Working, as I am, on a spreadsheet application, I often refer to the Open Formula specification which does a fairly good job of providing useful edge-case tests, and noting inconsistencies between the implementation of various functions across Excel, Gnumeric, Open Office, etc.

Recently I noticed that one of their test cases seemed to have a spurious minus sign in it. After confirming that I was correct in every spreadsheet app I had access to, I decided to be nice and report this to them.

But:

• There is no email address prominent in the document to send it to
• There is, however, a link in the PDF that takes you to a website that has a well-hidden link (that looks like text, but, for no obvious reason, isn’t, so that using Find in your browser doesn’t actually find it) to a feedback form.
• When you eventually find and follow the link, you are told: due to technical problems the public “Send a Comment” button on the formula subcommittee page currently DOES NOT WORK. We very much want to receive public comments — for now, please send comments to the formula subcommittee through the <OpenDocument TC comment form>.
• Upon following that link, you discover that the mention of such a comment form was a lie, and you actually need to subscribe to a mailing  list.
• But before you can even think of subscribing, first you need to download a PDF of a legal agreement (or, more accurately, a blank template of a legal agreement that isn’t even filled in for the right specification yet).
• Then you need to read and understand the 6000 word OASIS Intellectual Property Rights (“IPR”) Policy.
• Then, even if you were minded to sign such a document without first taking legal advice, you need to work out what to do with it. The “Guidelines for Mailing List” page implies that there’s a “Subscription Manager” tool that would presumably require you to submit the filled in legal document somehow, but that link is actually just a local link to <a href=”#subscribing”>, which simply takes you back up the page by two lines, as that’s the section that the link is already in.

Maybe there are more instructions in an auto-response email when you try to subscribe, but I don’t know, because I never got that far. Instead I emailed the committee chair – only to get a response (a real one, from a live person, not an auto-response) saying that paying any attention to my email without me going through the entire process would “potentially contaminate the work of the TC”.

Unsurprisingly I haven’t reported any of the 20+ errors I’ve found since.