Re: Organization Names

Skip to first unread message

Owen Ambur

Nov 29, 2022, 11:05:01 PM11/29/22
to gayaudeshani,
Gaya, the agency codes are the names of each of the files.  I've Googled unsuccessfully for a listing of them from the State of South Carolina but have been unable to come up with a listing that associates the codes with the names of the agencies to which they apply.

They appear at the top of each of the PDFs at  So unless we can come up with a better way, I guess I'll need to open each of those files, one at a time, and then copy the necessary information from it to include in the StratML file, using my XForm for Part 2.

Since each of the files is named by the agency code, it should not be difficult to determine which of the converted files to open in the XForm when I'm viewing the PDF on the State's site.

On Tuesday, November 29, 2022 at 09:47:17 PM EST, gayaudeshani <> wrote:

from where you get this 'SCAgencyCode', is that common to all the files in this folder?

On Wed, Nov 30, 2022 at 7:54 AM Owen Ambur <> wrote:
Gaya, I spoke too soon.  

Now having checked out several more documents in the .zip file you forwarded, I see that the first five of them -- K050, L040, L060, L080 & L120 -- all list the Department of Public Safety as the organization issuing the report.  That is also the case with the last of the 20 documents included in the file -- P330 -- so I assume it is true for all of them.

That's not a killer issue but it does apparently mean that I'll need to manually correlate the file names with the organizations to which they apply.  If that's true, it would be helpful to include the agency code (i.e., the file name) in the organization name element or perhaps in the organization identifier element to save me from having to manually keep track of it when I try to determine the name of the agency to which it applies.

Indeed, since the agency codes are identifiers applied by the State of South Carolina, it might make sense to use them as the organization identifiers in the StratML renditions of the reports, perhaps by appending SCAgencyCode in front of the code, e.g., SCAgencyCodeK050.  

If that's easy to do, it might be worth trying.  On the other hand, one potential downside is that it would not be considered a GUID or UUID.

Owen Ambur

Dec 2, 2022, 2:12:31 PM12/2/22
to Gayanthika Udeshani,
Gaya, see the attachment and let me know if you need anything further to include the agency names in the StratML renditions of their performance reports.

If that can be done, the next step might be to include information from the Partnerships tab.  It looks like that may be fairly easy to do.  The name of the partner would go in the Stakeholder name element at the Organization level of the schema and the description of the partnership would map to the Role Description element.  At some point in the future, it may be appropriate to map the partners more directly to the goals and objectives they support, but capturing them at the Organization level should be good enough for now.

I also took another look at the information in the budget tab and it looks more complex than we should try to take on, at least now.  Depending upon how well and easily the other data can be transformed, we can consider whether to take on additional complexity later.

BTW, w/re the thought of using the agency budget codes as their StratML identifiers, it occured to me that it might be good to include a hashtag indicator, like this: SCAgencyCode#K050.  What you think?  Would that be relatively easy to do and, if so, does it make sense?

----- Forwarded Message -----
From: Pincelli, Sara <>
Sent: Friday, December 2, 2022 at 08:46:20 AM EST
Subject: RE: [External] SC Agency Names & Budget Codes

Mr. Ambur,


The South Carolina Department of Administration (Admin) received your Freedom of Information Act request December 1st , 2022.


Please see the attached document responsive to your request.


Thank you,




Sara Pincelli
Project Coordinator, Office of the Executive Director

The South Carolina Department of Administration

1200 Senate Street, Suite 460, Columbia, SC 29201

O: 803-896-9540  | C: 803-673-6718



Careers email button


From: Owen Ambur <>
Sent: Thursday, December 1, 2022 11:50 AM
To: Quick, Elizabeth <>
Cc: gayaudeshani <>
Subject: [External] SC Agency Names & Budget Codes


Beth, can you tell us where we can find a listing of the agency names associated with their budget codes?


We're aiming to automate the process of converting the relevant data in the Agency Accuntability Reports from Excel to StratML format.



In his role as chairman of the Legislative Oversight Committee, Weston Newton has expressed interest in seeing agency performance data rendered in more readily comprehensible format.




