Hervé Pagès

Sure. Not duplicating the user input when it doesn't need to be modified sounds like a good general principle anyway, even without ALTREP in the picture. I will suggest a...

@Jiefei-Wang Hi Jiefei, This is done in IRanges 2.19.12: https://github.com/Bioconductor/IRanges/commit/6eb40e6a0d2dc9bccf8fdffc2b91f282bb023c94 Note the two-pass approach. The idea is to do a first pass in order to: - Determine whether the supplied...

Sounds good Jiefei. Thanks!

I was thinking in mimicking the approach taken by `data.frame()`/`DataFrame()`, which is to make it hard, but not impossible, to construct an object with no names, or with names that...

Thank Mike. I'll look into this. Note that it has never been a requirement for `z` to have the same row names as `se` when doing `rowData(se)

A compromise could be to only issue a warning that the row names differ and are being ignored. That won't break any code and people who have code that triggers...

Sounds reasonable. So to be precise, `colData(se)

This was considered but IIRC was discarded because of performance considerations. This was at least 7-8 years ago so I don't remember the details but I think that having the...

Ok I see. It looks like defining this coercion changes the class hierarchy. Before defining the coercion: ``` > showClass("SingleCellExperiment") Class "SingleCellExperiment" [package "SingleCellExperiment"] Slots: Name: int_elementMetadata int_colData Class: DataFrame...

Yes maybe. Best way to know is to try. Things have changed a lot in 8 years. A little bit of history: 8 years ago we didn't have SingleCellExperiment. VCF...