Add Relate / Collate to the software->infer section?

[hyanwong]

Should we do this? The RelateFileFormats program has a --mode ConvertToTreeSequence, but it makes a JBOT and loses information. Still, other people might want to use Relate for some things?

[petrelharp]

Hm, that's tricky - maybe? There's also the other simulators now, e.g. GSpace. But these are "other" software that we ("tskit developers") are not developing ourselves: e.g., endorsing the correctness or documentation of, for instance. But maybe we don't need to make that distinction? I guess I would like to be less exclusionary, but on the other hand we don't want to clutter up the list with lots of minor and abandoned projects that happened to use tree sequences. Another point here is that Relate does not have source code available; I'm not proposing that we make that a criterion for inclusion (but would be tempted to =).

[hyanwong]

I largely agree with you here @petrelharp. My main reason for mentioning Relate is that it is, AFAIK, seeing some uptake in the community (and indeed you used it in the stat paper, I think), and it might be useful to point it out as a piece of software somewhere. But I'm similarly tempted re source code.

[petrelharp]

I used it in the treestat paper because it had times, which we needed, whereas tsdate wasn't out yet...

[benjeffery]

We prob need a new section which is "other software that can produce tree sequences"?

[hyanwong]

We prob need a new section which is "other software that can produce tree sequences"?

Or maybe a subsection in each of the "infer", "simulate" etc sections that says "here's a non exhaustive list of other inference software that can output tree sequences"?