Google Says Your Desktop GIS Apps Will Be Irrelevant
Oh S#*t!
Why is it companies want to be so disruptive when all you want to do is make maps? Says John Herlihy of Google
“In three years time, desktops will be irrelevant.”
Bam, game over. We’ll be working our GIS magic on Androids and iPhones. I know, I know… We’ve heard this all before from Messrs. McNealy, Ellison, but now that Google is saying it (again?) we’d be get our act together or we’ll be irrelevant. Heck old Larry bought himself a trophy on the backs of those making desktop irrelevant.
Devil is in the Details
OK, so what does this talk really mean. We all know damn well we’ll be using our desktops to perform GIS analysis for years to come. Consuming data is what Google is talking about. Unlike Sun or Oracle, isn’t focused on content creation, but enabling people to discover and use it. The operative word is mining for data. Google is talking about the world (who most of the time is searching for information on their mobile devices) performing a search and then visualizing that information. Yep, you are thinking what I’m thinking, that is GIS in a nutshell, visualizing information. Content creation will still be performed on desktops around the world, but it will remain a niche marketplace. The “magic” we’ve been taking about will happen on mobile clients.
Cloud/Servers/SaaS
So right, ESRI is “cloud ready”. Everyone is either developing a cloud plan of action or executing it. We want to host our data with providers who scale as the need arises. We want to pay monthly (or hourly) for these services and not have those annoying maintenance agreements. But wait, where are we down Google’s path of mobile nirvana?
We’ve got some efforts into web visualization from Geocommons (who is probably at the forefront of web browser vizualization), but alas their client is flash based so it isn’t really designed for mobile apps. You can bet your last Amazon gift certificate that they’ve been working damn hard on one. ESRI has also moved into this space with two feed. Their Map a Map service is right up that Geocommons alley. I keep waiting for the the announcement that they’ll link Make a Map to their ArcGIS Online service and bam, ESRI users have a way to visualize their data online. But like Geocommons, their client is Flash (see a trend here) and isn’t usable on most mobile clients. Change is coming though. We’ll see their effort at the ESRI Developer Summit realized with their ArcGIS for iPhone API. Right to the front of the line goes ESRI with consuming geo-content on mobile devices.
Of course there are hundreds of others competing in this space. Cloudmade (I think this month they are mobile), SimpleGeo, Twitter (with their GeoAPI), and countless others.
Back to the Future
Yea so go about your business. Desktop Geo isn’t going anywhere. But clearly if you don’t provide mobile versions of your web apps you are going to be as irrelevant as every stupid MapGuide 6.5 Active X junk mapping site out there. All yawls standardizing on ESRI’s Flex API better be planning because Google is spot on, we’ll be visiting your website working with your data on our mobile devices. Good data not accessible on mobile is going to be devalued completely.
It’s dead Jim!
I love every image in this post.
I doubt that we’ll all be floating happily around the cloud in three years, but I am glad that Google’s there, pushing away, driving competition.
Laying blame on flash and other plugins is a cop out. It is a cop out for steve jobs to stand there and take the high road that flash is buggy and use his monopoly power to keep flash and other technologies like it out of the mobile space. Silverlight is coming on Windows 7 phones, which is really bad for Adobe, but it also makes Apple’s case of waiting around for HTML5 a really weak one.
James:
Just as I appreciate not having to open up Photoshop every time I want to look at a JPEG or PNG image, I look forward to the variety of ways to consume geospatial content quickly and easily.
(And the great thing about developing apps for mobile clients is that our industry’s impulse to “put a GIS on the web” will be so un-do-able, they might actually opt for highly usable “simple” applications…)
But that’s not really where the hard work is done, is it?
Coming back from ESRI PUG in Houston last week, the major take-away is that while everyone there loves their in-house Silverlight apps, the major headache continues to be data management. If the Cloud can crack this nut, then we’ll be on to something big.
Brian
P.S. I’d love to join you in dancing on the grave of MapGuide 6.5, but here we are almost 10 years later and its performance in streaming and displaying large volumes of vector data dynamically from a database over the web is still impressive…
I think Google is getting there with data management, for example here’s Zillow neighborhoods for San Antonio, TX : http://bit.ly/bFwJrB
You can use Google’s interface or API to perform CRUD operations and its starting to provide more database and geo functions such as joins and heat mapping.
We are working on transferring full featured GIS from desktop to web. For now we have implemented many features like: creating, editing, styling, some analyzes, WMS server and client, TMS with raster and vector tiles etc.
Our client is Flash based with vector and raster tiles, but system can also produce raster tiles, so map can be used in non Flash environment too. More info at http://www.giscloud.com/
Recently we have released features that enables serving tiles for Google Maps API. So you can easily create and upload your data on our Cloud and use it in Google Maps API (Javascript and Flash).
http://www.giscloud.com/about/blog/embed-gis-cloud-projects-into-google-maps-custom-raster-and-vector-map-overlay/
Well, just like in the investment world. Someone putting all their “eggs” in one basket will get burnt, possibly burnt very badly.
