[texworks] Unwanted side-effect of <tab> : how to avoid ?

[Philip TAYLOR (Webmaster, Ret'd)]

Alain Delmotte wrote:

> There is another solution, but you'll have to take the reflex to do:
> just add a space after {
>
> This {[return]
> [Tab]
>
> would create the problem
>
> This {[space][return]
> [Tab]
>
> not

Understood, and after faffing around trying other techniques,
I usually fall back on this one, but I hate it in principle :
setting a document with tabs can cause documents to look
different in different environments, but at least the implied
nesting is usually honoured, but setting a document with
a mixture of tabs and spaces can be an absolute disaster in
terms of portability -- implied nesting frequently flies
straight out of the window when one views such a document
in a different environment :-(

> Philip, I would like to create a completion file for you about XeTeX.
> Are there instructions in XeTeX specific and not part of Plain TeX?
> (Which already has a completion file.

Probably :-)  But please, do not do this : I do not use auto-completion,
do not expect to want to use auto-completion, and above all do not want
you to waste your valuable time writing something for me that I am very
unlikely to use.  So I am extremely grateful to you for your kind offer,
but I really prefer to create my documents entirely by hand !

P.S.

> Plain TeX? (Which already has a completion file.

In that case, why does a Beamer autocompletion kick in when
I have typed something that is pure TeX (i.e., 100%
Plain compatible, looking nothing like a call to Beamer)
rather than a Plain autocompletion ?  I type

	\def \foo
	<tab>{
	<tab>

and at that point autocompletion kicks in with a Beamer entity,
yet this is the entire document and there is nothing that
should lead TeXworks to believe I am creating a Beamer-based
document.  Of course, any Plain autocompletion would have
to be complete and utter guesswork : no editor in the world,
no matter how intelligent, could possibly guess on the basis
of

	\def \foo
	<tab>{
	>tab>

what the next intended construct is most likely to be !

** Phil.