User talk:Magog the Ogre/Maps FAQ

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

SVG Maps[edit]

I noticed you commented that "cropping an svg map is a difficult process." It can be, but have you considered just a clipping mask, or setting the page limits to be the area you need? That way you can keep the map intact, because the mask is non-destructive, and just output what you want. Just a small point, but if the maps were uploaded as svg they would be more scalable, and the possibility would exist for future changes by other editors.  Begoon•talk 12:00, 11 September 2010 (UTC)[reply]

I'm most certainly nothing near an expert at Inkscape. I will most definitely take a look into that next time I have an upload, thank Begoon. Magog the Ogre (talk) 12:16, 11 September 2010 (UTC)[reply]

You can use Adobe Illustrator too, for the clipping mask, if you have that, or, if you like, just upload the svg, and a cropped bitmap to show the correct area, and ping me, and I'll edit the svg for you to give you an example file you can look at. It's no big thing, but if we can have the maps as vectors then so much the better. If it turns into a problem, well, never mind in that case.  Begoon•talk 14:20, 11 September 2010 (UTC)[reply]

1) To respond, nononono, I definitely don't have Illustrator. Ha, I use Linux; you think I spend that kind of money on a product? 2) Take File:Giant Food footprint.png for example. That image would have been 10MB if I cropped, because the image file would have contained information for the entire United States (cropping doesn't remove it, IIRC, just changes the view window). That seems like a waste of Wikimedia's space, but also a waste of time to download and/or view from scratch. Magog the Ogre (talk) 22:39, 19 October 2010 (UTC)[reply]

Line to SVG?[edit]

Hiya, I was wondering if you considered using the various shape files the government gives away for free to make the maps? The datasets include (if you grab them all), area boundaries (ie: incorporated areas such as cities, towns, hamlets; as well as water bodies), streets, street types, etc. Since you'd be working with "shape files" where everything is a coordinate system, you could draw and scale any size map you wanted with full clarity, as well as easily do crop regions with simple math to simply cut a line on the bounding box. Just a thought...

If you're interested in seeing what it would take, let me know if there's anything I can do to help out (info, where to find what files, some simple calculation routines to get you started, etc). Best, ROBERTMFROMLI TALK/CNTRB 03:24, 17 September 2010 (UTC)[reply]

I did try to do this, but I ran into some problems with changing from shape to svg. I even tried to talk to the author of one of the software packages, but it just got to be too complicated, so I did the exhaustive work of taking the images on commons and mathematically calculating it to create an entire US map of the 48 states, with counties. Do you want me to upload it? I didn't because it's really big (~10MB), and takes forever to render. Magog the Ogre (talk) 21:13, 1 October 2010 (UTC)[reply]

C++ Program[edit]

You mentioned in the FAQ using a C++ program to plot the points. Is this publicly available software or something you created? If it's the former, can you name it. If it's the latter, consider this another request to distribute it. Cains (talk) 01:18, 27 January 2011 (UTC)[reply]

harris bank footprint map[edit]

your map for Harris Bank and Marshall & Ilsley need to be combined because the two banks have merged to become BMO Harris Bank — Preceding unsigned comment added by Mgoblue144 (talkcontribs) 23:25, 25 June 2012 (UTC)[reply]

Proposal[edit]

Hey Magog, I think the footprint work you're doing is great. I had an idea that might help you out in this process. My company has actually already done all of the hard work of collecting chain location data, and we do it in a similar fashion as you, by going to the company websites and extracting the data. You can see what we have at our website, aggdata.com. I'm wondering if you would be interested in using this data for your maps; we may be able to work out some sort of partnership. Feel free to respond here or email me directly at chris at aggdata dot com. Hathawayc (talk) 20:06, 6 December 2012 (UTC)[reply]

Hey there Chris. Sorry for the long time in my response. Just some brainstorming: part of the problem with my code is that, well, it's awful. I wrote it ad hoc for my own purposes, and used all sorts of ugly shortcuts that violate basically every principle of programming. So even if I provide the code for you, it's so bad you'd probably just pilfer parts of it for yourself and do a rewrite. What sort of ideas do you have in terms of a partnership? Magog the Ogre (tc) 23:52, 30 December 2012 (UTC)[reply]
Are you talking about the code for pulling location data or the code for creating the maps? We have the location data extraction down pretty well. My thought was that you could use our location data to create more of your dot maps on company wikipedia pages, and then there would be some sort of mention of AggData in the image credits. Thoughts? Hathawayc (talk) 05:25, 6 January 2013 (UTC)[reply]
No, I was referring to the code for creating the maps themselves. It's written in an awful mishmash of C++, which creates an executable that needs specific input variables, which then outputs an SVG (which is probably non-compliant with standards), which I then crop and convert with GIMP into a PNG. If you don't mind, I suppose I could provide you the code along with step-by-step instructions; you could cross your heart, say a prayer, and get it to work. Magog the Ogre (tc) 18:18, 6 January 2013 (UTC)[reply]
Hmm, we've actually got some internal code that can generate maps as well, using R. I think we're mostly interested in finding someone who would like to post the maps to Wikipedia, with all the proper markup and attribution. Is that something you would be interested in, or perhaps you know of someone who would be? Hathawayc (talk) 20:47, 7 January 2013 (UTC)[reply]