Wikipedia:Reference desk/Archives/Computing/2015 May 28

= May 28 =

Extracting a graph from a PDF
I have a proposal at

[ https://meta.wikimedia.org/wiki/Community_Tech_project_ideas#Start_a_project_--_a_real_project_with_measurable_goals_and_a_schedule_--_to_reduce_page_weight ],

and I would like to include the graph on page 5 (the second to the last page) at

[ http://www.itu.int/en/ITU-D/Statistics/Documents/facts/ICTFactsFigures2015.pdf ]

to support my argument. I am not much of a PDF wrangler; is there an easy way to extract that page so it can be uploaded to commons?

Optional tweak if it is easy to do: remove the word "persist: from the top line so it reads "Differences in broadband speed".

Note: Graphs, charts, and tables aren't considered "original works" and therefore don't meet the requirements for copyright protection. See discussion at Graphics Lab/Illustration workshop/Archive/Mar 2014.

It seems like there should be a simple way to do this. --Guy Macon (talk) 02:48, 28 May 2015 (UTC)


 * If you can't find any better method, you could always do a screen grab. One downside is that you are then limited to the resolution of the screen at the time of the grab.  Rotating the screen 90 degrees might help a bit, or you could do multiple screen grabs and then stitch them together.   You can then edit the pasted screen grab in any bitmap editor, such as MS Paint, which allows you to add text.  (If you don't know how to do screen grabs or use MS Paint, just ask.) StuRat (talk) 03:49, 28 May 2015 (UTC)


 * Another option might be to "print" page 5 to a file, then find a graphics program which can read that in. This would get past the resolution limitation above.  StuRat (talk) 03:51, 28 May 2015 (UTC)


 * I don't think you can upload this to Commons. Your source for the claim that charts are not original works is Hazmat2, who isn't a lawyer as far as I can tell, and who is wrong about this as far as I know. Information as such isn't copyrightable, but layout, color and font choices, and the like are. -- BenRG (talk) 04:42, 28 May 2015 (UTC)
 * I agree you should seek a second opinion relating to the specific work as it will depend significantly on whether the chart or graph meets the threshold of originality. Of the cases I can find that came up, 2 were kept Commons:Commons:Deletion requests/File:Alleged Luka Magnotta Murder Video View Statistical Data.png, Commons:Commons:Deletion requests/Graphiques SpritMonitor de Vi..Cult... (well multiple images there which interestingly remained tagged as CC licenced although I think the graphs were created by a contributor anyway, it only came up for discussion due to database rights) and 1 was deleted Commons:Commons:Deletion requests/File:Graph of ranks of first ten companies (1992-2013).jpg. Interestingly all these were dealt with by 1 individual, I don't know if that was a coincidence, due to my search terms, or that editor has experience with graph related copyright issues on commons. Nil Einne (talk) 16:07, 28 May 2015 (UTC)


 * Open the PDF in Inkscape, select page 5 as the one you want to import. Then delete all the stuff on the page you don't want and manipulate it as you like. Then you can save the result as an SVG or PDF, or you can export a region as a PNG. -- Finlay McWalterᚠTalk 07:14, 28 May 2015 (UTC)


 * I know that the normal reaction to the previous suggestion is "Holy crap on a stick! I can't install a program! I can't open a file! I can't delete anything!" However, this is the best suggestion. Inkscape is free. It does a great job at opening and saving PDF files. You will probably have to ungroup the elements of the file (you will see the ungroup function in the menus). Once you try this, you will be able to alter PDFs with ease. 209.149.115.214 (talk) 12:00, 28 May 2015 (UTC)


 * If the graphics in question were bitmaps (like photos) then I'd recommend Evince instead. But in this case the histogram isn't a bitmap, but PostScript drawing objects, so Inkscape is the better option. -- Finlay McWalterᚠTalk 14:30, 28 May 2015 (UTC)


 * I have no problem installing programs (that's what virtualization is for - to remove the fear that a new program will screw everything up) and will try Inkscape.


 * I will bring up the copyright issue on the appropriate noticeboard and will post a link here. Thanks to all for the examples. --Guy Macon (talk) 19:07, 28 May 2015 (UTC)

Some clients seem to be trying to connect to a password
Some clients seem to be trying to connect to a password. Scanning with airodump-ng I often see stuff like:

BSSID             STATION            PWR    Rate    Lost  Packets   Probes F8:XX:XX:XX:XX:XX E0:XX:XX:XX:XX:XX    0    0 - 0      7       39   ,1KNBKJNKJK2788348761,J7b7CPxpxePQ7yCFcjFK        ... ... ... .

Those 1KNBKJNKJK2788348761,J7b7CPxpxePQ7yCFcjFK look incredibly close to passwords. Are clients probing for passwords in the mistaken belief that these are networks? Is there an alternative explanation or are people that stupid? --Yppieyei (talk) 15:25, 28 May 2015 (UTC)

Data plan, contract limitations, counting the bites
If a data plan offers x GB for surfing, and y GB for downloading, what do they count as surfing and what do they count as downloading? If I use Skype, will Skype route calls through my device and therefore spend my data quota? Is uploading also counted as downloading? It sounds contradictory, but how can they limit users uploading big files, or seeding torrent files?--Yppieyei (talk) 15:27, 28 May 2015 (UTC)


 * An ISP can, to a limited degree, differentiate some kinds of traffic from another using packet inspection (particularly the port numbers and related stuff in packet headers) and deep packet inspection on the contents (particularly the protocol-specific header information) of packets. Which kinds of data they chose to put into which billing categories is a function of how they've configured their metering software, and you'd have to ask a specific ISP how that's done. One might imagine that "surfing" means HTTP and "downloading" means streaming and bittorrent - but those aren't technically specific terms, so this doesn't have a general answer.  Note that encryption, whether ssh, http-over-ssl, or VPN traffic should be opaque to the ISP's metering inspector, so will all show up as a generic traffic. Unencrypted BitTorrent traffic isn't especially hard to detect. -- Finlay McWalterᚠTalk 16:10, 28 May 2015 (UTC)
 * Beyond asking them, reading or searching the terms and conditions may provide some clues. Also it's worth remembering that there are still many what most would consider downloads done over HTTP. Nil Einne (talk) 16:13, 28 May 2015 (UTC)