-
Notifications
You must be signed in to change notification settings - Fork 183
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Style Guide rough draft #1192
Style Guide rough draft #1192
Conversation
fe0c807
to
85cfa20
Compare
@david-waltermire-nist as I noted in chat, there are (generated) files in this PR that could (should?) be removed: is there an easy way to do that? The only files in the commit should be the new |
…l commits Initial dump of outline with some contents Rearranged and built out more contents and links More slight adjustments Further touchups Various continuing enhancements Rearrangment and spiffup Many more adjustments to draft style guide Pushing up missing assets (thanks @david-waltermire-nist) More touchups; pretty close to a complete draft 1 for usnistgov#1185 Repairing broken links Undoing changes to generated files
00bcbca
to
3fe3536
Compare
More language on rights; Clarification touchup
f48656d
to
b76c378
Compare
The latest commit is about ready, so I am marking this for review. Small/final items to come back to:
|
Wendell, I quickly browsed through and found the following possible minor issues:
|
|
||
Capitalization is tricky. As a rule, a proper noun is capitalized in English, and hence a reference to a formal object (such as an OSCAL <q>Control</q> object or a Github <q>Issue</q>) should be capitalized, when it has its own name. Both abstractions and things that are not abstract may have names, such as the host of our web presence, which is named <q>Pages</q> (or <q>the NIST Pages site</q>). (Other examples include Avogadro's Number, the Second Law of Thermodynamics, the OSCAL Style Guide, and Markdown, a family of text encoding technologies.) | ||
|
||
But even a well-known abstraction such as <q>1 + 1 = 2</q> is not called an <q>Equation</q> (capitalized). To use capitals when referring to something by name is different from instantly or implicitly coining a new technical term out of nothing, by using capitals. To call a Github ticket an <q>Issue</q> is not incorrect if that is what Github calls it; but a ticket in another tracking system is not an <q>Issue</q> (capitalized). In an OSCAL context, even a technical term such as <q>profile resolution</q> should not be capitalized just because is used in a specialized sense. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
At the end of this line, you probably meant to write "because it is used" instead of "because is used".
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This is very thorough and impressive, thanks for working on this.
This is actually difficult to review. I have some minor specific feedback, but there are some larger questions I have in relation to the strategy and approach that are not really directed at individual content items.
Is it possible we review, as a group, the outline in the comments on #1185 and discuss as a whole team what are the specific goals we want out of this document? In the issue we cite some exemplary guides (like pl.gov) and others, we want to follow, so I wonder if we can choose what key elements of those we do and do not want?
@@ -0,0 +1,324 @@ | |||
--- | |||
title: OSCAL Style Guide | |||
description: (with links) Contributors to OSCAL can save time by scanning this early. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Question: what does "(with links)" mean in this context? Is it meant to be kept in the PR?
|
||
The coverage here aims to be helpful and suggestive rather than normative. Feel free to create an [Issue on Github](https://github.com/usnistgov/OSCAL/issues/new?labels=User+Story%2C+enhancement%2C+Scope:%20Website&template=feature_request.md) to raise questions or points of discussion regarding either style (covered or not covered) or the Style Guide itself. | ||
|
||
The styling on our web sites is enabled and limited by the back-end architecture we use to produce and maintain them, the [US Web Design System](https://designsystem.digital.gov/), which we deliver via an [implementation](https://pages.nist.gov/hugo-uswds-docs/) of [Hugo](https://gohugo.io/). See more below under [Hugo and Markdown](#hugo-and-markdown). Much of this guide assumes you are writing for a web site such as the [OSCAL web site](https://pages.nist.gov/OSCAL/) or for web-accessible resources such as `readme` documents in Github, and may be more or less applicable for other cases. Writing for OSCAL, we must rely on our readers to distinguish what advice is clearly not appropriate or useful to them, and the same rule applies to this page. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
So maybe the requirements and acceptance criteria has shifted for the user story of #1185, but I interpreted the main intent to be a style guide for the content, not for presentation style with HTML and CSS. We begin looping that in here.