On Thursday, December 1, 2022 at 05:47:37 AM EST, gayaudeshani <> wrote:



if we can find the names of the agencies associated with the agency codes, we can save it as a separated xml file and copy the necessary information automatically using XSLT.(matching information for the agency code)

11.29.22 Owen Ambur- Agency Names and Budget Codes - RESPONSIVE.xlsx

Owen Ambur

Dec 5, 2022, 12:30:47 PM12/5/22
to gayaudeshani,, Naval Sarda
Gaya, there is no deadline and I'll be happy to carry this effort on as long as you'd like.  However, for the sake of delivering result that are good albeit not comprehensive, it seems like we might want to decide when we think we've got something valuable enough that others, like the SC state legislators, might find useful.

I took a look at the budget tab and was concerned about how the budget figures might be associated with the goals and objectives they support.  As shown in the second and third screen shots below, the answer is in the last column of both the Budget and the Strategic Development tabs, i.e., the State Funded Program Number.

Given that database key, I trust that you could transform the budget figures into <PerformanceIndicator>s of the Input type in StratML Part 2 format.  However, that may be more complexity than we may wish to take on at this point.

On the other hand, I now see in column V of the Strategic Development tab (second screen shot below) listings of stakeholders to which each objective applies.  Ideally, each of the named groups would be separately listed as stakeholders under each objective and I suspect that you could do that by using the commas as separators.  However, a somewhat less complex alternative would be to copy the listings into a Stakeholder Description element under the objectives to which they apply.  (Since the StratML query service has been configured to search stakeholder names and descriptions together, the query results should be the same regardless of whether the stakeholders are named in the Name or Description element.)

BTW, I have not inspected the State Funded Program Number to see if it might be applied as the <Identifier> for each Objective.  I suspect that may not be the case, e.g., because one or the other may encompass more than one of the other.

On Monday, December 5, 2022 at 01:45:16 AM EST, gayaudeshani <> wrote:

SCAgencyCode#K050 -> {long Description column value from this excel} # {agency code}
yes that is possible, I'll look into the partners tab as well. Send me more details about the budget tab.

When do you need me to finalize these? 

Owen Ambur

Jan 4, 2023, 2:31:02 PMJan 4
to Kurt Conrad,, Gayanthika Udeshani, Naval Sarda
Kurt, my wife likes to eat around 6:00 p.m.  So an hour or so later would be better for me.  Would 4:00 p.m. your time (7:00 mine) work for you?

I trust will work but it's been awhile since I used it.  So if you have Zoom or another alternative, I'll be happy to meet you elsewhere if you'd like.

In the meantime, here are some additional thoughts responsive to your comments.

I don't care whether it's called a forms-based interface or not.  All I care about is making it as easy as possible for users to created and edit documents in valid StratML format.  I wouldn't expect many people to take on the complexity and cost associated with using oXygen, but I'd like to think that SyncRoSoft might see commercial value in making a specialized instance of WebAuthor available for StratML at low cost.  One of the biggest shorcomings of my (acually Joe Carmel's) XForms is the inability to expand and collapse sections in order to avoid a lot of scrolling to view/edit large documents.  No doubt, there are other usability issues to be addressed for newbies, who unlike me, don't have 20 years of experience with StratML files.

Regarding paragraphs, lists, and tables, the thought in the StratML standard is that paragraphs contained in the <Descriptor> elements should be relatively short and limited to one.  Brevity is the soul of wit, and if more text is needed, a new element(s) should probably be added to the standard to accommodate the type(s) of information it conveys. (Likewise, the <OtherInformation> elements are a catch-all for content that users may wish to include that doesn't fit neatly on elements already specified in the schema.) In most cases, it is the authors who want more; most viewers want less, with the capability to see more (e.g., via links) if they wish.  (I'm pretty sure I load most people down with more than they want, and I know that many people do that to me.  The World Bank did a study and found that no one was reading their PDFs.)

Likewise, if the content of bulleted lists is not already accomodated in the existing elements of the schema, such content may be a candidate for specification of a new element(s) of the standard.  In my experience, such lists are often evidence of attempting to include sub-plans within plans, when it would be better to keep them separate and use the stratml:Relationship links to connect them.

As far as tables are concerned, stylesheets can present <PerformanceIndicators>s in tabular format.  See, for example, the StratML community's Performance Indicator 1.2.1: Plans Converted.  Those tables can and are also presented in more visually attractive PDF, such as at, but unlike the XML cannot be externally referenced. (I understand the PDF community is working on such capability.)
That said, however, John Barker, who developed the PDF stylesheet for StratML, and I did come to an accomodation with respect to line breaks and bulleted lists.  In his PDF renderer, the ^ character produces a line break, ^^ a paragraph bread, and the * character renders as a bullet.  From my perspective, that's an accommodation for life as it currently exists, not as it should be in a better future.

