[pmwiki-users] Permissions conundrum, section behaviour
henrik.bechmann at sympatico.ca
Fri Oct 6 07:35:51 CDT 2006
> ...Only because you keep insisting that [[#comment]] is exactly the
> same as an "HTML anchor". It is not. Only the *first* such marker in
> a page generates an HTML anchor -- subsequent repeated instances
> of the same marker do not generate an HTML anchor, and thus do
> not break the "HTML rule to have unique anchor names."
Very interesting, certainly relevant, and news to me (and I fear,
confusing for authors)
> And to return to the earlier comment:
> "Where does that ordinal index come from? Surely not from the user...."
> Whatever is inserting the ordinal indexes in the [!textblock ...!]
> scheme above can just as easily be generating unique
> [[#comment_nnn]] markers instead. (Actually, it's slightly easier
> to use [[#comment_nnn]] instead.)
>> If further information was required by a recipe:
>> (:comments customcomment:)
>> [!textblock customcomment 1 time=10/5/06-12:46a author="Henrik Bechmann"
>> first comment text
> [[#customcomment_1 time=20061005T0046 author="Henrik Bechman" ...]]
> first comment text
> but I did want to get some quick observations
> about how I see [!textblock comment!] and [[#comment]] as really
> being equivalent.
> In fact, I think I can _prove_ that they're equivalent by noting
> that a simple pattern replacement is completely sufficient to turn
> one scheme into the other (and vice-versa):
> ## [!textblock comment args!] --> [[#comment args]]
> $y = preg_replace('/\\[!textblock (.*?)!\\]/', '[[#$1]]', $x);
> ## [[#comment args]] --> [!textblock comment args!]
> $x = preg_replace('/\\[\\[#(.*?)\\]\\]/', '[!textblock $1!]', $y);
> In other words, I'm saying that '[[#' is simply another
> (much shorter) way of writing what you have for '[!textblock',
> and that also has the mnemonic advantage of associating
> sections of text with url fragments.
Regarding the form of the delimiters, although I believe that textblock
url fragments could be available with a different markup, and that users
would find the lack of distinction between anchors and textblock
delimiters confusing, I think the application of the ADT concept in the
form of a structured extensible interface for generated textblocks (ie
the functionality) is far more important.
IAC let's see what bubbles up out of this in time.
Thanks for the rolicking discussion!
More information about the pmwiki-users