The corporation I work for uses Google geocoding for addresses of business partners' stores. Unsurprisingly, often those coordinates are wrong for a number of different reasons (same address + town name exists more than once, typos...). Doing research on the reasons, I found the strange case of two addresses getting the same, and totally wrong coordinates in a different US state. The addresses are:

• A1 Quick Stop, 1674 Us Highway 130, 08902 North Brunswick, New Jersey
• B-B Beauty Salon, 149 Joyce Kilmer Ave, 08901 New Brunswick, New Jersey

Both stores have gotten the very same geo coordinates of longitude -72.54565, latitude 41.45634, which can be identified at gps-coordinates.net as Ponset, 06441 Connecticut, an uninhabitated spot of Cockaponset State Forest. As the address names don't share any similarity, I don't even have the vaguest idea how this could possibly have happened. Has anyone ever experienced something like that? --KnightMove (talk) 11:02, 15 May 2019 (UTC)[reply]

And now several thousands people all over the world (including me, of course!) start pasting these addresses into Google Maps to see what has went wrong. Do we need to mention search statistics bump...? :) CiaPan (talk) 13:24, 15 May 2019 (UTC)[reply]

All sorts of errors creep into the gigantic map software databases. Without knowing all the intricate details of every piece of software in the production and delivery of these maps, it's fruitless for us to speculate why this bug occurred, let alone why it apparently occurred twice.

If you're using Google Maps, report errors directly to Google; or report the error to whoever administers your map or business database. :Nimur (talk) 13:25, 15 May 2019 (UTC)[reply]

You're probably right. But to explain my question a bit more: There are also cases where the error can be deconstructed. For example, a German address Görna 10 was written without umlaut Goerna 10, and so the address of a Spanish store called Goerna Mecanizados was coded. My examples could have included a subtle pattern. I could have missed it, but it could have been clear to someone here who has encountered similar problems. --KnightMove (talk) 15:23, 15 May 2019 (UTC)[reply]

I think you're seeing a more benign version of the error described in this article : https://theweek.com/articles/624040/how-internet-mapping-glitch-turned-kansas-farm-into-digital-hell

Addresses that the computer can't locate are redirected to some default value for that region with a large margin of error. Unfortunately, most tools don't indicate the margin of error in the result, so to the user it looks like that default value is being returned with the same authority as any other response.

Nowadays, they supposedly make an effort to place those default values on uninhabited land, so that if there is a misunderstanding, at least it won't involve cops or pizza delivery men showing up at somebody's house. ApLundell (talk) 00:10, 17 May 2019 (UTC)[reply]

KnightMove, there is a significant problem with the addresses you gave — easy for a human to understand, but a computer might be confused. The zip code always appears at the end of a US address, not before the town name. Here's what you wrote (left) and what Americans would write (right):

Capitalization is fine, and Google Maps knows how to handle United States postal abbreviations, so I suspect that it sees the zip code and interprets it as part of the town name, and since there's no New Jersey town called "08902 North Brunswick", it rejects the location as invalid and returns its default value. (And yes, there are occasional places with numeric names, e.g. 84, Pennsylvania or 76, Kentucky.) Could you try searching for A1 Quick Stop, 1674 Us Highway 130, North Brunswick, New Jersey 08902? Thank you. Nyttend (talk) 22:18, 17 May 2019 (UTC)[reply]

@Nyttend: Thanks for the infos. The wrong format is actually my bad. We have stored the address data in a DB, they are uploaded in a CSV file, and the order for those two is the same as for all the others, most of which get the correct coordinates.

@ApLundell: Thanks, this is interesting. But I wonder: The "center" of what entity including New Jersey could this spot in Connecticut be? --KnightMove (talk) 06:53, 21 May 2019 (UTC)[reply]

Can an Excel function modify the cell it's in?

This is prompted by a colleague who keeps sending me population statistics with decimal values. I know the reason is that they are computed estimates but they should know that we need them to be integers, so I keep having to round them to the nearest integer. I do it by making an equivalent matrix of =round() values and then copying the matrix back 'as values'.

