This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
Section 5.4.4 says that the current mode is "the mode in which the current template rule was invoked". However, section 6.7 specifies cases when the current template rule (CTR) becomes null, and the current mode, according to section 6.5, preserves its value in those cases. I found it confusing, and, in my opinion, it might help other readers if the spec clearly says that the current mode is still defined even if CTR becomes null. Also, I found two appearances of the term "current template", I think they should be changed to "current template rule" for the sake of consistency.
I think this point is largely editorial. I'm not sure how best to respond to it. The sentence in 5.4.4 is intended to give a brief introduction to the concept of "current mode", and it is clearly linked to the more complete definition of the term in section 6.7. Perhaps I should just add a word in 5.4.4 that emphasises that the explanation is informal, like "generally" or "normally". I agree that it would be a good idea to use the term "current template rule" consistently, rather than "current template". This also means (see comment on your other bug report) fixing the problem that the current template is set to the initial template even if it isn't a template rule. Michael Kay (personal response)
The XSL Working Group asked the editor to propose revised wording for the relevant rule to clarify the intent. The WG also accepted that "current template" should always be "current template rule" (2 occurrences found). Proposed revision: change The current mode, which is the mode in which the current template rule was invoked: see <specref ref="modes"/>; to The current mode, which is the mode set by the most recent call of xsl:apply-templates (for a full definition see <specref ref="modes"/>); If there are any problems with this resolution, please reopen the bug. Michael Kay XSLT 2.0 editor