Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Thank you everyone for your efforts with this! -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-698976288 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Closed #243 via #294. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#event-3806215352 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Reopened #243. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#event-3705525148 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Some final rewording agreed above (https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-597152790) didn't get pushed to PR #244 before it was merged. Reopening this issue and linking new PR (#294). -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-683132191 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
It looks like consensus has been reached on the text as it stands - I'm starting the countdown and will merge this PR, barring objections or further changes, on 2020-04-16. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-604313430 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@ethanrd : yes, dropping "assigned" is a good idea and I like your latest reformulation. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-597187583 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@graybeal @martinjuckes @JimBiardCics - Thanks for your input. @martinjuckes : I agree the first sentence in the [Data Type section](http://cfconventions.org/cf-conventions/cf-conventions.html#_data_types) could be stronger. I would suggest simplifying your alternate sentence by dropping "assigned" and being more specific that these are all netCDF external data types supported by netCDF-4: > "Data variables must be one of the following data types: ... _list of data > types_ ... (which are all the [netCDF external data > types](https://www.unidata.ucar.edu/software/netcdf/docs/data_type.html#external_types) > supported by netCDF-4)." -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-597152790 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@ethanrd I like your verbiage. I think you could simplify it a bit by saying In many situations, any integer type may be used. When the phrase "integer type" is used in this document, it should be understood to mean **`byte`**, **`unsigned byte`**, **`short`**, **`unsigned short`**, **`int`**, **`unsigned int`**, **`int64`**, or **`unsigned int64`**. This covers both "an integer type" and "any integer type". -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-593511337 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
(Just an idea, take it or leave it.) I would add the following text after the next-to-last sentence in the paragraph (which first uses the phrase "integer types"): "In this document, the term "integer type" includes all of the following types: …" Could be in parentheses, or not. I think that means the same as what your text expresses. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-593059990 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Expanding to ... When the phrases "an integer type" or "any integer type" are used ... Anything else? Or again, better wording all around? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592685709 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
I am adding the following sentences to the end of the first paragraph in Section 2.2 "Data Types": In many situations, any integer type may be used. When the phrase "any integer type" is used in this document, it should be understood to mean **`byte`**, **`unsigned byte`**, **`short`**, **`unsigned short`**, **`int`**, **`unsigned int`**, **`int64`**, or **`unsigned int64`**. The wording seems a bit awkward, so please make suggestions and I will update the PR. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592684824 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
While I agree with @Dave-Allured on keeping PRs focused, I can also see @martinjuckes point that the "must be type integer" statements are linked to this PR. It is also a pretty minimal change, so I will update the "must be type integer" statements and push that change to this PR. The flag_values section might, I think, need some discussion so I will break that into a separate issue/PR. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592669704 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
> perhaps you'd be interested in contributing those to @ethanrd 's PR? No thank you. I believe that the original proposal is sufficient and elegant. I would prefer to leave further tuning to those contributors who are unsatisfied and have changes to offer. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592648776 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
I think readers of the document would probably want to understand "integer" to mean any integer type (any number of bytes, signed or unsigned). There is one reference to integer in regard to `flag_values` and `flag_meanings`, which must be of the same type as the data variable (whatever that is). All the others relate to discrete sampling geometries, in various roles. I think any integer type would serve those purposes as well. I agree with @martinjuckes that it would be sensible to replace "type integer" with "any integer type" (or something like that) everywhere, and it would be helpful in sect 2.2 to list explicitly all the datatypes which we mean by "any integer type" in the document. I too support the change being made. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592499989 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
As noted by @Dave-Allured, @ethanrd is attempting to remove restrictions on data types in general. This has broad support. The concern of @martinjuckes is that it would be possible to lift the current restriction on index variables, as is used for ragged arrays. However, this could also be addressed in a separate issue - under the current proposal, ragged arrays would still require integers, but we could change that in the future and in the meantime make use of the new types for other variables. Is this OK for the group? Barring objections I will merge this after the three week deadline, namely 24 March 2020. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592455682 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
# Summary 3 weeks have passed since any contribution and it seems we're pretty much at consensus. Supporting voices, including some from the Conventions Committee: - @ethanrd (Conventions Committee) - @erget - @Dave-Allured - @czender In my opinion one item should be addressed before this merge, however: [the issue raised by @martinjuckes](https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584521272) concerning ensuring internal consistency in document. @JimBiardCics agrees with this concern. @ethanrd could you respond to this please? -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-592064405 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
I support this, it will be very welcome. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584844080 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@ethanrd The CF documentation regarding constants of various types references the NUG, which gives some examples of CDL syntax, but which is by no means a rigorous description of how to represent each data type in an attribute. The **ncgen** man page has a discussion of this, but that sure seems like a multiply-removed place for a basic definition to reside. Perhaps we need to have the NUG or CF go into greater detail about this. (I think the NUG is probably the more appropriate place.) -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584737639 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@martinjuckes Notice that the examples where **`flag_values`** are given as 1b, 2b, etc, have the 'b' indicator because the variables in the examples are of type **`byte`**. It would be good to change an example to use a different integer type so that this is clearer. We should also make sure the language in the text is clear about type agreement between the attribute and the variable. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584731535 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
I wholeheartedly approve of this proposal. @martinjuckes makes an excellent point regarding "must have an integer type". -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584729984 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Hi Ethan, I support this .. but we need to check other references to "integer". There are 9 statements in the convention that something "must be type integer" (e.g. "In this representation, the file contains an index variable , which must be of type integer, and must have the sample dimension as its single dimension"). Perhaps all occurrences of this phrase can be replaced with "must have an integer type"? It might be worth stating the link between `flag_values`, which are expressed as, e.g. `1b, 2b, 4b`, and the new types. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584521272 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
@ethanrd I'm happy to moderate this if you like. I also support this proposal. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584514450 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Me two. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584438861 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.
Re: [CF-metadata] [cf-convention/cf-conventions] Add new integer types to CF (#243)
Meant to ping a few others involved in original discussions: @JimBiardCics @Dave-Allured @rsignell-usgs (didn't find user names for all from email discussion). I wasn't involved in earlier discussions and happy to be moderator. -- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/cf-convention/cf-conventions/issues/243#issuecomment-584394767 This list forwards relevant notifications from Github. It is distinct from cf-metad...@cgd.ucar.edu, although if you do nothing, a subscription to the UCAR list will result in a subscription to this list. To unsubscribe from this list only, send a message to cf-metadata-unsubscribe-requ...@listserv.llnl.gov.