#geowebchat transcript, 1 October 2013: Techniques and conventions for geoweb cartography


12:01 PM – 1 Oct 13 Starting now: about techniques and conventions for web cartography. More info here:

12:02 PM – 1 Oct 13 Try as a way to follow along w/ the conversation. Make sure you always include the hashtag in your tweet!

12:03 PM – 1 Oct 13 And if you’d rather ignore the conversation, see if your twitter client lets you mute the hashtag.

12:04 PM – 1 Oct 13 Today our big question is: does web cartography follow the same rules as traditional paper cartography?

12:06 PM – 1 Oct 13 Bad cartography knows no bounds – general rules of thumb hold true for digital and paper

12:08 PM – 1 Oct 13 Are we talking neocartography vs. traditional cartography? Professional vs. non-professional?

12:09 PM – 1 Oct 13 Yes, bad cartography is everywhere. We could argue whether the geoweb makes bad carto easier or harder, but…

12:10 PM – 1 Oct 13 I’d agree – web cartography give a huge toolbox that is not always used carefully. Wind map is a good example

12:11 PM – 1 Oct 13 Is the wind map an example of the “huge toolbox” or the “not always used carefully”?

12:11 PM – 1 Oct 13 We could talk about both axes. But can we talk about non-professional mappers (on or offline) having “conventions”?

12:11 PM – 1 Oct 13 there are print v web conventions, e.g. you are forgiven for no scale bar on a slippy map, right?

12:13 PM – 1 Oct 13 both. I found the wind map to be (beautiful) eye candy, nothing else. It was an animated texture choropleth.

12:13 PM – 1 Oct 13 wind map violated the ‘leather pants principle’. Just because you can, doesn’t mean you should.

12:13 PM – 1 Oct 13 some conventions are instilled by tech choices up early on: see the over-use of web Mercator

12:13 PM – 1 Oct 13 I think being forgiven for lacking a scale bar should depend on the use of the map. Some slippy maps probably need them.

12:14 PM – 1 Oct 13 I’m inclined to agree. Scale on screens, much less web maps is tricky territory.

12:15 PM – 1 Oct 13 I’m guilty of teaching the mantra of always-include-a-scale-bar to print carto students, but it isn’t always needed.

12:16 PM – 1 Oct 13 Scale bar on something you really don’t measure?

12:16 PM – 1 Oct 13 scale bar is a skeumorph

12:17 PM – 1 Oct 13 I think we can say that web maps should never have a representative fraction scale, but a bar could be meaningful

12:17 PM – 1 Oct 13 I tell my students it’s like Miles Davis on Jazz: you got to learn the rules before you can break’m

12:17 PM – 1 Oct 13 (Who knew we’d have a all about scale bars!)

12:18 PM – 1 Oct 13 Scale bar only makes sense in a equidistant projection

12:19 PM – 1 Oct 13 Yes, in all cases you should be free to break rules when needed, but should we teach different rules for the web?

12:19 PM – 1 Oct 13 I hate when js heat map libraries use zoom level to control kernel size, but not sure how else to do it optimally

12:25 PM – 1 Oct 13 So true Go to Zoom in an go up and down to see scale ajust to the map center

12:26 PM – 1 Oct 13 A comment from the audience at : Flea from the Red Hot Chili Peppers never learned to read music… who needs “rules”?

12:28 PM – 1 Oct 13 – Fairly certain that’s not true – he just never learned to read music for bass (he learned for trumpet)

12:30 PM – 1 Oct 13 Do we want to drop the word “rules” & use “conventions” instead? Is that less loaded? And this learning can take place anywhere.

12:30 PM – 1 Oct 13 – but to spin that out, some web designers are great cartographers (same field, diff. instrument)

12:31 PM – 1 Oct 13 “Mussolini had rules” -another comment from the audience

12:32 PM – 1 Oct 13 Very good point about design. If you’re a good designer can you figure out what makes good maps work?

12:31 PM – 1 Oct 13 Let me ask. what are the carto conventions by which you judge web maps?

12:32 PM – 1 Oct 13 So what conventions hold true across print and web cartography?

12:32 PM – 1 Oct 13 Most carto guidelines hold for both digital and paper: color, font, layout balance. Unique to digital = zoom level design!

