Wikipedia:Reference desk/Archives/Computing/2013 November 22

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 22

Google Hummingbird page

I gave a blog link in the External links section of http://en.wikipedia.org/wiki/Google_Hummingbird and it was deleted. But many other blog links can be found in your references. Is it Wikipedia paid that the moderator can put their friends links only or paid links only.

The link was - http://technologiesinternetz.blogspot.in/2013/11/how-hummingbird-has-changed-definition.html

First I put this link under - References - the link was deleted

Later on I put this link under - External Link - Again it was deleted.

Is it biased that only moderator or checker can do anything. Other blog links are visible but when I put my favorite blog they delete it. — Preceding unsigned comment added by 182.74.24.62 (talk) 04:56, 22 November 2013 (UTC)

This same question has been asked, and answered, at the Help desk. Maproom (talk) 07:39, 22 November 2013 (UTC)

video capture

i need to know if it's possible to take a high quality video of my screen the same restitution as my screen at 30fps in real time (with software.) if so, i need a free program to do so if anyone knows of one. thank you. 70.114.242.17 (talk) 14:06, 22 November 2013 (UTC) p.s., i have a 4-core 2.8ghz amd cpu, 4gb ddr3 ram, 1280*1024 moniter with 24bit rgb color and 60hz refresh rate. 70.114.242.17 (talk) 14:18, 22 November 2013 (UTC)

If you want to capture DirectX/OpenGL stuff (games), which your framerate requirement would suggest you do, you need something like Fraps or DXTory. Fraps has a free limited version. -- Finlay McWalterჷTalk 14:39, 22 November 2013 (UTC)

Otherwise take a look at Comparison of screencasting software, which lists both opensource and freeware screen recording programs. But most are aimed at recording the ordinary screen (where things like browser windows are drawn) - to capture video game footage it's usually necessary to reach into the graphics buffers themselves, which is a lot more challenging to write. -- Finlay McWalterჷTalk 14:50, 22 November 2013 (UTC)

Useless cookies

I grudgingly see the need for (some) cookies - like remembering my user id or personalised settings for various things, but I find the targeting of adverts quite annoying. What is really bugging me though is that I might look around the internet to buy some product or service and then make my purchase, only to be stalked with adverts for the very same product or service for the next week or two. Why do targeted adverts not take notice of the fact that you actually bought something and are obviously no longer interested in that particular thing? Astronaut (talk) 20:15, 22 November 2013 (UTC)

I agree, but the cookies don't know whether you went through with those purchases. It would be even more spooky if they did. It is good practice to delete all your cookies at regular intervals. Having to then re-enter your passwords is a minor inconvenience.--Shantavira|^{feed me} 20:49, 22 November 2013 (UTC)

[1]--Aspro (talk) 20:51, 22 November 2013 (UTC)

AJAX website page ID in URL

Hello, My website hosting company gives really horrible addresses to pages of websites they host, eg. example.com/#!PageName/xxxxx. It's that last bit that looks really unprofessional, but they claim (http://www.wix.com/support/forum/html5/other/other/scrappy-web-nameaddress-bar) that this 'page ID' is necessary for AJAX websites. I may be mistaken but I think Facebook uses AJAX and I don't see such horrible page IDs at the end of their page URLs, so I think it's poor/lazy coding on the part of the service provider. Is there a way to hide the page ID? Thanks 81.101.120.9 (talk) 21:14, 22 November 2013 (UTC)

These are sometimes called "hash-bang urls" - their purpose, and problems with them, is discussed pretty well in this post. -- Finlay McWalterჷTalk 21:21, 22 November 2013 (UTC)

Oh, awesome,thanks! 81.101.120.9 (talk) 21:50, 22 November 2013 (UTC)

"Document Expired"

Does anyone remember the good old days of the 1990s, where you had to open every single page in a browser in a new window because if it went out of sight, the information and all your changes went away? Well, I just had a blast from the past getting hit with an edit conflict on Wikipedia - I did what I've always done, namely back-arrowed out of their useless "upper/lower text box" thingy so that I could grab my content and hit the Editing Section tab in the history browser. And guess what! "Document Expired". On Wikipedia I actually could go to the accursed lower box to retrieve my text, but if it had been another page, another time, I could have lost an hour or more of work, so it's important for me to stake this vampire, salt and burn the bones, bury it in its native earth, etc.

Now I've seen this plague rear its head more than once, and changed browsers more than once to get rid of it, and I'm amazed that something so universally derided by readers is still making appearances. The browser I was using for this was the latest update of Firefox on a Windows system. I went hunting and I found various threads claiming that you may or may not be able to fix it with a specific about:config setting - they're about evenly divided. [2] [3] and elsewhere. Oddly, some of those are from a period when I didn't notice that problem. Anyway, some general questions:

• Why anyone think this good idea, ever? It stands to reason, back arrow, forward arrow, flip, flip, flip. What's so complicated about saving a page the way a person expects it saved?

• Specifically, is all this rooted in some capitalo-utopian's freakish ideology of how the site owner owns your eyeballs and has the right to keep you from re-viewing a page, all sense be damned, even if he has to blow up the building, or is this truly apolitical?

