[krylon]

OTOH, sharing code between different components developed under an umbrella project is not bad per se - if they require the functionality, re-implementing it from scratch for each component would not be a good idea, either. Duplicating and manually syncing the code also has its share of problems.

[JdeBP]

Is there any chance -- any chance at all -- that we cannot do the same dance of ineptly-phrased-objection followed by rebuttal-that-misses-the-point when it comes to this particular facet of the discussion?

Both sides of this are wrong. The people complaining about "monolithic" are groping in the dark for ideas such as low coupling. The people saying "but count the binaries!" are not addressing the questions of fully documented interfaces between said binaries, composability, and interoperability.

The "uselessd guy" explained this quite well, as have many other people. Can you please advance to not dancing this same old dance every time?

* https://news.ycombinator.com/item?id=12879771

* https://news.ycombinator.com/item?id=12652071

* http://uselessd.darknedgy.net/ProSystemdAntiSystemd/

[d33]

It's not only about whether you do it, it's also about HOW you do it. I once saw a slide from FOSDEM which listed lots of CLI tools that systemd replaced. Some of them seemed quite trivial and if they all depend on low-level systemd, I believe it really is bad design.

[wott]

But this tends towards Windows model, not a form of UNIX-like model. If I wanted a Windows-like, I'd use a Windows.