ESRI UC Response to Questions
June 30, 2009 49 Comments
Every year ESRI sends out a questionnaire to attendees of the UC and they’ve just posted the results from this year. Some of the questions are C-level and probably not interesting to this crowd, but it does get into some of the 9.4 (as was seen at the DevSummit), map automation, file geodatabase, and new technology. A couple that caught my eye:
Q: When will ESRI support direct access to the spatial types within relational databases (i.e. SQL 2008, Oracle Spatial, etc) that are not geodatabases (i.e. does not utilize ArcSDE technology)?
At 9.4, ArcGIS introduces new functionality called “query layers†to allow users to directly access spatial type data stored in a database that is not a geodatabase. Query layers allow users to use a SQL query to access spatial type data and create a new (read-only) layer in ArcMap. This layer will allow users to map, query, and analyze data from spatially enabled databases such as SQL Server, Oracle, PostgreSQL, DB2, or Informix without registering the spatial information in a geodatabase or installing and configuring ArcSDE.
Additionally, geodatabase users who store their spatial information in spatial types can use this functionality to work with their data using complex SQL queries.
Q: Will ESRI support the iPhone?
Yes, we will support the iPhone as a mobile platform to get maps from ArcGIS Server and do queries and edits on data from ArcGIS Server. We plan on releasing an application for the iPhone later this year and then adding additional capability as part of our 9.4 release. In addition developers can build their own solutions for the iPhone using the REST APIs available from ArcGIS Server.
Q: Is ESRI moving into any new business lines?
Our fundamental business is building ArcGIS and supporting our users and partners with effective technical support and professional services to help them implement their technology. In the last year we have extended ArcGIS with online services. Fundamentally, this is ESRI’s implementation of GIS in the cloud. This environment involves a large deployment of ArcGIS Server and an extensive library of GIS content. This direction is principally focused on providing Web services to our software users. Generally speaking our users have been very pleased with these services. ESRI now receives several million requests a day for the use of these services worldwide, and we believe it’s just beginning.
Q: How will ESRI support professional standing through certification?
Professional certification provides proof that an individual has attained competence at a defined level of performance. With the ArcGIS 9.4 release, ESRI intends to offer a technical certification program designed around the use of our products in successful ArcGIS implementations. This program will be available to ESRI users, partners, distributors, and staff. While it will focus on the use of our products, it will be complementary to the GISP certification offered through the GIS Certification Institute, which focuses on the practice of GIS. We hope this will help create an active and qualified user community that can expand the reach of GIS in solving problems around the world.
About James Fee
If true that first one is a bombshell. All our ArcMap users can use PostGIS as read-only and we can upload our production datasets in batch mode. For on-line editing we will buy zigGIS as needed for those with permissions. Huge step forward for ESRI – again if true….
I wonder how this will affect the Data Interoperability Extension.
“…and create a new (read-only) layer in ArcMap.”
So you can access the data, and display it on a (perfectly wonderful ESRI) map but you can’t have any spatial fun with the entities unless you (buwaah ha ha ha ha) drop for the magic decoder…in one form or the other.
I’m not sure how I feel about an ESRI certification program. I have two degrees and a GIS certificate, I’m working toward GISP certification, will I now also have to be ESRI certified as well? Will this mean more classes (most of which will probably be incredibly basic information) and more money? ESRI never has liked giving anything useful away.
@Gus
I have mixed feelings for many of the same reasons as you. The one thing I’d say from the perspective of a GIS Manager who has to make occasional hiring decisions is that a person’s experience and background far outweigh any certification. I too am finally working on getting my GISP application together, but I don’t think for a minute that it in any way carries more credence than my experience in terms of career advancement. My gut feeling is that an ESRI certification could very well end up being clearer than the GISP in expressing what skills and background its holders actually possess, but that’s probably jumping the gun a bit since we don’t have any details from ESRI.
I realized an easy benefit for us right away. We have more than a few departments that “have GIS” who probably do more harm than good with the people they put on enterprise GIS tasks. They refuse to hire anyone because they insist they have enough training in house. You know these people, the ones who take $10,00 of training classes then edit everything in the wrong projection at 1:300000 scale, build unindexed unkeyed flatfiles, and then want edit access to update the corresponding enterprise dataset (even if it lines up on the wrong continent).
