We should start a rational redesign of \vardef and \varseq, now that we are using it more:
[ ] \varseq should also support subsequences: If we have declared a sequence \varseq{seqR}1n{\comp{R}_{#1}} then we can easily talk about the full sequence \seqR!.
But when we work with \seqR, very often want to talk abut the subsequences R_1,\ldots,R_{n-1} and/or R_2,\ldots,R_n. But this does not seem to be supported.
I would like to e.g. have a generated macro for subsequences so that I would writen them \subseqR1{n-1}and \subseqR2n. Note that this is not necessarily the final name or form for such a macro.
[ ] \varseq should have prececedences: $\inset{\seqw{i}}\RealNumbers$ currently gives brackets around w_i)
[ ] should we have infinite sequence variables as well at the sTeX level?
[ ] we sometimes have variables indexd by something different than numbers, e.g. for Constraint networks
They should behave like sequences as well (I think)
We should start a rational redesign of
\vardefand\varseq, now that we are using it more:\varseq{seqR}1n{\comp{R}_{#1}}then we can easily talk about the full sequence\seqR!. But when we work with\seqR, very often want to talk abut the subsequencesR_1,\ldots,R_{n-1}and/orR_2,\ldots,R_n. But this does not seem to be supported. I would like to e.g. have a generated macro for subsequences so that I would writen them\subseqR1{n-1}and\subseqR2n. Note that this is not necessarily the final name or form for such a macro.\varseqshould have prececedences:$\inset{\seqw{i}}\RealNumbers$currently gives brackets around w_i)They should behave like sequences as well (I think)