Also, with respect to ease of use, I've thought about prospects for developing a voice-prompt interface for authoring website about us and relatively simple strategic plans.  See, for example, the dialog by which I prompted ChatGPT to disclose the elements of its plan, at  I posted about it on LinkedIn.

BTW, Gaya is working on transforming South Carolina's agency accountability reports from Excel to StratML and Naval is developing a StratML-enabled query service for me.  If you haven't already done so, you can check it out at  Unfortunately, a Stakeholder query on your name turns up no hits.  However, a full-text query reveals your plan at  When the query service is good to go, I will feature it at

On Wednesday, January 4, 2023 at 12:35:18 PM EST, Kurt Conrad <> wrote:


A video call sounds appropriate.

I'm scheduling these conversations outside of my core work hours. Would sometime after 15:00 pst, 18:00 edt work for you? I'm free today at 15:00/18:00. If Thursday/Friday works better, we would need to start a bit later.

I could also show you the combined training / knowledge capture / final report document that I created in oXml.


/s/ kwc 2023.01.04 09:35

At 2023-01-03 20:16, you wrote:

Kurt, I'm not happy with my local ISP.  It took them a week to get my Internet access back. 

I'll take a closer look at your comments tomorrow but would be happy to have a televideo conference with you if you'd like.  We could meet at

Owen Ambur

On Thursday, December 29, 2022 at 04:54:58 PM EST, Kurt Conrad <> wrote:


I'm puttin' it in print: I don't like public discussions with people that I don't know and haven't been introduced to. I have no clue who these people are and how to communicate with them. Whatever this is, it isn't an organization. It feels more religious. God, what are the shared values?

Anyway. There is important work to be done, and I will trust that my thoughts won't come back to haunt me...

Oh, and another thing, as we're ringing in newness. Owen, your emphasis on all communications being text-based is very expensive and limits the progress that I can make with the time and attention I can devote to these matters.

For example, I clicked on the links that you provided and found data bricks that camouflage whatever you wanted me to read. I didn't waste the time and was out in a couple of seconds.

I'm seriously thinking about writing a Balisage paper about the hazards of treating documents like data. Jean's getting included, because I got this idea from her.

Responses embedded below...


/s/ kwc 2022.12.29 13:53

At 2022-12-28 17:48, Owen Ambur wrote:

Kurt, my experience with oXygen's forms-based interface for StratML is pretty dated and was fairly thin even back then. 

It is documented at as well as in these notes in the StratML history at
  • On October 29, 2013, George Bina provided a forms-based UI for StratML in oXygen.
Good to clarify. This uses oXml forms controls, but I don't consider it to really be a forms interface. This,, I consider to be a forms interface.

And it really gets to what I think is a fundamental challenge when trying to tune oXml Author and WebAuthor to work with StratML:

You've built a data model, using a data modeling language, and when expressed as a form, it makes perfect sense. Could I quibble about some choices? Of course. But the form is perfectly usable. Even the ID blocks make sense.

xsd was developed by database folks to handle the xml expression of complex, often relational, schema. The language of choice for designing and maintaining document models was and still is sgml/xml dtd syntax. For one thing, it's a modeling language, so you can more easily read and understand more complex data structures, more info on the screen, more signal, less noise.

Also dtds differentiate the roles of elements and attributes, based on decades of lessons learned and best practice. The databasers smooshed them together. In documents, it is very rare that ids are visible, especially when they're machine-generated. My prodoc system has controls to  make them visible and editable, but they're off by default.

Oh and another feature that speaks to the different performance targets: In documents, you don't want the validation rules for a structure to magically change. It limits reuse.

In my feeble brain, you've developed an interchange standard. Cool. Do you want to talk about an authoring standard?

I author and publish strategic plans. I was introduced to the government performance and results act in the early 90s and have incorporated many activity-based management and StratML concepts into my personal oXml doctype: prodoc.

But I would never author or publish a strategic plan for human consumption using StratML. Why? I believe that the sequencing reinforces bad, conflict-amplifying habits.

