I don't have any interest in gnome, however it is the best illustration of the problem.

Systemd is becoming a web of inter-dependent bits for which more and more stuff will depend on. Even at this early point in its adoption, running a non-systemd system, even when it isn't the default, involves way more effort than it should and excludes you from a list of packages that it shouldn't.

As systemd takes over more and more system functions, this is going to become more pronounced, until systemd and all it's non-init related functions become wedged in so tightly that it becomes a nightmare to use anything else.

It may become a very elegant solution and may result in a better functioning system with more mass-appeal, however it'll be the only practical choice anyone has.

Yikes, I don't know what happened in your past, but I hope someday you can come to terms with it.

Personally, as a programmer I've had to deal with various system and networking issues and am reasonably comfortable diving into that stuff, but I appreciate that being a sysadmin isn't just configuring switches and installing software, and if you threw me into that role I'd probably make a lovely mess of things.

It has been a fairly long slide.

I feel like at some point mass adoption became the big goal, and a lot of the things that really drew my to Linux in the first place have eroded away.

I always loved the variety of choices for just about everything and the general "if you don't like it, change it or make your own" mindset. The new thinking seems to be centered around agreeing on standards and rallying around a subset of options in order to make a more presentable solution to present to the masses. This is probably great for humanity and all, and even if we just end up with an open source version of Windows it's probably a good thing, but it's not what made me love Linux.

Systemd is just another step down that road.

That sounds very encouraging :)

Luckily for me I'm a desktop user, so wifi is a non-issue (and seems to be something that gets brought up a lot as being a sticking point, but then, the same was true of Linux for a long time).

Definitely glad to hear there is VirtualBox.

Neat, I actually figured FTB would be no problem and OpenRA would be the problem. Then again, for as much as I love minecraft and FTB, I'm in constant awe of how terribly implemented everything is. It wouldn't surprise me at all that they'd manage to make something in java platform dependant.

And yeah, when I discovered OpenRA, it basically became my life for like a month. Huge nostalgia trip. Got a small group of people I play with on pretty much a weekly basis now. When they get Tiberian Sun working I'm probably going to regress to university levels of binge gaming.

I think a lot of people tried OpenRA back when it first came out, said meh, and then forgot about it. When I was told to go play it, my reaction was basically "yeah I tried it, it wasn't that great", but It has really matured into a very playable game (albeit with it's fair share of quirks).

Anyway thanks for the response! :)

How practical is this for the desktop?

This is actually a serious question. I'm not overly familiar with BSD but have been thinking about giving it a shot on the desktop. I've been a Gentoo user for many years and am reasonably comfortable diving into stuff, so I don't anticipate user friendliness being a show stopper, more likely something I can currently do on Linux won't be available or will have poor support in BSD.

The main things I'm concerned with are Minecraft/FTB, mplayer, flash, VirtualBox, OpenRA, and jack/rakarrack. I'm open to alternatives as long as they actually work.

Flash I could probably live without, but much as I hate it, browsing the web sans-flash does still pose the occasional problem.
jack/rakarrack I could also probably live without. I currently use my desktop as a quick-n-dirty guitar amp/effects stack.
OpenRA is the thing I anticipate having the most problems with, but I play it somewhat obsessively so very much desired.

At some point I'll probably just try it and see, but I'm curious if any other slashdotter has gone this route and has anything interesting to say about it.

As a Gentoo user, I can kinda relate to this.

I've had to use the blacklist for the first time in a long damn while because (said in McBain voice) the USE flags, they do nothing!. Gnome indeed would seem the biggest offender. I don't even use gnome (openbox/xfce), but I have enough libraries and little utilities that were pulled in as dependencies for various things that I still had to spend hours dealing with stuff trying to pull in systemd.

So even on a source based distro where systemd isn't the default it takes effort to have it not end up on your system. I can only imagine it's worse and will get worse on Debian.

Identifying in the summary that dice owns slashdot would be apt. They used to do that back in the geeknet days when they'd post something that was even related to sourceforge or anything else they owned.

The tech "articles" posted on dice are generally shit. There is basically no reason for them to appear here except dice owns the place.

This is a two person operation?

Nerval's Lobster works for slashdot, and from his comment history, his entire job is to submit dice.com stories (this is not an exaggeration, as was pointed out to me, go look, it's literally nothing but dice.com posts).

However, he can't actually directly post the articles? So he is literally paid to _submit_ articles to slashdot, but can't directly post them himself? Isn't that a little silly?

Is it really?

It _used_ to be, but then at some point (seemingly unnoticed and unannounced) they stopped publishing the code. The old stuff is still there, and it's what was used to start up soylentnews, but I don't think slashdot in it's current form is open source any more (feel free to correct me/point me at source code if wrong).

I think it comes down to frequency increasing the risk (and thus adding liability).

Sure, I may cook for friends and family once in a while, which carries some risk, which society and the law have decided is acceptable without any special licensing or insurance. If I cook for tends or hundreds of people every single day, the risk increases greatly.

The same can be said about driving with passengers. Sure I might give someone a lift occasionally, and we've decided that this carries a low enough risk that standard personal drivers licensing and insurance covers that risk, but if I do it many times a day every day on a continuous basis, that risk goes up.

If this all sounds familiar, it's because the blackberry guys already went down this road.

They were smartphones for business. If you needed a smartphone for business, it was almost always going to be a blackberry. They sat on those laurels for years until all of a sudden those competitors that had been slowly nibbling away at their market share started taking bigger and bigger bites. Enough people said "fuck it, I want to use my android/iphone for work and ditch this blackberry shit" and they suddenly found themselves fighting for their life.

Microsoft has the desktop PC (especially at the office) and they have word processing, but mac and Linux are nibbling away at the desktop, and a lot of people are moving to other tools for word processing and spreadsheets. If they just sit their for the next 10 years saying "meh, we've still got all the workstation and office market share", I honestly think they'll be in the same situation RIM is in right now (passport aside, they are still kinda boned).

Google and Apple have been doing it right. It would take an awful lot to kill either right now because they've got substantial things going on in lots of diverse industries.

I gotta say, the C# version kinda reminds me of perl.

The Java version is annoyingly verbose, but I honestly think I prefer it (except I'd probably rename temp to something sane). Also google collections provides a lot of utilities to eliminate a lot of this kinda boilerplate (you could probably do something very similar to the C# example with iterables).

To me the strength of Java has always been the tool stack and libraries around it.

Most languages usually have something available with regards to dependency management, continuous integration, static analysis, code quality, unit testing/test coverage, etc. Java just seems to have multiple well supported and very polished versions of all that.

And one thing Java seems to have as an exclusive is a consistent coding convention. Yes people sometimes deviate from it, but the vast majority of Java code you'll see follows the same conventions. It seems like a trivial thing, but it makes a huge difference when working with 3rd party code and tools. Working on a c++ project that involved more then three 3rd party libraries and the code will be a complete mess (and writing adapters for everything is kinda impractical).

I don't know if it's necessarily irrational.

Microsoft did some very shady things in its past. They may not seem like the threat they used to be, if anything they have become the new IBM, but I think the old Microsoft which left a trail of corpses as it tore through the computer industry and scared the living shit out of everyone back in the day is still laying dormant.

I think it is still reasonable to view any apparently positive thing they do with a great deal of suspicion.