Is that not a separate goal?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The scope probably stretched a little however I discovered that: (a) the mechanics were not covered elsewhere, (b) they more or less had to be, among other reasons because they were sometimes (not always) implicated in matters of style; accordingly, a bare-bones treatment was judged to be in scope. Arguably a page on the mechanics could be carved out but TIASD (that is a separate discussion) raising questions of its own.
I suggest raising this as an Issue IFF it is found, shown or discovered to be a real issue for users.
|
||
### Accessibility | ||
|
||
Since materials for the web site are authored using generic encoding (Markdown) and produced with an automated production pipeline, we are able to encapsulate and manage page design within our build. This takes accessibility and accessibility compliance (such as ADA) out of the hands of authors, for the most part. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree this information is important, but is it relevant to the style guide? We already link to the global NIST footer about accessibility.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
See above. This could arguably at some point be separated out; for now we need a stake in the ground. (Metaphor without scare quotes.) Don't think a Style Guide today should be without some level of treatment.
|
||
Please be so kind as to provide `alt` values for all images to save us effort in meeting Accessibility benchmarks. Otherwise, authors should be able to rely on the publishing process to handle technical requirements related to accessibility. | ||
|
||
At the same time, the builders of the back end support for these publications will be interested to know of any accessibility-related requirements that are not already adequately addressed. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I agree this is important, but should we not express this elsewhere?
Perhaps set aside development effort in the backlog for automated and/or manual accessibility testing?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Same as above. We should also express this elsewhere, not instead (in my view).
|
||
For example, whenever writing about <q>validation</q>, be specific as to what kind of validation is meant, since the word has both generic and specific senses. | ||
|
||
Similarly, <q>control</q> has both an OSCAL sense, and an RMF sense (Risk Management Framework as defined by NIST SP800-37); a document that uses the term specifically in one of these senses, the other, or both (when OSCAL is being used to model RMF), might indicate this. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
The point made in this paragraph is excellent, and I believe Dave suggested we will apply what we quickly described as "use a specific formal noun, at least two words rule: OSCAL Control, RMF Control, but not control; XML Schema validation, OSCAL model validation, but not just validation" Can we reference that heuristic in the project? I happen to like that as it is a simple measure for these kinds of things.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I like it well enough but I think it is best as a "folk heuristic". I don't think people should be penalized or ostracized for saying "control" after they've said "OSCAL Control" six or eight times in a conversation....
|
||
Similarly, <q>control</q> has both an OSCAL sense, and an RMF sense (Risk Management Framework as defined by NIST SP800-37); a document that uses the term specifically in one of these senses, the other, or both (when OSCAL is being used to model RMF), might indicate this. | ||
|
||
Yet not every use of an overloaded term must be qualified. Various senses might be left implicit if no confusion is caused (again, considered in context). In accordance with a more general rule that plain common nouns are to be preferred over jargon, terms such as <q>control</q> or <q>profile</q> might be used without further explanation when context makes it clear enough what is meant *even if a term is also overloaded* (means other things in other contexts). Within the context of describing a processing pipeline for OSCAL profiles, <q>valid profile</q> might be taken as shorthand for <q>XML \[or JSON] document known to be valid using a \[profile] schema</q>, especially when this is stated outright at the top. The fact that other kinds of validation may be implicit, or not relevant, does not have to be explained. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I like the sentiment, but I would disagree with this statement. Is this something we want to encourage in the style guide? I defer to the greater group.
