SpatialTau v2.1 - Scripting

SpatialTau is my weekly newsletter that goes out every Wednesday. The archive shows up in my blog a month after the newsletter is published. If you’d like to subscribe, please do so here.

GIS and scripting go hand in hand.  With its roots in command line mainframe tools in the 70s and 80s, GIS required scripting to perform any analysis/cartography.  The hardest thing was that each GIS package required a different scripting language; AML, SML, Avenue, MapBasic, Magik, LISP, VB-Script, JavaScript, Python, Perl, TCL, etc.  In the past 5 years though, everyone from Autodesk and Esri to open source uses Python.  It’s at the point that you write Python scripts to perform GIS analysis (well unless you’re a wizard-based GIS professional or what I like to call button pusher GIS).

I could spend hours writing about how Python has changed GIS analysis but I think we’ve talked about that enough.  No one needs to sell Python to GIS professionals, it is just assumed as a requirement.  I doubt I’d hire any GIS person who couldn’t write Python scripts to perform analysis.  We all use it to the point where we don’t even think about using it.  That said, I think JavaScript could be more important for the future.

When I think GIS and Python, I think classic GIS Analysts doing their thing with large multi-monitor displays and huge loud desktops.  I hate to stereotype people but generally Python is old school GIS analysis.  The world has moved to a JavaScript focused environment and we’re starting to see some amazing things being done with JavaScript (D3).  I don’t think one can replace Python with JavaScript just yet but the tempo of innovation is clearly with JavaScript.  If you’re in an Esri centric universe, Python is still the weapon of choice, but outside of that silo, JavaScript is where people are doing amazing things.  I used to say, Just learn Python” but I think that needs to read, Learn JavaScript and Python”.

That doesn’t mean there isn’t some really cool stuff going on with Esri.  Just 2 months ago, Esri said they were supporting the SciPy Library for spatial analysis.  This is a huge shift and one that I think everyone should be excited about.  SciPy is a great analysis tool that you can use anywhere, just not Esri.  You don’t need to worry about an Esri license being available for ArcPy, just use it wherever you wish.  Now short term this is probably just a niche thing as only the bleeding edge ArcGIS Pro supports it, but the plans are to back port it to the classic ArcGIS Desktop enabling you to work with your existing projects.  I can’t stress enough that people should learn how to use SciPy with Esri if only to broaden their marketability.

So all this cool cutting edge Python/JavaScript means we’ll be standardizing on one or two scripting languages?  Possibly, but I was just looking at my text editor to see what scripts I’ve used/written in the past few months.  For me I see mostly .pl and .rb as extensions.  That’s right, I seem to write mostly Perl and Ruby scripts.  But does that mean I’m not following my own recommendation above?  Possibly, I hold on to Perl out of habit as I’ve been using it since the early 90s.  But I don’t use it for GIS analysis, I can clearly see that Python is the choice for that.  Same with Ruby, but I blame that on my stint at WeoGeo which was a Ruby/Rails shop.  But what is my point here?

I think it is that workflows we use daily have tools that best help get them done.  I write this newsletter in Markdown and then use Gruber’s canonical Perl library to convert it to HTML.  I also use Perl to slurp data off websites that don’t allow downloads.  I use Perl to manage files on my computer and push to Amazon S3.  Here is the kicker, I could do all of this with Python but I haven’t.  I think the key is always to use the best scripting language to get the job done.  That still means Python and JavaScript with GIS.  Ruby and Perl are best left to where it makes sense and for most people I can simply say it doesn’t make any sense to use it.

Now let me go ahead and run the Perl script to convert this newsletter into HTML and send it off.  I wish everyone a super 2015 and be ready for some amazing new spatial tools coming out in the next few months!

January 7, 2015