Some responses from MS experts re DRs
charlie.clark at clark-consulting.eu
Wed Sep 5 17:51:54 CEST 2018
Am .09.2018, 16:51 Uhr, schrieb Rex Jaeschke <rex at rexjaeschke.com>:
> 17-0029 SML: Quoting of Worksheet Names Containing Periods
> 2018-09-04 Alfred Hellstern:
> Alfred Hellstern: For the new questions in Charlie’s last comment, it
> would be better to open a new request since they are more broad than the
> original request and represent an increase in scope.
I agree that a detailed discussion of release notes could be a separate
issue. But it remains relevant to this issue as to whether it can be
acknowledged that a particular release solved the problem: was the issue
as described a problem with the XML or MS Excel 2016?
> Response from our product team:
> In his most recent comment, Charlie mentioned the original ask has been
> resolved. One clarification that can be made is to the EBNF to indicate
> that the first character of an unquoted name should be further
> restricted to also exclude ‘.’ and 0-9.
Thanks very much for this but unfortunately it seems, alas, still
incomplete. For example, if the % character is used, the worksheet name
also seems to require quoting. If the specification is still incomplete
then perhaps the solution is to require quoting for all worksheet names
and remove additional distinctions.
Also, I may have missed it, but shouldn't the use of EBNF be explicit?
Does a sample implementation and test suite exist for this? I think this
issue highlight the importance of being able to validate such rules,
preferably at the schema level. Indeed, given the use of XML schema it
seems odd to have additional requirements that cannot be checked against
the schema. The EBNF might make sense at the formal level but should these
not be accompanied where possible by relevant regular expressions which
can be enforced in the schema?
Also, who edits and validates the EBNF?
Furthermore, there has been no information provided about reserved names
for worksheets, ie. names that while formally correct, may cause problems
between implementations and which should, therefore, be avoided.
Clark Consulting & Research
This message has been scanned for malware by Forcepoint. www.forcepoint.com
More information about the sc34wg4