- From: Marcos Caceres <w3c@marcosc.com>
- Date: Sun, 8 Oct 2017 22:53:18 -0400
- To: spec-prod <spec-prod@w3.org>
Hi ReSpec Users, tl;dr: After discussion with the community [1], we've had to make a breaking change to how ReSpec generates some fragment identifiers. # What happened? Although ReSpec was generating conforming URLs (including conforming fragments), we found that a number of applications have problems with fragments that included special characters - even when those characters where URL encoded. This included validators, email programs, and Twitter. This meant that it became difficult to share URLs reliably amongst ourselves and the wider developer community. # What did we break? ReSpec now strips non-alphanum characters when generating fragment identifiers. So, for example, the following fragment: "#the-foo()-method" Becomes: "#the-foo-method" # What does this mean for me? For most people this will have zero impact. If you are cross-referencing another ReSpec document, ReSpec automatically rewrites the fragments to the new format for you. However, in order to "heal" any links that might have broken in this process, we advise running the link checker and, where possible, republishing working drafts on /TR/. # Help! I'm having problems! It this has affected your spec in a significant way, please feel free to reach out for assistance. Happy to have a look at your spec. Otherwise, please file a bug: https://github.com/w3c/respec/issues Kind regards, Marcos [1] https://github.com/w3c/respec/issues/1353
Received on Monday, 9 October 2017 02:53:47 UTC