Wikipedia:Reference desk/Archives/Computing/2015 September 29

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.

September 29

Edit box font size and Chrome

Right now I am using a 2009 Gateway running Widows 7 to edit WP. When I use the Chrome browser, the edit box font size for pages like the ref desk display at something like 6 to 8pts. I have looked for a default minimum font display size, and cannot find one. Is there eaither a Chrome or WP setting I should be looking for? The edit bozes with IE 11 and Firefox display normally. Thanks. μηδείς (talk) 00:46, 29 September 2015 (UTC)

Just to clarify, it's just the edit box? The rest of Wikipedia and other sites when displayed in Chrome are at a good size?

If it is indeed all text, then try hitting CTRL-0. Additionally, CTRL-- (control and the minus key) will decrease the text size and CTRL-+ will increase it. CTRL-0 will bring everything back to the default size (cancelling changes made by CTRL-- and CTRL-+). Dismas|^(talk) 16:03, 29 September 2015 (UTC)

No, bizarrely, as far as I can tell, it is only the edit box at wikipedia, and only with the Chrome browser. The print is so small I can't make it out, so I have to use CTRL + to expand it, and then everything else of course gets huge. I am assuming there's got to be some weird glitch. μηδείς (talk) 20:31, 29 September 2015 (UTC)

The following steps might help tell what's causing the problem:

1. Right click in the edit box and choose Inspect Element. The Developer Tools will appear.
2. In the right side, click on Computed. This will show the final styles used for the edit box.
3. Look for an item called font-size. Click the triangle to expand the item. This will show the source of this style.

What size is shown? What source is shown? This information might provide clues about what to check next. --Bavi H (talk) 01:55, 30 September 2015 (UTC)

Thanks, Bavi H, it says font size = 16.5 px and all the other font values are "normal". I use WP monobook too, in case that's relevant. The actual size is smaller than 8pt Calibri as displayed on MS word in the default viewing mode, I would guess 7 or maybe even 6pt, although MS Word doesn't go below 8pt for normal text text. I am editing in firefox right now, and everything is normal. I also did click on the restore default settings in the inspect element window, but nothing changed. μηδείς (talk) 20:55, 30 September 2015 (UTC)

I'm using Internet Explorer, but I'm assuming the Developer Tools in both browsers are close enough to the same. I set my Wikipedia preferences to use MonoBook, then examined the computed style of an edit box. I clicked the triangle next to font-size to see what's "inside" it. I see:

 ▼ [√] font-size: 13.33px
      [√] div#globalWrapper- 127%   load.php (1)
      [√] body- x-small             load.php (1)

This says the font-size is originally set to x-small in a body rule, then modified by 127% in a div rule, yeilding a final value of 13.33px. You're likely getting something similar. (I was hoping you might see a more obvious problem. For example, if you saw the final computed size was a small number, and saw some rule from a file like "user.css", that might suggest you need to disable a user preference somewhere in Wikipedia or a browser add-on.) If the computed styles don't suggest a cause, then I guess Chrome just renders the text box font size differently that other browsers.

• If the problem started recently, maybe a recent update of Chrome changed how the textarea font size behaves. You might search for the exact version number and terms like textarea font size to see if others have reported the issue.

• Maybe Chrome is only applying the x-small rule to the edit box text. Try unchecking the box next the x-small rule and see what happens. Or uncheck the 127% rule. Or uncheck both. This won't really solve the issue, but might help describe the problem in a bug report.

• You might work around the issue with user CSS. In Wikipedia, you can edit one of the user CSS files in your preferences, but this will also affect other browsers you use. To just make a user CSS in just Chrome, you might install Stylish and write a new rule. Either way, trying entering something like textarea { font-size: 120%; } or textarea { font-size: 16pt; } and see if that helps.

--Bavi H (talk) 02:08, 1 October 2015 (UTC)

Thanks, I spent about an hour toggling the settings last night. Nothing under the inspect element windows showed any weird smalling factors. What I did eventually find was that Chrome was set at a 67% zoom factor for some reason. Setting that back to 100 made the edit box legible, if not still undersized. I think what I need to do is find out how to smallen the size of default text so that when the zoom is at 125% the normal text appears, well, normal.

As for recent changes, this 2009 Gateway is my backup computer, my preferred computer has some keyboard issues I probably won't budget to be fixed for a while, unless I decide to do so myself. Since the 2009 was slow and crashing all the time I reinstalled it from the rescue disks, and at that time and for the first time installed Chrome. Chrome works okay on the 2013 ASUS with the dead keyboard, also running windows 7, but it simply works horribly with this computer, with hesitant scrolling and a whole lot of other problems. In any case, The edit boz text is legible, so at this point I just have to spend the time getting the default text smaller. μηδείς (talk) 00:09, 2 October 2015 (UTC)

The cost of Video Conference Business Meeting

Why is video conferencing so horribly expensive? And why is Skype free of charge (for most services)? Is is so difficult to stream real-time information reliably?