I assume that if there is a way to modify the original cells directly, it will be applicable to most or all functions. Rounding is just my example. Hayttom (talk) 16:27, 16 May 2019 (UTC)[reply]

It's not entirely clear what you want to do, but you have some options. If I have a value in cell A1 that reads as 56.1, I can make that appear as just 56 in a few different ways. I can do as you are doing, I can go into each cell and do the same kind of thing (i.e. type =round(56.1) in cell A1), or I could just alter the appearance of the cell through "Format cells", flagging them as numbers and reducing the decimal value to zero from the default of two. That last one is the easiest to do, though the value of 56.1 still exists there in the cell - it's just that you only see 56 being displayed. Matt Deres (talk) 12:28, 17 May 2019 (UTC)[reply]

Thanks Matt. I'm sorry I wasn't clear, but I need to change the real numerical content - not leaving a formula and not relying on formatting to hide the decimals. I suspect it isn't possible without the the process I use of placing a =round() function elsewhere and then manually copying its value back as the specific value. Hayttom (talk) 13:00, 17 May 2019 (UTC)[reply]

@Hayttom: A possible solution: you can just add another column (or row) of rounded values and then hide the column (or row) of original values. Benefit: both columns are in the spreadsheet available for further calculations or exports, and you can control their visibility by hiding and displaying the chosen column. This is important if actually any further calculations are needed, e.g. for values 0.333, 0.333 and 0.333 rounding to integers would yield three zeros which sum to zero, whilst the former sum up to almost one. Limitation: it's only possible if you have just one table in a spreadsheet; otherwise hiding the additional column may spoil a structure of data lying below or above the table you want to modify. --CiaPan (talk) 14:09, 17 May 2019 (UTC)[reply]

• Yes, and you should never do this.

You don't need to. Find a better way. Usually by having a hidden column alongside - the "basic" value goes in one and is hidden, the "massaged" value goes alongside it, selected by some complex selection function (or whatever). This is closer to concepts of functional programming.

To actually do this self-modification (as self-modification), you need VBA or a cruder macro. They're also a nuisance to debug. Andy Dingley (talk) 14:46, 17 May 2019 (UTC)[reply]

I was going to use my sandbox and the history, but the content has nothing to do with Wikipedia and some of it may in fact be a copyright violation.

I sent emails to myself and they were quite long, so when I was trying to delete the information, I couldn't copy and paste on my computer at home. The easiest thing was just to send myself a reply to the original email and figure it out later. Now, a lot of the emails have the same content but there's so much of it I can't see it.— Vchimpanzee • talk • contributions • 22:14, 16 May 2019 (UTC)[reply]

WWhat you're looking for is a diff tool, AKA a File comparison tool. (Wikipedia did not invent the concept.)

If your need is infrequent, and doesn't need security, I'd recommend one of the many online diff tools.

https://text-compare.com/ seems to work well.

A more heavy duty tool might be WinMerge, and if you're on a Unix-style machine, you could always use the command line tool [[diff].

ApLundell (talk) 00:01, 17 May 2019 (UTC)[reply]

Thanks. I never suspected Wikipedia was the first, but it's where I was introduced to the idea.— Vchimpanzee • talk • contributions • 15:14, 17 May 2019 (UTC)[reply]

I am working on a project (academic; non-profit) and have identified about 30 people who might be interested in it. I don't know most of them personally but I know their email addresses and have good reason to think that they might be keen to be involved. I have these details on a google spreadsheet but I've not contacted anyone yet. I want to write a semi-personalised email to each one of them, asking if they're interested. They might say yes, or no, or offer advice, or ignore the email, or ask for further information, or suggest I write to a colleague instead. Or maybe some other unforeseen response that might involve back-and-forth email exchanges. Quite often I won't be able to respond immediately for one reason or another, and in this case the problem is that I will forget to reply.

I want a system that records the various communications, and maybe do things like list every person who has not replied within a week, or every person that has said "yes".

What zero-cost software exists to help me manage this? This must be a common situation but I don't know how to describe it. I'd like to see what actions there are on me at any time. Maybe someone will email me and I'll want to see all previous correspondence with that person. Is project management software good here? Emacs org-mode? Does gmail have some facilities here? Robinh (talk) 03:19, 17 May 2019 (UTC)[reply]

What's "zero cost"? To buy, or to learn? Almost certainly, something that you already know would be more useful to you than something you adopt from fresh (even if free).

Personally I do this with MediaWiki. I run a couple of MediaWikis just for organising business and domestic stuff. Draft WP articles too. It's easy to install and you already know how to edit it. Other options would be Excel, a decent email client with good tagging, or just a text editor, lists and regular expressions.

