[chrisbridgett]

It sounded good to me initially, but then I thought... those users that are interested in testing their application (hopefully most) will already have built tests for their API.

I guess it's a question of whether you can provide a service that is good value for money (i.e. it saves time), when I can already whip some tests together relatively quickly during development of my API.

[gggarnier]

Thank you for your answer, but I think I haven't described it clearly enough: it wouldn't check for broken API links but it would be a REST API (a service) for checking broken links on any website - something like W3C checker.

[chrisbridgett]

Ah I see what you mean now. My mistake.

That sounds more useful... but it's also something that most developers could probably build for themselves in a few minutes?

1. Grab page 2. Parse and find links 3. Make a request to each link 4. Return any links that are returning 4xx or 5xx.

Maybe there's a use for it where people want to test a particularly large set of pages/links making use of a data center's typically superior connection? shrug

[gggarnier]

Yes, exactly. It would also be able to crawl linked pages, take screenshots of broken pages, show the path how to get to the broken pages, provide reports, ability to schedule checks like say every 24 hours, alerting (SMS, email), ...

Plus it would check not only links but also images, stylesheets, scripts, submitting forms - something like a 'website testing platform'.