Re: package manager guix on Windows and OSX

[indieterminacy]

There are many bawdy jokes Id like to make concerning that citation. 

June 25, 2021 10:56 PM, jbranso@dismail.de wrote:

> That and their software is top notch! Linus called their developers
> "masturbating monkeys", because of their obsessive pursuit of security!
> hahaha. At every shutdown, the OpenBSD kernel is re-linked. It's the same
> kernel when you reboot it, but the binary is re-ordered. That's amazing!

Yes, they are highly capable and ambitious.

I picked up that information, having researched Hyperbola... based upon your 
prompt.

>> Additionally, while I understand that MIT is in many ways deficient
>> compared to GPL licenses, if not pernicious and counterproductive I do
>> empathise regarding why networking engineers may prefer having a licence
>> which permits encapsulation more readily.
> 
> Well, what is interesting, is that the HyperbolaBSD developers intend to
> rewrite 20% of the BSD kernel and license the whole project GPL. :)
> 

> My personal feeling is that GNU should adopt Org mode as their documentation
> standard. It's slightly easier to use than texinfo. Thought texinfo is
> pretty rad. :)

I love Latex, Context, I feel a bit weird for not having dabbled with Texinfo. 
Im not sure Texinfo is going to sway enough younger programmers (Im neither 
young nor old), I fear too many have been malconditioned into accepting delible 
communication techniques - Texinfo may no longer cut it.

I would consider Org mode to probably be the most acceptable default, though in 
many respects Skribilo could be more of a purer expression of a complete Guix 
approach. Are the aforementioned all different ways of dissuading people from 
considering Guix or documenting for it?

FYI, I have been wading into the Gemini protocol the last two months. Beyond 
its more noticable security and publishing advantages, I have been entranced by 
the terseness of its Gem .gmi (minimalist MarkDown) format. I consider it has 
crossover appeal (as least between documenting power users across OSes). FYI, 
the OpenBSD crowd seem to have the lead in the Gemini space - but this is 
presumable for the protocol rather than the markdown.

Since then I stopped annotating in Orgmode and will be building workflows to 
(eventually?) approximate a lot of Orgmode functionality. Obviously Orgmode is 
awesome but I wonder if it is too designed around individual workflows and 
procedures - where greater payoff comes from pooled workflows and procedures.

I had success/pleaseure converting from .gem to .org formats with this 
experimentation (concerning annotations for a Guix CWL blog post)
=> https://git.sr.ht/~indieterminacy/q1q20hqh_kq_oq_parsing_gem_zsh/tree

>From the tree you can see that it is feasible to output to *tex* or *html* 
>formats, using simple REGEX foo.

Additionally there is an unfinished attempt at exporting to (sic) Skribilo.

(You may want to ignore the potentially impenitrable annotations, which 
concerns a 'Recursive Modelling Language' Ive been working on - it would 
certainly confuse this topic)

I would be happy if Guix writing was done with minimal Gem markup but with 
heavy Lisp usage for interpretation, synthesis, collection and publishing of 
content. I had originally taken the approach that there should be Tex heavy 
markup first and then simplified transposing into other formats later. Now Im 
on the other end of the horseshoe.

I miss experimenting with regards to Tikz as a mechanism for generating 
graphics. I understand why other tools are used and ho programmers tend to 
seemingly think in terms of characters. It bothers me that I do not have 
beautiful graph displays representing my environment - to consider things from 
an impressionistic viewpoint and a contrast to text-editor/browser dualism. I 
suspect it isnt insurmountable and could allow visually minded people to not 
feel aggrieved by TUIscapes.


> What do you mean by:
> 
>> empathise regarding why networking engineers may prefer having a licence
>> which permits encapsulation more readily.

I mean: the MIT license allows you to operate in a commercial setting, whereby 
only the binaries are provided, without the   requirement to provide the source 
content. While I normally am against this, an OpenBSD networking head has 
explained to me how there would be usecases where this would be useful - if 
only to provide the commercial breathing space for niche projects. I probably 
should stop paraphrasing this person now.



Jonathan McHugh
indieterminacy@libre.brussels