The game changes if you want to share that information amongst a group of people. Then you might look at Trello, which is web-hosted. Andy Dingley (talk) 16:34, 17 May 2019 (UTC)[reply]

Also consider Dokuwiki which supports groups well. The syntax is a little different from MediaWiki, but the great thing is that it avoids the overhead of having to run a database system – all storage is in plain text files. Martin of Sheffield (talk) 08:38, 18 May 2019 (UTC)[reply]

Thanks for this. "Zero cost" means no cost to purchase, a steep learning curve is no problem. Everything I want to do is for my own reference to keep myself organised, so there is no issue with sharing information. I do like the idea of a personal mediawiki though. thanks, Robinh (talk) 03:48, 18 May 2019 (UTC)[reply]

(OP, somewhat confused...someone suggested I look at customer relationship management but the edit seems to have disappeared). Anyway, the phrase I needed was exactly that: customer relationship management. Thanks! Now I can google "open source CRM software" and see what's out there. The software I can see looks somewhat OTT for my use-case, but that's good to know. Knowing that little phrase makes all the difference. Thanks, really. Robinh (talk) 05:23, 18 May 2019 (UTC)[reply]

1. How can I check to see if I have messages from people on my twitter account? I hardly use twitter by the way. 2. On facebook when I see the chat thing that has contacts for some weird reason it shows when people were on facebook a few minutes ago and not current people logged into the chat. Please help 204.239.8.205 (talk) 03:25, 17 May 2019 (UTC)[reply]

This was originally posted at WP:RDL, but it didn't seem related to natural or "normal" constructed languages, and it's definitely computing, so I decided to move it here. Nyttend (talk) 22:03, 17 May 2019 (UTC)[reply]

I saw this in "Azby-Shiftwords". Like "loopwords" on wikipedia, I want a list of words that Atbash plus Caesar to eachother.

A cipher would be: first you Atbash cipher a word, then you Caesar cipher it.

The shift of 5 for example would be plain - abcdefghijklmnopqrstuvwxyz coded - edcbazyxwvutsrqponmlkjihgf

A program would be like the program for Caesar cipher pairs(https://github.com/duckythescientist/CaesarPairs/blob/master/CaesarPairs.py) modified for Atbash. My modification of it didn't work:

1. !/usr/bin/env python

'rots.py -- find caesar shift pairs in american-english' 'uses dictionary found in /usr/share/dict'

def getpairs():

   pairs = []

   try:
       f = open("/usr/share/dict/american-english", "r")
   except IOError, e:
       print "*** file open error: ", e
   else: 

       strs = f.readlines()
       raw = [x[:-1] for x in strs]
       raw2 = [x.replace("'s", "") for x in raw]
       raw3 = [x for x in raw2 if len(x)>2]
       vocab = set([x.lower() for x in raw3])  
       
       for word in vocab:
           for shift in range(0, 26):
               atbash = "".join([chr(97 + (27 - (ord(x) - 96) % 26)) for x in word])
               caesar = "".join([chr(97 + ((ord(x) - 97 + shift) % 26)) for x in word])
               if caesar in vocab:
                   print word, caesar, shift
                   pairs += tuple([word, caesar, shift])

   return len(pairs)/3

if __name__ == "__main__":

   print getpairs(), "pairs found"
   print "note, each pair has a reverse shift as well"

Can you post a list of pairs or a working program to do this?

I had to look up Atbash but it is pretty simple. The basic idea is to use a fast lookup table (Python dictionary or set) to store all the unenciphered words from your wordlist, then run through the wordlist, encrypt each word, and see if the encrypted version appears in the lookup table. I'm assuming your goal is to learn Python rather than just get the resulting wordlist. So I'd say you are basically on the right track. But you compute the atbash and caesar encryptions separately (using the word as input to both) and then do nothing with the atbash. You want to use the atbash output as the caesar input instead. The main other suggestion I'd make is split the program into separate functions (read and clean the wordlist, do the atbash cipher, do the caesar cipher) so they can be studied separately. That makes the program more modular and easier to understand. 173.228.123.207 (talk) 03:35, 18 May 2019 (UTC)[reply]

I'll be away thru the weekend so will post my version below. It's not polished or anything but just followed what seemed like a natural breakdown of the problem. I avoided using too many slick python-isms like generator comprehensions so it's less concise than it could be, and various optimizations are possible like using byte arrays for the words. It's python 2 and doesn't attempt to do anything sane with unicode.