• Is there a name for a browser model/philosophy that sticks with the notion that what you had is what you get? Are there browsers that strictly adhere to never giving you "document expired" except as a warning for your voluntary decision making purposes? (I see the Firefox thread mentions Opera, for example)

• And by the way, does that fix (setting browser.sessionstore.postdata = -1 in about:config) work, why do some people say it doesn't, and is there some dread exploit/virus/Knock At The Door From the Copyright Police that will happen if you make it?

Pardon my rhetoric, but this one... it's just so absurd... Wnt (talk) 21:45, 22 November 2013 (UTC)

There are at least two legitimate technical obstacles here:

• Browsers can't realistically treat every page in the history like a separate tab because there could be hundreds or thousands of pages in the combined histories of all tabs, and if you think browsers are bloated now imagine if they always acted like there were hundreds of additional open tabs.
• If a page was returned in response to an HTTP POST request, and the browser has discarded it (probably for legitimate reasons—see previous point), then it can't retrieve it again without resending the POST data, which can have side effects that shouldn't be repeated (such as charging a purchase to a credit card).

I do think it's silly that if a remote site has instructed the browser to immediately discard a page (as shown in the thread you linked), browsers will keep it around indefinitely as long as it's the latest page in a tab, but then forget about it instantly when you navigate away.

If I understand the browser.sessionstore.postdata documentation correctly, setting it to anything other than 0 instructs Firefox to automatically resend POST data in some circumstances, which could easily come back to bite you some day, so I advise against it. Also, it apparently only matters when quitting and restarting Firefox, not when hitting the back button. But if you're getting that "document expired" page with the "try again" button as shown in the thread you linked, and you don't mind resending the POST data in this particular case, clicking the "try again" button is safe and should solve your problem. -- BenRG (talk) 06:20, 23 November 2013 (UTC)

I didn't try it in this particular case, but my previous experience was that "try again" means that you lose all of the information you typed into the form, which is the whole reason why this is a problem. I'm not sure if hitting it would also mean that I'd lose the information in the next window (the one with the two edit boxes) also. I don't want the browser to hold all the data on every page I visit - only the ones with post data. I mean, before now I was counting on it to do that 100% of the time, so not only doing it 99% of the time doesn't seem like a huge burden. Wnt (talk) 14:48, 23 November 2013 (UTC)

I see that the copy of Firefox that did this is also doing something related: my list of contributions on Wikipedia is being changed while it is off-screen to reflect the more recent ones. While this isn't as obviously annoying, it also varies from the desired behavior of being able to count on the backarrowed and forwardarrowed pages to remain undisturbed. Wnt (talk) 18:48, 23 November 2013 (UTC)

I sent myself an email; why is it white space unless I reply?

Rather than carry around a flash drive I email myself stuff if I don't have time to read it at a library.

I forwarded the email to User:Nil Einne to see if he saw just white space too and if anyone can figure out why.— Vchimpanzee · talk · contributions · 22:49, 22 November 2013 (UTC)

Hi, the email doesn't show up as whitespace (I take it you mean it's totally blank for me). That said, I didn't expect it to. You'll probably find if you forward the email to yourself, the forwarded copy will likewise show up with text. I had a brief look at the source for the email, but didn't see anything interesting however I didn't see much point as forwarding the email as almost definitely meant a lot of the formatting could have been modified etc.

If you really want someone to look in this for you, you'd need to use 'show original' or whatever in your email client, and copy the original version and paste that somewhere like [4] or the millions of other paste bins out there. But you might want to be careful, while the email itself doesn't appear to be particularly private, I'm not sure if you want your email address etc permanently in a paste bin. The headers will contain your email address, you can exclude most of them since I doubt they are relevant (but do include anything MIME related including content type etc) but it's also possible your email address is somewhere else. The email itself is very long, probably only the first few lines are needed but I can't be sure of that.

You could send it to someone, the reason I'm not suggesting that to actually reliably send it so it isn't modified, you'll either need good control over your email client (which you obviously don't have), or you'll need to do something like compress it.

Anyway it's easy to guess what your problem is. The email is likely malformed. Perhaps it's a HTML email with such poor formatting it results in nothing being displayed. Alternatively, it's also possible both a HTML version and plain text version were sent using MIME multipart/alternative but the HTML version was actually blank (I'm pretty sure I've seen this before) and your email client, as with many, displays the HTML copy by default (of course the alternative in all ways is possible but I doubt it here). When you reply, your email client has to do some reformating. If the HTML formatting is borked, most likely it will fix the main problems resulting in the text displaying. If the HTML version is blank, probably replying will result in the client realising there's something wrong and using the text version.

To be honest, I've had a few of these myself before, but I've never really looked in to the precise cause of most of them. Unless you're composing the emails yourself and need to work out what's going wrong, it's unlikely to be worth your bother particularly if you lack the experience to figure it out yourself. The only case would be if you're strongly suspect the emails are fine, and your client itself is the problem but that's generally unlikely.

Nil Einne (talk) 20:53, 24 November 2013 (UTC)

Thanks. I forwarded it to another email address of mine and got the same result. I don't even remember how, but I finally got made a copy that showed up without my having to click on reply. I don't know what an email client is. I save a link to a web site in an email to myself, in the email service which comes up when I first turn on my computer. I rarely use that one. But I use the email service I sent the one to you from at home, and once the web site comes up, I click on "email". It is otherwise the closest to ideal. I created the email where I am now, Windows 8 and whatever version of IE is here. At home, I have Vista and IE9.— Vchimpanzee · talk · contributions · 18:26, 25 November 2013 (UTC)