Version 3.9.0: Add short year formatter and fix year of day computation

[ChristopherRabotin]

What's Changed

• Relax UNIX test + release 3.8.6 by @ChristopherRabotin in Relax UNIX test + release 3.8.6 #266
• Introduce %y formatter by @gwbres in Introduce %y formatter #268
• Update tabled requirement from 0.14.0 to 0.15.0 by @dependabot in Update tabled requirement from 0.14.0 to 0.15.0 #269
• Update der requirement from 0.6.1 to 0.7.8 by @dependabot in Update der requirement from 0.6.1 to 0.7.8 #252
• [Maybe breaking change] Fix day of year computation by @ChristopherRabotin in [Maybe breaking change] Fix day of year computation #273
• docs(fix): fixed reversed description of duration magnitude by @ethanmsl in docs(fix): fixed reversed description of duration magnitude #271

New Contributors

• @ethanmsl made their first contribution in docs(fix): fixed reversed description of duration magnitude #271

Full Changelog: 3.8.6...3.9.0

---

This discussion was created from the release Version 3.9.0: Add short year formatter and fix year of day computation.

[gwbres]

Awesome Chris 🚀

side note: it takes me at least 20 mins to upgrade all my tools on new hifitime releases

Corrected Day Of Year (DOY)

RINEX filename conventions uses %yy or %YYYY, and {:03, %DOY} all the time.
The previous faulty DOY behavior did not help in supporting that.
We can now easily generate standardized filenames:

See the footnote for the remaining tweak I still have to do to achieve that.

New %yy Epoch formatter introduce

Will most likely simply some of our parsing procedure

Next step ?

The next step that would fix all the little tweaks we have to do, is to provide a method to decompose an Epoch in other than UTC timescale.
If we stick to RINEX filename generation example: RINEX files are most of the time expressed in one GNSS time system like GPST.
When formatting this file, we need to retrieve %YY %HH %MM of the first observation.

Doing this

let (yy, _, _, hh, mm, ss) = epoch.to_gregorian_utc();

returns all fields in UTC, while we want them in GPST in this example.
See rinex/src/epoch: fn epoch_decompose for the logic I need to implement as a work around

[ChristopherRabotin]

I'm glad this works.

For the next step, I think that's a good idea. I'll let you create an issue so we can keep track of it.

Now that there's the minor breaking change warning in the readme, I think that version 3.10 can integrate what we had initially planned for version 4.0.0, including changing the behavior of keeping the duration in the initialization time scale.