import string
wordfile = '/tmp/words'

def get_wordset(wordfile):
    ws = set()
    with open(wordfile) as f:
        for line in f:
            word = clean_word(line)
            if len(word) > 2:
                ws.add(word)
    return ws

def clean_word(line):
    w = line.strip().lower()
    if w.endswith("'s"):
        return w[:-2]
    return w

def atbash(letter):             # letter must be ascii lowercase
    return chr(ord('a') + ord('z') - ord(letter))

def caesar(letter, shift):      # letter must be ascii lowercase
    return chr(ord('a') + (ord(letter) - ord('a') + shift) % 26)

# encrypt word using both the caesar and atbash ciphers
def encrypt(word, shift):
    result = []
    for char in word:           
        # chars might be non-letters, so only encrypt the letters
        if char in string.lowercase:
            d = caesar(atbash(char), shift)
        else:
            d = char
        result.append(d)
    return ''.join(result)

def main():
    wordset = get_wordset(wordfile)
    n_pairs = 0
    print len(wordset),'words'
    for word in wordset:
        for shift in range(14):
            if encrypt(word, shift) in wordset:
                print word,shift,encrypt(word, shift)
                n_pairs += 1
    return n_pairs

if __name__ == '__main__':
    print main(),'pairs found'

173.228.123.207 (talk) 06:25, 18 May 2019 (UTC)[reply]

This helps, but you need to change "14" to "26" since each shift reverses itself and the pairs for let's say 14 are different from 12, and change "/tmp/words" to "/usr/share/dict/american-english"

Hi there, I'm trying to setup mediawiki and am trying to implement math typesetting. But:

wget https://extdist.wmflabs.org/dist/extensions/Math-REL1_23-d0e998f.tar.gz
--2019-05-18 20:53:34--  https://extdist.wmflabs.org/dist/extensions/Math-REL1_23-d0e998f.tar.gz
Resolving extdist.wmflabs.org (extdist.wmflabs.org)... 185.15.56.49
Connecting to extdist.wmflabs.org (extdist.wmflabs.org)|185.15.56.49|:443... connected.
HTTP request sent, awaiting response... 404 Not Found
2019-05-18 20:53:35 ERROR 404: Not Found.

can anyone advise?

Where did you get that URL? If you look at the directory on that server (https://extdist.wmflabs.org/dist/extensions) you will see that there are seven versions of the Math tarball, labeled REL1_23 to REL1_32 (with some gaps). You seem to be trying to download REL1_23, but the filename on the server is different than the one you have. Try https://extdist.wmflabs.org/dist/extensions/Math-REL1_23-5d564a6.tar.gz if you really want the old 1_23 version. CodeTalker (talk) 16:52, 18 May 2019 (UTC)[reply]

(OP) thanks for that, I now have the tarball. I got the URL from following the "how to install mediawiki" page: https://www.mediawiki.org/wiki/Extension:Math/advancedSettings It would be good to let those guys know they have an out-of-date URL there but frankly I do not have the confidence to edit the page. Thanks again, Robinh (talk) 20:20, 18 May 2019 (UTC)[reply]

(OP again). thanks again. But https://www.mediawiki.org/wiki/Extension:Math/advancedSettings says to untar the tarfile, then tells me to compile the sourcecode with "Make". But there is no Makefile, as a banner says. How then to proceed? Robinh (talk) 22:38, 18 May 2019 (UTC)[reply]

I'm in an apartment complex where Comcast basic cable is included in the rent. There is a TV in the living room whose HDMI input is connected to the HDMI output of a cable box, so if you want to change channels, you use the cable box remote rather than the TV remote.

Someone here wants to install an additional TV in the bedroom. Will that need a second cable box? Do users normally have to pay Comcast an extra monthly charge to have the second box? The bedroom TV will sometimes be used for listening to music and news shows (i.e. audio-only). Is it normally possible with a TV to turn off the screen (so the room is dark) while leaving the audio turned on? The second TV will have to be purchased, so if that's a special feature to look for, we can do that. Sorry if these are dumb questions but as a computer junkie I'm quite unfamiliar with the TV world. Thanks. 67.164.113.165 (talk) 22:20, 18 May 2019 (UTC)[reply]

Added: what the person really wants for news and music is an FM radio, but for some reason the radio reception in that room is terrible, so installing another TV set is plan B. But it will probably get some use as a TV set as well as an audio device, so it is fine. 67.164.113.165 (talk) 22:22, 18 May 2019 (UTC)[reply]

