Wikipedia:Reference desk/Archives/Computing/2013 November 4

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.

November 4

Twisting lines in SVG

I have this small SVG file:

<?xml version="1.0"?>
<svg xmlns="http://www.w3.org/2000/svg" version="1.1" width="500" height="500">
	<g id="Rings">
		<circle cx="250" cy="250" r="249.5" stroke="black" stroke-width="1" fill="none" id="Ring_1"/>
		<circle cx="250" cy="250" r="218.25" stroke="black" stroke-width="1" fill="none" id="Ring_2"/>
		<circle cx="250" cy="250" r="187" stroke="black" stroke-width="1" fill="none" id="Ring_3"/>
		<circle cx="250" cy="250" r="155.75" stroke="black" stroke-width="1" fill="none" id="Ring_4"/>
		<circle cx="250" cy="250" r="124.5" stroke="black" stroke-width="1" fill="none" id="Ring_5"/>
		<circle cx="250" cy="250" r="93.25" stroke="black" stroke-width="1" fill="none" id="Ring_6"/>
		<circle cx="250" cy="250" r="62" stroke="black" stroke-width="1" fill="none" id="Ring_7"/>
		<circle cx="250" cy="250" r="30.75" stroke="black" stroke-width="1" fill="none" id="Ring_8"/>
	</g>
	<g id="Lines">
		<line id="Line1" x1="0" y1="250" x2="500" y2="250" style="stroke: black; stroke-width: 1"/>
		<line id="Line2" x1="0" y1="250" x2="500" y2="250" style="stroke: black; stroke-width: 1" transform="rotate(45,250,250)"/>
		<line id="Line3" x1="0" y1="250" x2="500" y2="250" style="stroke: black; stroke-width: 1" transform="rotate(90,250,250)"/>
		<line id="Line4" x1="0" y1="250" x2="500" y2="250" style="stroke: black; stroke-width: 1" transform="rotate(135,250,250)"/>
	</g>
</svg>

What I'm looking to do is twist the lines around the center in such a way that each point along the lines will be rotated by an amount equivalent to the distance from the outer ring multiplied by a given constant X° divided by 250. How do I apply such a transformation? — Melab±1 ☎ 04:37, 4 November 2013 (UTC)

So this is going to turn your radial lines into ... what? A spiral maybe, but is it? Astronaut (talk) 19:31, 4 November 2013 (UTC)

You might find having to do all the calculations yourself and draw the lines as many small segments. On the other hand there is a set of transformations in the (proposed) SVG 1.1 standard which might take some of the hard work out of it. Astronaut (talk) 19:47, 4 November 2013 (UTC)

For something like this you will want to use polar coordinates. In that system, each point is defined by it's radius from the center and rotation angle. This makes it simple to do the math. However, presumably you will need to convert the radial coords back to the rectangular coordinate system (X and Y) to actually draw them. You could either draw a series of very short lines or a series of points. The points might be a bit quicker to draw, but could leave gaps, if the rotation angle changes too much at each step. You might also find that the lines look thicker near the center. I can tell you how to solve this problem, but it's a bit involved, so let me know if you're interested. StuRat (talk) 20:14, 4 November 2013 (UTC)

(edit-conflict)

First, is hand-crafted SVG the right tool for your job? You could represent your plot in a mathematical form - say, as an analytic equation represented in Python code, and a series of plotting commands in matplotlib - and then export to SVG whenever you desire. This will allow you to modify the equations in a form that might be closer to the style of equation you're familiar with. The plotting library will parameterize the SVG, and you can control the details of the SVG output. If you are particularly concerned about the mathematical accuracy of the plot, then raw, hand-written SVG is definitely not the best tool for the job - because although SVG graphics are represented as scalable vectors, they must be composed from a set of standard path elements - like straight lines and specific mathematical curves. Not every mathematical form is expressly representable as an elemental data type in SVG.

There are numerous other ways to represent path objects in SVG; many are better choices than the line element when you want to modeling a piece-wise approximation for a curve. You can hand-craft piece-wise spline segments that approximate your curve (or let a plotting software do that for you). If you really want to do this by hand, you probably want to brush up on your spline math: the cubic Bézier curve is probably the most simple of the general-purpose smooth curve segments supported by standard SVG renderers; you can use a set of these splines to approximate the entire length of an arbitrary curve. Given an arbitrary number of spline segments, you can produce a curve that is arbitrarily close to the mathematical function you are approximating. Depending on your function, you might be able to do this using a very small number of spline segments.

Splines of order n are intended to enable a mathemetically-inclined individual (or robot) to convert a series of data-points into a straightforward, smoothly-varying representation whose n-th order derivative is also continuous. Whether you find this process straightforward depends on your mathematical aptitude. Nimur (talk) 20:16, 4 November 2013 (UTC)

This is emphatically not the right way, but I should note that modern browsers seem to have little trouble displaying 100,000 1x1 divs. You can actually use Javascript to add all those divs to plot out any given function (preferably appending them in a DOM model while display:none is set, then unveiling at once when you're finished; it seems to work faster). Wnt (talk) 06:59, 5 November 2013 (UTC)

Resources on - What factors are the most influential in the usability and user experience of a mobile application, that utilize crowd sourced information

Hi im Luke,

i'm currently working on my dissertation and i'm trying to answer the following question:

What factors are the most influential in the usability and user experience of a mobile application, that utilizes crowd sourced information?

