Remove an unneeded import std/formatFloat

And clean-up release status to be tip.

Add CLuser config support

That's it; Nothing more substantitve, but also no reason not to release.

Install old fitl/dists program; Add new estMI program

While I should still really hook fitl/estMI into the IID testing in fitl/gof and that IID testing into fitl-proper for residuals, these programs/modules are both useful enough on their own as-is to justify a release.

The VC log and code has more details / references on mutual information estimation which is the only extant approach with error estimates/p-Value capability on the MI statistic I know of.

Actually update .nimble file

Oops. Even with nrel, one can forget to add -u. :-( Sorry.

Track addition of `maxTry` to `spfun/studentT.ccPv`

This was a breaking API change in spfun-0.7.0.

Fix fitl/gof non-compilation error

Oops. Needed a float conversion.

Add k-sample homogeneity & asymptotic ACF IID to `gof.nim`

Add k-sample Anderson-Darling homogeneity test & Sarantsev IID on ACF tests wired into gof.nim helper/demo program. (Must still Bonferroni correct this test suite someday since with so many it's now gotten much too likely that some will just fail by chance.. as with any "suite".)

Also adapt to and update new spfun & retire Q.

Add `fitl/cds` an interpolated resampling module with gnuplotter

It's really a pretty trivial application of qtl, but might be useful.

The gnuplot test program is for visualizing a swarm of resamples - an infinite swarm of infinitely thin curves overlaid being roughly the "most accurate" confidence band/error envelope that the inbound bu/edplot point-wise & simultaneous confidence bands "bound" from below & above (if that makes sense to anyone besides myself).

Track new simulation-based p-Value testing of serial autocorrelation

This is mostly to assess model appropriateness via fit residuals.

Just update spfun dependency

'nuf said.