Since you are talking about a service included with the rent, the apartment manager would be the first person to talk to. They will be familiar with people wanting a second TV and ought to know how it is done there and can tell you about any extra cost. --76.69.46.228 (talk) 04:02, 19 May 2019 (UTC)[reply]

Does the cable box not have an alternative output? If not, then you could use an HDMI splitter. Dbfirs 16:32, 19 May 2019 (UTC)[reply]

Thanks all, and sorry for the slow response. Yes I guess I should talk to the building people, good point 76.*. Dbfirs, hmm, the cable box might have an alternate output, but the bedroom location is maybe 20 feet away, past a hallway. So it sounds like there would be no way to change the channel from the other room? And with an HDMI splitter it sounds like both TV's would have to be on the same channel (maybe that is ok). I'll look into this but will also check out further alternatives, like maybe a fancy FM antenna. 67.164.113.165 (talk) 08:20, 20 May 2019 (UTC)[reply]

Yes, sorry, my solution of a splitter is not what you want. Dbfirs 10:35, 20 May 2019 (UTC)[reply]

I use a computer and a Samsung tab 3v phone for my gmail. How can I hide the display of gmail notifications displayed on the home screen. thank you2402:4000:2180:919A:4480:2E98:C195:E7E9 (talk) 06:11, 19 May 2019 (UTC)[reply]

I don't know about that specific phone, but normally you would go to Settings > Notifications > Gmail, and turn off "Allow notifications".--Shantavira|^{feed me} 08:29, 19 May 2019 (UTC)[reply]

I have an XL spreadsheet containing a list of dates, mapped as 3 numerical columns by day, month, year, e.g. today 19 May 2019 is mapped as:

I want to produce a 4th column with a number containing all of the digits in the first 3 columns, in ascending order, and ignoring zeroes. In this example the 4th number would be 112599.

How do I do it? -- Jack of Oz ^{[pleasantries]} 10:40, 19 May 2019 (UTC)[reply]

Row, not column.
Sleigh (talk) 13:32, 19 May 2019 (UTC)[reply]

This is pretty ugly, but the following formula should work (assuming the values are in B2, B3 and B4):

=REPT(1,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,1,""))) 
&REPT(2,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,2,""))) 
&REPT(3,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,3,""))) 
&REPT(4,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,4,""))) 
&REPT(5,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,5,""))) 
&REPT(6,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,6,""))) 
&REPT(7,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,7,""))) 
&REPT(8,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,8,""))) 
&REPT(9,LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,9,"")))

To explain this from the inside out:

B2&B3&B4 concatenates the values, giving 1952019

SUBSTITUTE(B2&B3&B4,1,"") replaces every 1 in the string with the empty string, giving 95209 (and similarly for the digits 2-9)

LEN(B2&B3&B4)-LEN(SUBSTITUTE(B2&B3&B4,1,"")) gives the difference in the lengths of the string with and without the digit, which gives the number of time the digit occurs

REPT(1, <complicated expression>) produces 1 repeated <complicated expression> times

Finally, the resulting strings for each digit are concatenated.

And I have to ask: why do you need to do this? AndrewWTaylor (talk) 15:05, 19 May 2019 (UTC)[reply]

Brilliant! That works beautifully. Thanks, Andrew.

All will be revealed in due course. Patience, my child. :) -- Jack of Oz ^{[pleasantries]} 21:27, 19 May 2019 (UTC)[reply]

Resolved

This question arose in my mind due to the recent bot-disruption. But my question is by no means limited to the attack on the refdesk.

I know spambots are a common creature on the internet, unfortunately - and on Wikipedia too, inevitably. Ditto for malwarebots. My question is not about these sorts of bots. Spamming has a clear commercial motive (and sometimes other motives - you do get the odd crank).