ESRI certification could mean that we could set a bar for access to enterprise GIS. Want your department to participate at a high level? Get one of your people ESRI certified (or hire one) to run your department’s program. GISP unfortunately is too high of a bar, but if there are different levels of ESRI certification, it might be a good solution for that problem. (This has always been one of my big gripes with Manifold – we need more than a training video for people who don’t even understand the difference between a coordinate system and a projection.)
I personally don’t see the value of the GISP program and it’s a waste of $200, IMHO. The scoring is very heavily weighted towards academics and conference attendance. Using myself as an example, when I filled out the forms a couple of years ago, I had about 12 years experience with 10 of those as a GIS PM managing local/state/federal GIS projects up to $8M. I’d been published in Public Works and the Military Engineer magazines. I’d presented at 4 conferences. However, I did not have enough credits in education and conferences to qualify except under the grandfathering clause since I only hold a BA and didn’t attend enough conferences (I’d attended at least 1 a year).
Now I’m the enterprise manager at a major city with 350+ GIS users. There are probably a dozen with GISP credentials here. Maybe 3 I’d put in the ballpark of GIS experts. A few of the GISP’s have such poor GIS skills that I can’t begin to guess how little vetting goes on when the applications are reviewed other than making sure the check doesn’t bounce.
I personally do not know of one manager that would give preference or more money to a candidate with a GISP. Having now seen the disparity in quality among those that have a GISP, I definitely don’t put any stock in it whatsoever.
The conferences and degree both fell under the education requirement. A BA + 10 credits of geography or computer science undergrad course hours would give you enough without attending a single conference. An 18 credit hour GIS certificate + 5 credits of computer programming is enough to meet the education qualifications.
I personally thought it was heavily weighted towards work experience, since it required 6 years full time experience to even be eligible regardless of any other education or contributions to the profession; especially academic work experience since it was virtually impossible for someone employed 6 years in postgrad work to not qualify.
(To add to that, 1 year of GIS analyst work experience counts the same as a PhD or 250 days of conference attendance. I don’t see how that would weight heavily towards academics and conference attendance.)
I guess it depends on where your experience comes from. If you spend your life in academia, it would appear to be weighted towards work experience but if you work for a living it appears weighted towards academics. I was a geography major but only took 1 GIS course and 1 CS course. I’ve only taken 1 course since then (Intro to ArcInfo workstation back in ’96). Unless the requirements changed in the past couple of years since I looked at it, I didn’t have enough education/conference credits.
Some of the best GIS experts I’ve ever seen have never taken 1 geography course in their life and wouldn’t qualify.
Yeah, it’s more of a GISCience then GISystems certification in that respect. There’s an awful lot of geography courses that fit inside the GIS&T BoK. Seems like it would be pretty difficult for a Geography major not to come out with enough education credits. I think the real people at risk are the ones who do completely unrelated degrees with no statistical analysis components who switch to GIS later in life.
http://www.aag.org/bok/AAGKnowledge_Flyer.pdf
But any system you create is going to leave out some GIS experts, depending on what you consider a GIS expert. There are going to be people that can do extensive systems design and program complex applications that have no clue how to read a simple semi-variogram. You’ll have people who can do geotemporal component analysis with confounding variables and relate it back to a field and region description, who can’t merge two shapefiles. Both of them are GIS experts in one sense, but not another. GISP covers one aspect of that, ESRI Certification could cover another.
I agree, real world experince far out ways a GISP
I studied Engineering and geology and then studied engineering drafting. Went to work for a large consulting firm whom specilized in offshore oil exploration and mapped alot of the sea floor. Learned so much in Geotechnical engineering, enviro, water, etc. 30 years now in our industry I understand the needs and results from GIS but have no desire to get a GISP.
sounds like sour grapes to me.
and I’d like to meet some of these “GIS experts” you speak of. Chances are I’ve seen their kind before. Usually the ones I have to spend time explaining basic things like datums and ellipsoids.
I agree with Jack D., some of the best GIS people I know never had a single geography course. In my circle they are engineers and scientists.
I’ve have met the requirements for the GISP for a few years and never bothered with it. And I work for in a field where the more initials after your name the better. Furthermore, my company would even reimburse me for it.
