Catch up on stories from the past week (and beyond) at the Slashdot story archive


Forgot your password?

Slashdot videos: Now with more Slashdot!

  • View

  • Discuss

  • Share

We've improved Slashdot's video section; now you can view our video interviews, product close-ups and site visits with all the usual Slashdot options to comment, share, etc. No more walled garden! It's a work in progress -- we hope you'll check it out (Learn more about the recent updates).


Comment: Re:Driver model (Score 1) 199

by hobarrera (#49186063) Attached to: The State of Linux Gaming In the SteamOS Era

So you ADMIT that Linux can't even KEEP ITS OWN DRIVERS FUNCTIONING for a lousy 5 years, just HALF the life cycle of Windows?

No, I never admited anything alike (I've no idea how you misinterpreted my response). I've also seen drivers that have been in kernel well over a decade. Generally, as long as someone is interested, there's no reason to remove drivers from the kernel.

Comment: Re:Driver model (Score 4, Informative) 199

by hobarrera (#49144403) Attached to: The State of Linux Gaming In the SteamOS Era

Then riddle me this...why does NOBODY, and I do mean nobody, not in FOSS nor in proprietary, support Torvalds driver model? After all if it was good there is absolutely NOTHING stopping them from adopting it, right? And what about BSD, why does it not follow the great Torvalds driver model?

"nobody" migth have been an exageration. Intel does. As do plenty of others (logitech, realtek come to mind, but there's a lot more). But I think naming Intel should prove that it's not just just one man.

Also, BSDs follows an extremely similar model: In the kernel tree. Most OpenBSD don't support binary blobs either, I've no idea about the rest.

The reason why is obvious, its because its shit that just won't scale. Hell basic math will show you that "let the kernel devs handle it" utterly collapses when the number of drivers reaches 5 figures because there simply is not enough kernel devs to keep up with all the hardware that is already out, much less the hundreds of new devices released this and every other quarter. It really VERY simple, in 1993, when the entire OS could fit on a single floppy? Then sure letting the kernel devs handle it made sense, they had MAYBE 30 drivers all told to deal with, now how many is there? 100,000? 200,000? Even if you pumped up the devs on coke and locked them in a room with NOTHING to but but deal with drivers they would have MAYBE 5 minutes every 3 years for each driver!

The devs just check that everything is the tree is ok, The drivers themselves are written by the hardware developers. When I had an issue with a Logitech mouse on PowerPC, it was a Logitech dev that submitted that patch to the linux kernel. That model does scale.

But if you truly believe what you are saying? Then put your money where your mouth is and take the Hairyfeet challenge which just FYI only requires Linux to run HALF, I repeat HALF as long as a Windows lifecycle. Surely your OS can do half of what Windows can, right? I look forward to seeing your video posted here and the complete vid on Dropbox. of course we'll never see it because if you actually attempt to take the challenge you'll see what I saw countless times and that is Torvalds.driver.model.doesn' and it all comes down to his driver model being made of fail.

The hairyfeet challenge is stupid. Is someone is stupid enough to invest money on something without knowing what it is or any previous research to see if it fits their purpose, they deserve what they get. Even if you know nothing about PCs, you can ask someone that does.

The problem is not related to the driver model at all (which is actually far better than the MSFT one), but to the fact that microsoft has a huge amount of money, has held a strong monopoly over a very long time, and there's a lot of money motivating manufacturers to just write windows drivers. It's money, there's nothing technical about that.

Comment: Re:It's totally superfluous (Score 4, Interesting) 164

by hobarrera (#48658357) Attached to: NetworkManager 1.0 Released After Ten Years Development

> It's hard to see how one could simplify Linux networking. It requires one ip addr command to set an IP address and one ip route command to set a default route. And on IPv6 even those are unnecessary, it's automatic.

And a dns, too. And the wireless network name. And the wireless network username+password.

And then, I have to do it all again in two minutes when you walk out of range. And then again when you get home. And then again at a cafe.

NM might not be the nicest of things, but it sure beats the hell out of running several commands every time I relocate myself/my laptop.

Comment: Re:what's wrong with ifconfig? (Score 3, Insightful) 164

by hobarrera (#48658345) Attached to: NetworkManager 1.0 Released After Ten Years Development

On Linux? To connect to WPA2 networks (including WPA2+802.1X). That's an everyday scenario for a pretty much every laptop user.

Sure, you can also do it via cli (with more tools than just those you mentioned), but, do you remember all the steps? Can you teach them to your mum? Can you automate it?

Comment: Standarization (Score 1) 928

by hobarrera (#48284743) Attached to: Ask Slashdot: Can You Say Something Nice About Systemd?

When jumping on a machine with a distro different to my own, I'd have to read for a while before understanding how to start/stop services. This was a pain if I needed to quickly help someone out on how to do something.

Now all distros run systemd, so it's just "systemd start nginx" on any gnu/linux distribution. Except gentoo, but I don't see a gentoo user asking me for help on how to do X. I also haven't come across clients with gentoo-based servers.

Standard service unit files help too. The work is done once, (generally upstream) instead of having to write service configuration files for every single distro (and keeping them up to date!). This may sound trivial, but the amount of effort reduced is immense!

Comment: Re:Repeat history (Score 1) 184

by hobarrera (#47972979) Attached to: KDE's UI To Bend Toward Simplicity

From TFA: Thomas prefers a layered feature exposure so that users can enjoy certain advanced features at a later stage after they get accustomed to the basic functionality of the application.

I assume that they'll keep the options around, just not mixed with the very basic options. A big issue with KDE right now, is that the settings windows of any application has half a dozen tabs, with dozens of options each. The very basic options and most common options should be grouped together, an advanced settings slightly on the side. Otherwise, it's just too intimidating.

Old programmers never die, they just branch to a new address.