According to [1] 2 hours would cost an estimated total of $2,070.

OK, short answer is because people, or business and governmental clients, pay for it. However, couldn't other IT companies provide the same service at a much cheaper rate and skew the market prices?--Jubilujj 2015 (talk) 07:17, 29 September 2015 (UTC)

If you're looking for a market analysis that accurately explains why professional video systems can command a high price ... prepare to be amazed! It costs loads of money to get a good answer to that type of question.

Although free information is a pervasive societal trend, there are a few areas - like well-researched business analysis - where information is decidedly not free. Who is out in the marketplace paying six figures for a video teleconference system? What motivates their purchase? How many transactions per year are taking place in this market? Companies like Gartner can sell you answers at market-price.

Although you might assess a product like Cisco TelePresence, and consider its service equivalent to a free alternative, evidently you are not the target customer for a hundred-thousand-dollar product. Those customers do exist, and they willingly pay large fees. Understanding who they are, why they pay, how much they can spend, is a valuable piece of information that you probably can't get on the cheap, in the clear. If this information was available, it would throw off the entire game-theoretic basis for this type of marketplace. Nimur (talk) 16:32, 29 September 2015 (UTC)

OK, maybe the issue could be focuses from the perspective of what computing hardware and software we need, not how the market works. After all this is the Computing RD. If someone pays more for a 1 hours business conference than it costs to travel to Europe, there must be a market reason, but from a computing perspective this is irrelevant. The question is more what hardware, software and access to infrastructure do we need to hi-def low-latency video conference across an ocean. What item in there makes it not payable to lil' people like me? --Jubilujj 2015 (talk) 19:28, 29 September 2015 (UTC)

I cannot achieve hi-def low-latency video conferencing with my internet connection, even to the nearest town, so I would need a dedicated link, probably delivered via a series of specially installed microwave links which would cost many thousands of pounds. Skype is a joke here, almost unusable. Are there places where Skype isn't annoying in its delays and lags? Were I a business with highly-paid employees, I would pay the high cost of proper video conferencing just to avoid the irritating delays in cheap "skype-type" connections. I assume that the companies who provide such links have software that sends and receives multiple-redundant packets to avoid stutters and lags. Perhaps we have some experts here who know the technicalities. Dbfirs 21:15, 29 September 2015 (UTC)

Do you know that these very-expensive video conferencing systems have not had a huge slump in sales in recent years? Some businesses do use Skype. Some of them curse it daily, when it doesn't have near the reliability and robustness that some paid services have. My employer now uses Vidyo, which does seem a decent bit more reliable than free services, but I have no idea what they pay for their license. You might want to read up on the history of some of these "free" services like Skype. How do they make money? Do they make any money, aside from securing additional rounds of venture capital? Sure, someone made a bunch of money when Microsoft acquired Skype, but that doesn't meant that the unit by itself is in any sense profitable. It could just be an attractive loss leader for MS. I also recall that Skype started out as a distributed service, so that when you used skype, your computer was also being used as free infrastructure for other calls. But I think that has changed now. SemanticMantis (talk) 21:09, 29 September 2015 (UTC)

The $1000/hr price is for a "telepresence suit", like this one: [2]. A remote group meeting would almost feel like a face-to-face meeting. A company would save lots of time and money using it. The service provider even cares that rooms in both ends have the same decoration, lightning and so on, for a more realistic immersion. There are also less expensive systems, that look like as if both participants were watching each other on a hi-def TV without network hiccups for $2,500/month flat. Or you could use a desktop video-confering like Skype (which is only free for basic uses) or ooVoo, for something less than $10/month.

It looks like the whole market spectrum is covered by some type of service of varying price and quality. Scicurious (talk) 22:18, 29 September 2015 (UTC)

It's worth mentioning that "free of charge" doesn't really mean "free". If you read the so-called "privacy" statement that goes along with Skype, you'll see that the only privacy they really guarantee is that the actual content of your call will be kept private. Who you talked to, who is in your address book, your name, email, etc, etc - are all theirs to do pretty much whatever they want with. They go to a lot of pains to ask you to connect up with your Facebook page so they can suck more information out of there. That kind of data is pretty valuable. Now, imagine you're a business - all of your customers and service providers are talking to you via video conferencing - now Microsoft know who you are, what you do, AND who all of your customers are. Would one of your competitors like to know who your customers are? When you talk to them? For how long you talked? Which of your employees were talking to them? What those employees' private email addresses are so they can poach your best people? I think they'd quite like to know that stuff...and Microsoft (it seems) will be happy to sell them that information. How much is it worth to you to keep your competitors ignorant of this? Probably more than the cost of a video conferencing system that guarantees actual, useful, privacy.

