Is it recommended to use REFramework for the simple projects that don’t involve multiple transactions or rows? If so, how can we modify the framework to match our needs. Any sample project file with simple open notepad, type hello and close using REFrameWork would be of great help.
It is up to personal and/or company preference.
If I’m doing something extremely simple and it isn’t transactional, then I won’t use the framework and just have a standalone flowchart or sequence(s)
I recommend that you use some kind of framework for all projects that can handle error logging, multiple retries, and environment cleanup at the start and end (like after the job failed you don’t want to leave stuff open).
You could probably eliminate the Get Transaction block. That way you go straight into the Process block, where you can place any flowchart or invoke.
But it’s up to you, just as long as the project has robust process management and logging, because that will give you a headache in the future.
What do you suggest for the following projects like:
Download an excel file containing two columns with numbers, if both numbers are equal write true in notepad and if false write false.
Even the case you described above contains transactions. However, rather than inserting the rows from excel into a Queue in the orchestrator your “transaction queue” would be the result of your excel read (read range) i.e a Data Table.
The first assignment of the advanced training alters the REFramework project to use a DataTable of “transactions” rather than a queue.
That being said, I agree that some jobs are too trivial to build on the REFramework. We have a bot that calls a web API and updates an asset in the orchestrator and that’s it. But still as @ClaytonM says standardization at some level is a good thing even if you’ve rolled your own.
In some projects where we use two workflows, one for dispatching and other for processing. Is it recommended to use REFramework for both the workflows?
Based on your previous conversations, i have a process where i need to:
- navigate to a web page
2)download an xls report,
3)download an xls report from outlook and
4)process both reports with different criteria.
Do you have any suggestions/ideas on how this process could be adjusted to the reframework template?
There are many approaches one could take. You could have a bot that downloads the reports and creates two transaction items in a Queue; one for the first report and one for the second. Or, depending on how different the criteria for your report processing is, you could have four robots, built in the dispatcher performer pattern and REFramework. Two for downloading the reports and creating Queue Items (dispatchers) and two for the actual processing (performers). Rather than a single queue item for each report you could have a queue item for each line of each report. This would provide better fault recovery and would also provide more impressive transaction throughput metrics on the dashboard… As in most things, the answer is “it depends”.