- Re-order and modify page on fixels and dixels to emphasize that 'fixel' is the dominant term, and 'dixel' may not be relevant for many readers.
- Couple of spelling errors fixed.
- Add instructions on how to deal with 'too many open files' error.

This page is not "Crashes, errors and performance issues".
In addition, some entries that were duplicated between two pages were removed, and some pages from FAQs have been moved into this page.

I'm getting questions about this all the time, and the most convenient and accurate solution has always been to use this algorithm. If this for some very specific reason didn't work well, msmt_5tt didn't work either; and either the -fa parameter of dhollander could be tweaked for a good solution, or, worst case, manual selection of responses was the last resort.  Offering a quick, convenient and accurate solution has always been my intention here, with the aim of multi-tissue csd to be more accessible, and hence to encourage its adoption. It's also good for the perception of msmt-csd that it produces a good quality output; which does depend on good quality responses, obtained without hassle.

This section about setting the CPU architecture was included in the
standalone installer instructions due to the fact that we used to run
with -march=native - this was likely to cause issues for this type of
installation specifically. Now that the default is up to the compiler,
there is no longer any reason for this section to be where it is.

More to follow, this is a work in progress at this stage.  Just committing now to check the readthedocs output for layout.

- tck2connectome: Add references.
- FAQ: Add descriptions of some common issues encountered, including: tcknormalise requiring warpcorrect to be applied beforehand; 5ttgen fsl failing at the FIRST segmentation step.

…ect to the Tournier et al. 2013 publication, differences with respect to 'tax' algorithm

In use case where input image header does not contain a gradient table, and user instead relies on the -grad or -fslgrad option, dwi2mask would fail due to the absence of a gradient table.
Instead have dwi2mask explicitly import from grad.b, which is guaranteed to exist regardless of how the baseline gradient table was provided.
No need to import the gradient table into the image heder for the rest of the script, since tckgen is always run with an explicit -grad or -fslgrad option based on a permuted gradient table.
Supersedes #1151.

Closes #1065.

…fields

DICOM: fix bug introduced in previous commits

… not binary"

This reverts commit 068f4bc.

…tibility

to allow mercurial to access the Eigen repo.

…y_fix

population_template: python 3 & no numpy fix

SH peak finder: fix behaviour at pole for z-aligned delta

dwigradcheck: Fix dwi2maskcall

Note that the testing data, essentially noise.mif, is also used for the mrcrop tests; so it wasn't deleted.

Further changes to the interface, functionality and default behaviour will happen at some point in the future.

…w Year's

Also includes a change of mention of MRtrix to MRtrix3, in line with the phrase MRtrix3 developers, which was already in there/
Also includes removal of 2 .s, because they hindered clickability of the links in certain environments.

…arting with a fresh revision