Open tsalo opened 4 years ago
@jcolomb wrote:
I have just spent 6 hours to debug an application developed on windows but that should run on linux because of a capitalisation problem: my 2 cents: do not accept any capital letter.
I am in favor of route (2).
@tsalo would it then also cover entity's values?
or those should be fine as long as they are not coming from controlled (in BIDS) vocab (like _task-rest
)
I am asking because not sure how Windows would behave if there was e.g. _task-one
and _task-One
-- would it be different/the same name? I think we in DataLad and git in general had similar gotchas , e.g. https://www.hanselman.com/blog/git-is-casesensitive-and-your-filesystem-may-not-be-weird-folder-merging-on-windows .
I don't particularly want to restrict freeform values (i.e., the label
pattern for certain entities), but I can see why having case-insensitive unique values would be a good restriction. I personally want to adopt a single convention for suffixes, since some are currently capitalized, while others are not.
I should clarify, though, that the original proposal was @jbpoline's. I just copied it over from the old BIDS 2.0 Google Doc. Though I do support it, in the form of JB's option 2.
oh -- and that reminded me about
This came to mind recently, particularly thinking about https://github.com/bids-standard/bids-specification/issues/1862 and https://github.com/bids-standard/bids-specification/pull/1831.
My initial instinct before coming across this existing Issue was that capitalisation could apply exclusively to:
This would arguably then extend beyond suffices to directory names also; eg. sub-01/anat/
, sub-01/DWI/
, sub-01/func/
.
Not 100% sold on the idea, but considered it worthwhile listing.
oh boy stop pointing out at inconsistency here .. @yarikoptic s' baby for BIDS 2.0 lol
should we here just lower case all 4 cases?
should we here just lower case all 4 cases?
I think from this comment you may have erroneously associated my comment here with the ongoing discussion around DWI scanner-generated derivatives on BIDS 1.x. It is however intentionally associated with BIDS 2.0 discussion. What to do about capitalisation of specifically DWI scanner-generated derivatives in 1.x would be better placed at https://github.com/bids-standard/bids-specification/pull/1864.
See: https://groups.google.com/forum/?utm_medium=email&utm_source=footer#!msg/bids-discussion/yOYaLNTh-_A/rPLd3JpsAgAJ
I see 3 possible routes
I must say 2 and 3 seem at first sight preferable to 1 - but I would probably go for 2 if there are no strong rationale for 3 (which there may be)
Original authors: @jbpoline