This idea putting everying solely on or even relying on the web is fool’s gold.
Is there a back-up plan if the “network” isn’t corroborating? I bet most don’t have one if they are going this route.
@KoS
A GIS Manager from Shell gave a keynote at ESRI PUG last week and estimated that 90% of the data they managed was 3rd party, and that only 10% was truly proprietary and in-house.
His pitch was to vendors to utilize the Cloud/SaaS so he could consume that 90% as services without taking on the management burden.
So by all means protect that proprietary 10% offline–but there has to be a better way to handle the other 90%.
Brian
Understandable…but what happens when that 90% is unaccessable for a period of time? Does it greatly impact their work? Or can they carry on without the 90% for a reasonable period of time? The question are not direct at you, BT. Just thinking out loud.
Where I’m coming from is…I work with soils data. Currently there is and has been a push to put everything to the web for our customers(internal and external). They can view spatial/tabular data through the web soil survey(WSS) and can download tabular/spatial data through WSS or our soils datamart.
Which is fine. But what do they do if the access is removed for a extend period of time? Our servers get hammered by nerfious individuals all the time.
Some states have create soil survey CDs, which has a stripped down ArcView viewer and all the necessary data. , If a long outage happens, at least having those CDs gives our customers another option. They are not helpless while waiting on the web system to come back up.
In the end, I’m a proponent of have a redundant system offline. To what extent, is a matter of debate.
I dont’ want to be entirely beholden to the magical web for my work.
Redundancy and backups are a must have in any critical system.
Bad things can happen offline too. Hard disk can fail, usually workstations are much less redundant and secure than professional servers in data centers.
GIS will have its future on the web and mobile, this is inevitable. Critical clients will always have redundancy in offline or online way. For others the cost/benefit ratio of a web based system will be in most time satisfactory.
Hi KoS. Good points, and interesting to see someone who works on (NASIS?) posting on James’ blog. While we are talking about moving GIS functionality to mobile devices, here is an example of how to get the very soils data you mention onto your iPhone (well sort of):
http://casoilresource.lawr.ucdavis.edu/drupal/node/886
On a side note, I have found that a PostGIS DB full of soils data to be one of our most useful resources. To hell with the cloud!
Interesting.
Maybe I’m ole school. I like my phone to only be a phone. NOt a “swiss army knife”.
Ugh NASIS!! Thankfully I don’t have to enter data into that beast.
We are not allowed to use un-approved software. NO postgis for us.
And why do I keep thinking that is part of the problem? ESRI’s system of data management is awful. Add in the fact that they get us to think that there is two kinds of data: spatial and tabular, when in fact there is only data with spatial attribution. I have been using alternative methods like SQL only, no spatial columns, or postgis with spatial columns but still not ESRI object oriented models (uggg) and the management and performance and portability has been fantastic.
“In three years time, desktops will be irrelevant.”
“Open mouth, insert foot” is the first thing that comes to my mind reading that statement, although I totally agree that the implied message remains very real as James points out with the better statement of:
“Good data not accessible on mobile is going to be devalued completely.”
“Completely” is farther than I would go, but it’s a far more realistic prediction.
I’m glad Google is driving innovation as well as competition. Let’s just hope the competition can keep up so all the driving keeps up.
Did ESRI check who’s network is under an iPhone before they thought this idea up? http://www.youtube.com/watch?v=uCVNAPlQgiA
Ugh. Flashy (oops, did I say that?) looks with weak performance. Reminds me of so many things I see from Redlands lately. Why not BREW or java or i-mode or (heavens) even dotNET? Even better, XHTML? iPhone is like, sooo last year.
Not that the Nexus is doing any better, in relative terms. But anything that has a data plan can hit a Google site. Or yours, provided you’ve coded your ‘wapp’ for the wireless world.
In 3 years, will desktops be irrelevant to those who are developing the mobile apps, managing the databases that house the data that these apps run on or managing the cloud infrastructure? Nope.
In 3 years, will desktops be irrelevant to the rapidly growing segment of society using technology for most other purposes? I think that there’s a chance. We’ll see.
However, John Herlihy’s “…desktops will be irrelevant” comment covers a much broader spectrum. It certainly helped to grab some headlines, but was that at the expense of taking the focus away from his real point?
My AML from 1996 still works.
Yea baby yea!! I have a few as well…
Live long and prosperous AMLS!!
Mobile schmobile.
I can’t see the stinkin’ screen since my eyes went old on me – and read a map? Hmph.
Now – give me an app that has a map underneath it, and returns useful text in BIG BLOCK LETTERS – or reads it aloud – and I might think about using something like this.
or maybe I should shaddup and get me one o’ them nice big iPad things since that’s going to work with the ESRI apps as well. Wait, there’s no AT&T anywhere near me. Never mind.
“Now – give me an app that has a map underneath it, and returns useful text in BIG BLOCK LETTERS – or reads it aloud – and I might think about using something like this.”
