Latest Contributions by MarcoBoeringa

‎04-30-2016

I would also recommend Mansour Raad's interesting blog posts about ESRI's Big Data tools and more generic Big Data problems and solutions: Thunderhead Explorer

‎04-26-2016

I have the strong feeling there are some bad memory leaks in ArcGIS Pro that need fixing (ArcGIS Pro 1.2). I have worked on the conversion of a tool that now runs in both ArcMap and ArcGIS Pro, and that needs to generate a couple of hundreds of layers in the TOC, setting symbology and definition queries. This runs fine in ArcMap, but it simply won't finish in ArcGIS Pro. It gets stuck somewhere in the process, with memory soaring above 6-7GB. When I limit the number of the layers to generate, it may finish. But after generating a small number of layers (+/- 15), I accidentally opened the attribute tables of all the layers at once... (which reference a relative small dataset of +/- 50k records). ArcGIS Pro dutifully started to try to open all of them, but it was terribly slow, despite limited number of records and fully indexed fields. Not wanting to wait for this, I hit the close button of the last opened table still in the process of opening, and subsequently, ArcGIS Pro became totally unresponsive and memory went up to 13GB before I decided to kill it (machine has 16GB).

‎04-15-2016

Dan, I know, I have read through these pages dozens of times in search for different types of functionality and code examples that I needed for the conversion I was working on. In fact, I have succeeded in making my ArcGIS Renderer for OpenStreetMap "future prove" by re-writing parts of the code to be both Python 2.7/.3x compatible, and ArcMap / ArcGIS Pro compatible. I now have one advanced toolbox that can run both in ArcMap and Pro. That was no small feat, because it required quite some minor and major adjustements, only part of which could be deducted or was documented in the Pro Help pages like Python migration for ArcGIS Pro—ArcPy Get Started | ArcGIS for Desktop This is just one of those last minor hurdles or annoyances that I can live with, but it would be nice to see a better support in Pro. Like Joshua wrote, we probably just need to be a bit more patient for ESRI to catch up in Pro.

‎04-15-2016

Thanks, that was what I was already afraid of. Hope ESRI will add something in a future version. Otherwise, all arcpy tools being upgraded from ArcMap, and using mxd.activeDataFrame, will likely need an additional "Map" parameter to allow user to set the appropriate map, as there is no real way from arcpy to get it.

‎04-14-2016

Hi everyone, I have wondered if there is currently an equivalent of df = mxd.activeDataFrame in ArcGIS Pro? I can't find anything up to now. I would have expected something like aprx.activeMap, but that doesn't exist on the ArcGISProject object. activeDataFrame is such a crucial tool in working with arcpy in ArcMap, that I am slightly stumped to see no obvious equivalent in Pro. Yes, you can use listMaps to get to the maps, but that doesn't tell you anything about which one is currently active / in front view. Of course, with multiple possible maps and layouts, there may also be a layout "active", but there should be ways to deal with that. Maybe a more generic aprx.activeFrame? ... with some possibility to test what type it is. Anyway, is there anything I missed?

‎03-02-2016

I have decided to re-install and downgrade to 10.3.1. Now my tools run fine again. It may be that the 10.4 installation somewhere failed, or something else caused the issue, but for now I will stick with 10.3.1 as I don't have the time to look into this further. I noticed at least one other issue as well, that the familiar "scripts" and "tools" icons were not displaying as well in models, so an install fail may actually be likely (although I was able to run tools). Anyway, once 10.4.1 or so is out, I may revisit this and try re-installing 10.4.

‎03-01-2016

Hi all, I just today upgraded from 10.3.1 to 10.4. Now all of a sudden, arcpy.SaveToLayerFile_management in one of my vital scripts, throws an arcpy.ExecuteError. I can catch the error with a "try... except arcpy.ExecuteError:" block, but using print(arcpy.GetMessages(2)) in the except doesn't return a thing. There is no message, the tool just fails. Anyone else who has problems with arcpy.SaveToLayerFile after upgrading? Marco

‎02-23-2016

One thing already mentioned in the other posts here, but that I would like to emphasize more, is the fact that ArcGIS for Desktop and ArcGIS for Server are almost completely database platform independed as regards to the user functionality offered from within the software packages. This is one of the major achievements of ESRI regarding database support in GIS, and often under-appreciated or not well understood by people less familiar with ESRI software. Essentially, you can simply swap out any database platform and replace it with another, e.g. throw out Oracle and install Microsoft SQL Server. Or throw out SQL Server and replace it with PostGIS, DB2 or any of the other supported platforms. You can even "mix" database platforms. Replicate from an Oracle to a SQL Server database, or even to a "simple" File Geodatabase... ArcGIS for Desktop or Server simply don't care, you can keep doing what you did yesterday on the old platform if you make a switch! (and that is a great thing). ESRI really is on its own with such diverse database platform support, yet unified user experience. The only "but" is that you need to stay within the ESRI Geodatabase context... Start messing at the database level, and of course you will need to be fully aware of the multitude of differences between database platforms, even if all of them are supposed to "talk" "standard" SQL.

