`gutils` not available on Anaconda, and other questions

[VeckoTheGecko]

Hey there. I was very happy to find a FOSS package to interact with data output from Teledyne Webb Slocum Gliders.

Following the setup instructions in the readme, I found that gutils isn't on Conda forge. I assume this method of installation doesn't work (yet?)

It would be great to know:

• how feature complete this package is
• whether this package is in active development/what the current focusses are

If contributions are welcome, I'm happy to open some PRs tackling different areas (assuming my supervisor gives me the go-ahead).

[kwilcox]

Hi @VeckoTheGecko! There are a few groups of us using gutils operationally to process both real-time and delayed mode *.*bd files from slocum gliders into a set of standard in-memory and on-disk formats.

We (@jr3cermak) are actively working on a few things:

• Replacing the slocum binary tools (dbd2asc, etc.) with python versions to remove the compiled binaries as a dependency,
• Creating a new intermediary on-disk format (parquet with embedded metadata) over the existing netCDF on-disk format. The current netCDF file is lossy and makes some assumptions about the layout of the data that is better suited for a data-product.

You are correct that there is no PyPI or conda package. The gutils name was not available and we all know naming things is hard. You can install by cloning the repo and installing via pip install -e /path/to/clone for now and I'll work on packaging up the library as another name on pypi and conda-forge in #25.

We currently run gutils to process the real-time Slocum data via the Docker file that is built in CI on each push to the master branch.

[VeckoTheGecko]

Hi @kwilcox , thanks for the quick response and the overview of the state of things. I'll get more familiar with the package as I continue with my work and see how things go.

How is the state of the documentation for this package? I couldn't find any quickstart instructions on the documentation site (instead resorting to poking around the source code). I'm also a bit perplexed as to how the GH pages deployment is updated as I couldn't see any GH action workflows which did that (although I may have missed them).

If there are any parts of the documentation you feel need attention, let me know and I'm happy to take a look (I'm quite familiar with Sphinx).