Closed ghobona closed 2 years ago
@ghobona Why has there been no review on this PR? There are several issues with it, one being that it is very bad practice to commit auto-generated content (here even PDF files) to the main repository. There are also lots of /cache files committed in this PR which don't look like they belong here, but I may be wrong on that.
@ghobona +1 on really establishing a policy on avoiding to commit auto-generated content to Git repositories, especially large PDFs.
Specifications repositories containing ASCIIDoc document would be just a very few megabytes each if we follow this practice.
On the other hand if we commit auto-generated content and PDFs, they quickly blow up in hundreds of megabytes as the repository size include its full sets of revision (because Git is a distributed version control system).
When one contributes to a few of these specifications, this really starts eating up disk space. It also further discourages additional contributors when cloning a repositories takes up a lot of time and space.
Yes, I agree with @letmaik and @jerstlouis - it seems that there are a lot of files in this PR that shouldn't be in version control, although I appreciate that there was a need to get the job done quickly in time for submission and I'm grateful to @ghobona for doing the metanorma conversion.
When there is time, I think we should tidy up the repo - I think there are ways of removing files from version control (including the history) to reduce the repo size.
there are ways of removing files from version control (including the history) to reduce the repo size.
Yes, you can always amend the commits and force-push, but that should be used sparingly -- and the sooner it is done, the less likely it is to cause problems with others who have fetched the previous versions being replaced by something else.
Towards https://github.com/opengeospatial/CoverageJSON/issues/75