Until a certification program has a standard technical competency exam I see little use in it.
I just don’t see how a standard technical competency exam is going to be very feasible. Some of the most obvious topics to include are probably going to trip up a good 1/3 of candidates, and not many people are going to be able to cover all of them.
Just off the top of my head, some of the topics you would have to cover are:
(Techniques) COGO, orthorectification, ANOVA, component analysis, assisted classification
(Concepts) Coordinate systems, projections, and planar topology
(Models) Object, field, region, vector, raster, TIN, coverage, and dynamic segmentation/linear referencing
(Map algebra) binary, ranking, rating, and weighted rating analysis models
conceptual, logical, and physical design phases
(spatial allocation and routing) Location allocation, point optimization, path optimization, and weighted geometric networks
(terrain analysis) Slope, aspect, flow, catchment, viewshed, and rugosity
(interpolation) Kriging, co-variance, IDW, minimum-curvature splines, and trend surfaces
(density estimators) Fixed kernel smoothing and k-means partitioning
(fragmentation and similarity) Tobler’s 1st law, contagion, interspersion, Moran’s I, and Geary’s c
These are just fundamental topics without getting into field specializations at all. How many GIS professionals can nail all of these topics cold? I know I can’t, but maybe I would have a lot of studying ahead of me before I could pass the exam.
Brett,
I think that is a great list of topics.
I work with engineers and have seen how much they study for their engineering license. I’m a geologist and my geology license was not easy either (yet not as difficult as the engineering exam I understand).
My point is, it should be difficult if it means something!
I apologize for getting a bit off topic. I will keep an eye out to see where ESRI goes with it.
This one’s kind of a big deal also:
“This API will be free and will provide direct read/write access to the file geodatabase for non-ESRI software. ”
http://events.esri.com/uc/QandA/index.cfm?fuseaction=answer&conferenceId=2A8E2713-1422-2418-7F20BB7C186B5B83&questionId=2578
That is big, but the wording is different from what you quoted. Did it change?
“We have been working to answer user requests for a way to access and use the file geodatabase without using ArcObjects. With the release of ArcGIS 9.4 next year, we will provide an open API for the file geodatabase.”
No mention of “free” (speech or beer) there. The $1M question is also “what technology will the API support”. A .Net API isn’t much good to Java or vice versa. Hopefully a C/C++ API, so the other technologies can be bound on top.
James:
Well things continue to be bright & sunny in Redlands. Perhaps the prime marketing benefit of it being a private company is that you can make any old statement about what you promise for the future. And, true to form, salvation is just one dot release away…
But your readers should read the whole Q&A because there are some real gems, unintentional and otherwise.
For instance, I learned that ArcLogistics delivers the largest ROI of any GIS application. Who knew?
Though the not-reassuring 2-sentence response to “What is the value of GIS in the economic downturn?” makes me wonder if I should seek re-training in some other field.
Brian
B-ri,
If you find something good that will make me forget about 13+ years of GIS with a clear conscience, please let me know!
Maybe we can have a beer at the ESPNZone and kick it around, oh wait no we can’t!
Point 1 sounds good, although I still don’t understand why it’s taken so long or proven so difficult for ESRI to persuade ArcMap simply to read from a database. Motto: Yesterday’s technology – today! Who knows, maybe in a decade or so they’ll finally figure out how to write from ArcMap v.15 to a database as well…
“Additionally, geodatabase users who store their spatial information in spatial types can use this functionality to work with their data using complex SQL queries.”
Does this mean I can finally use “Group By” and other SQL fun? So long, XTools…And it’s good that ArcGIS is finally getting some Manifold functionality.
It’s annoying to have to open up another program whenever I need to do some heavy lifting.
Yeah, no kidding. Maybe better questions to Jack would have been:
Q: When will you support full spatial SQL in ArcGIS Desktop?
Q: When will ArcGIS be a true 64-bit application?
Q: When will ArcGIS utilize parallel processing capabilities using NVIDIA video cards?
…Hey, Manny, your surname is “Fauld”, right?