My question is about the use of bots for trolling or vandalizing. either wikipedia, or other internet forums. Just how common are the use of bots for this purpose? Are genuine "trollbots" or "vandalbots" a common phenomenon on the internet? How common are they compared to the spambots? (I would speculate that there would easily be 50 or more spambots for each trollbot or vandalbot, but I'm curious if any studies have been done on the use of bots merely for trolling or vandalizing internet pages/forums). I'm also curious to hear (anecdotally) from those who are regular editors on Wikipedia: How common is it to encounter a trollbot or vandalbot vis-a-vis a "common spambot"? Eliyohub (talk) 13:49, 19 May 2019 (UTC)[reply]

We had a trollbot on the refdesks here whose purpose was to libel regular users, especially those who had reverted vandalism. It was difficult to keep up with the edits until it was eventually blocked and the edit summaries revdel'd. I don't know of any research. Perhaps someone else does? Dbfirs 16:27, 19 May 2019 (UTC)[reply]

Yeah, I'm aware of the refdesk attack, which is what made me think of this question. I'm wondering if it was an isolated episode, or if trollbots and vandalbots are a broader threat - either on or off Wikipedia. Anyone encountered them, besides here on the refdesk? Eliyohub (talk) 17:26, 19 May 2019 (UTC)[reply]

Proper vandalbots are exceedingly rare. I could probably count them on one hand. Most things that we might call vandalbots are actually spambots or other automated things gone wrong instead of a deliberate attempt to disrupt. People also call repeated vandalism on dynamic proxies vandalbots, but they are not always automated. Vandalbots are of course usually just a means of self-promotion, which is not too different from spam. We can basically treat them the same as any other spamvandalism. And I don't think I've ever even heard of a trollbot around these parts. -- zzuuzz ^(talk) 17:51, 19 May 2019 (UTC)[reply]

zzuuzz thanks for your answer. What do you make of the mass attack on refdesk, referred to earlier by Dbfirs? Was it your first encounter with an actual trollbot? Or would you classify it as a vandalbot? Or do you believe it was not automated? Eliyohub (talk) 18:32, 19 May 2019 (UTC)[reply]

I've seen various bots before. This had all the characteristics of a vandalbot, IMO, and I could even tell you what script language it was written in. When someone mentions trollbots, I think of Twitterbots or Twittertrolls. I guess there was an element you could call trolling, but that's not the term I would use. -- zzuuzz ^(talk) 18:55, 19 May 2019 (UTC)[reply]

• Most 'bots are content neutral. There's a 'bot framework, then there's an adapter to tailor that framework to a particular host platform, such as WP, and then there's a vandal- or spam- 'bot app which runs on top of that. Some of the chatbot frameworks now also take information from other sites, such as a control site of "topics to slide into conversations" or a respectable news site such as CNN or the BBC, so that they can produce enriched content replies such as <Alabama is mentioned> – "What do you think of Trump's comments about it?" – <CNN story of Trump and Alabama>.

Once the platform adapters are in place though, or even (in complex 'bots) some natural language processing, then choosing whether the 'bot acts for good or evil, and which sort of evil, is a simpler task. Andy Dingley (talk) 10:17, 20 May 2019 (UTC)[reply]

The issue with Wikipedia isn't bots. That is part of the system - tons of bots scan and update Wikipedia. The issue is automatically creating a bot account that appears to be a full-fledged user. So, you have to automate the CAPTCHA and then automate necessary trivial edits to get the account authorized to make edits. When I was first shown this whole reference desk thing, there was a person who said he was developing a tool specifically to do that because he was upset about the owner or something. I assume it isn't just one guy. There must be multiple people developing tools like that. 12.207.168.3 (talk) 12:16, 20 May 2019 (UTC)[reply]

Can't make video calls on this laptop on Facebook cuz it says please turn on the camera/microphone to allow. But doesn't show how. It only links to a Facebook help page with no instructions. A quick look at the Control Panel, don't see a icon for it. 2601:246:4100:C687:20A3:96EF:A5E1:8B90 (talk) 14:37, 19 May 2019 (UTC).[reply]

Does your laptop actually have a camera and microphone? Most do, so if you tell us the make and model, we might be able to Google the instructions for you. Alternatively, you could do this yourself? Dbfirs 16:21, 19 May 2019 (UTC)[reply]

I don't see a name for the webcam, but the laptop itself is a Dell. Windows 7. 2601:246:4100:C687:20A3:96EF:A5E1:8B90 (talk) 16:29, 19 May 2019 (UTC).[reply]

I don't understand PARI/GP and I'm trying to understand and translate some code. I'm stuck on the line

R = concat(R, p^k*select(x->(x%p),invsigma(n\d,d)) );

I don't know what the select function does (I couldn't find it in references) and I don't know what (x->(x%p) does. Can someone help? Bubba73 ^{You talkin' to me?} 01:24, 20 May 2019 (UTC)[reply]

(01:23) gp > ?select
select(f, A, {flag = 0}): selects elements of A according to the selection
function f. If flag is 1, return the indices of those elements (indirect
selection).

