Post #895.

We are approaching 1000 replies in this thread...

...of which about 10 had any value. This thread is now just a feedback loop.

2 members found this post helpful.

you do not say the truth because the answer was in #896
and I still wait for your technical prove that you can reproduce all the functionality in the same quality, and it does not count that you leave out what you find pointless on functionality.

And still no clear or validated benefit to having it either. Go figure. In my opinion, I think the best possible answer is, it offers no benefits we don't already have, and people just want it to turn Slackware into a more mainstream distribution by wanting it, and just don't value simplicity nor correctness.

To my point of view, I understand why it's wanted, but I don't understand why it's needed.

1 members found this post helpful.

thats true, they say it right in the readme that windows support is not taken seriously or something similar. so unless you want to write non-portable apps for only 1% of the desktop market look elsewhere.

good luck to them finding developers that want to work on native gnome3 applications with there spare time.

Name a part of systemd? Let's see...

Logind? Try ConsoleKit2 or systemd-shim. Even then CK is still actively supported and equally used.
Networkd? Try dhcp, dhcpcd, and networkmanager and networkd lacks ipv6 support.
Parallel init service loading? OpenRC, Runit, s6, perp used in conjunction with sysvinit, GOD, etc. We even had daemontools doing parallel loading since 1999 used in conjunction with sysvinit.
Journald? Syslog-ng, rsyslog, and sysklogd and last it was known journald isn't 100% reliable due to a bug.
CGroups? Docker and libcgroups with init just fine.
Udev? eudev, extracted-udev, udev-classic, and even mdev with hal can work. You might not get evdev for X with mdev but it works just fine from tests I've seen and xfce has a plugin named genmon to do mounting, plus there's autofs without udisks or udisks2.
A building block set for Linux? Try the GNU operating system or BusyBox. We've had them collectively for over a decade or more.

And actually systemd isn't being duplicated. It is doing the duplication itself of long existing services, functions, etc.

So other than one big monolithic project, the benefits are negligible other than being new things.

So now that I've answered your question, answer mine, or did by chance your answer got answered by mine already? There is no benefit. There's nothing except more of the same under a cathedralistuic monolithic project claiming new, cool, and groundbreaking.

Dan Bernstein innovated parallel service loading long before Lennart even thought of systemd. Why isn't Dan ever praised for his work, but we're forced to worship Lennart like a golden god.

2 members found this post helpful.

Ha! I remember compiling Gnome when it first came out. Not only was it a PITA to build, there was the download over 56K modem taking hours or days in some cases.

I like it much better when you use a proper argumentation like you gave in this post!

Eric

I asked for a working sample, not for a repetition of your opinions.
So I accept that you are not able to bring a working system and that you can not prove anything what you are saying.

yes, and no one, not even the systemd devs denies that systemd duplicates and brings together on one place a lot of functionality,
and you are not able to do so, systemd devs are, its that easy.
that is why you and others like you spend your times in forums crying aground, telling not always the truths and stating opinions that are not provable.
and systemd devs implement one new functionality after the other and they have solutions people can test
and the world is using their solutions and they get even paid, while your non provable opinions are just .... noise in the internet.
Bring a working solution that people can test, you might even become famous and earn a lot of money, but if you are not able to do this ..., its that easy.

Edit:

I missed your edit
your post was better without this paranoid part.

the irony of gnome/gtk/glib is there are so many dependencies and tightly coupled components you could probably make a convincing case that it only makes practical sense to distribute the whole thing a giant binary blob.

its a shame because I find the opposite with a lot of the lower level gnu projects. they seem to understand why portability matters.

systembsd

Working examples already exist as viable existing stand-alone projects. Since when are those opinions? Those have been fact for years. Where have you been? If those aren't examples, then exactly how has GNU/Linux existed all these years before systemd came along? How did Slackware work for so long? A miracle from Bob Dobbs himself? Seriously... where have you been, and what have we been using as GNU/Linux for all this time?

I didn't know coding in free open source software was about making money and getting rich. I thought it was about contributing to promoting free open source software without the need for profiteering, monoplositic markets, and price gouging. If that's the case Linus should charge for using the kernel, Stallman for GNU, and every other developer get their $15 of fame.

If that's the case, then why hasn't Bill Gates open sourced Windows yet? He's famous and rich enough, so why isn't Windows open source?

1 members found this post helpful.

Coding in free open source is all about getting things done the way the developers want, sometimes with a political background (the creation of GNOME as a reaction to the Qt licenses), sometimes because developers aren't satisfied with existing solutions (Wayland, systemd, countless WMs, ...), sometimes just for fun or to support other hobbies. For many people open source is just the better development model, not a political statement. It is just the easiest way to get other people involved in your project.

It would be great if you pro-systemd advocates could consistently obey each others' rules. systembsd does not meet a4z's criteria for "a working system". At this time it is only an unintegrated GSOC project in a personal repository that hasn't been updated since mid-October.

Putting that aside for one moment, if one GSOC student can create an unofficial portable implementation of the systemd interfaces in one summer, why can't the whole systemd cabal of experienced professional rockstar coders do that in four and a half years?

Maybe the answer to that question involves what motivates them...

2 members found this post helpful.