update dc2_object_run2.2i_dr6_v2_with_addons

[yymao]

In this PR:

• The original dc2_object_run2.2i_dr6_v2_with_addons was renamed to dc2_object_run2.2i_dr6_v2_with_addons_v1
• dc2_object_run2.2i_dr6_v2_with_addons_v2 was created with the new truth match addon
• dc2_object_run2.2i_dr6_v2_with_addons was changed to an alias config, currently still pointing to dc2_object_run2.2i_dr6_v2_with_addons_v1 so that it doesn't break anyone's one. A deprecation message was added to encourage users to update to dc2_object_run2.2i_dr6_v2_with_addons_v2.

Thanks @johannct for suggestion.

[yymao]

In aefba29

• dc2_run2.2i_truth_merged_summary was added as an alias to dc2_run2.2i_truth_merged_summary_v1-0-0
• dc2_truth_run2.2i_summary_tract_partition was deprecated

[yymao]

In 812e6f8,

• desc_dc2_run2.2i_dr6_v4_object, desc_dc2_run2.2i_dr6_v4_truth, and desc_dc2_run2.2i_dr6_v4_object_with_truth_match were added. Note that v3 is skipped to match the release note version.
• desc_dc2_run2.2i_dr6_object, desc_dc2_run2.2i_dr6_truth, and desc_dc2_run2.2i_dr6_object_with_truth_match now point to v4 instead of v2.

[JoanneBogart]

I haven't tried to load these, but overall the changes and additions look reasonable. I have a question about the update to dc2_truth_run2.2i_summary_tract_partition.yaml. Base directory is ..truth/tract_partition/raw, which includes rather large parquet files. The suggestion is to use dc2_run2.2i_truth_merged_summary instead. That points to the new, much smaller summary files. Is that because the files in tract_partition/raw include all the columns from galaxy truth? But we don't expect people to actually be referencing those extra columns?

[yymao]

Checking the size:

• /global/cfs/cdirs/lsst/shared/DC2-prod/Run2.2i/truth/tract_partition/raw is 73 GB
• /global/cfs/cdirs/lsst/shared/DC2-prod/Run2.2i/truth/truth_merged_summary_v1-0-0 is 57 GB

Neither of these includes all the columns from galaxy truth. I think the difference mainly come from (1) the smaller footprint of the star truth catalog and (2) the small total number of columns (mostly from removing flux_noMW).

[JoanneBogart]

I was comparing only individual files; I think 2) is the reason for the difference I saw.

[JoanneBogart]

This all seems fine.