That said, if you use Skype to get free video calls to your mom, who lives halfway across the planet (as is the case for me) - then having Microsoft know that is not a big deal. So pick your battles! SteveBaker (talk) 14:59, 30 September 2015 (UTC)

I work for a global enterprise organization and have used Cisco Telepresence on many occasions. We have several rooms around the world that have 3 60"+ televisions that make up one end of the room, you sit around a table that terminates at the screens, which displays the corresponding room you are conferencing with, with a similar set up it looks like you're all sitting around the same table. You can be conferences to more than one room at a time, the system automatically "focuses" on who ever is talking. You can have people who don't have access to a room dialed into the meeting and you can have people watching the meeting online, we've had meetings with 300 people in them.You also have a projector in each room which a presenter can share their computer screen to. I believe the rooms are not just "plugged into the network/internet" but use several dedicated ISDN lines. Also, keep in mind that for business travel, the cost is far more than just the airfare, there's transfers, accommodation, per diem. Vespine (talk) 23:28, 30 September 2015 (UTC)

GTA San Andreas save files

Why all GTA San Andreas saved game files I downloaded from the internet start from the game's beginning in the airport and how to fix that (I'm running licensed game version)? All have the same file title as my original save file, GTASAsf1.b. Changing file name doesn't help, even though the slot's title is correctly displayed in the game. My path is C:\Program Files\Rockstar Games\GTA San Andreas\data where saved games are stored (in Win 7). Reportedly this may have something to do with game slot number, but I don't how to fix this either. Thanks in advance.--93.174.25.12 (talk) 15:56, 29 September 2015 (UTC)

PC Gaming Wiki's page on GTA:SA has some info about incompatabilities between saved games generated by older versions and the newer ones. It also has links to converters which it says fixes these problems. -- Finlay McWalterᚠTalk 16:15, 29 September 2015 (UTC)

You must be running the "Second Edition" - all GTA San Andreas SKUs published since the Hot Coffee scandal have slightly different game logic (main.scm, script.img and several others were changed besides a file check for edited assets) hence why a downgrader is needed. Blake Gripling (talk) 10:02, 30 September 2015 (UTC)

Strength of password

Is the password '++++++++' equally strong as 'dJ+dhg3*'? A cracking algorithm won't know that I decided to repeat the same character 8 times, and when I decided to do that I had all letters and special characters available. --Scicurious (talk) 20:24, 29 September 2015 (UTC)

See Password_strength#Entropy_as_a_measure_of_password_strength. It depends on the threat model, but the entropy is fixed by the string length and the size of symbol list. Attacks designed by humans against human-generated passwords may well start with strings of constant symbols, string including sequential substrings, etc. SemanticMantis (talk) 20:46, 29 September 2015 (UTC)

Maybe I should try to incorporate the following into the section that SemanticMantis linked, but I'll put it here for now.

Basically the strength of your password is (the log of) its index in your adversary's dictionary. The problem is you don't know what's in your adversary's dictionary. It may not contain ++++++++ early on, but you can't be sure.

What randomly chosen passwords buy you is a guaranteed high probability of not being in the dictionary. If the attacker is going to test, say, 2⁵⁰ words (and this is something you can estimate, based on the price/performance of current hardware and how much they'd stand to gain by cracking your password), and your password is chosen uniformly at random from 2⁸⁰ words, the chance that they will crack your password is at most 1 in 2³⁰, no matter how cleverly they choose their dictionary. (This is similar to the fact that you will correctly guess 50% of random coin flips no matter what strategy you use; the attacker here has exactly 2⁻³⁰ odds if their 2⁵⁰ words are a subset of your 2⁸⁰, and worse odds if they aren't.) In contrast, if you invent a password in your head, there's no way to bound the probability that the password is on the list. It's dangerous to rely on your adversary not having a good model of people's psychological preferences for certain passwords over others.

Note that what matters is how you generate the password, not what the password is. You might randomly generate ++++++++. That's okay because the chance of doing so is small and is included in the 2⁻³⁰ upper bound. -- BenRG (talk) 03:01, 30 September 2015 (UTC)

True in theory, but in the real world, if your random password generator randomly picks "password" or "12345678", run it again.

