I mostly agree with David Pirotte here. For a program, it is not always relevant what it is implemented in except if it is meant to go with several other programs (say it is part of a framework). Now, for an all new program, mentioning what it is implemented in in the name is a reasonable choice. Changing the name of a program to do it is a bit different circumstance.

However, there is a major catch in this situation. Christian Brunello mentioned in the first email, if I read it correctly, that there are plans and intentions to extend it beyond just fdisk. Then renaming to something like [something] diskutils makes a lot of sense. If it is just a bunch of programs we get back to the argument earlier about including guile in the name or not. However, if those plans for extension include libraries and/or packages for use by other 3rd party programs, then the implementation possibly matters a lot. If this is part of the plan, then Guile Diskutils is a fairly reasonable name in my opinion. I really hope this is in the plan, by the way. There are a lot of interesting things that could lead to.


Freja Nordsiek

On December 13, 2017 1:01:41 AM GMT+01:00, David Pirotte <david@altosw.be> wrote:
Hi Christian,

 IMHO the programming language/compiler a utility is written with is an 
 implementation detail that should not manifest itself in the utility's 
 name.  In this case, I think "GNU Distutils" would be better.  

  "GNU Diskutils"

 1+
 David  

 I'm sorry but I do not agree. Guile is not an implementation detail in
 this case. It means that the package is based on Guile. It's like xterm
 (a terminal for x window), gnome-terminal (a terminal based on the GNOME
 framework) and so on.

It is, by definition, an implementation detail: whether your user know (or not ) in
what language the tool is implemented won't make _any_ difference on how they will
use it, neither will it change the output of their usage ...

Besides, having Guile in the name may even be a 'blocker': 'out there' most do still
fear scheme, in general, and Guile is no exception.  They will question, because of
prejudgment mostly, if it 'works well', why the hell is it not implemented in php,
python, go, java ... in a 'real language' ...

If it was a library for Guile, it would be different. As an example, I'm the author
and maintainer of GNU Foliot, not GNU Guile-Foliot, because it is an app (even
though users could extend it, but that is anther story...) ... users (most users)
are not interested to know i what language it's been written , they want to know if
is good, if it does the job, if it is well maintained ...

My 2c,
Do as you wish of course,
David