Comments on PWG MFD Scan Service Model and Requirements 8.7.7

Page:Line		Comment

6:170			"... scan instruction sheet ..." what is this? should we define it? use case?

7:209			Mailbox.  Are we out-of-scope?

7:228			coping -> copying.

8:254			append "... and selects the target Scan Service."  Required for Step 2.

8:261-265		ISSUE:	I think if a validate template operation is provided, it needs to 
				specify which scan service(s) to validate against.  Note the template repository 
				may be located on MFD1 and all other MFDs may refer to it.  Therefore, no 
				assumptions can be made as to any relationships between templates and services.
				
9:296			Validation is optional.

10:304			Validation is optional.  Change last sentence to: The template's storage location 
				is specified by an URL.
				
10:309			All request messages may not be internal.  PutTemplate to a remote repository.

10:331			ISSUE:  Implementation specific (?)  Maybe a period of time when user can continue
				the job and, after timeout, the job completes.
				
11:338			At bottom:  GetScanJobElements or GetScanJobStatus?  No QueryScanJobStatusResponse.

11:349-352		This seems to be a hold over from print service.  Printing deals with inbound data and 
				so "push" and "pull" make sense.  Scanning deals with outbound data.  The scan service
				always "pushes" data to some location and the scan service client always "pulls" the 
				data.  Not to say there couldn't be some plumbing to allow a tightly coupled transfer,
				but that seems out-of-scope and implementation specific.
				
11:355			JobHoldUntilTime:  Deleted on exceeding time?  Is "indefinitely" valid?  If so, do we 
				need a DeleteHeldJob operation?
				
11:364			Okay, the "push" model sends the scan data as an attachment to an event.  I thought the
				end result of the scan operation was a glob of data residing at some location specified 
				by an URL.  This seems like orchestration.
				
14:432			What's the difference between StartScanJob and CreateScanJob?  (StartScanJob pushes the 
				Start button!)
				
15:450			What does this mean? (line 458)

16:472-491		Needs some work -- does not flow well (IMAO).  At least copy Step 8. to Step 10.

17:497			See 16:476 Step 5.

17:500			It may not be possible to complete the scan without storing the data during the scan.  
				Imagine a very large scan on a resource constrained device.
				
18:504			The "... and Retrieve ..." seems out-of-scope.  This, to me, is a composite of scan service 
				and delivery service.  
				How does this differ from scan2email, which is explicity out-of-scope?
				Okay, the "mailbox" could be an URL, but the retrieval is (or could easily be) some other 
				application/service.  The scan application UI would orchestrate.
				
19:532			The scan service understands the concept of mailboxes?  Are mailboxes local to the service?
				If we say a mailbox is just a storage location and retrieval is via some other service, 
				then I think we can and should do away with this use case.
				
19:554			Authentication information ... credentials ... realms ... proxy ... headache ...
				
19:557			Scan Policy Template needs to be defined.

23:592-596		Since the user can change the destination, will the scan service know about all possible 
				storage locations?
				
26:612			Why specific to LDAP and DNS?

30:656			New term "ScanJobReceipt"

30:866			If we have a HoldUntil attribute, should we not also have a PendingScanJobsQueue?

31:707			UML would be nice, but is not required.

32:747			So, "Other" would be qualified to account for scan states like "Paused for Copy/Print"

36:823-827		I thought we resolved this.  I think we should do both.

38:864			Change "server" to either service or device.

38:867			not -> NOT

38:866			Paused.

41:988			Yes.