Yesterday I visited a Smallworld Conference where I was present at a presentation on mobile data capturing combined with of speech recognition software. So people are working on this one (the company is called SmallCases and they’re located in Germany).
In the next few years we will probably see best practices for mobile cartography develop. It’s not simply a matter of you built your mapping site using Flash and no one can use it on their iPhone blah, blah, blah. We have to decide what is appropriate to display on mobile devices. Maps in general are not designed to fit within the small screen real estate offered by mobile devises, and for a very good reason, it’s damn hard to read due to the lack of context. Do some mapping sites need to be mobile ready? You bet! Do all? Certainly not.
I think mobile mapping is fantastic, but when I am out hiking or visiting a new city give me a nice big paper map I can unfold – I want context. Context comes at a cost, and that cost is space. Mobile devices just don’t have it.
Yes. A map. An extremely lightweight, rugged (assuming suitable substrate) analog mobile device with no power requirements and excellent contextual referencing.
See < http://www.snap-map.com/ourmap.php > for an example.
Helps if you put your !@#$%^& phone down while you’re using one, as well. Much easier to hold and fold without that bothersome chatter in your ear.
If everything is going mobile why do we all keep getting bigger and multiple screens on our desktops? Mobile maps are great and will continue to grow but to be truly useful they need to be focused on specific tasks. The desktop will continue to rule the domain of ad hoc and unfocused analysis. Also I am wondering what effect Flash showing up on mobile devices later this year might have on this whole situation?
Remember, we aren’t “normal”. We are special niche cases when it comes to computers.
I would have to disagree here James. From desktop computers for personal use to home entertainment systems the trend is for larger screen real estate. It is clear that people want a large view, but it is just not practical for mobile devices.
laptops and notebooks outsell desktops (and have so for years).
http://news.cnet.com/PC-milestone–notebooks-outsell-desktops/2100-1047_3-5731417.html
While I agree, the average screen size on desktops has increased, the trend is mobile and smaller overall and each year mobile devices increase their marketshare over desktops. iPad, iPhone, Android, BlackBerry all show internet consumption trending away from “desktops”, not to mention all the netbooks running around.
Yes, but people don’t like small screens, they are simply willing to make sacrifices for mobility. Only time will tell if a solution to make larger views more portable will take hold that is not too dorky.
I don’t think we want all our maps optimized for tiny screens, we just want maps that function on both, and probably not all maps.
I think we need most of our maps optimized for mobile content. If you don’t, the next generation won’t be using them.
http://www.minonline.com/news/Ignore-the-iPod-at-Your-Peril_13550.html
I would point to the trend in paper maps.
I happen to have a hefty collection of old and very old road and highway maps.
They started out very small; like postcard side. Then they were bound into road atlases about 5×7. Then they moved up into folding maps, around 11×17 and continued to grow in two directions: bigger and bigger fold ups, and thicker and thicker large format road atlases.
Today? You see three things, enormous foldouts, inch think letter size, and half inch thick large format. No more postcard maps, few 5×7 atlases.
Mobile devices are still linked to phone, and people don’t like slapping a legal sized book next to their face to talk on the phone. As mobile becomes detached from phone service (and it will), phones will stay small, but you will see mobile get creative on how to stay light but get bigger (just like paper road maps did).
@Jarlath
I’d wager maps are most useful when consumed where the people need them. Not attached to your gaming rig in your parents basement.
Maps and geo are probably most suited for mobile use than most anything else out there.
Agreed in the case of Google Maps and other consumer applications, but I think there will be certain apps better viewed on the big screen. We live in a world where everyone has a mobile device yet there is also growing demand for 3D home entertainment systems. I am not saying ignore mobile, it is probably the most important segment, but it’s not the only segment. And it is quite clear that when given a choice it is not the preferred medium. If all maps were optimized for mobile devices we would loose a lot.
I’m not sure Jarlath that it has to be either/or.
It should be both, mobile optimized and of course full screen goodness.
Agreed, but a mobile version requires an investment. Just curious what you are thinking of doing over at WeoGeo? Are you guys going to have a mobile version (I could not get it to work on the mobile version of Opera). Do you think it will receive heavy use? For browsing and ordering data I would rather be at my desktop. When it comes to finding directions the mobile versions of Google and Bing are more useful.
So right, consuming geo-content on a mobile device just isn’t happening yet.
We are exploring mobile technology and how users of our content management system can share data with mobile users. I’m thinking devices like the iPad probably lend themselves more than blackberry or android devices.
I may not have done the best job illustrating my point, which is that something is not worthless just because it does not work well on a mobile device. I wouldn’t want to watch Avatar on my phone, nor would I want to browse WeoGeo.
I noticed WeoGeo’s latest release was an ArcGIS Desktop toolbar, not an iPhone app. I think this was an excellent move. I would not care one bit if WeoGeo allowed me to order aerial photos on my phone. I do, however, enjoy the Bing/Google GPS tracking feature, something that for obvious reasons, would be worthless on my desktop.