Integrate grunt-check-pages to automate link testing, prevent link rot

[dforesman]

Many of the newly-revised sections of the Standards Document have been stripped of the useful content links and source citations that made the original document so broadly applicable - as it stands now, it is more a "list of things to google" than a collection of distributable documentation.

My understanding is that the drive to remove links from the content is related to the problem of link rot. In order to combat this, a tool like grunt-check-pages should be integrated into the build to automatically check links in the document and report any that are dead, enabling us to more easily stay on top of keeping links current within the doc.

grunt-check-pages

• repo: https://github.com/DavidAnson/grunt-check-pages
• demo / launch article: https://dlaa.me/blog/post/gruntcheckpages

alternative: grunt-deadlink: https://www.npmjs.com/package/grunt-deadlink

[roblarsen]

(Two serious comments in a row from me!)

Tools are great and there are many ways to automate checking for link rot beyond a build tool. I'm betting that you could sign up to an email service that would send you broken link reports once a week or whatever.

That said, if you (and by you I mean the entirety of the Isobar team involved in this) are so worried about editorial resources going forward (a valid concenr! This new version has been in the works since just after I left isobar!) that you've created a bad web document (the fundamental defining characteristic of the web is the use of hyperlinks, full stop) about web standards, I think the whole thing is at this point, a questionable exercise. I'd hate to see this doc become less useful in order to defend against the inevitable (ongoing?) resource crunch.

( @rcherny I swear my next comment will be some weird minor troll.)

[dforesman]

@roblarsen : Thanks for putting into words what I was feeling and not quite able to express - the root problem here is that if we are so resource-starved as to be unable to keep a handful of links updated over a several-year period, I question how we expect to maintain this documents' relevancy in the fast-moving world of front-end.

From personal experience, it was the wealth of links and straightforward resources in the Standards Document that originally drew me to it, years before I joined Isobar. I worry that in an effort to reduce maintenance overhead, we may be destroying the most useful aspect of the document. I understand that one of the goals for this project is to have living documentation that can be delivered to clients - but without the critical sources and links, what we have is a lot of words about front-end dev but very little usable content. I doubt many clients would consider this to be sufficient documentation for delivery to client development teams, who likely have a much lower baseline of front-end knowledge.

I've filed this issue as an attempt to alleviate the immediate concerns around link-rot, but there may need to be a larger discussion around how this document will be maintained and updated post-launch, and how we can ensure it gets the attention it needs outside of the occasional bench time.

[rcherny]

@roblarsen Don't go and get all serious on me! I'm kidding, of course.

All of these are valid concerns and of course the Web is built on a foundation of linking to things. I'd never sit here and say "we need to stop linking to things". I am just trying to refocus us from infrastructure and conceptual discussions back to actual content generation so we can just refresh the darn thing.

... and yeah it better be maintainable or yeah, we've done the wrong things.

I'm just saying we need content. That's my mantra these days...

Anyone can feel free to issue pull requests with links to content that has merit. It's a pull request. Let's not get bogged down in the how's and why's and actually do something here :-D

[rcherny]

I'm marking this for Standards.Future for the time being. That's not to say someone resourceful could change that or just do it ...