James Fee GIS Blog

30 responses so far ↓

• 1 Grinder // Aug 21, 2007 at 9:56 am

  I’ve done a lot of soul searching the past couple of weeks. With my ESRI renewals coming up next month, I had to sit back an really rethink my strategy. I’m not sure I can provide cost effective solutions anymore using ESRI products. SharpMap, Manifold, and OSGeo seem to provide all the tools I need at a significantly reduced cost. What does this mean for me? More profit and cost effective solutions to my customers.

• 2 JC // Aug 21, 2007 at 12:04 pm

  Are there any production or demo sites using SharMap that I can visit?

• 3 Micah // Aug 22, 2007 at 8:51 am

  At this point, most of the Open Source GIS projects are just a notch above novelty. Qgis, sharpmap, you can’t make a cartographically sound map with these. We all like to be Anti-Establishment about ESRI (myself included) but until I see more functionality come out of these products, I will continue to read Jack’s silly little Quotes every time I install one of his products.

• 4 Lefty // Aug 22, 2007 at 9:33 am

  Micah, to be fair…..

  How cartographically sound was MapObjects?

• 5 Micah // Aug 22, 2007 at 9:44 am

  Ok, you got me there
  I’m thinking less development/server-side more layout/end-product.

• 6 Grinder // Aug 22, 2007 at 9:45 am

  How cartographically sound is most ArcIMS projects?

  The best maps I’ve seen are generally created in Adobe Illustrator. ESRI can only get things so far and Illustrator can take things the extra mile.

• 7 Micah // Aug 22, 2007 at 10:59 am

  You bet, Illustrator makes great maps. It’s not GIS.
  You bet, IMS doesn’t. It’s not a layout software.

  My only point was that until ‘desktop’ Open Source GIS provides the same functionality as ESRI provides with ArcMap, Open Source GIS will be just a little bit better than a novelty.

  I can’t believe I’m playing the role of ESRI fanboy here!

• 8 James Fee // Aug 22, 2007 at 11:04 am

  Micah, I don’t think anyone disagrees with you on the desktop side. ArcGIS is way beyond anything in cartographic terms, open source or otherwise.

  As far as SharpMap, I don’t view it as a replacement for ArcGIS or even ArcGIS Engine. To me it is a continuation of MapObjects.

• 9 KJ // Aug 22, 2007 at 11:46 am

  … but if you’re programming with a non-ESRI continuation of MO, or even still with MO for that matter - you’ll always feel pressure from your customers to program basic and new things that are in ArcGIS Engine, like dynamic zoom and dynamic display - the things that make a map app look halfway modern. You can do dynamic zoom in MO really well in like a couple hundred lines of code (that you will have to test, debug, fix, maintain) - or you can do in zero lines of code in ArcGIS engine. You’ll end up burning developer time programming broad horizontal functionality instead of your unique vertical market value add on top of the GIS engine - which is what your competitors will be doing with ArcGIS engine. So in my soul searching it seems you can save a little $$ up front with a cheapo engine, but it will end up costing you more over time. At least that’s how it sized up for me. I don’t want to be competing with developers at ESRI on the horizontal stuff (they only have like a thousand more programmers than I do), I want to be spending time on my secret sauce. Now the only thing that does tick me off when staying on the ESRI technology stack is when the cheapo engines do stuff out of the box that ArcGIS engine does not do - like support open formats like GML, or common proprietary formats like MapInfo TAB. It frustrates me to no end that you have to add $$ data interoperability extension to ArcGIS Engine for that when the cheapos are already doing it for nothing extra - but fortunatley there is not too much of this, at least that impacts me.

• 10 Grinder // Aug 22, 2007 at 11:50 am

  I’m certainly not looking at SharpMap or any other open source project for a replacement for ArcGIS. While I gasp at the next statement, my plan is to use Manifold and some other products to feel my desktop needs. When I glance over at my ESRI maintenance & renewal bill sitting on my desk, I’m having a hard time justifying the expense.

