David Widmann

I think it might be more consistent to apply the pattern in https://github.com/JuliaNLSolvers/Optim.jl/blob/7cc8328a590a0c815e23005321459712382b5b42/src/multivariate/solvers/first_order/bfgs.jl#L81 to https://github.com/JuliaNLSolvers/Optim.jl/blob/7cc8328a590a0c815e23005321459712382b5b42/src/utilities/perform_linesearch.jl#L18. Generally, isn't the error caused by an incorrectly typed `initial_stepnorm` provided by the user? If...

This is the old webpage (AFAIK nobody has access to) which also only contains old and outdated docs and tutorials. The Turing docs and tutorials are hosted at turinglang.org nowadays.

AFAIK unfortunately really nobody has access to the old turing.ml domain which means it's impossible to add a redirect.

I have no idea :shrug:

The page seems to be served by cloudflare (nameservers audrey.ns.cloudflare.com and wilson.ns.cloudflare.com according to a lookup), maybe the old content is just cached there and there's no intentional copyright violation...

> Is there a reason why we're doing a lot of progress reporting during tests @yebai @devmotion ? It makes it quite annoying to determine which tests are actually failing...

The keyword argument is called `initial_params` since Turing 0.30.0/DynamicPPL 0.24/AbstractMCMC 5.

Seems that https://github.com/TuringLang/Turing.jl/pull/2115 is problematic - it added support for AbstractMCMC 5 (without removing support for AbstractMCMC 4 even though the API is incompatible!) and DynamicPPL 0.24 without updating all...

Yes, I think it's redundant and confusing, now that `initial_state` is an official AbstractMCMC keyword argument.

> This PR reenables the support for _PriorContext_ in optimization, which was disabled by #2022. AFAICT support for `PriorContext` has never existed and has never been tested. The check in...