從ESRI總裁Jack的一貫言論，ESRI一直將Microsoft、Google定位于消費(fèi)類空間服務(wù)提供商，也一直不愿意提供與Virtual Earth，Google Map整合的工具或技術(shù)，但形勢所迫，這種現(xiàn)象有所松動(dòng)。
不過，客觀的說，從ESRi龐大的產(chǎn)品線、解決方案來看，要和ESRI競爭還需要付出很大的努力，而目前的競爭對手也短時(shí)間難以達(dá)到這種水平，特別是在傳統(tǒng)專業(yè)GIS領(lǐng)域。
下面是相關(guān)Blog的原文：
ArcGIS Image Server – Author, Serve, Use…
Random
notes about ArcGIS Image Server - Released with ArcGIS 9.2 late last
year. Recall, uploadfilery is a natural background for many GIS apps. Useful
for direct interpretation, statistics and analysis, used for
vectorization (80% of vector data collected using uploadfilery backdrop),
data verification after GIS analysis. Diff. apps have diff demands on
their uploadfilery – time, quality, sun angle, etc…
Why Image Server??
Use of uploadfilery growing exponentially
Available from many sources (aerial cameras, scanned maps, satellites)
Depth
of uploadfilery increasing, more bands of uploadfilery, higher resolution,
overlap in uploadfilery (same uploadfile from multiple dates or diff. angles)
Imagery is often available but simply not accessible (obviously Google Earth has addressed this)
Fast access to uploadfilery and metadata
Much more! See ArcGIS Image Server – Author, Serve, Use… Random Notes from the ESRI UC for more on this
Cool new AGX toys
At the 2007 ESRI UC today, we heard cool news at the ArcGIS Explorer sessions. The next rev of AGX (they are aiming for a release every 12 weeks) will contain several look-small-but-act-big improvements and one fairly-big improvement. And I put in another vote for multi-threaded tile retrieval, but I'm sure they're tired of hearing about it. Thanks to the always cheerful and friendly ESRI developers at the ArcGIS Explorer island.
- More natural support for geocoding. The geocoding tasks (e.g. Find Place, Driving Directions) will allow input in a single text box, rather than requiring you to separate addresses yourself into street number, street, city, etc. Doesn't seem like much, but it's a usability boon.
- Task Results: Hierarchies, multiple links & views.. If you execute a task that results in dozens of results, you can now put them in folders and associate multiple links and views with them. This is nice when returning, say, all the exploration wells in an area. There might be 50 of them (or more), but you can group them by category (e.g. oil, oil/gas, dry, etc.) and for each create a pre-canned view of the well's very local context or of that well in a larger play-wide view. Less work for the user.
- Open GL. Details still a bit fuzzy, but a nice demonstration of wind vectors across southern California showed the possibilities. As with ArcGlobe, you can volunteer your own OpenGL objects for rendering. AGX will also (at some point) provide primitives for helping map geographic coordinates to drawing coordinates. It's a nice feature that could allow for all sorts of interesting extensions. (Especially for one of the projects I'm working on, where we are already integrating with a code base that makes extensive use of OpenGL... could we simply reproject those objects' vertices and throw them in AGX or ArcGlobe?) Or how long before someone makes a game in the thing?
Apparently an internal AGX once included a 2-D map viewer as well as the globe control it's got now. For global datasets, a map is a better option than a globe which hides half the planet. However it was apparently scrapped because the navigation paradigms are too dissimilar. This makes some bit of sense. The "post 9.3" nirvana is when they are working on a combined 2-D and 3-D rendering engine, and that's when we might see a map control in AGX. I'm sympathetic to the problem—these renderers grew up on their own and have all sorts of organic cruft on them making their merger difficult. Furthermore, once you're 2-D you're in the confusing world (for most end-users, certainly) of coordinate systems. Given the sex-appeal of the globe and the genuine usefulness of 3-D scenes, I can't imagine this being an issue in the public market. But I'd sure like it.
So now I've gone and raved about upcoming features when I just got finished complaining that ESRI was, in general, introducting too many new features at the expense of fixing bugs. For AGX, I think their 12-week cycle puts them at an advantage. As always, the developers at the conference gave excellent, well-practiced presentations and were deeply knowledgeable when asked questions. I hope ESRI's opening up the bug database will provide the sort of positive pressure their management probably needs to let them fix more bugs.
