Latest Contributions by DeanChiang

Nuria, Your experience seems even more pervasive than I have encountered.  I would first try to find all files named schema.ini in all your folders and delete them if you find any.  My experience has been that ArcGIS Pro can add it to any folder it has permission, including the main AGP folder parallel to the APRX file, under any subfolders, under Users, separate data folders on another drive, even under OneDrive (in the cloud).  The schema.ini is a plain text file you can open in any text editor.  ... View more

A workaround is to rename the CSV file differently every time you refresh the source CSV or change its data structure, but that is no way to build a reliable, repeatable workflow or automate a task. Here is a potential solution (need others to verify): Trying to figure out why/where ArcGIS Pro is persisting data format for these external files, I noticed there was a file called schema.ini in the source folder for the CSV files that I tried to load into Pro. I deleted that file (which I did not create).  Open the APRX project again - Voila!  The CSV file now loads with the correct field name and type, and converted to FGDB table correctly. I cannot believe it took me this long to figure this out, or that Esri does not have this documented somewhere.  ... View more

It is June 2023 and this is still happening (v2.9.2).  Remove the CSV file from map. Save, close, and open APRX. Click Refresh on Project, Folders, the data source folder. Dozens of times.  Still the old *ghost* data field names persist in the latest CSV file (view by Open or under Fields in Table To Table conversion tool).  If it was just the field name, I can change that after converting to file geodatabase table, but the data type is interpreted wrong too for me not to be able to import the data.  How is it a native Windows Desktop software program cannot read a text file on the system dynamically correctly?  Boggles the mind.  Not to mention hours of time wasted.  It's a question for ChatGPT4 - but I don't have time to do fancy GIS stuff. Cannot even import a text file into ArcGIS *Pro*.   ... View more

The basic OAuth sample worked for me too.  Basically I want to use the sample method for users to switch between secure mode and public view. For example, without login in, an anonymous user sees a non-editable public view layer that is a view of a protected editable feature layer which can then be added and edited if user with proper permission signs in.  The bug comes up when you build on the sample with more functionality. After signing in, the portal loads. I see a fully populated portal object, with all the right properties , except for portal.user is null. IdentityManager.checkSignedInStatus returns a credential in Session Storage. If you replace portal.user.username with credential.userId you can even do a PortalQuery and list your AGOL items like the sample does. The token is valid because I can string it behind a secure service URL and open the REST endpoint. But my AGOL secure layer that I want to add to a map (after signedInStatus checked and portal loads) fails to load.  I did not want to "pre-load" the secure, editable layer by default and prompt every user to login immediately upon opening the web page. I did discover that if you set esriConfig.request.useIdentity=false at the start of require() it will produce this bug.  Obviously there are probably other conflicting configs or custom modules in my app that created this issue, and I am now forced to go through my entire source code to debug it.  I was just wondering if this is a situation that others might have encountered.   ... View more

Stephen,  1. Open PyCharm and navigate to File/Settings/Project:{a_project_name_or_Deafult}/Project Interpreter... 2. If the Project Interpreter is blank, click the gear icon next to the blank box to Add... 3. Point it to your local ArcGIS Pro installed Python location, most likely   C:\Program Files\ArcGIS\Pro\bin\Python\envs\arcgispro-py3\python.exe (Tip: you can see the ArcGIS Pro Python 3 installation location by opening the Python Command Prompt under Windows/Start/ArcGIS) Another place to configure this is under  1. Run/Edit Configuration... 2. Under Python/Environment there is a box for Python Interpreter. 3. If it is not set you can add it there. Once this is set you have both arcpy and conda (for working with ArcGIS Python API) available in PyCharm. #arcpy #pycharm ... View more

Robert,  I noticed the new Image Service Identify Task in JSAPI 4, but my target layers are not in image services.  We have hundreds of raster layers published from raster datasets spread across dozens of dynamic map services on-premise. It looks like Esri is gesturing toward Image Sever as the future online raster solution. The trouble is Esri has recently changed its Image Server licensing (took it out of the bundle of ArcGIS Server). We really do not want to take it out if their existing platform. I wish they would just come out and make it clear to the customers regarding imminent and breaking changes in the latest API.  Apparently the solution to upgrading our web apps from JSAPI 3 to 4 is a complete re-architecture of the entire data hosting platform. ... View more

Hi Kelly,  In this case I am not trying to display thousands of feature on a map, but show a tabular result from a query that may return up to a couple hundred rows from a table layer within a hosted feature service. The feature service contains a primary feature layer (index 0) consists of a few thousand polygons, and a related table (index 1) that has no feature/shapes and contains over a million records based on one(polygon)-to-many(records) relationship.  Since I limit my application functionality to only query one polygon at a time, the related data returned based on querying just one polygon ID is not overwhelmingly large. Perhaps this is not the best model for display our dataset, but in uploading and testing the related table, it begs that question, what is the size and performance limit of even simple tabular data on ArcGIS Online? Seems like we've hit the limit. I understand there are certain output limits to a hosted feature layer, such as the 2000 features limit, but I wonder if you guys have done any internal testing as far as how LARGE a single dataset hosted can be, in terms of number of features, rows, and fields? I would be nice to have some guideline so we can design our catalog to work most efficiently, without the need for manually curating each dataset for different resolutions and scale displays. We work often on California statewide dataset scale. It is difficult enough to keep track of publishing/utilizing 3 difference types of services for each dataset already (tile layer, feature layer, and dynamic map image layer).  Thank you,  Dean ... View more

This idea is critical for serious workflow in ArcGIS Online because it is basically a Production environment only, and for most AGOL customers, I would guess they do not have a parallel on-premise Portal to use as the Dev or Test environment. Often prior to a public release, you would want to have draft data or web maps reviewed (or previewed) without requiring user login, but you don't want the general public to easily find the data or map or app still in beta.  ArcGIS.com search engine should follow the basic Internet convention of using the noindex  metadata keyword to block the item from being indexed and not display the item when we tag it as such. Other possible tags may be  no-index‌ nosearch‌ no-search‌ testonly‌ #draft‌ ... View more