12:33 PM – 1 Oct 13 !!! Sorry I’m late to the party. Did anyone respond to this point. It’s one of my pet peeves too.

12:33 PM – 1 Oct 13 I've been thinking a lot about interactivity. What features do I expect to be clickable, identifiable

12:34 PM – 1 Oct 13 …and can good design be taught? How much of our “cartographic conventions” are just prescriptions to prevent bad design?

12:34 PM – 1 Oct 13 There aren’t rules. Just “what worked well before that might work well now.” Sometimes newbies make the coolest stuff though.

12:34 PM – 1 Oct 13 I agree that interactivity is one of the key differences. Static maps have to explain a lot more up front.

12:36 PM – 1 Oct 13 phew glad I’m not the only one… searching for a heat map library w good parameters… so far nil

12:37 PM – 1 Oct 13 Conventions are also about what the map reader expects. Web maps are new enough that we haven’t developed clear expectations

12:37 PM – 1 Oct 13 Yeah, all forms of ‘design’ have overlap. Figure-ground is key to maps and often ignored/difficult to apply

12:38 PM – 1 Oct 13 For ex: Eric R at hates scroll-to-zoom on embedded maps because behavior is unexpected to someone scrolling the page…

12:40 PM – 1 Oct 13 RT : One convention should be not having a hundred page disclaimer to agree to before being allowed to see the map.

12:40 PM – 1 Oct 13 . heat maps are the perfect "in between" problem that no one works on…

12:41 PM – 1 Oct 13 Partially agree. A web map that doesn't zoom is really a print map on the web, or a d3 thing…

12:42 PM – 1 Oct 13 Agree that D3 is opening up a whole new set of web maps. We can’t assume that “web map” = “slippy map” anymore.

12:42 PM – 1 Oct 13 . Even though stats, cartos, pub. health, journos, etc. use them a lot, no one field has done work on toolbox.

12:43 PM – 1 Oct 13 Where the movement starts is also a mean to show what's is important

12:43 PM – 1 Oct 13 but then must ask if D3 has any special design considerations independent of the usual colors, fonts, layout…

12:45 PM – 1 Oct 13 What’s the best paper out there on dot density maps? Seeing them a lot for demographic things these days.

12:45 PM – 1 Oct 13 I correct people who say heatmap, by saying, "you mean kernel density map"? and they look at me strangely

12:46 PM – 1 Oct 13 Good point! Brings up ?s on how we disseminate our conventions, what fields might use them?

12:46 PM – 1 Oct 13 You mean specifically the “census dot maps” that have been coming out lately?

12:46 PM – 1 Oct 13 :) But, but, but….140 character limit.

12:47 PM – 1 Oct 13 . right. This one is probably the best?

12:47 PM – 1 Oct 13 good cartography is good cartography ..bad cartography can be just iterated out more quickly online

12:47 PM – 1 Oct 13 . Although, a web map that only pans & zooms is basically a print map that can pan & zoom. Interactivity w/ data is key.

12:48 PM – 1 Oct 13 Thoughts on handling clustering of data points & how user zooming can easily going from region view to overview w/o confusion?

12:49 PM – 1 Oct 13 The census dot maps are cool because they leverage some of the strengths of the geoweb: zoomability + lots of data & processing

12:49 PM – 1 Oct 13 So even though those census maps are really the same as a massive print map you can zoom/pan, it still /feels/ different

12:50 PM – 1 Oct 13 . I’m forgetting: Is there anything in MacEachren about dot density visualization rules?

12:51 PM – 1 Oct 13 Yeah, clustering (and declustering) points interactively based on zoom is a new problem on the geoweb.

12:52 PM – 1 Oct 13 we had a stakeholder get mad when his town disappeared on the map upon zooming out – hard to answer

12:52 PM – 1 Oct 13 …and I haven’t seen any satisfying solutions. Don’t like google’s radioactive-looking red and yellow clusters.

12:54 PM – 1 Oct 13 Been thinking a lot about web cartography “rules” lately. Most seem to be print rules modified for web, not truly new.

12:54 PM – 1 Oct 13 Yikes. Is it ‘better design’ to make town visible on all zooms?

12:54 PM – 1 Oct 13 Not sure we’ve clearly understood map readers yet (thinking of user studies). Or do you mean something else?