ESRI User Conference - general impressions
Apologies
for not blogging more during the ESRI User Conference, I required all
my mental energy to avoid being completely consumed by the Dark Side of
the Force :) !! I held out (for the moment at least), but this picture
of me towards the end of my three days there shows some cause for
concern! After my earlier post on "The Force", Ed Parsons
emailed me with the following quote: "Don't be too proud of this
technological terror you've constructed. The ability to destroy a
planet is insignificant next to the power of the Force". And I have to
say that the Force was indeed very powerful in San Diego.
So
anyway, this was my first ESRI User Conference after twenty years in
the industry, as I have always worked for ESRI competitors, and
therefore been on the uninvited list. Others who are more into the
specifics of the ESRI products than me have shared lots of details, so
I thought I would just share some more general impressions here (mainly
for others who are uninitiated), and I will talk about a few more
specific things that interested me in subsequent posts.
As many
people had told me beforehand, the plenary sessions which fill the
whole first day really are a very impressive show, with 15,000 (ish)
attendees in a huge hall with 3 giant screens. Jack presided, talking a
lot on his usual theme of all the good things that GIS can do for the
world, and his talks were interspersed with many presentations and
demos from other ESRI staff and customers, all of which were very well
rehearsed and choreographed. There was an excellent talk by Nobel Peace
Prize winner Wangari Maathai.
The message that ESRI customers are doing great and important things
was repeated a lot. It is easy to see how people get caught up in all
this and catch the "ESRI religion" which I have seen throughout my
career, from the "other side". The exhibit floor is also huge and very
impressive (and a great place for networking if you're trying to work
out your next move in the industry!). And there is also a vast "map
gallery" exhibition, where customers show off what they have been doing.
So all that was great, but posts like this one from Sebastian Good
help remind you that everything that is presented in the choreographed
sessions may not correspond to the real world. There was an admission
in the general session that there had been significant support issues
with ArcGIS 9.2, and a discussion on what they plan to do about it. And
based on my conversations with both customers and implementers who have
worked with multiple systems, I think that my former companies,
Intergraph and Smallworld (GE), continue to have technical advantages
over ESRI in some specific areas including scalability (many concurrent
users), workflow, network modeling, and robustness. And of course ESRI
faces growing competition, as do all the established geospatial
vendors, from the rise of Google, Microsoft and open source solutions
(not in all aspects of what it does, but in some significant respects).
But
having said all this, when you look at the scale and scope of what ESRI
is doing, the religious devotion of its customer base, and the huge
effort that it is putting into product development, it remains a
daunting task for its competitors to make significant inroads into its
Microsoft-like dominance of the industry, especially in its core
"professional GIS" space. Though in the area of distributing and
sharing geospatial data and certain categories of applications, there
is an interesting battle shaping up with the "neogeography" systems,
and in some specific markets there is greater competition - for example
in the utility market, where it remains a close fought battle between
the three major contenders (ESRI, Intergraph and GE Smallworld -
General Electric not Google Earth!).
So now the big question for
me is ... should I consider going over to "the dark side" with my next
career move? For a little bit of fun, I have put together a geospatial
poll at pollmappr so you can give me your input. Let me know what you think :) !!
ESRI integration with Google and Microsoft
I thought that one of the more significant announcements in the plenary session at the ESRI User Conference was the functionality in 9.3 relating to integration with Virtual Earth and Google Maps / Google Earth. Up to this point, as I've commented before, ESRI has seemed a little reluctant to integrate with these systems, and third party software like Arc2Earth has filled that hole. The Microsoft Virtual Earth blog talks in more detail about the integration capabilities with Virtual Earth. In the plenary, there was a brief demo which showed a nice looking analysis from ArcGIS Server overlaid in a Virtual Earth environment. Given the results of the ESRI customer poll in this area, I guess this type of integration was inevitable, but I still think it's a significant step. Jack consistently tried to position Google and Microsoft as "consumer" products in his talk, but it is clear that they are already being used in many business-oriented applications. Once these easier integration capabilities are available, it will be interesting to see whether that accelerates the move of these "consumer" systems into the application spaces traditionally occupied by the established geospatial vendors. This is scheduled to be available from ESRI next year, while Intergraph plans to provide similar capabilities this year.
{{item.content}}