You could ask the 3rd question to Manifold as well. Until their vaporware product comes out, it doesn’t do much parallel processing either
@Manny Tard
“…it doesn’t do much parallel processing either”
Manifold 8.x uses Nvidia GPGPUs top perform 20 or so operations on arrayed data, so-called “surface transforms”. This is 20 more CUDA-parallelised GPGPU-based operations than ESRI products can perform. Indeed 20 more parallelised operations than is provided by any other GIS currently available.
Now, if only we used Nvidia graphics cards…
“Now, if only we used Nvidia graphics cards…”
Obviously, as CUDA is a proproprietry language of Nvidia and Manifold only currently supports CUDA. However, I am led to believe that Manifold is not coded in such a way that it can could make use of ATI GPGPUs via OpenCL, even the myhtical Larabee from Intel too. It does not yet, because OpenCL is not as evolved as CUDA and Larabee is still in dvelopmemnt.
“Now, if only we used Nvidia graphics cards…”
A CUDA-enabled card costs $80 or $150 for a really fast card. Won’t spend $80 to run 20 times faster? That’s dumb. Guys do in minutes with Nvidia what takes days without.
Like my login? Its pronounced “Envy Idea” LOL!
The reasons we don’t use Nvidia cards have little to do with cost of the cards (and more to do with the cost of unix-on-windows enterprise software that does not support Nvidia drivers).
Perfect. “unix-on-windows enterprise software” that prevents common sense computing. I bet it cost an arm and a leg and you can’t run Windows 7 either. LOL!
It is not the cost of the cards but of your time. Compared to that it costs nothing to buy an extra PC to do your heavy lifting with Nvidia. Run Windows 7 x64 too. Life is good.
You lost me at “run windows 7″. No sane person runs critical business processes on beta software unless they can review the code to ensure its compatibility.
I guess you don’t use ESRI software then! LOL!
Windows 7 x64 is great. It’s better than XP, 2003 or Vista. No need for Captain Crashalot pain like Arc puts you through.
Mission critical means getting the job done fast and reliable. On the desktop today that’s Windows 7.
You live loose, good luck with that.
“It is not the cost of the cards but of your time. Compared to that it costs nothing to buy an extra PC to do your heavy lifting with Nvidia. Run Windows 7 x64 too. Life is good.”
I already pointed out it was not the cost of the cards. My time is only $30/hr (and I can always run operations overnight). System downtime on an emergency response system for a metropolitan area costs a lot more than that. Believe it or not, that unix-on-windows setup is a lot more stable than, well, any 64-bit windows system. It’s really not worth it to introduce unsupported hardware to a critical system just to save the GIS guy a few hours a year.
Brett I got to help you out here.
“Believe it or not, that unix-on-windows setup is a lot more stable than, well, any 64-bit windows system.”
Running something on top of windows can’t make it any more stable than windows. It can only make it less stable or no change. It can’t make it more stable. 64-bit windows is more reliable than 32-bit windows. Windows 7 x64 is more reliable than any previous windows.
“My time is only $30/hr (and I can always run operations overnight).”
You got me there Mr. Slowsky. If you don’t mind taking hours to do what could be done in minutes Nvidia is not for you. LOL!
For everybody else, Nvidia is 100% supported by the vendors that use it. If you don’t care how long anything takes Nvidia is not for you.
James, it’s not living loose to use the most reliable windows ever. A Microsoft RC is better tested than anything Arc. If you use windows you should use the most reliable windows, Windows 7.
Okay, I’ll spell it out for you. Our 911 center has desktop software components that run on unix (note: unix, not linux). This is very common in 911 centers; from my understanding it has something to do with how phase II works, but I don’t really know. I’m the gis guy, not the 911 guy. So, we have to run unix on windows, aka x-vision. X-vision is way more stable than running the same components in any windows emulation of unix. X-vision supports matrox cards to run at its 2d fastest. We have no 3d applications. 3d cards would essentially be a waste. It is not worth it to have 30 computers with 3d cards for one person to run a handful of operations on CUDA a few times a year.
Meanwhile, since the desktops are actually windows desktops, I can run HEC-RAS on them. I do not have a stable computing environment; I could be using any of the computers on any given day. I am not nearly as important as 911 operations. Having 911 go down for 5 minutes because I wanted to install unsupported cards would easily offset all the time I would save running CUDA enabled operations.
yeah and I’m sure running any of those commands invalidates your trial of manifold and allows them to foist a copy on you. Manithugs.
like I said, “doesn’t do much …”
Nice quote from Manifold marketing, fan boy. So, where is the new Manifold? ESRI has never kept their users in the dark like this.