‎09-18-2015

My hunch here is that this is caused by the fact that true curves in geodatabases traditionally are stored in a separate table or "CAD" binary field type column as described here (scroll down to the section "Feature classes stored using the Microsoft geometry or geography type in a SQL Server database" , and the CAD side table mentioned there), and that the SQL view, being based on SQL Server backend definition rather than a geodatabase Feature Class object, can not incorporate this data. AFAIK, the base table of the feature class stores a secondary geometry in the form of a "simplified" (maybe better said densified) version of the true curve in the way you describe here: a set of straight lines stored in the Geometry or Geography spatial column. It is this that is likely returned in your spatial view, rather than the true curve stored in the CAD side table. Of course, SQL Server 2012 introduced native support for at least circular arcs in Geometry and Geography storage, but I doubt if this is yet supported by ArcGIS (maybe reading but probably not editing), as it doesn't fulfill all the needs of parts of ArcGIS (e.g. no bezier curves). By the way, I encountered these interesting articles related to curves and SQL Server: New Spatial Features in SQL Server 2012 | Lenni Lobel on SQL Server and .NET Development Geospatial Support for Circular Data in SQL Server 2012 | Lenni Lobel on SQL Server and .NET Development

‎09-16-2015

I now actually found some hints to an answer of this question in the Help, at least related to map services: 10036: Label class value has an SQL query that references field names that are not indexed 10037: Label class value has an SQL query that is not optimizable It does seem that the label classes SQL Query usage at least in some cases gets optimized, although in the second case, this does not so much seem related to indexes, but to the number of database queries a SQL Query causes through the drawing engine. It does seem though, that with the default "spatial index first" setting for rendering by the drawing engine, the benefits of an additional indexed field, may be minimal, but I don't see any real reference to this in the above links.

‎09-16-2015

I have been wondering if the Maplex labeling engine actually benefits from indexing. Since ArcGIS for Desktop layer files work with "label classes", that can have an associated SQL Query, I wonder if indexing the specific field(s) referenced by the SQL Query, speeds up Maplex labeling. E.g. let's say I have a SQL Query set like: "highway IN ('motorway','trunk')" is it worth indexing the "highway" field in this case, assuming I may have large layouts (A0), potentially containing hundreds of objects to label? I assume the primary filtering is simply spatial, so is the hassle of indexing attributes worth the effort for Maplex?

‎03-04-2015

Thanks Vandana. As I wrote, I can't share this bigger model right now. It runs fine in 10.2.0, which is OK for me at this moment in time (although eventually, it will need to work in other versions too...).

‎02-27-2015

Well, only trying to open the model (as in Open, not Edit), crashed ArcGIS instantly, even without the normal ESRI crash reporting window opening indicating it had been caught as such. Really boom, gone... I even looked in the Windows logs, but couldn't find a trace of it... (Windows itself kept running). Do note though, the model I am talking about contains considerably more than this tiny demo model I created here, so the problem may well be somewhere else. I can't share this model with you at this point in time, I am afraid. It runs fine and stable in 10.2.0. I would appreciate though, if you could try accessing the model by both choosing Open and Edit in 10.2.2, that would at least exclude (or make less likely) this problem as the cause of the instant crash.

‎02-27-2015

Thanks Vandana, I do appreciate the input. However, I have decided to work around this by using a simple Python string "replace". It is bit tedious, as I needed to include a submodel with iterator, and script, to get it working in my particular situation, but at least it works. The need to use substitute codes to do this, is not really an option in my design. Well, at least, I don't want to go there yet if it isn't really needed. And as it seems inline variable substitution simply skips the strings in case of such problem, that may be OK for now. I will think about it a bit more keeping your suggestion in mind. With maintaining % signs, there is of course a very small risk that users will enter something in a LIKE statement, that matches a variable in the model, but since I tend to use rather long descriptive variable names, and knowing the context in which it is going to be used, I think this chance extremely remote. *** By the way: *** could you test this simple model for me in ArcGIS 10.2.2 if you haven't already done so? I have tried to upgrade, but noticed issues with my main model using a setup similar to the small demonstration model I included (crash of ArcGIS). I am now slightly wondering if it has to do with this. It was at a point where I didn't want to spent time to figure out the cause, so I stuck with 10.2.0.

‎02-27-2015

Hi Vandana, I have attached a small toolbox with one model in it that demonstrates the issue. Also included in this post two screenshots of the output as I see it. Notice the difference on the output variables when hovered over with the mouse. I actually consider this as a possible bug, as I see no reason why the inline variable couldn't be replaced. However, it would be nice to have some confirmation. My environment is ArcGIS 10.2.0 (patches installed) on Windows 7. *** FAILED variable substitution *** : *** GOOD variable substitution *** :

Online Status	Offline
Date Last Visited	Monday

My Ideas

Latest Contributions by MarcoBoeringa

Re: Publications on Spatial Framework for Hadoop?

Re: Has anyone else noticed the inordinate amount of memory Pro takes up?

Re: Equivalent of df = mxd.activeDataFrame in ArcGIS Pro?

Re: Equivalent of df = mxd.activeDataFrame in ArcGIS Pro?

Equivalent of df = mxd.activeDataFrame in ArcGIS Pro?

Re: Error when using arcpy.SaveToLayerFile_management in 10.4

Error when using arcpy.SaveToLayerFile_management in 10.4

Re: ST_Geometry vs SDO_Geometry?

Re: Curves converting to straight lines in SQL spatial view

Re: Does Maplex labeling benefit from indexing attributes?

Does Maplex labeling benefit from indexing attributes?

Re: Escape "inline variable substitution"?

Re: Escape "inline variable substitution"?

Re: Escape "inline variable substitution"?

Re: Escape "inline variable substitution"?

Re: Hinduja Swiss (Switzerland) IT-Professional-Ho...

Re: Issues with Editors using the new threading op...

Re: Issues with Editors using the new threading op...

Re: Issues with Editors using the new threading op...

Re: Issues with Editors using the new threading op...