Also, the doctype doesn't allow me to acces the data structures that I expect when communicating complex material to humans: paragraphs, lists, and tables.

Back to the current oXml stylesheets. The form controls are used to add and remove structures. I recently worked with this feature in a production environment. Most of it got turned off.

Personally, I find it noisy and distracting. While data models are often hard to learn, most document models have been designed to both be learnable and easy to type. They reflect agreements around the natural languages used in a complex, multi-agent knowledge flows.

I find that these natural language considerations of often in data models. I consider most document markup to be semi-formalized. The tags are very explicit, but the containers (especially with regard to datatyping) are usually more flexible. And compared to the fully-formalized ontology-based systems, much-less formalization, much-lower costs.

Hmmm... Owen, have you defined performance targets for the oXml Author version? Have you collected feedback from folks who've worked with the current system? - video of George Bina demonstrating it
  • On May 14, 2015, George Bina of Synchro Soft used StratML as an example in his demonstration of guided XML authoLOGNULL NowTransReader::ReadIt() JJFileMT::Truncate(21918513) LOGNULL NowTransReader::ReadIt() JJFileMT::Truncate(0) ring in oXygen 17 - another video demo

Yes. Forms controls. Not a forms interface.

BTW: Phenomenal engineering. The systems that generate those little buttons are really sophisticated and use code generators to generate huge css stylesheets that don't seem to slow oXml down, that much.

Cool stuff, but someone actually engaged in authoring is likely to consider them training wheels.

I thought I had only used oXygen to create or edit a few StratML files, like these:

More like documents, but I want more whitespace. Why aren't there paragraph breaks in descriptions? (I'm working from memory). You're headings are weird, bouncing around inside the text. It's hard to read and it's hurting my head.

I see a bunch of stuff linked to itself. Huh? There's another convention for fine-grained linking, purple numbers. It comes out of the Doug Englebart school. (Again, documents, not data).

While in qa mode. Bold text is easy to read, but body text isn't. Point size if fine, but the stroke is too light and turns into a monotonous gray, especially in the large data bricks.

This is important stuff. I wish it were more writable and readable.

However, when I reveal the source and do a word-find search for "oxygen" at, there are more of them than I remembered.  (Note: In conjunction with development of the query service, Naval recently converted all of the files in the collection to apply the most recent stylesheet and they are now in a new /docs/ folder, meaning that the previous file locations will go away and the association with oXygen will be lost.  See, for example,  I'll plan to maintain the previous versions indefinitely on my PC but, in order to avoid confusing the full-text indexing services, will most likely remove them from the site when I'm confident the transition has been successfully accomplished.)

More recently I have used oXygen to find and fix errors created when I copied, edited, and inserted text in StratML files using Notepad.  I have occasionally done that when long listings of stakeholders are available in a format that can be copied into Excel for the insertion of the appropriate tags on both sides of the text, after which I have used Notepad to insert the tagged text into files I had created with my XForms.

This sound similar to one of the things that I've been doing with prodoc. There's a document-level generation utility that extracts and summarizes all of the agents, artifacts, and behaviors in the document for downstream processing.

My XForms are available at &

Again, this is more what I consider to be a forms interface. Putting borders around topics in oXml works well, I have incorporated a number of paddings and border treatments in prodoc to help communicate the hierarchical context for where you're typing.

Owen, in closing, you used xsd to build an xml schema. You didn't build an xml doctype. I think that's a strategic challenge if the goal is to improve  usability and adoption.

Off to my weekly Community engagement study group. This could be a project...


Owen Ambur

On Wednesday, December 28, 2022 at 07:14:16 PM EST, Kurt Conrad <> wrote:


Sounds like I should see what you're doing in xforms. I don't really consider any of the current oXml styles really form-based.

I'll start thinking in that direction.


/s/ kwc 2022.12.28 16:13

At 2022-12-28 10:31, Owen Ambur wrote:
I'm copying Kurt because he expressed interest in improving the forms-based framework for authoring/editing files in oXyen.

Kurt Conrad
LinkedIn .com/in/KurtWConrad
Zoom: 509 581 0832 / sagebrush

Kurt Conrad
LinkedIn .com/in/KurtWConrad
Zoom: 509 581 0832 / sagebrush

Kurt Conrad
LinkedIn .com/in/KurtWConrad
Zoom: 509 581 0832 / sagebrush
Reply all
Reply to author
0 new messages