<ellipse id="ellipse7" fill="#EE82EE" cx="400" cy="600" rx="150" ry="150">
 <animateColor attributeName="fill"
    values="#FF0000;#EE82EE;#EE82EE;#FF0000;#FF0000"
    keyTimes="0;.42;.5;.92;1"
    dur="24s"
    repeatCount="indefinite"/>
</ellipse>

Incidentally, since rx=ry, you could as well use circle instead of ellipse. As for the second question, you can animate gradients just like any other elements, for example

<svg xmlns="http://www.w3.org/2000/svg">
<defs>
  <radialGradient id="grad" cx="50%" cy="50%" r="50%" fx="50%" fy="50%">
     <stop offset="0" stop-color="white"/>
     <stop offset="1" stop-color="#FF0000">
        <animateColor attributeName="stop-color"
           values="#FF0000;#EE82EE;#EE82EE;#FF0000;#FF0000"
           keyTimes="0;.42;.5;.92;1"
           dur="24s"
           repeatCount="indefinite"/>
     </stop>
  </radialGradient>
</defs>

<circle id="ellipse7" fill="url(#grad)" cx="400" cy="600" r="150"/>
</svg>

—Emil J. 18:11, 13 October 2010 (UTC)

more gradient fun

<svg xmlns="http://www.w3.org/2000/svg"> <defs> <radialGradient id="grad"> <stop offset="0" stop-color="#F00"> <animateColor attributeName="stop-color" values="#F00;#EE82EE;#F00" keyTimes="0;.5;1" dur="12s" repeatCount="indefinite"/> </stop> <stop offset="0" stop-color="#F00"> <animate attributeName="offset" values="0;.5;0;.5" keyTimes="0;.5;.5;1" dur="12s" repeatCount="indefinite"/> <animateColor attributeName="stop-color" values="#F00;#EE82EE" keyTimes="0;.5" dur="12s" calcMode="discrete" repeatCount="indefinite"/> </stop> <stop offset=".5" stop-color="#EE82EE"> <animate attributeName="offset" values=".5;1;.5;1" keyTimes="0;.5;.5;1" dur="12s" repeatCount="indefinite"/> <animateColor attributeName="stop-color" values="#EE82EE;#F00" keyTimes="0;.5" dur="12s" calcMode="discrete" repeatCount="indefinite"/> </stop> <stop offset="1" stop-color="#F00"> <animateColor attributeName="stop-color" values="#F00;#EE82EE;#F00" keyTimes="0;.5;1" dur="12s" repeatCount="indefinite"/> </stop> </radialGradient> </defs> <circle fill="url(#grad)" cx="400" cy="600" r="150"/> </svg>

—Emil J. 16:49, 14 October 2010 (UTC)

Thank you, that is more help than I could actually have hoped for and the effect you have achieved in the last section of code is amazing. I shall be having an experiment with that to see what I can come up with. Good spot on the ellipse/circle too. I hadn't actually seen that simple error. Cheers :D 195.49.180.89 (talk) 13:46, 15 October 2010 (UTC)

Why is Word awful at saving documents as HTML?

For the first time in years, this weekend I tried taking a Word document and saving it as HTML in order to post it to a web site. I used Word 2007. The results were atrocious. There was nothing particularly complicated about the Word document — some inline photos with text set to wrap around them was the most complicated bit. The resulting HTML file was awful! The images were placed all wrong, the spacing between lines was wrong, the place where word wrap occurred on each line was wrong. It substituted Verdana for my font of choice, which was expected.

When thinking about why the premier document creation software from the world's premier software company (yes, that point is arguable) was so poor at this task, I could only think of a few logical possibilities: (a) The feature is crippled in order to try to get consumers to purchase a different Microsoft product that excels at this. However, when I tried importing the Word document into Publisher, the output was also awful. It just wasn't *as* awful. (b) Microsoft puts "Save as HTML" very low in its priority list and is satisfied with its current performance relative to the quality level of the entire product. (c) "Save as HTML" is a Hard Problem. I don't believe either (a) or (c) — I know that HTML must be more difficult than RTF because of the need for browser independence, variable page width ... but you should have seen how bad it was! Could (c) really be correct? Comet Tuttle (talk) 15:37, 11 October 2010 (UTC)

I think "converting to HTML" is not a "hard problem" as much as it is a paradigm shift. A Microsoft Office document is a (proprietary, controlled) file format that specifies a lot of things - text layout, image positioning, exact details of the way a page should be rendered. That is its purpose: it is designed to be a desktop publishing tool and not a purely-digital format. So when you try to convert from a fixed-form document into a "free-form" format like HTML, you have a lot of difficult decisions to make. To what extent should you (attempt to) exactly preserve formatting? Knowing that individual browsers may override your layout, font, and positioning instructions, can you even try to preserve the document's flow without mangling it horribly? On the other hand, if you punt on the export and strictly copy/paste the text and images and let the browser render them however it so desires, you have violated the user's expectations of formatting. So, it is not a problem of difficulty in translating the document format, it's trying to figure out exactly what the user wanted when switching from a controlled-format rendering engine to a free-format engine with text and graphical reflow. The two types of documents serve different purposes - there is not an exact one-to-one mapping between formatted- and free-format content; so the export engine does the best it can to heuristically decide which formatting decisions are relevant, preserve those, and free-flow the remainder. Nimur (talk) 16:04, 11 October 2010 (UTC)

