For Mark A: I am continuing to take up your invitation of asking really basic questions, in the certainty that the answers will be helpful to me and in the hope that they may be useful to some onlookers.

Here is the kindergarten-level question I have right now.

The text of the export template "HTML item" is shown, in its entirety, below. This is directly cut-and-pasted from the template.
Code:

<h3>^title^</h3>
^source^
^text^
^children(/Templates/HTML page/HTML item) 

My goal is to have the export first show the item's title, then its $Source (a user attribute of the Set type, in practice a person's name), then the text.

What I get instead from this template is shown in the screen shot of a sample item.



That is, it is treating ^source^ as a literal string to be included, rather than as a reference that in this case would bring up "Zelikow." Que pasa?

Quote:


Thanks! That solves the operational question. And then, for me, raises a conceptual one:

Why does Tinderbox understand ^text^ and ^title^, rather than requiring ^value($Text)^  , but does not understand ^source^  ?  

Is it that ^title^ and ^text^ are unique exceptions (because of their ubiquity) to the ^value()-based approach? Or is there some other structural rule for us to be aware of here, comparable to the rule that .contains() will work with regex for string values, but not for set values?

Thanks on both practical and conceptual fronts.  

Way, back, TB was much simpler and there were fewer things for TB to figure out in terms of export code.  We now have c.250 system attributes and as many custom ones as you like.

^text^ and ^title^ are special as they don't just export the $Text and $Name data but they check it for stuff that needs turning into HTML. Also in 2014 is much more accommodating of things like accents and non-Roman alphabet characters than it was in 2001. Probably in most cases ^value($Name)^ would work as well as title. ^text^ remain special - for now anyway - as it has to do things like turn web and TB links into HTML link mark-up, handle images (I know that's no functional right now) and do things like make bulleted list mark-up for lists in your source - all so you don't have to know or understand how that's done.

I'll also bang the drum for closing export codes with a ^. I know it's not a requirement but I can see no sensible reason why a learner of export code would want to increase the chance of TB getting confused as to where the codes stops> so is ^text^title implying ^text then ^title or ^text^ followed by the text 'title. Using ^text^^title^ is unambiguous. Apart from when I forget (!), I allways close my export tags.

It's all buried in aTbRef but I totally acknowledge there's a lot to read and you need to assimilate a lot before it swims into focus! Threads like this I think help a lot - I often have to go and look up/test things I thought I knew.

Long-term TB users will be conscious this is 'artisanal software' as opposed to something built on a corporate frame. As such, action code has grown in slightly irregular fashion and not all similar-looking codes behave the same way. Not being able to use == with a set or list is one such oddity. If not sure, rather than guess it's best to check a reference.