• 11 KJ // Aug 22, 2007 at 12:31 pm

  … manifold’s good too, I like that Manifold IMS can connect to GIS data in a commercial grade DBMS like SQL Server standard and enterprise editions, unlimited users, and supports editing, on say a server with dual quad core CPUs - for under $500!! For the same thing in ArcGIS server you need ArcGIS server enterprise advanced, and double the cost becuase of the dual quads, and if you want to sell the app like an ASP to your customers… well you know what happens next! Of course, just like with the engine runtimes, the ArcGIS Server Web ADF built in tools mitigate this a bit - think of the time you will spend programming nice pop-up map-tips, and tiles, for your Manifold IMS app - not to mention trying to figure how to scale a Manifold IMS web application to multiple web servers - all of which ArcGIS server does nicley built in. I know what it costs to run a couple of developers on a project for six months! But the case is definitley more compelling here than on the desktop engine side for me. For a serious web app where you want to sign up hundreds or thousands of users (e.g. SaaS) - you need lots of servers and lots of CPUs - and even a modest ArcGIS Server system like this could scale to $$millions in licensing costs (standard list prices at least). I mean, this has to be killing ArcGIS Server for web application hosting right? I mean, for $millions you could buy a ton of custom programming to make Manifold IMS or MapGuide do exactly what you like. While I want it to work for me, I am having mega problems making ArcGIS Server work simply because of the cost. For the SharpMap stuff - opensource is fun to watch grow, but - what if the two guys developing on that project get tired of it, or who do you call when a critical bugger is mucking up a mission critical app. I guess its just about your approach to risk management, and the type of market you are in.

• 12 Bill Thorp // Aug 22, 2007 at 1:46 pm

  Oi! Its a library. You can use it *AND* / *OR* other libraries.

  Let pretend that the SharpMap developers all decide to make it explode instead of making maps, and you can’t figure out how to use Subversion to get an older copy. You simply replace SharpMap with ArcManifoGuide, which surely does everything that SharpMap does.

  And “who do you call when a critical bugger is mucking up a mission critical app?” Who do you call now?! How long does it take to get fixed? Give me source-code over customer-service any day of the week.

• 13 KJ // Aug 22, 2007 at 2:21 pm

  … well I call a live breathing human who is connected to a CRM, knowlegebase, and developers of the product who can give me a workaround or a fix. I can understand your point about liking source over support. But still, you see no risk using an opensource thingy from 2 developers to base a product on? “Simply replace”… does not exist. For example, try taking a 500,000 line MO app, and pulling MO out and putting a different engine in. You can’t simply do that, and in fact have to some redesign, and a lot of programming. Imagine programming for 12 months just to swap out GIS engines - that is some huge cost! So I am just pointing out what I came across when I did the same analysis: MO is dying, should I switch to ArcGIS engine, or a different and more open engine. My view no doubt is not right for everyone.

• 14 Grinder // Aug 22, 2007 at 2:40 pm

  KJ - What about the cost of deployment? The ArcView engine runtime, or whatever ESRI calls it, was priced around $1500/seat. For a product that has 500 users, that is a serious chunk of dough. One or two good .NET programmers could certainly add a lot of functionality to SharpMap for that type of cash! Another alternative would be the Manifold runtime. Priced at a much more reasonable level. GASP. Did I say Manifold again? I CAN Manifold.

• 15 KJ // Aug 22, 2007 at 3:23 pm

  Well I definitley see your point. But still its not for me. It depends on your own situation.

  ArcGIS Engine runtime retails around $500 a unit. I have heard of lower costs based on specific vertical markets. Also ESRI has an OEM program you can get into - you would negotiate your engine price based on your business plan, number of deployments forecasted, vertical market, and so on, and of course there are price breaks on volumes.

  So yeah you can run a couple of developers and testers on that $$ for a year trying to make sharpmap do dynamic opengl display for in vehicle navigation, web map services support, 3D support, multiple label class rendering, dynamic zoom, swipe, a measure tool that lets you measure areas and lines and the user change units on the fly, I could go on and on - so you could spend the money on programming all this yourself, and at the end of it not have a GIS app that does what ArcGIS Engine can do, drag and drop out of the box w/o a lick of code. Or you could just buy all that stuff in the engine that already has it, spending a little bit for it with each deployment, and spend your money programming what you know that is special that no one else knows.

  So for me we had a product with enough value in it, and a volume and OEM to get the engine runtime cost down, so that its not all that different from MO or Manifold.

  Or course that would not work with a product with a very low price point (like $100) - but if that was the case you were never in MO anyways - and you could always call ESRI and inquire about an OEM and price that fit your market niche (what the market is willing to pay for a GIS app in your vertical).

  Now for me, the same logic does not apply to ArcGIS Server. With sever for seriously hosting web GIS apps, the standard list prices are working out to $$millions - now there you could run more than just a couple of guys on it, and really do something special on mapguide or something else. Of course, I don’t have $$millions exactly burning a hole in my pocket, but for me the case for switching is more compelling on the server side than the desktop engine side, even though I do really like the features in ArcGIS Server.

