Turing 1.0 Meta Meta Issue

[penelopeysm]

We need to reshape the following topics into meta issues.

Please feel free to edit the list to:

1. Add links to the meta-issues that you've already made;
2. Remove the corresponding topics / edit them to exclude any meta-issues that were already made.

---

General

• Figure out a consistent and robust way of defining and testing interfaces #2434

DynamicPPL / JuliaBUGS

• Modernise the tracing data structure for DynamicPPL #2423 (depreciate VarInfo in favour of new designs which are more modular, intuitive and easier to work with)
• Significantly simplify the tilde-pipeline in DynamicPPL #2422
• Simplify the context mechanism in DynamicPPL #2424
• Better performance for JuliaBUGS and integration with DynamicPPL models, e.g. via DynamicPPL.to_distribution  #2425

Inference

• Remove (duplicate) samplers being defined explicitly in Turing.jl #2413
• Significantly improve the Libtask library using ideas from Mooncake / ReverseDiff #2427
• Integrate SSMProblems & GeneralisedFilters with Turing.jl #2428
• AD Meta Issue for 1.0 #2411

Documentation

• Update our documentation for v1.0 #2429
• Introduce a docs FAQ section #2431

Misc

• Do a pass of all open issues on Turing.jl/DynamicPPL.jl/JuliaBUGS.jl/TuringGLM.jl and the docs repos.
• AdvancedVI 0.3 and compatibility with Turing.jl #2430

[mhauru]

I made empty shell issues of all the above bullet points and collected links to some related subissues in them. Except for one or two, they all need to be spelled out in much greater detail. Everyone should feel free to edit them. @yebai obviously has the high level picture though, so Hong could you please take a look and write a summary for most of them, or tag whoever you think is best placed to do it? @torfjelde might have a good overview of many of them too, would you be happy to spell some of them out?