Surjit's Questions

Skip to first unread message

Owen Ambur

Mar 29, 2022, 9:52:48 PM3/29/22

Documenting my responses to Surajit's questions.


-------- Forwarded Message --------
Subject: StratML Query Service
Date: Tue, 29 Mar 2022 14:04:33 -0400
From: Owen Ambur <>
CC: Jeff Maynard <>

Surajit, see my responses to the questions you posed on copied below.


Hello thank you for sharing more details. I have some questions to better understand the work scope.

12:47 PM
Q1. Do you want users to get/view XML data after login/signup?

A1. No.  Definitely not.  I want users to be able to query and view results anonymously.  I don't want to require them to register nor do I want to get into the business of managing UIDs & passwords.

12:53 PM
Q2. Do you have a diagram or flowchart? If possible could you share an elaborate demonstration of it?

A2.  There is no particular "flow" to diagram -- unless you consider this to be one: 1) The user submits a query, 2) a listing of relevant results is delivered, and 3) clicking on links in the results listing takes the user directly to the element they wish to see, like this.

Eventually, another possible "flow" might be the ingestion process, i.e., how files get into the database.  Something along these lines might be appropriate: a) the user provides the URL for the file to be imported, b) the application checks to see if the file validates against the StratML schema, c) if not, the system sends notification of the failure to the submitter's E-mail address contained in the file; d) if the file validates, the system notifies the submitter to confirm that they actually made the submission; e) they confirm the submission by clicking on a link in the E-mailed confirmation request; and f) valid/confirmed submissions are indexed.

However, initially, all I want to do is import all of the >5K files currently in the collection, all of whose URLs are listed in sitemap format.  I DO NOT want the development of a minimally viable query service to be held up with security-related complications.

No, I don't have an "elaborate demonstration" of the capability that I'm hoping you might be able to provide.  However, I can point to several prototypes that either failed to meet the requirements or were not maintained, including:

  • MarkLogic's prototype came closest to meeting the requirements but they password protected it so nobody got to see it.  Then they abandoned it when the developer left the company.  Recently, he suggested that I consider using MongoDB instead.  All that remains of MarkLogic's prototype are these screen shots.
  • Andre Cusson's StratML portal is still live but hasn't been updated in recent years.  Its indices are static hyperlink listings and do not support querying by example (QBE) capabilities.
  • Ximdex's StratML portal is also still in operation and is reasonably attractive.  However, it hasn't been updated to include the many more files added to the collection in recent years.  Worse, it doesn't take advantage of the structure and semantics of the StratML schema and, thus, doesn't meet the requirements.
  • Chris Fox's StratNavApp enables registered users to query their own files and provides direct links to the relevant sections, albeit in edit/forms-based rather than read-only text mode.  So it doesn't meet the requirements either.
  • Jorge Sanchez put a lot of time and effort into his StratML portal, running on BaseX, but it was never fully functional and he apparently abandoned it when a software update required him to start over again.

1:00 PM

Q3. Do you want this data
To represent like this

A3.  I want the data at to present like this:  The stylesheet is already available, at  So you shouldn't need to worry too much about that.  However, beauty is in the eye of the beholder and improvement is always possible.  So while I wouldn't want presentation of the data to hold up the project or increase its cost very much, I am open to considering other presentations as well.  For example, Turn-Key Systems is providing a StratML-->PDF conversion service.  However, PDF is best suited for print purposes and does not support direct referencing of the queried elements.  So while it would be more than fine to offer PDF conversion as an added service, it does not meet the requirement for direct referencing of queried elements.

Reply all
Reply to author
0 new messages