You won’t be able to use even those paltry 20 functions if they aren’t in business anymore.
“ESRI has never kept their users in the dark like this.”
Hah. Saw the new thing in Denver. They spent hours showing it and answering questions. They took a vote at the meeting to add whatever the group wanted. Just to show off they could do anything extra users want, commit on the spot. Impressive. Don’t recall ESRI ever doing that.
“Paltry functions” – ? Some great presentations in Denver by users. A guy showed flood planning for a city that took 3 minutes, used to take 3 days, using those “paltry” Nvidia functions.
They got lots of money, amigo, and all the time in the world to do the job right.
I think he was doing it wrong if he took 3 days before. I work in emergency management; during floods last year I had about 15 minutes every morning to run new projections and slap together a presentation covering 500+ sq miles (Denver is 155 sq mi).
They knew what they were doing. It was a very sophisticated, dynamic model. Not just a simple waterline or contour thing. They used it to plan maintenance and construction on storm drains and sewers. Saved the city over $500K per year. Not bad.
Speedup using Nvidia GPUs is the real deal. No excuses. Nothing else comes close.
Yeah, all I used was HEC-RAS.
“Yeah, all I used was HEC-RAS”
I thought you said you used unix-on-windows enterprise software. HEC-RAS is straight windows. LOL! Forgot to keep your stories straight!
If HEC-RAS could use GPU your 15 minute job would be done in 15 seconds.
“Manifold fanboy”.
Definitely. Absolutely. I am with good reason. I’ll take that as compliment.
@Bruce — RE: Query Layers and the Data Interop extension — it is true that the Data Interop provides direct read to the spatial databases in question, and so this would seem to be a replacement for that functionality. I have a hunch that each method of getting data from databases like PostGIS will have its own pluses and minuses and so we’ll be doing testing to see what the best practices scenarios are for each. In any case, I feel that the largest value of the Data Interop extension is the transformation capabilities it provides via the Spatial ETL tools and the Workbench integration with ModelBuilder, so I don’t think there will be too much an impact on the Data Interop from the Query Layers. In fact this new method of accessing data could make for very interesting transformation workflows — the input would be the result of a Query Layer and the output any Interop supported format, including Geodatabases, with additional transformation on the way by.
(Advertisement — we are doing a 1 day hands-on Data Interop training at the upcoming ESRI UC — visit the Preconference Seminar page and sign up for “Data Integration for ArcGIS Users”, should be a good time and a great way to get the most out of Data Interop).
Dale
I for one am in no hurry to get ESRI certified. I’ve met too many folks with similar certs for all sorts of tech related products/protocols/platforms that lacked many basic skills necessary to actually work in a professional environment.
Experience will continue to trump all in this and most other technical fields. I’ll give the GISP program credit for doing everything to ensure that a person is a ’rounded’ professional.
I just can’t ever see either being a requirement, or even an edge in searching for employment in the field.
Desmond
ESRI Certification is fine, but I’d like to see some recognition of real MVPs like Microsoft. ESRI current MVP program is forums based and most of the really smart folks I know rarely venture in there.
probably because the “really smart folks” you know have ESRI premium support and just pick up the phone. Really smart folks don’t have time to wait days on the forum for an answer. They have to get things done, which in turn makes them really smart.
As a GIS newbie I’ve not used them myself, but are the ESRI forums really that bad?
Obviously, forum support only works if there are enough “really smart” people around to answer questions from, ahem, less smart people like yours truly, and to share their different areas of expertise with each other. Otherwise you get into the vicious circle that nobody ever finds any help there, so they never bother looking, so they can never help anybody else, and so on.
But the forum support model seems to work pretty well in many mainstream areas of IT (e.g. Oracle or Java) and for some of the open-source GIS tools e.g. I’ve had excellent and very prompt advice from GeoServer forums/mailing lists, and I also got prompt and helpful answers to questions on the Manifold forums, despite their relatively small user-base.
But from what you say, it sounds like the ESRI world hasn’t really taken to this particular collaborative model. Maybe they’re missing a trick?
Pingback: map butcher » Certification