ipp 2021.10 build fix

[casperdcl]

Describe your changes

• fix IPP includes (closes fix missing ipps.h by pinning ipp #1581)
  • ipps.h has been moved to ipp/ipps.h
  • ipp.h already (indirectly) includes the above anyway
• fix recipe/build.sh include dirs

Describe any testing you have performed

Please add any demo scripts to CIL-Demos/misc/

Link relevant issues

TODO in follow-ups:

• fix markdown errors fix borken whitespace #1589
• fix/update/unify scripts/*.sh tidy dev scripts #1590
• fix sphinx build on recent python versions update sphinx #1597
• windows CI CI matrices #1598
• macos CI CI matrices #1598
• GHA GPU runners CI matrices #1598
• conda build matrix... or just get on conda-forge CI matrices #1598

Checklist when you are ready to request a review

• I have performed a self-review of my code
• I have added docstrings in line with the guidance in the developer guide
• I have implemented unit tests that cover any new or modified functionality
• CHANGELOG.md has been updated with any functionality change
• Request review from all relevant developers
• Change pull request label to 'Waiting for review'

Contribution Notes

Please read and adhere to the developer guide and local patterns and conventions.

• The content of this Pull Request (the Contribution) is intentionally submitted for inclusion in CIL (the Work) under the terms and conditions of the Apache-2.0 License.
• I confirm that the contribution does not violate any intellectual property rights of third parties

[lauramurgatroyd]

Just had a quick look at the changes to see if this could solve our problems in other repos, and noticed some requests for reverting headers etc.

I will re-review fully when the actions are working 😄

[gfardell]

There's a lot of unrelated changes in this so it's hard to check the impact. Some builds are failing on GHA and Jenkins but not all so it would be good to know what's different between these.

Was the ipps.h being added to the C++ code the only fix here? I don't like including conda requirements of '>' as this is impossible to restrict in the future if there are issues. If we need to pin a particular version we should.

The build scripts are conveniences for creating environments are conveniences for us and I think having more options is better than less. We Can support the environment file for users? This is something that we need to sort out and simplify in the developer/contribution guide.

[gfardell]

The GHA and Jenkins issue is the high priority thing here. And we need to locally run conda build on linux and windows to check the fix works for each version combination.

[casperdcl]

unrelated changes

moved to separate PRs; please re-review :)

Some builds are failing on GHA and Jenkins

can you post links (URLs) to failing builds?

I don't like including conda requirements of '>' as this is impossible to restrict in the future if there are issues

Please explain "impossible to restrict in the future if there are issues".

Given that '>' is used a lot already before this PR, please do open a separate GH issue to discuss why you think they should all be removed.

[paskino]

It looks good to me. Although Intel may break again compatibility in the future, I don't think having >= is detrimental. At least we know that it won't work with version lower than 2021.10.

[paskino]

did you try on Darwin?

[casperdcl]

no and I'm also uncertain about whether it may need e.g. dylib. I think I'll add a GHA macos test in a follow-up. wdyt?

resolved

[gfardell]

The windows builds all worked. If Linux worked if happy to merge.

[gfardell]

You should add yourself here: https://github.com/TomographicImaging/CIL/blob/master/NOTICE.txt

[lauramurgatroyd]

Please update the Changelog https://github.com/TomographicImaging/CIL/blob/master/CHANGELOG.md