Re: Objed maintenance

[Philip Kaludercic]

Adam Porter <adam@alphapapa.net> writes:

> Hi Philip,
>
> On 5/2/24 12:09, Philip Kaludercic wrote:
>> Adam Porter <adam@alphapapa.net> writes:
>> 
>>> On 5/2/24 01:02, Philip Kaludercic wrote:
>>>> Adam Porter <adam@alphapapa.net> writes:
>>>> 
>>>>> Hi Philip,
>>>>> 
>>>>>> Abstractly: My advice is my advice, it is inherently biased.
>>>>>> I take that position, because of my experience, which is why
>>>>>> I refuse to install packages with more than 1-~2 transitive
>>>>>> dependencies (I was recently once again shocked by "ement").
>>>>> Would you please explain what you mean here?
>>>> Just that I recently wanted to try out ement, but decided not to
>>>> do so when I saw the list of dependencies.
>>> I don't understand.  There are only a few, and they're all on GNU
>>> ELPA. Which ones do you object to, and why?
>> I am trying to give Amy an example of why to avoid dependencies,
>> because people like me don't look at
>> Requires: emacs-27.1, map-2.1, persist-0.5, plz-0.6, taxy-0.10,
>> taxy-magit-section-0.13, svg-lib-0.2.5, transient-0.3.7
>> and say it is only a few, if my limit is at 1-2 /transitive/
>> dependencies.  (In addition to that, you know that I have strong
>> opinions on Emacs package names and do not what to install a package
>> called "taxy", even though I know that this is probably just my
>> opinion).
>
> Thanks for explaining your view.
>
> Of note, one of those dependencies is Emacs itself, and two others are
> included with Emacs, i.e. map and transient.  That leaves 5 ELPA
> dependencies: persist is commonly used and works very well (and until
> multisession is improved a bit further, it seems like a better
> choice); plz is now a stable, reliable library that's becoming more
> widely used; svg-lib, by Nicolas Rougier, is very useful for what it
> does, but it could be made an optional dependency, I suppose.
>
> That leaves taxy and taxy-magit-section, which are fundamental to the
> room list feature and its dynamic, user-programmable grouping.  The
> name "taxy" is short for "taxonomy" (which might seem to imply a
> biological purpose), or "taxonomical-programmable-grouping" (which
> would be a bit long).
>
> If the name puts you off so much as to not even let it be installed as
> a dependency, and to explore what functionality it offers (which IMHO
> is unique and powerful), that is a bit disappointing to hear--if I
> removed the software with unusual or non-obvious names on my system,
> we wouldn't be having this conversation--but to each his own.

I refused to use grep at first, and it took me a bit to get over that
inhibition.  There have been multiple packages I decided not to install
after apt/dnf/... displayed the dependencies it would install.

This is not to defend my position, or a call to change my mind.  I just
want to illustrate that people like me exist and that I'd appreciate it
if these preferences aren't totally ignored.

BTW. I checked on my one non-development laptop, and all the recursive
dependencies I have installed were by Magit.

> --Adam

-- 
        Philip Kaludercic on peregrine