• 16 Joe // Aug 22, 2007 at 3:45 pm

  I’m not a developer but find it somewhat troubling that I understand what the photo caption “1337 h4×0rs” is all about.

• 17 Chris C. // Aug 22, 2007 at 4:07 pm

  James likes Urban dictionary

• 18 wed // Aug 22, 2007 at 4:56 pm

  Native….Dot….Net

  No….COM…Interop

  Nuff…said

• 19 Grinder // Aug 22, 2007 at 6:32 pm

  So wed— What about the commercial dot net products like Tatuk or ThinkGeo?

• 20 wed // Aug 22, 2007 at 6:56 pm

  i’ve got no problem with those. or any other commercial product either. i’m not a zealot. there are a lot of really clunky open-source products out there. sharpmap just happens to not be one of them.

• 21 GeoMusings Is SharpMap Viable? « // Aug 23, 2007 at 1:48 pm

  [...] Bill Dollins It seems there’s been a little more buzz about SharpMap lately and some folks have been doing some fact-checking with James offline. He deftly spun the question of “viability” into “popularity” by citing [...]

• 22 Bill P // Aug 24, 2007 at 8:30 am

  Guys, I think this has been a no brainer for some time now. For the desktop stuff, choose your poison (ArcGIS, Geomdia, MapIno, Manifold etc). Theres nothing in the open source world to compete. On the web side its quite another story. The open source webmapping tools are not the ’science projects’ they used to be. I think the big players are safe and secure in the desktop realm, but they are slowly realizing they are losing any grip they had on the web. Autdesk saw the writing on the wall. Take a look at the current MGOS release. You can build very large modern sites with robust functionality that connect to real spatial data databases using your choice of platforms and development environments without writing a sinlge line of code. Of course you may still need to customize it for your needs, but thats certainly no different than any ArcIMS/ArgGIS Server site. The point is, even if they were both free, the open source stuff is better in a lot of cases and is no longer the realm of ‘code geeks living in their parents basement’.
  All the large sofware vendors (gis and otherwise) have taken notice. Why else would Oracle have lowered their prices by up to 70% on their database offerings. I think if Oracle can bite the bullet, it might be time for Jack to do the same. Im just waiting for a powerhouse like Google to buy up a buch of smaller GIS vendors and start offering an enterprise desktop GIS for $300 and then see how smart Jack thinks he is.

• 23 mfdusr // Aug 27, 2007 at 9:35 am

  Im just waiting for a powerhouse like Google to buy up a buch of smaller GIS vendors and start offering an enterprise desktop GIS for $300 and then see how smart Jack thinks he is

  sorry to bring this up, but for $495 Manifold can build an enterprise GIS with native spatial connections to Oracle spatial, DB2, and rumor has it perhaps Katmai when it comes up.

  So, this is the sub $500 realm for true entrprise GIS with a real database (i.e. Oracle, SQLServer).

  Scarey stuff…

• 24 Lefty // Aug 27, 2007 at 9:59 am

  Even when its not about Manifold….

  it always is.

• 25 Chris C. // Aug 27, 2007 at 11:26 am

  ‘Even when its not about Manifold….’

  Lefty:

  Have you actually read any of the comments here?

• 26 Lefty // Aug 27, 2007 at 11:33 am

  I’m not going to get dragged into a Manfold debate when any of you. SharpMap has ZERO to do with Manifold.

• 27 Petz // Aug 28, 2007 at 1:57 am

  Getting bought up by Google seems to be the wet dream of any niche sector software company these days. Come on ! Why would Google want to buy any legacy GIS package, including ArcGIS, Mapinfo or even Manifold?

• 28 Chad // Aug 28, 2007 at 9:55 am

  Because it is cheaper than starting from scratch?

• 29 oakfish // Aug 28, 2007 at 11:54 am

  SharpMap, folks… the subject is SharpMap

• 30 Richard // Sep 3, 2008 at 4:55 am

  We are using sharp map to generate maps on the fly based on our internal data sources + ITN GML data.

  It (and our product) is royalty free and all web based. Plus we have to integrate with Esri, MapInfo and other formats. It gives us complete control over out product and means we can extend/improve/fix as required.

  For me those features are a big plus. We have plugged it in to OpenLayers using WMS and it is just as good as GoogleMaps.

  Our roadmap includes a XAML based feed to Silverlight or desktop components with rendering happening on the client…

Leave a Comment