12:54 PM – 1 Oct 13 ah I think the recent race/pop dot map that went around rescaled people per dot with each zoom

12:55 PM – 1 Oct 13 By the way, a couple of us are very interested in discussing these very topics at NACIS next week if anyone’s going!

12:56 PM – 1 Oct 13 Yes, more user studies are needed. Always surprised to see people still use the pan controls on a slippy map…

12:56 PM – 1 Oct 13 No way, they do?

12:56 PM – 1 Oct 13 good point(s)!

12:56 PM – 1 Oct 13 And when there are few conventions about how interaction works on the web, of course users will be confused!

12:56 PM – 1 Oct 13 And for folks coming from the GIS/dev side, both interaction and carto design are often afterthoughts to functionality

12:56 PM – 1 Oct 13 . even web maps already have their (burden of) traditions

12:57 PM – 1 Oct 13 I signed up for at the last minute. I’ll see you there! Anyone planning a ?

12:58 PM – 1 Oct 13 But bad design leads to nobody using it, thereby not functioning. Lesson is to consider design from A1.

12:59 PM – 1 Oct 13 We’re coming to the end of this , and we still haven’t talked about label placement, sorry ! (see: )

12:59 PM – 1 Oct 13 I have lots to say about spatial interpolation & many Qs about how to rep point/clusters responsibly

12:59 PM – 1 Oct 13 there is historical precedence for this- the business incentive to put towns on wax-cut maps c1880s

1:00 PM – 1 Oct 13 cosign

1:00 PM – 1 Oct 13 Agree more with : zoom is only one of the interaction. Most popular due to web mapping evolution: slippy map > dynamic map

1:00 PM – 1 Oct 13 We’re waiting for you to write the definitive essay about when not to use heat maps. Why isn’t that your thesis topic?

1:01 PM – 1 Oct 13 !Yay a research agenda!

1:01 PM – 1 Oct 13 Dark/black map design is a current web trend. Not used so much in print due to Ink/paper restrictions

1:01 PM – 1 Oct 13 I hope to see some of you at NACIS next week & the rest of you can follow vicariously through the and hashtags

1:02 PM – 1 Oct 13 The next will be in two weeks, on October 15, at the same time. Topic suggestions welcome. Guest hosts welcome too!

1:03 PM – 1 Oct 13 Agreed. I think our conventions for traditional cartography are also designer’s best guess, typically.

1:04 PM – 1 Oct 13 oh but cool now maybe it will ;)

1:04 PM – 1 Oct 13 b/c if it were my thesis it would never get done…. wah waaaaa

1:04 PM – 1 Oct 13 & those conventions have been reinforced by generations of readers who learned to read maps using those conventions…

1:04 PM – 1 Oct 13 I didn’t see this one coming : landsat8

1:05 PM – 1 Oct 13 I’ll post the chat transcript here: Talk to you all next time! Thanks to all the new ppl this week!

1:09 PM – 1 Oct 13 Yes this is a good point (as all traditions, I guess). Reminds me of this scene:

1:11 PM – 1 Oct 13 RT : Label position is a topic, also proximity for placement is important. Want to get labels when near in geolocation

1:11 PM – 1 Oct 13 RT : Label placement in dynamic maps is another topic that can be informed by print conventions, but research needed.

1:54 PM – 1 Oct 13 there is a limit to how many data points can fit on screen at once and be useful.

1:52 PM – 1 Oct 13 indeed!

2:07 PM – 1 Oct 13 They said they used ‘s code that scales brightness, not number of dots, by zoom level.

3:22 PM – 1 Oct 13 thanks for the correction! ;)

3:24 PM – 1 Oct 13 Sorry, I hope that wasn’t obnoxious. I am having a lot of trouble with tone today.

3:27 PM – 1 Oct 13 no way, genuinely appreciate & should’ve fact-checked myself before tweeting ;).

3:58 PM – 1 Oct 13 The windmaps proliferate! RT : Live and beautiful windmap of Tokyo

3:59 PM – 1 Oct 13 Honestly surprised that nobody at stood up in defense of the hint.fm wind map Do cartographers revile it?

This entry was posted in geowebchat, Mapping, Networks. Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.

Post a Comment

Your email is never published nor shared. Required fields are marked *

*
*

You may use these HTML tags and attributes <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>