Harry Beck drew this iconic 1933 London Underground map, based on his experience with electrical circuit diagrams:
Really beautiful. Instantly recognizable. Legible. And there’s the Thames for reference; but that’s the only geographical feature. (Another view of the same map.)
The current London Underground map (.pdf) encodes a lot more information, not as beautifully as Harry’s I think.
More on Harry Beck.
[Related: NYC Subway Porn]
Aaron Benson, who heads up our identity management practice at Novell Consulting, is going fishing, supposedly for the legendary Blue Walleye, long thought to be extinct (Wikipedia calls that extinction on par with the demise of the passenger pigeon and the near-demise of the bison.) Check out the map of his destination, Pedlar Lake in Canada. The map is like something out of a boyhood fantasy, complete with “uncharted depths” and the “treed swamp.”
There’s a new-to-me startup, Ohloh, which looks to be putting a layer of analysis and interpretation on top of SourceForge or CollabNet. They want to be a directory of open source projects; the idea is that you go to Ohloh to search for the right scratch for your itch.
They have statistics and summary descriptions of open source projects including language and license, all the Web 2.0 stuff (Google Maps mashup: check, tags: check, etc.), and some automated rudimentary analysis. For example, warning signs on a project might include:
I poked around a bit and liked what I saw; I did a search for collaboration projects and ended up looking at the interesting Mindquarry project, which, like Ohloh, was new to me. Ohloh also has a project cost calculator that estimates how much it would cost to hire a team to write projects from scratch, given certain (very broad) assumptions. In Mindquarry’s case, Ohloh figured that it would take 47 person-years and $2.5m to hire programmers to recreate the code. This is only useful in a rough-and-ready kind of way, because it’s based on lines of code, a notoriously imprecise metric. Average annual salary for those hired programmers in adjustable, though.
Programming Language Popularity by Monthly Commits
But the feature that caught my visualization eye was a preview of programming language popularity graphs. Since Ohloh is already collecting information on programming languages by project and contributor over time, it makes sense that they put some interpretive tools on top of this data. You can assess open source programming language popularity by month for the past ten years or more and look at the data either as a percentage of the total or by the raw values. You can look at commits, contributors, LOC changed, and by project
Here’s a view of the decreasing pride of place for C/C++ (the orange line; sorry about the lack of a legend, it’s there at Ohloh) as a percentage of monthly commits. Interestingly, scripting languages — all the other lines — individually show no particular pattern, although together they are taking share away from C/C++.
Below is the same graph, with C/C++ still in orange but with Java instead of the 4GLs (PHP, Ruby, Python, etc.); Java has been and remains a key open source programming language with a steady 20% of the ‘monthly commits market.’
Here’s the same graph as above, but this time on a count not percentage basis. It’s clear that even though C/C++ is decreasing as a percentage of the total and Java’s remaining steady, the trend — in terms of the raw commits — is up and up.
Likewise, if we look at the 4GLs versus C/C++ on a raw commits basis, the same upward trend is clear:
From Jon Bultmeyer, tools for mapping complex networks. And from the folks at Big Sky, a post about how to choose among many decision making visuals, including a link to a periodic table of visual elements.
The current NYC Transit subway map is well-known for compressing an enormous amount of information into an easily readable very effective format. But the design has its constraints, and people have designed other New York subway maps that have different priorities, including fidelity to surface accuracy (see this geographically correct subway map) and another that gives you *all* track information, even if it’s superfluous. So, for example, here’s the track schematic for 14th – 28th Streets.
See also abandoned stations; Eighteenth Street, for example.
