[Pkg-rust-maintainers] Bug#1077511: please compare against dfrs in long description

[NoisyCoil]

Package: dysk
Followup-For: Bug #1077511
X-Debbugs-Cc: noisycoil at tutanota.com

Hi Jonas,

I am having a bit of trouble understanding if your concerns are limited
to the description alone, or extend to the package itself, so I'll try
to cover both.


> I have difficulty understanding the purpose of this package.
> I do understand what is described in the long description, but what I am
> missing is how "df but better" is not already covered by the package
> dfrs.

> Personally I installed the package, tried it out and read its manpage,
> but still failed to find any benefit of dysk over dfrs, and a few
> drawbacks

I am not sure whether you are only talking about the description here,
or more widely about the usefulness of the package itself. I'll assume
it's the latter so I give you an answer if that's the case, and come back
to the description afterwards. In case it's the former you can just ignore
what I'm about to write, if you want.

As for the package itself, its purpose is to bring dysk to Debian.
Apparently there had been some interest in this, the first RFP (#1006367,
under the old name 'lfs') having been filed in 2022. Asides from the
RFP/ITPs, dysk is a moderately well-known and appreciated tool which I
believe to be a good addition to Debian.

As for the purpose of dysk, as I'm sure you've gathered, it is the very same
as that of df. And of dfrs, as you mentioned it. I had never heard of dfrs
before. I installed it and tried it and I don't particularly like it (no
offense to the developers of course, this is largely a matter of personal
taste, and I'm sure the same can be said of dysk) mostly because of features
you pointed me to yourself:

> expansion of device paths

I don't like that dfrs points me to devices which do not exist on my system.
E.g. it displays a '/dev/debian/root' which plainly does not exist, its actual
name and location being different,

> argually needless use of box drawing characters

I like it better with boxes. Again, personal taste. (As for the critically
filled partitions, I have none and can't check, will take your word for it. It
could be an interesting whishlist bug for upstream, especially if "critically"
is configurable). There are other things I don't like but let's not get into
further detail as it would be pointless.


Back to the description, and to the subject of this Bug Report, when you say

> please compare against dfrs in long description

and

> ... I recommend to improve the long
> description to compare not only against df but also dfrs.

I assume you mean feature-wise? I.e. that you are not asking to explicitly
mention dfrs in the description as is currently done with df? If so, there are
at least two measurable (i.e. not subject to personal preference) features that
dfrs doesn't have: 1. the ability to export data in the CSV and JSON formats
and 2. filters, in case you want to ignore some filesystems based on size, type,
etc. Both of these are documented in the manpage and in the --help (filters
could be better documented TBH). I believe these may be good additions to the
long description and I am more than willing to work on this.

Best,

NC