Latest Contributions by VinceAngelo

The blog announcement states: Other limitations of the File Geodatabase API at its initial release: Rasters (Raster Dataset, Raster Catalog, Mosaic Datasets and Raster Attributes) are not supported with the initial release of the File Geodatabase API. I believe the intent is to extend the API to support rasters, but the initial release development must be completed first. - V ... View more

The doc on ClearWorkspaceCache says "Note: Clearing an ArcSDE workspace from the ArcSDE workspace cache does not guarantee the connection to the ArcSDE server will be disconnected..."  The example shows using env.workspace = "" *before* arcpy.ClearWorkspaceCache(), so I'd suggest: # when done doing some stuff, try to disconnect: gp.workspace = "" gp.ClearWorkspaceCache(connection) gp = None - V ... View more

The 64-bit ArcSDE SDK will *NOT* link to 32-bit applications. This is true on *all* platforms -- 32-bit app with 32-bit SDK with 32-bit DLL on 32-bit _OR_ 64-bit OS or 64-bit app with 64-bit SDK with 64-bit DLL on 64-bit OS *only* If you uninstall the 64-bit SDK and install the 32-bit, your applications will resolve the sde.dll/sg.dll/pe.dll dependency chain. - V ... View more

If the linker isn't finding the DLL, it might not be looking for the one that's been installed. Do you have a 64-bit ArcSDE SDK installed?  ArcGIS 10.1 will be mostly 64-bit, but 10.0 is mostly 32-bit. Unix has this great little utility called 'ldd' which reports dynamic library dependencies and resolution for an application or library.  The closest windows has is Dependency Walker, which I've used to track down obscure dependencies. If the time_t size was an issue you would have gotten a different error. - V ... View more

Good point.  There isn't a "one Oracle per server host" requirement (I run 2-6 Oracle instances on almost every box I control). Still, it can't be argued that 2-6 Oracle instances have a comparable impact on system resources as 2-6 databases do on the other ArcSDE-supported database products, especially when the intent is to only exercise one or two customer datasets at a time, but all of them must be available without delay.  In some respects, six virtual server hosts, with one oracle instance each, might be an efficient way to manage the equivalent resources (keeping in mind that this would be more of a dev/test deployment and not a production one). - V ... View more

Windows ArcSDE services are made dependent on the Oracle service to ensure proper start order.  If the Oracle service is not local, the install instructions call for manual install to make use of the '-n' switch to disable service dependency. - V ... View more

Your title is a bit misleading, since the issue is VS2005->VS2008 migration. I've been using  vs9 for several years, but I write 'C' programs with the 'C' API. I found incompatibility with the default time_t upon upgrade to VS2008, requiring addition of a _USE_32BIT_TIME_T macro on Win32 projects (x64 ArcSDE uses the default 64-bit time_t). The static/dynamic issue seems like an artifact introduced by Microsoft with the new compiler.  I've found that rebuilding my solution from scratch can sometimes change dependencies, so if you can't live with a static DLL, then that should probably be your first step. - V ... View more

What is your purpose in loading this data?  Under most conditions, you can isolate data by user in a single instance, then export by user to transfer to clients.  Sometimes you can isolate just by naming scheme. ArcSDE for Oracle has the capability for user-schema databases, which is as close to SQL-Server as you'll get, but since they live in one database instance, there can be both unintended interaction and unintended non-interaction (e.g., this will not substitute for using a standard set of usernames for all logical datatbases).  You'd have to prototype procedures for using database table transfer tools quite carefully, since there are a number of potential problems down this path. - V ... View more

The 'insert' mode overrides the last file's contents.  'import' mode would clobber any previous 'inserts' with file contents, so it's rare that you want to do both. I'm very conservative in my DBTUNE management: + I *never* edit the %SDEHOME%\etc\dbtune.sde file, leaving it as installed. + At each new major release I always copy the dbtune.sde to dbtune.{instance_name} and recreate the keywords I added to the previous release (taking note of any added parameters not reflected in the old build). [For SP updates I just move (or hard-link on Unix) the previous release's file.] + I only use 'sdedbtune -o import' to load the entire dbtune.{instance_name} file, so I know the database content is accurately reflected in a single ASCII file. + I only use 'sdedbtune -o export' if the SDEHOME has been corrupted (because I prefer to maintain documentation and formatting) in my dbtune files -- and then I just use the export data to reassemble (and re-import) dbtune.{instance_name}. - V ... View more

An 'sdeimport' *also* maps SE_STRING_TYPE to NVARCHAR2? That's bizarre. How is your character set defined? As SYSTEM: SELECT value$ FROM sys.props$ WHERE name = 'NLS_CHARACTERSET'; and/or SELECT * FROM NLS_DATABASE_PARAMETERS; Did you change the UNICODE_STRING in the dbtune.sde file or in SDE.DBTUNE? (Or both, by running 'sdedbtune -o import'?) - V ... View more

The documentation is much better than any explanation I could provide. - V ... View more

No, you cannot use ROWNUM. The value must uniquely represent the row, not any row returned in some order. This is necessary to tie the attribute table to the map display (click on one, flash in the other). You should drop out the UUID column from the view. Relationship classes exist to resolve the 1:M parcel/owner problem. - V ... View more

I have more access to the defect tracking system than non-employees, but not as much as the support analysts. I realize it's a pain to put together a bug report, but *everyone* who's experiencing a particular "feature" really ought to file bug reports.  There are several benefits to this: 1) Sometimes they look like the same behavior, but aren't; you don't want the other fix to not fix your problem, 2) Even if it is the same problem, it's easier to fix something if you have more than one real-world case with which to test, 3) In the realm of limited budgets, the squeaky mouse gets the cheese; hotfixes are generated based on a triage of impact -- less severe problems are more likely to be rated critical if more sites are associated with the NIMBUS record. - V ... View more

I submit bugs on reproducble things that I encounter, but I can't submit if I can't give a detailed breakdown of how to reproduce the issue. - V ... View more

I'm not a fan of having multiple copies of the same data. In a perfect world you could have the view changed so it contained a unique key which is seen by ArcSDE as INT32 [e.g., NUMBER(38) or NUMBER(10)]. If you don't need realtime access to changes you can create a table which contains the natural compound key to the many view, plus a unique ID, and drive the view access through this table (updating the table with a regular process which identifies INSERT and DELETE requirements).  Or you can just configure a relationship class (ArcView can read them but it takes Editor or higher to create one). - V ... View more