Also, by international law, we can't discuss this without adding the following links:

• https://xkcd.com/936/
• https://xkcd.com/792/
• https://xkcd.com/538/

--Guy Macon (talk) 07:22, 30 September 2015 (UTC)

Per the last one, rubber hose cryptography is an interesting read. One Julian Assange worked on some of the early Deniable_encryption software to counteract this, named rubberhose... SemanticMantis (talk) 15:31, 30 September 2015 (UTC)

Well responded, Citizen Macon. And in this context we might also mention https://xkcd.com/221/ . —Steve Summit (talk) 13:52, 3 October 2015 (UTC)

@BenRG: I was personally a little disappointed in how short and unreferenced that sections is. I for one encourage you to add anything you can to it. SemanticMantis (talk) 15:33, 30 September 2015 (UTC)

BenRG's statement is true more than just "in theory", but it's the a priori probability that is known. After (it happens that) the generator has chosen "password", the a posteriori probability that its choice (i.e., that string!) is in (and near the beginning of!) the attacker's dictionary is rather larger, even if never known precisely. --Tardis (talk) 18:44, 2 October 2015 (UTC)

Those are all good answers. Let me just add that it's quite likely that the string "++++++++" just made it into at least one cracking dictionary. —Steve Summit (talk) 13:40, 3 October 2015 (UTC)

In this case, I'd change my password as soon as I get a free minute.--Scicurious (talk) 19:30, 3 October 2015 (UTC)

https problem?

Starting a few days ago, Firefox (41.0) started giving the error

“ Secure Connection Failed. The connection to www.EXAMPLE.com was interrupted while the page was loading. The page you are trying to view cannot be shown because the authenticity of the received data could not be verified. Please contact the website owners to inform them of this problem. ”

Error trips on many/most new connection attempts, to several urls (e.g. facebook, WP, google, etc.). Page always loads correctly when I click "try again", or manually reload/re-navigate. Problem has not manifested yet with other browsers, though my testing has been very limited. Problem persisted through an upgrade to Firefox. I have not noticed any similar issues with other devices using my same home network. Questions:
1. This has to be on my end, right? The error makes it sound like the server is configured incorrectly, but I can't imagine that so many servers would start having the same problem at once.

2. Is this an https problem?

3. Could this be some sort of malware or man-in-the-middle shenanigans?

4. Any ideas how to resolve the issue? Firefox is set to use https whenever available, a setting that I prefer to keep.
Thanks, SemanticMantis (talk) 20:55, 29 September 2015 (UTC)

This can occur for many reasons. One (usually benign) reason would be a corrupted certificate database. That database is a file, cert8.db located in your Mozilla user profile. You can delete your entire profile (taking a sledgehammer to the problem); or simply surgically delete only this file "cert8.db" from your profile directory; that will cause Firefox to "start over" with the default certificate database. This is the most probable cause, and the solution is directly-actionable.

The database may have been corrupted by a bug in Firefox, or an extension. It may have been caused benignly because of a crash or hang in the application. It may also have been corrupted by - and here's the bad news - malware or malicious behavior (internal or external to Firefox). (However, if it actually was malware, it didn't accomplish its job; if it had successfully modified your certificates file to inject bogus data, you'd never have seen any error message and never known a problem even existed!) Malicious corruption, in this case, is less likely than file-corruption-due-to-buggy-Firefox.

It is also possible that you actually have a different problem unrelated to the cert database - like a weird network firewall that started being misconfigured very recently. It is also possible that a lurking bug in Firefox is incorrectly showing you these warnings.

Obviously, the error message alone isn't enough to root-cause the problem... but on a hunch, try removing the cert database, and see if it helps.

Nimur (talk) 04:27, 30 September 2015 (UTC)

@Nimur: So far, so good; thanks! SemanticMantis (talk) 15:27, 30 September 2015 (UTC)

Facetime for android?

Hello everybody. I have one question: Can i download facetime for android device? My girlfriend have iphone and facetime app ;/ I want too, but i have samsung galaxy 3. — Preceding unsigned comment added by 188.68.229.62 (talk) 21:16, 29 September 2015 (UTC)

Edited: I found amazig website, and now i have facetime app on my android, this is the website: (WP:SPAMLINK redacted) Regards. — Preceding unsigned comment added by 188.68.229.62 (talk • contribs)

That appears to be a link to malware. FaceTime is an Apple product and is only available on iOS and Mac. It is not presently available on other systems. Nimur (talk) 04:06, 30 September 2015 (UTC)