I'm finding it difficult to find any resources specifically on this matter, their is tones of publications on usability and user experience of mobile application but not specially around application that utilize crowd-sourced. So any publication, debates, journals, Authors and studies around this matter would be great.

I hope i have provided enough information.

Luke H. — Preceding unsigned comment added by 194.83.248.5 (talk) 09:35, 4 November 2013 (UTC)

I think our WP:HOMEWORK policy applies to dissertations, although ironically, you have managed to craft a question such that even linking to our policy on not doing your homework for you is actually doing your your homework for you. I'm struggling to come to terms with the multiple meta-levels on which this question is predicated; you are asking us to crowdsource the research for your dissertation about factors that affect the user-experience of crowdsourcing information retrieval? On Wikipedia, that would be our policy on WP:HOMEWORK, which outlines the guidelines and conditions by which our volunteer contributors will help crowdsource information for you. Strictly, it is not a policy or guideline, but is a community consensus. Nimur (talk) 13:40, 4 November 2013 (UTC)

Ok, thanks for the help, i dont want anyone to answer the question, merely point me in right direction.

Luke H. — Preceding unsigned comment added by 194.83.248.5 (talk) 14:13, 4 November 2013 (UTC)

Breaking it donw into chunks, I looked at this from the other end. Crowdsourcing#Criticisms might have something to say on the latter part of your question. Astronaut (talk) 16:36, 4 November 2013 (UTC)

Much like here at Wikipedia, it comes down to an issue of quantity versus quality. You need both. Let's say your app tells you where to get gasoline cheaply. If it has all current and correct info, but for only 1% of the gas stations, then it's almost useless. Similarly, if it has 100% of gas stations listed, but much of the data is weeks old or just plain wrong, then it's also useless. I think the key might ultimately be to pay contributors. Let's say I find the info supplied by a particular person to be correct and useful, I could give him some small payment, say a cent. If thousands of people do this, then that person will be happy they were appreciated and likely to contribute more. Alternatively, if transferring a cent can't be done without get tax agencies and national security agencies involved, you could give them non-financial rewards, like list them in the "Hall of Fame for Excellent Contributors".

Also, once you've established that a given contributor gives correct info, then more weight should be given to their contributions than from unknown contributors, who might just be vandals. And, of course, checks should be done for absurd info, like if somebody enters that gasoline costs $99.99 a gallon. StuRat (talk) 20:24, 4 November 2013 (UTC)

Why does this work? I didn't sign in

Try clicking on this link. It doesn't work for non-subscribers, right?

I saved that link because of the difficulty of reaching that page, something I asked about in a previous question which I'm not going to try to find.

But this link came from a Google search. Once I clicked on it, I tried the above link and it worked.

At one time I could see everything in an actual newspaper. Now I have to go to this site for most of it, and until a few weeks ago, clicking on a link from The Houston Chronicle worked.— Vchimpanzee · talk · contributions · 16:39, 4 November 2013 (UTC)

Both links seem to work for me and I am not a subscriber. I get the most recent issue of The Mini Pages, but the second link gives the fun and games section. Perhaps when they are changing the issue the link fail temporally. Try this one. Richard-of-Earth (talk) 20:34, 4 November 2013 (UTC)

If I change the order they both work. The link you gave me won't help because the library's computer disabled Adobe Flash and I have to remember how to enable it temporarily. I don't want to mess with anything like this at home.— Vchimpanzee · talk · contributions · 22:21, 4 November 2013 (UTC)

I clicked on the first link and am told I'm blocked.— Vchimpanzee · talk · contributions · 22:22, 4 November 2013 (UTC)

I've seen this before at New York Times - you had to go through Google to be welcome. As a rule with newspaper links that don't work, paste them through to the magic Better Than You and click on the first hit, and they will welcome you with open arms. It's all very peculiar. Wnt (talk) 06:46, 5 November 2013 (UTC)

If I had to guess, I'd say they wanted Google to be able to read the page so it shows up in searches, but that they messed up. I've seen other screws up like this before- and other related varieties. Then again, I don't know anything about the site, so it could be something completely different.Phoenixia1177 (talk) 09:16, 5 November 2013 (UTC)

Probably this, but I doubt they "messed it up". Google will de-list a site if it shows one page to the google-bot, but a significantly different page to a real person who clicks the Google link.

It was probably a very calculated move to let people see the site "for free" if they click through from google. The alternative is having that page indexed as a "log-in" screen, which is no better than not being listed on Google at all. APL (talk) 01:36, 6 November 2013 (UTC)

I contacted The Mini Page and was promised something like a subscription which I wouldn't have to pay for. At least for a while. They sent me some logos and I contacted them again and it took quite a lot of effort to finally get one. And that one was one I had already seen, at a library (which I traveled nearly 200 miles to get to, so I can't do that again) that subscribed to the actual paper that actually printed the whole thing, unlike one paper I can read which always prints about half except for last week with election coverage, one that prints one out of four pages most of the time, and one that prints two pages if they feel like it.

Given all these promises which they never came through fulfilling, it's only fair if I see it for free.— Vchimpanzee · talk · contributions · 18:42, 6 November 2013 (UTC)