Latest Contributions by VinceAngelo

The other option is to use a true JOIN clause in a VIEW, but doing so is tricky, since you need to preserve a unique rowid column. - V ... View more

Just for clarity, what data storage format do the two feature classes share?  If enterprise geodatabase, which RDBMS? - V ... View more

There isn't really an answer for this, and asking for an Esri response from the user forums isn't the right approach -- Even if employees do respond, they aren't speaking for the corporation (and asking them to actually reduces participation). That said, the recommendation is to use a server adequate to the task. Is that a standalone server? Well, maybe. But give the DBA the benefit of the doubt, since requiring that they add a box they don't want to use (and that has to be paid for) isn't likely to make their life easier. In the end, the "best" solution is the one that works for everybody. If the shared server is overloaded, adding GIS to the mix isn't going to help, but you'll be able to benchmark that. Sometimes everything seems fine in the TEST system, but moving into PRODUCTION demonstrates a scaling issue, even with a standalone server. And increasing the number of servers doesn't always improve performance, either, because sometimes the network or the SAN is the bottleneck. Each deployment is different, and can have wildly varying "best" solutions. - V ... View more

The third option is an ON INSERT OR UPDATE trigger to preserve the contents of the actual column. - V ... View more

This is more like a religious issue than a technical one. The DBA should have a say in the process, since they'll be the one supporting the configuration. Esri supports both (provided you're using the release documented as supported [and recognize "and higher" where appropriate]). Benchmarking both for your requirements is not a bad idea. - V ... View more

Okay, it could be the Esri DLL that's at fault, but it's more likely the FME code. The best way to explore this is with FME tech support (Safe Software -- https://support.safe.com/knowledgesubmitcase ). It's best to build as small of a test dataset as exhibits the issue (and if size impacts it, be sure to let them know where the cutoff is). ... View more

Is this the FME extension to ArcGIS or FME which is crashing? Esri doesn't support FME software. Does it work with a file geodatabase target (which you could then process with less complexity)? - V ... View more

If the user is really "User_20230925" and database is really "Test_20230925", you might be having issues due to the upper-case characters. The connection code recases object names to lowercase. I've never been able to connect to mixed case named databases or as a mixed case username. - V ... View more

Are you using the "Append" tool? Is the "geodatabase"  file geodatabase or enterprise (and if so, which RDBMS)? Even if it would allow it, I wouldn't be terribly comfortable trying to project on the fly as an append. - V ... View more

Possible? Yes.  Supported? Big no. Yes (a change since I last attempted it, at 10.4) If you're familiar with how PostgreSQL works, hacking in via conf file changes might get you in,  [or you can use the supported tools to show the user/password and enable external connections], but it's possible that it won't get you too far, because you'd then need to traverse the geodatabase architecture from SQL (in psql or pgAdmin) You really want to work with Tech Support to resolve the root problem. - V ... View more

There are better, more useful ways to lock down a Python install than to just ban Python use. Doing it this way is so simplistic it doesn't bode well for "business reasons" arguments. This is more a management issue than a technical one. You need someone to lean on the IT folks to create a workable solution, not to just punt the problem to the end-users. - V ... View more

While possible, that's not what I was recommending. Using a binary loader is the fastest way to get all features into a staging table. Then it's a simple join to identify features not present: INSERT INTO table1 ( {long_list_of_columns}) SELECT {long_list_of_columns_prefixed_with_"t."} FROM temptable t LEFT OUTER JOIN table1 p ON p.keyfield = t.keyfield WHERE p.keyfield IS NULL You would, of course, need an index on the keyfield (or keyfields) in table1, and have a mechanism for generating a reliable objectid value (this was dirt easy with a serial column in PostgreSQL, but there should be a function you can use with Oracle). Detecting the changed rows is a similar UPDATE statement. Identifying the rows to mark for deletion would flip the table order, and require an index on temptable(keyfield). I've found that it's an order or magnitude or two faster to load the parallel table, then manifest deltas in a single commit, than to deal with variable duration downtime on the data during a TRUNCATE/INSERT/REINDEX  outage. - V ... View more

"TRUNCATE" is an unlogged removal process, so you probably mean "UPDATE" or "DELETE". "Append" is generally an INSERT. Your update criteria isn't very clear, so it's not easy to make a specific recommendation. I've had good luck using FeatureClassToFeatureClass or FeatureClassToGeodatabase into a temporary table in the database, then using SQL to UPDATE changed rows, INSERT new rows, and DELETE removed rows. This was not in a versioned feature class, so I didn't have to deal with versioned views. arcpy.ArcSDESQLExecute() can be used to issue the SQL from ArcPy. Not sure what the Pro SDK equivalent would be. - V ... View more

Hexagons can't be tessellated themselves. You need to move on to triangles after the initial layer. I've actually got code for this in my Blog post. Squares (aka diamonds) can be sub-tessellated, as can triangles themselves. It's not clear what a unic layer might be. - V ... View more

The value in the computed length field depends on the map projection used in your canvas. Web Mercator is useless for planar measurement (the poles are infinitely far from the Equator, so all measurements not along the Equator itself are skewed, potentially infinitely), as are all geographic coordinate systems (Cartesian degrees are meaningless). Some measurement calculations in the UI are smart enough to not trust Web Mercator, and to do geodesic computation with a geographic coordsys, but the table rows are going to return Cartesian results (the units of which may not make sense). If you use a coordinate reference which is not the same as the data source, any stored values in the original coordref may not agree with the values computed "on-the-fly" by reprojection.  The amount of error may vary by location (e.g. over Africa vs. over Iceland). - V ... View more