If A is a vector or list, then select(f, A) is the vector or list of the elements x in A for which f(x) is non-zero.

x->(x%p) is the function which takes x to x%p, where p is some value. In other words, x->(x%p) is the function f defined by f(x) = x%p, but "x->x%p" doesn't name the function.

select(x->(x%p),invsigma(n\d,d)) is the elements in invsigma(n\d,d) which are not divisible by p. Example:

(01:30) gp > select(x->x%3,[1,2,3,4,5,6,7])
%4 = [1, 2, 4, 5, 7]

PrimeHunter (talk) 23:44, 20 May 2019 (UTC)[reply]

Thanks, I think I see. Bubba73 ^{You talkin' to me?} 00:04, 21 May 2019 (UTC)[reply]

I use Google Maps, but can barely read the text after I take a snapshot with my (non-smart) phone. While still on the PC, I try to zoom in using Google Map's own "+" sign, but that doesn't increase the text size. Zooming in on the Chrome window works ("CTRL +"), but reduces the portion of the screen occupied by the map window to a postage stamp on the lower, right corner. So, is there a way to get large text on a full-sized Google Map, or with Mapquest or any other utility ? SinisterLefty (talk) 23:42, 20 May 2019 (UTC)[reply]

This is a common problem that has been around for years. There is absolutely no way to increase the font size. In the past, you could trick it into using a font that was slightly larger, but that "bug" was fixed. You are stuck with microscopic text and it will not be fixed. The issue is posted regularly to Google's bug tracker and immediately closed as "known - not a problem." If you have a smart phone, you can turn on disability mode which lets you triple-tap the screen to zoom in the display. Then, you can read the name and triple-tap to close the zoom-in mode. On a computer, you can use the snip tool to take a screenshot, open it in something like Paint, and zoom in. I hope that once there are enough developers at Google who are over 40, they will realize that microscopic text is not useful. If they don't want us to read the street names, just don't clutter up the display with them. But, right now, the cool script kiddies think that a 6-point font is cool. Anyone who doesn't agree is lame. 97.82.165.112 (talk) 01:10, 21 May 2019 (UTC)[reply]

Thanks. And yes, I've used Paint to magnify the map. But even at only 2X it then fills over 4 screens, requiring multiple screen shots on my non-smart phone. I want the text bigger, but not the roads. (This could cause overlapping text, but I'd be willing to risk it.)

Another option is to upload the PC image to my phone, so I can zoom in there without it resizing the text back down or becoming blurry. Lots of extra steps there, though, and lots of zooming and panning on the phone to get text to a readable size. Not the type of thing I'd like to do while stopped at a traffic light. Still, this is the best option I've found so far. SinisterLefty (talk) 06:28, 21 May 2019 (UTC)[reply]

I was wondering if I got a new mobile phone but kept the same number would I have the same cookies tracking me? I am getting a new mobile regardless of the answer. (78.17.116.81 (talk) 20:21, 21 May 2019 (UTC))[reply]

A cookie is a specific type of tracking technology, related to the HTTP protocol, that would not propagate to the new device. In this context, "cookies" are specific to one instance of the web browser on the original phone, and do not transfer to new browsers on new devices.

Various other methods exist for web tracking, including but not limited to associating you (as a person) to a specific individual device; or to a specific telephone number; or guessing or inferring that you are the same individual human based on ancillary activities that you perform. In addition, there are many "cookie-like" software features that might migrate to the new device depending on how you set it up or transfer accounts or other data. Furthermore, one or more new HTTP cookies might be created on the device that can somehow be associated to the old cookies on the old device.

It is quite difficult to prevent tracking, because so many different methods of tracking exist, including various passive tracking methods that can occur even if you and every invisible piece of your device technology actively try not to cooperate with the trackers.

Our main article is Web tracking, and you may also be interested in reading the website of the Electronic Frontier Foundation's privacy advocacy organization.

Nimur (talk) 23:02, 21 May 2019 (UTC)[reply]

Which was the first mass-produced home computer with a terminal ? 223.191.16.212 (talk) 08:41, 22 May 2019 (UTC)[reply]

When you say "with a terminal" are you literally meaning a standalone computer attached (probably by RS232) to a separate terminal? Martin of Sheffield (talk) 08:49, 22 May 2019 (UTC)[reply]

Yes .223.191.16.212 (talk) 09:33, 22 May 2019 (UTC)[reply]