You may have an option to save as type Web Page, Filtered. This will cut down on the CSS formatting. Also word 97 is much simpler in its HTML output if you still have that around. Graeme Bartlett (talk) 10:06, 13 October 2010 (UTC)

Microsoft SQL Server

What is "Microsoft SQL Server" for? I have this in task manager taking up 70mb of ram and I've never ever used it. Is it doing something important in the background that's necessary to run Windows, or can it be turned off? If it can be turned off, how do you turn it off? I never installed it so I don't know where it's settings are or how to turn it off. The OS is Windows 7. 82.44.55.25 (talk) 18:49, 11 October 2010 (UTC)

Microsoft SQL Server is a database; it is a general-purpose utility that might be supporting some other program. It is often used by web-servers, developers (computer programmers), and "industrial users" to store data. But it can also be a support utility for a lot of other programs that you might have installed. You can try to stop or uninstall the server and see if any program stops working. Unfortunately, the number of programs that might use SQL Server is too numerous to list - it really could be anything - games, word-processors, email clients, music library programs, .... do you recall what you recently installed before you noticed SQL Server was running? Nimur (talk) 20:36, 11 October 2010 (UTC)

Thanks for the info. I only noticed the SQL server today but then again I only looked through the processes in process explorer today; so it could have been there since I installed Windows 7. I guess I'll just leave it alone 82.44.55.25 (talk) 22:31, 11 October 2010 (UTC)

I assume it's the compact version of SQL Server? I always remove it when from my customers' computers. Nothing bad ever seems to happen. It doesn't come with clean-installations of Windows, unless your computer's manufacturer installed it with all the other bloatware they put on your computer. It's included alongside certain extra programs, like Microsoft Visual Studio. I would remove it. You can always download it and install it again, if you need it. The compact and express editions are free. I bet it degrades the start-up time of your computer, and could also create security problems, given the fact that SQL Server is a networked program.--Best Dog Ever (talk) 00:16, 12 October 2010 (UTC)

Security is greatly improved in the newer versions. MS SQL Server 2008 (and 2005 too, if I recall correctly), a great number of features are disabled by default, including remote connections. This is to reduce the surface area for attacks, requiring one to specifically enable the features that they want. In any case, if you wish to just disable it, open up Services and set the Startup Type for each SQL Server service to Disabled. The next time that you reboot, they will not be started. If in the event that you later find that you need it for something, you can always go back and set it to Automatic. 180.11.188.56 (talk) 09:24, 12 October 2010 (UTC)

Coq help

I have a fairly specific question about Coq, probably too specific for this desk. (Unless there are any experts here?) Is there a good (active) forum online where I can get good help? Thanks- Staecker (talk) 20:21, 11 October 2010 (UTC)

The Coq web page lists something called the "coq club" -- that would seem like the natural place. I somewhat misdoubt me that experts are in plentiful supply. Looie496 (talk) 20:36, 11 October 2010 (UTC)

You might try the Math Desk - there are a lot of people there with surprising levels of familiarity with obscure mathematical tools and techniques. Nimur (talk) 20:37, 11 October 2010 (UTC)

Google Earth / Streetview Crashing

Is it just me? If I click on a camera icon in Google Earth to fly into a Streetview picture, and then try to follow the road by successively clicking on subsequent icons, after repeating three or four times Google Earth crashes on me. I'm using Windows XP. Rojomoke (talk) 21:45, 11 October 2010 (UTC)

Is it laggy, or does the computer shows symptoms of CPU and memory stress? That is probably why. Sir Stupidity (talk) 07:10, 12 October 2010 (UTC)

attributes and simpleType in XML/XML Schemas

Simply put, how do I make the following code from an XML schema I am working on valid:

<xs:element name="ISSN"> <xs:simpleType> <xs:restriction base='xs:string'> <xs:pattern value='[a-zA-Z0-9]{4}-[a-zA-Z0-9]{4}'/> </xs:restriction> <xs:attribute name="IssnType" type="xs:string"/> </xs:simpleType> </xs:element>

The problem lies with the attribute - simpleTypes cannot have attributes. This element needs to be data type checked for that restriction/pattern, as well as have an attribute attached. xs:all, xs:simplecontent, xs:complexcontent and xs:sequence cannot have restrictions. I have agonized over this issue for hours trying to figure out a solution. Please advise.

--Baalhammon (talk) 22:05, 11 October 2010 (UTC)