4D Execute on sever command vs method checkbox differences
4D Execute on sever command vs method checkbox differences
I was troubleshooting some long executing code when I discovered the difference between Execute on server
command and it’s sibling checkbox in the method properties.
What you need to know:
Use checkbox if you care about the results of the executed code, you want the server to handle the processing and don’t care that the client waits for the response.
Use command to detach the wait from the client and process the code asynchronously.
I won’t go into much detail as it is well documented, but if you do some deep reading.
Tim Penner on “Command vs. Property”
One important difference is that the Execute on Server command always creates a new process, whether it is called in Client/Server mode or in single-user mode; the Execute on Server command still creates a new process.
In contrast the Execute on Server method attribute will not create a new process on the server, but will instead use a “twin” process of the client process that requested the execution. In Single-user mode, this method property has no affect and the method runs in the same process that requested its execution.
More Information
Execute on Server attribute:
http://www.4d.com/docs/CMU/CMU40945.HTM
Execute on Server command:
http://www.4d.com/docs/CMU/CMU00373.HTM
Stored Procedures:
http://www.4d.com/docs/CMU/CMU40975.HTM
Author’s Note
Initial md
Generated using https://github.com/jsr6720/wordpress-html-scraper-to-md
Original Wordpress categories: [‘4D’]
Original Wordpress tags: “4D”, “4D”, “execute on server”, “stored procedure”
Original Wordpress comments: None
Significant revisions
tags: 2011, wordpress, txcowboycoder, 4D, execute-on-server, stored-procedure
- May 6th, 2024 Converted to jekyll markdown format and copied to personal site
- Jan 31st, 2011 Originally published on txcowboycoder wordpress site