Make readers aware of broken-symlink-and-BIDS-validator-issues

[adswa]

Fixes #560 with an explicit FAQ on the BIDS validator and a more general note on troubles with broken symlinks for some tools.

[Remi-Gau]

Message to past Rémi: "RTFM or to be more precise just even read the --help section of a command!!"

Turns out that the command line bids validator has a couple of ways to "get around" this issue but I think it would be worth putting them in the datalad handbook.

Let me know if you think that some of this should be added in the sections you created.

---

All the following were done on a valid BIDS dataset that has been dataladed (dataladified? datalified?) but with all annexed files dropped.

the bad

The following will make the bids-validator scream

:~$ bids-validator gin/my_sweet_sweet_bids_data/rawdata/  

1: [ERR] Quick validation failed - the general folder structure does not resemble a BIDS dataset. Have you chosen the right folder (with "sub-*/" subfolders)? Check for structural/naming issues and presence of at least one subject. (code: 61 - QUICK_VALIDATION_FAILED)

the sort of good that doesn't work but tries to help

The following will "work"

:~$ bids-validator gin/my_sweet_sweet_bids_data/raw/  --ignoreSymlinks 

But you will get the extra message:

2: [ERR] This dataset contains remote files. If you would like to validate with remote files, use the --remoteFiles option. (code: 114 - INCOMPLETE_DATASET)

the uglies

Contrary to the previous message using --remoteFiles did not help in any of the combinations I tried.

This fails:

:~$ bids-validator gin/my_sweet_sweet_bids_data/raw/  --remoteFiles

1: [ERR] Quick validation failed - the general folder structure does not resemble a BIDS dataset. Have you chosen the right folder (with "sub-*/" subfolders)? Check for structural/naming issues and presence of at least one subject. (code: 61 - QUICK_VALIDATION_FAILED)

the good

The following actually worked:

:~$ bids-validator gin/my_sweet_sweet_bids_data/raw/  --ignoreNiftiHeaders --ignoreSymlinks

[adswa]

Ha, thanks for following up on this! Maybe also I should have RTFM ;-) That's very useful information. Feel free to add it in, else I'll do it later today after a streak of video calls. Thank you!

[Remi-Gau]

Won't have time to do the edits that quickly. 💨

I will let you do the honors.

[adswa]

[an overly polite clarification ensues to overcome the inevitable deficiencies of communicating online instead of face-to-face]

Didn't mean to rush you. I can also wait, if you want to do it. I wouldn't mind doing it myself, or waiting, so just stop me if you want to take a whack at it :)

[Remi-Gau]

[an overly polite clarification ensues to overcome the inevitable deficiencies of communicating online instead of face-to-face]

Didn't mean to rush you. I can also wait, if you want to do it. I wouldn't mind doing it myself, or waiting, so just stop me if you want to take a whack at it :)

hahaha I did not feel like you were being pushy 😄

Just wanted to let you know. And also my 2 hands are so full, I wish I were a Hindu deity to help me make progress on this many fronts.

So I just wanted to say, I would not be able to make a decent addition very soon.

[adswa]

cool, got it :) thx

[yarikoptic]

Re the bids validator not informative error message, see bids-standard/bids-validator#792

The problem though is a bit less benign -- bids-validator is used as a initial step in many bids apps. Some do not have an option yet to skip it (e.g PeerHerholz/BIDSonym#35) . And also it might be .json etc files which are also annexed so ignoring niftis would not be sufficient -- I am getting ORPHANED_SYMLINK on scans.tsv .

Filed: bids-standard/bids-validator#1050

[adswa]

Some short PR's have been open for a really long while now. Given no-one complained, I will merge them.