Help talk:Recursive conversion of wikitext

Parser functions

If the substitution call is in a parameter of a parser function which is also substituted, two iteration steps are carried out in one save.

• I thought two saves are needed. What is one save?--Hillgentleman 18:33, 19 April 2007 (UTC)
  • I see (template talk:square root A).Hillgentleman 18:47, 19 April 2007 (UTC)

If

Thus it does not seem possible to put the construct of the previous section inside an if-construct.

• Are you referring to these? Template:Eraser-left A, Template:Eraser-right A

--Hillgentleman 08:42, 24 April 2007 (UTC)

No, I mean that e.g. in Template:Square root with defaults I had to put the branching to distinguish a defined and undefined parameter value at every occurrence of the parameter, instead of having "if" with a substitution call in the then-part and another in the else-part.--Patrick 12:41, 24 April 2007 (UTC)

• Do you mean this Template:Square root with defaults A?--Hillgentleman 18:41, 24 April 2007 (UTC)

• Thanks. I removed "it does not seem possible".--Patrick 13:29, 25 April 2007 (UTC)

new preprocessor

Expansion limit is to be replaced by "preprocessor node count": [1].

Notice also that there is an ongoing discussion on parser development. [2]

Template:subst

It can be created using Template:subst containing "{{subst:"

Except, on the English Wikipedia, that template contains something different, making this documentation wrong for there. I barely understand this page enough as it is, so I'm not going to edit it myself, but someone needs to do something about that. -- Hex (talk) 09:52, 24 September 2012 (UTC)

Also, I might as well add that this page is excruciatingly poorly-written and difficult to understand. -- Hex (talk) 09:55, 24 September 2012 (UTC)

This is an advanced topic. The reader is supposed to understand that the template name is an example.--Patrick (talk) 14:42, 24 September 2012 (UTC)

It fails completely at conveying that impression. It needs rewriting; the "advanced" documentation here that I've encountered is written like UNIX documentation from the 1980s, and based around massive assumptions about the reader. -- Hex [^t/_c] 09:37, 26 September 2012 (UTC)