Looking for feedback on a new MarcEdit feature

* Posted to the listserv, but putting it here since the community is larger than the list.

Before I go too far down the rabbit hole, I wanted to toss an idea out and see what folks thought.  I’ve been thinking a lot lately about how MarcEdit could be modified to cut some steps out of workflows.  And one idea I’ve been noodling is this idea of having MarcEdit process data as it comes in, and without user intervention.  Here’s kind of what I’m thinking —

• During installation, in addition to MarcEdit’s primary installation, you can install a service.  Within the service, you can configure watch folders.  Within these watch folders, you can set actions that simply occur as soon as data is added to the folder, with finished data moved to a “processed” folder.
• Users could setup a watch folder and have MarcEdit process all files or files that match specific names and assign tasks to them. 
• Within the Watched folders, MarcEdit would be able to perform any action that you can currently perform with a Task or with the batch processing functionality (character/format conversions).

The workflow that I’m trying to specifically improve happens at least once a month.  Each month, libraries (at least, I’m assuming mine isn’t unique), receive multiple vendor files.  We have tasks that perform the vast majority of the data processing separated by specific vendor.  The idea is that rather than having to open the file and run the task, you could just save the files to be processed into a “watch” folder, and MarcEdit would just process the files for you.  You’d receive a notification (in windows 10 or macros in the notifications area, in other versions, within the task tray most likely) as files were processed (the goal would be to make notifications as unintrusive as possible), but you could also just check the processed folder knowing that any data there, was finished.   You could essentially save data to a folder throughout the day, go home, come back the next day to find your files have been processed.  If all you need to do is run a task on a file, then you would be done.  If you need to do other work, you could do it on a nearly completed file.  Ideally, I could even work on developing a process that would allow users that have ILS integration with MarcEdit to directly upload their processed data into their ILS if they wanted to close the processing loop.  I could see how this could be all pretty slick. 

The other workflow however that I’m thinking about are really long running ones like reconciliation processes.  Libraries wanting to do linked data work have a real disincentive because it takes a while to do it.  This would allow you to include that processing in a task, and just have it done using a low priority background processes that shouldn’t impact system performance and would just complete for you using your machines unused cycles. 

I wanted to bring this up because in theory, I’ve already got most of this working — at least the skeleton.  I have a watcher I run on my laptop that just does work for me.  I’d need to turn it into a service, and I’d need to make some changes to the installer (because services would need to be able to be shutdown and started up).  I’d also need to investigate the creation of user only services (I’d prefer to keep this out of the administrative space — regardless of installation type).  So, before I pull the installer a part, I wanted to toss this idea out there.  It saves me a lot of time, but it could be that my workflows aren’t typical.

Anyway — let me know what you think.

–tr