Here’s what my code looks like
and here’s how it renders
It’s finding terms from a prior experiment that are no longer used in the text. I can only assume that the filter has cached the older terms and now is using them instead of the actual specified macro.
I opened and shut the tiddler several times, but the results are still there.
Is the code complete? It looks like it’s missing the definition of <<toc-lorem>>.
Do you get the same behaviour if you save and reload the wiki?
This is on TiddlyWiki.com. The definition of .toc-lorem is provided by TW. The behaviour is quite persistent. I’m sure it will restore after I save as a local file and reload.
…
Whoa! It persists even after saving to a local file and reloading.
I was using Chrome. I just recreated the error on Firefox.
On TiddlyWiki.com, I first use this code:
<$list filter="[[This is an example tiddler]splitregexp[\b]trim[]!is[blank]limit[5]]">
</$list>
<<.toc-lorem>>
That works. Then I replace it with this code:
<$list filter="[<.toc-lorem>splitregexp[\b]trim[]!is[blank]limit[5]]">
</$list>
<<.toc-lorem>>
It continues to display the previous text, as though it has retained a cached memory of it.
I am probably missing something. Shouldn’t the output be identical in those two examples?
What output do you expect from the following?
<$list filter="[<.toc-lorem>splitregexp[\b]trim[]!is[blank]limit[5]]">
</$list>
You’re right. My original text was too much like the output of .toc-lorem. My bad memory confused them. Sorry! 
No worries, we all get a bit cross eyed staring at code at times 
@Mark_S one informal rule I have is if a filter does not start “[all[current]]” I expect to get the content from elsewhere or everywhere (all tiddlers). Then I put extra effort into where exactly the content comes from.