Here's the significant changes of this release
FDO Toolbox is now 64-bit only
You should all be running a 64-bit version of Windows by now, so there's no real point trying to make a 32-bit version available.
Bulk Copy Enhancements
The Bulk Copy feature of FDO Toolbox has undergone many enhancements and quality-of-life improvements to make it ever more robust in getting spatial data out of one spatial data store and into another. This post covers all the significant changes.
New FdoCmd command-line tool
This release includes FdoCmd, a much more powerful and flexible command-line tool that replaces the existing FdoInfo.exe and FdoUtil.exe tools.
FDO 4.1
This release of FDO Toolbox ships with FDO 4.1 (r7973). This is pretty much equivalent to the FDO that ships with MapGuide Open Source 3.1.2 with extra PostgreSQL and MySQL provider enhancements made after the 3.1.2 release.
Improved file extension to provider inference
Most of you are probably accustomed to dragging and dropping a .sdf file or a .shp file into the Object Explorer of FDO Toolbox and it automatically creating a respective SDF or SHP FDO connection.
Unfortunately, the list of file extensions that work like this was a hard-coded list. Drag/drop a:
- .geojson file
- .csv file
- .tab file
- etc
And expect a FDO connection to be created? This was not possible until this release. For this release, we no longer hard-code a list of file extensions, we delegate that out to a new and external FileExtensionMappings.xml file. This file (which you can edit) defines all the file extensions a FDO connection can be created from and this file helps to service:
- The file drag and drop functionality of the Object Explorer
- The --from-file command line argument of any FdoCmd verb that requires FDO connection details
5 years is a lot of time in the world of technology so naturally when coming back to this codebase, I took a look at things that were either obsolete or done real half-heartedly and gave them the axe. This includes:
1. Removing specialized connection UIs for Autodesk Oracle and SQL Server FDO providers. I have not touched an Autodesk Geospatial product in many years (where these providers are included), so whether this feature still works or not I do not know, nor do I want to invest in the resource to know (I'm long gone from the Autodesk reseller/partner game, so it's not like I have easy access to these products to find out). The existing open source SQL Server and Oracle providers are more than adequate for this task.
2. Removing specialized connection UI for the legacy PostGIS provider. If you don't know what this is, this was the FDO provider for PostGIS that was replaced by the more robust OSGeo.PostgreSQL provider around the FDO 3.5 timeframe. The legacy provider itself has long since been removed, but the UI for this was still present. Not anymore.
3. Removal of Sequential Process support. This feature (which I probably never documented, but you may see references to it in various UI menus) was an XML-based definition around wrapping calls to the old FdoInfo/FdoUtil CLI tools. Obviously, with the consolidation of these tools into FdoCmd, the Sequential Process support was broken and since it is much simpler to just write your own powershell wrapper around the FdoCmd tool, the choice to axe Sequential Process support was an easy one.
4. Removing all semblances of scripting and extensibility. My original ambitions for FDO Toolbox was for it to be a fully customizable and scriptable Windows GUI application. These ambitions were better realized in MapGuide Maestro, but were left half-baked in FDO Toolbox. Having the opportunity to revisit this codebase, I've come to the conclusion that FDO Toolbox doesn't need customization or an integrated scripting engine.
It is a tool with a singular set of purposes:
- I want to peek/inspect some spatial data
- I want to query some spatial data
- I want to get data in/out of some spatial data source
And in this frame of reference, it was clear that adding in a half-baked IronPython scripting engine was overkill. Maestro has legitimate use cases for an integrated scripting engine, whereas for FDO Toolbox I struggle to find such use cases that cannot be satisfied with this release. In terms of scripting/automation, FdoCmd + powershell is already a combination that can address this niche. So as a result, this release of FDO Toolbox no longer include the scripting engine UI and has closed off the addin system. There just isn't a strong need for such features.
This release also no longer includes API documentation for the FDO Toolbox core library. Coming back to this codebase, I've found that this library is really just a thin-wrapper on top of the FDO .net API and there isn't much value that the library adds on top. You are better off just using the FDO .net APIs directly or using the new FdoCmd + powershell combination.
In closing
This release has addressed all of my personal pain points and frustrations built up in the last 5 years since the last release of FDO Toolbox, and hopefully it does the same for you!
No comments:
Post a Comment