When working on our informational overview document regarding the emerging rules construct and the validation documentation in #1169, this was one of the key issues we had in discussion with community: many misunderstandings in OSCAL occur due to the use of a common noun, and we presume the audience knows which meaning of the 1/n
meanings of the overloaded common noun we mean.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In view of your resistance I am making small edits on this paragraph. The advice is for me, actually (but if the shoe fits, wear it).
|
||
In general, technical terms (and even formal names) should be avoided in favor of non-technical language except when necessary for specific clarification *within the context of use* including your intended audience. For instance, a tutorial intended for beginners does not (and should not) exhaust every topic: it should be technical only as appropriate and necessary, and technical language not actually useful to understand problems or concepts at hand, can often be saved for elsewhere. | ||
|
||
Nonetheless, technical terminology is often necessary. For documents reaching non-technical audiences (or non-domain experts), this often requires offering brief definitions for terms at the head of a document, or even a glossary. Even technical audiences can be helped by a link to a specification that defines formal terms. In any case, formal terminologies can be called out and identified as such. For example, both OSCAL and RMF (Risk Management Framework) terminology may combine in a description of a use case for a new feature. In this instance a brief notice of what is meant by terms such as <q>profile</q> or <q>baseline</q> is not out of order. |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Resurfacing the previous comment about acronyms, can we use this to drive our own use of the terminology page, the CSRC glossary, and a hierarchy of more general resources we ask people to rely upon? (And to be fair: I address this to the whole team.)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We can include links to glossaries / acronym pages; I have already included a couple.
|
||
In addition to maintaining web platforms, the OSCAL project publishes resources and tools through the formal NIST data publishing process, for release through NIST's [Science Data Portal](http://data.nist.gov). This is appropriate for (NIST-authored) contributions that should be indexed as scientific publications, or archived as discrete data sets. | ||
|
||
The formal review process this entails will presumably not be a problem for works the site editors are also reviewing. Mechanically, the process can be lightweight, requiring nothing more than a structured metadata record with a link to a published (web-accessible) OSCAL resource. This record is provided and maintained in NIST's metadata repository for [Management of Institutional Data Assets](https://midas.nist.gov/)). |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I do not believe MIDAS is a public resource accessible outside of NIST staff's intranet, so do we reference it? I know we should remind people there is a data management policy, but Google dorking to search for it outside of NIST intranet I struggle to find mention of it. I find a lot about another scientific project named MIDAS though.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Edited to remove the links requiring credentials. Readers will just have to follow instructions on finding them.
|
||
Even if submitting work only for the web site, NIST authors will want to be aware of NIST-specific guidance for its formal publications: | ||
|
||
- [Guidance for Authorship of Scholarly and Technical Publications](https://inet.nist.gov/adlp/directives/guidance-authorship-scholarly-technical-publications) (requires NIST authentication) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Like MIDAS, ditto re access to NIST internal resources on the intranet, unless we really deem it valuable.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Okay, but if I get slammed for removing the helpful links....!
Co-authored-by: Alexander Stein <[email protected]>
@aj-stein-nist has a couple of questions/issues to float up, but not AFAIK blockers at this time. Our long-term strategy is still evolving.... |
This has been ready and not touched. Is there a way for me to put it back into the Triage bucket? @david-waltermire-nist @aj-stein-nist |
I know I had initial feedback but I hadn't circled back on my review comments or how I can help after I returned because I checked #1185 and this work is slated for 1.1.0 (not 1.0.4). That said, I definitely would like to get it in place sooner rather than later if that is possible and desirable! Let me know how I can help. |
b7d376f
to
d8682eb
Compare
09e6eee
to
7aa46b6
Compare
c4de2fe
to
0a6189a
Compare
fa5250c
to
e4f710c
Compare
e5a3198
to
a37c965
Compare
eff9ce3
to
dd6486a
Compare
This PR is good, but we have not revisited this work in recent sprints and it is unlikely to get merged in the next sprint, perhaps not the following. I will close the PR, but the branch will remain. We can revive once we bring the related issue back to the scope of a sprint and take it from there. |
Committer Notes
Rough draft towards a web page addressing #1185. (PR targets
develop
pls tweak if wrong.)This may have to cycle a couple of times to get the links to work, because it links to its own source.
All Submissions:
"?
Changes to Core Features: