Become a fan of Slashdot on Facebook


Forgot your password?
Compare cell phone plans using Wirefly's innovative plan comparison tool ×

Comment Re: Does not replace mount (Score 1) 538

Firstly: I deliberately said supporters, not users.They are not the same thing. It very much is a choice, whether you want to call it political, that is up to you. I am a systemd user, I have it on my Fedora laptop, two Ubuntu servers and soon on one or two desktops, probably also Fedora or maybe Mint. That I use it does not make me a supporter or proponent. I use it for certain purposes. For other purposes on other machines, I do not use it and never would. Do not mix those two up. Most people who use UNIXoid systems professionally also have to use Windows machines at some point. That does not make them Microsoft supporters. Having a Dell monitor does not make me a Dell supporter.

Secondly: Your grand-standing about computer literacy is not exactly a sound counter-argument to what I wrote. Most system admins are incompetent (not as a judgement but literally lacking the technical competence to properly manage all aspects of running systems), I agree. To some extent I belong to them, I luckily do not run servers for a living, I can rely on others to do that for me. But people like me are not the benchmark here. People worth listening to do not complain about systemd in the belief that stuff never breaks. Quite to the contrary! Assuming breakage is the baseline from which you work up everything else. The issue I have with systemd is that it treats breakage as normal, working around it and trying to apply band-aids automagically. There are use cases for that, ie. the desktop (in most cases) and a certain portion of the server user base. But on my servers I do not want the system to route around damage unless I specifically set up fail safe mechanisms and fall backs myself. I want my machines to fail. So I know where trouble is brewing. And from my experience systemd has made it harder for me to debug issues. It is an opaque layer between me and the basic system functionality, and punching through has been more trouble and work than it should be. I like to compare it to NetworkManager: When that was introduced, it sucked. By now I love it on the desktop. On the server it is the first thing I purge from the system, because it has consistently caused problems and made things a lot more complex to set up and debug.

Thirdly: Do not confuse resistance to systemd with resistance to change. sysvinit has been showing its age for a long time. As have X11 and a few other parts of the ecosystem. I am happy to leave init scripts behind for a more sensible mechanism. But systemd is not the only kid in town. There are other initiatives to solve the issues we face, and most of them have taken a much less radical approach and far smaller scope. This is the way for me on the server side. This does not diminish the benefits some use cases draw from systemd right here and now. But it is not the holy grail some of its proponents make it out to be.

Comment Re: Does not replace mount (Score 1) 538

There is actually a second aspect that I missed: I know, systemd is more or less modular, not one monolithic binary etc. Still from what I have seen in recent distros, if you get systemd, you get the whole shebang in one go. On a server the common philosophy is not to have anything that you do not explicitely want on it. Whether this auto-mount and some other desktop-oriented systemd features are trivial to disable is beside the point: On a server they should not even be there without someone actively installing them. This is not really systemd's fault but the distros', since they should package it according to system scope or role; still I have seen this change in mindset coincide with the introduction of systemd. There does seem to be a systemic 'corruption' of the ecosystem going on where we all unlearn the lessons of the past.

Comment Re: Does not replace mount (Score 3, Insightful) 538

No. My argument is that on the server, many of systemd's components are either solutions in search of a problem or trying to solve real problems the wrong way. It started with things like accepting log corruption in return for faster performance. On most desktops, that is a justifiable trade-off. On a server, it is a minority use case. With a really tiny minority on whose servers I for one would not be willing to rely.

As I said: systemd is acceptable for desktops or other user-facing systems. Things that you expect to break anyhow because of user dumbness, hardware failure or spilled coffee, where reinstallation or replacement is cheaper and more practical than investing in reliability. There it brings you net benefits due to its design trade-offs. On a server I want to be able to retrace why something failed, not just have the system go back to some mostly functional default state or take a guess at what might be the best way to proceed. On a server I need to trust the system to do exactly what it is supposed to do, and to do it not just once or twice but every single time.

I know that it has become a bit of a trend to treat servers like just another machine and simply redeploy instead of fixing issues. For many business cases this may be the more economic solution. It is not the one I consider sensible and future-proof.

Comment Re: Does not replace mount (Score 5, Interesting) 538

[...] I get that the Unix way is to have lots of little utilities and services doing specific things, but it actually turned out to not be the best model. [...]

This is where most of the disagreement lies: It has not turned out to be the worse model. SystemD supporters are mostly concerned with the desktop (re. the user-does-stuff-with-thumbdrive rationale given for the mount manager). On the desktop, the SystemD approach makes a certain amount of sense. Though I have to strongly disagree on the notion that its implementation is anywhere near clean, hardened or tested, but given the timeframe and the ever-increasing scope that is to be expected and will likely improve over time - hell, all the alternatives it is trying to replace were shit when they came out. On the desktop SystemD is an improvement over the status quo, not the only possible venue, maybe not the ideal one, but it is here now and it mostly works.

But many Linux users care first and foremost about one use case, and one alone: the server. And on the server the UNIX way is the right way. The only sensible way, actually. On the server things like auto-mounting a thumbdrive so a user can diddle with it are not a thing. As are most other things SystemD is trying to do. Here SystemD is only one thing: a superfluous, possibly dangerous OS on top of the OS.

The balance between desktop and server has been turned over. I think it is great that the desktop is receiving more attention, don't get me wrong. But not at this cost.

Comment Re:Wayland bashing (Score 3, Interesting) 151

Bandwidth does not matter nowadays [...]

Bandwidth does matter, and in some ways more than ever, for two reasons:

  1. By now we have highly portable devices with enough computational capacity to do out in the field what would only two decades ago have required a few shelves full of Crays – but in the same time we have managed to unlearn how to efficiently use bandwidth. Many applications of computing or telecommunications still work in environments where bandwidth is a scarce resource. Not every service goes through fibre or LTE.
  2. We have so many devices out there that every modest reduction in bandwidth use per device runs up to a sizable gain.

Comment Re:So glad I don't work with her (Score 1) 290

I fully agree with you, except for one point:

d) A guide can, in a pinch, be automatically translated.

No. Hell no! English is my second language. When dealing with technical products I have made it my SOP to simply throw anything but the English manual directly into the litter. Even the typical half-Taiwanese pidgin found in there is infinitely more accessible than the same thing run through a crappy translation engine. I regularly cringe when I have to look something up in Microsoft's knowledge base and the website 'helpfully' gives me an automated translation.

Every language has its own approach to technology, using subtly different images or looking at processes from different perspectives. Culture plays a large role here. Properly translating a technical document is an art form.

Comment How can anyone? (Score 1) 290

Seriously, how can anyone work productively with such a system? And frankly, what do these people's mailboxes look like? I receive about 100 mails a day, sans spam. Of those, ~70 are automated reports and simply get filed away by filters, 15 are digests from forums or mailing lists that I glance over if I have the time and the inclination, and the remaining 15 I actually have to read. Of those, ten are one- or two-liners, mostly of an informative nature, and five might actually have more than one paragraph and require a close reading.

For me Young's way of communicating comes off primarily as a way to offload work onto her subordinates: Figures for speed of production and reception of communication in different media led me to that conclusion. Typing on a physical keyboard is well below 40 words/minute for most people; since she mentioned WhatsApp I assume she is communicating primarily via a mobile device, where typing rates are even lower. Speaking, on the other hand, can go up to 200 words per minute (especially in a language like Spanish, which I assume she uses, that lends itself well to rapid speech). The recipient, on the other hand, loses the very powerful ability to skim. You don't know if you can skip a portion of audio unless you heard it, in most cases. Reading works vastly differently.

And her very first quote sums up why this idea is significantly flawed for professional communication:

The practical benefit of saying an awful lot without having to turn your slightly inarticulate thoughts into an articulate email is obvious [...]

The article goes on to quote personal communication and the conveyance of emotion as supportive arguments. Yes, I do prefer voice or video when talking to my fiancé or my sister – because the person matters as much as or even more than the content, and time there is not a scarce resource divided between several activities that all have to come together for me to get any work done but something I willingly set aside. But frankly, virtually all of my personal digital communication has long moved to messengers like WhatsApp and Threema. I can count the truly personal mails I receive each year on one hand.

I have been gently nudging the institute where I work towards more formal and more permanent forms of communication, precisely because of the drawbacks of this 'slightly inarticulate' nature of speech.

And, in general, there is nothing that I hate as much as pre-recorded voice messages. I have fought my mobile provider tooth and nail to get their oh so helpful super-duper voice mail box ("It's digital! And there's an app for downloading your messages! And it's digital!") disabled. Call me, or text me. Or GTFO.

Comment Re:I am fascinated by all the genius in this world (Score 1) 118

Computers are complex, that is true. But firstly, quite a lot of this complexity is superfluous, the result of crappy engineering, feature creep and backwards compatibility to standards that should have been laid to rest decades ago. And secondly, yes, we are perfectly capable of building machines with several million parts of which a sizable portion moves which are highly intolerant of errors and still bringing down the defect rate to a manageable level. The IT world is the shanty town of the industrial sector, with few exceptions. Even the lucky players with more or less closed ecosystems like Apple make headlines every other week with blunders that in other industries would get people burnt at the stake. Yes, we make a conscious trade-off between crappiness and costs. But I am not in the least convinced we are drawing the line in the best place.

Comment Re: ink vs pixels is still a thing (Score 2) 117

[...] anything else is just a conversion from rgb.

Absolutely true. But only in a universe where "just a conversion" is defined as "that one step in the whole weeks-long process that will make or break the final product and decide whether you will be throwing away a truckload of stuff because QA/your boss/your customer says the colours do not match what they want".

Comment Re:That huge cost (Score 1) 1145

Figure in savings in the criminal justice and healthcare systems, and those 3 trillion are a real steal. From a real-world experiment:

Mothers with newborns stopped working because they wanted to stay at home longer with their babies, and teenagers worked less because they weren't under as much pressure to support their families, which resulted in more teenagers graduating. In addition, those who continued to work were given more opportunities to choose what type of work they did. Forget found that in the period that Mincome was administered, hospital visits dropped 8.5 percent, with fewer incidents of work-related injuries, and fewer emergency room visits from accidents and injuries.[9] Additionally, the period saw a reduction in rates of psychiatric hospitalization, and in the number of mental illness-related consultations with health professionals.


Savings in decreased hospitalisation and prison time alone should be substantial. UBI done right goes way beyond a monthly payment.

Comment Re:Cue the lawsuits. (Score 3, Informative) 443

And your answer is relevant how? MS has no God-given right to their users' telemetry data. They may politely ask their customers to volunteer such information (as they did with that Office improvement dialogue thingy, as far as I remember). But the way they are clawing for it shows they do not understand who they are and what they produce. There are many environments where it is undesirable or even illegal to simply throw around data so detailed as what I have seen in Windows telemetry to a third party over the internet.

And do not forget that those issues that need fixing are not God-given, either. They are defects. Some in Microsoft's products, some in someone else's products. It is not the customers' job to help fix them. If anything, the whole tech industry should be infinitely grateful for the incredible leniency it receives regarding product faults. Imagine the smoldering ruins in Redmond and Cupertino (and some other places) if Microsoft and Apple had to operate under the same regulatory regime as GM and Volkswagen. I know, we as consumers get to play with shiny toys that would otherwise be prohibitively expensive and incredibly boring. Still I think especially Microsoft urgently needs to be reminded that they do not rule over a lawless wasteland but operate under the same laws as everyone else.

Comment No, it did NOT! (Score 1) 180

This must have been submitted by someone who does not know how law and politics work in Germany. The federal government has forwarded a proposal to regulate fracking. Not to ban it. And it ultimately leaves the decision to the individual states, most of which are in dire need of more funding. Some of the state governments have already expressed support for fracking – this is the enabling piece of legislation they have been waiting for.

And in contrast to the praise some AC showered over our social democrats, I am delighted to be just as disappointed as I expected to be by the Traitors' Party (as the moniker for that particular party goes around here) for the sell-out they committed here. Again.

If only someone invented a machine that recovered electricity from broken political promises, all our current and future energy needs would be fulfilled by year's end.

Slashdot Top Deals

The ideal voice for radio may be defined as showing no substance, no sex, no owner, and a message of importance for every housewife. -- Harry V. Wade