Want to read Slashdot from your mobile device? Point it at m.slashdot.org and keep reading!


Forgot your password?
Take advantage of Black Friday with 15% off sitewide with coupon code "BLACKFRIDAY" on Slashdot Deals (some exclusions apply)". ×

Comment Re:Odd choice (Score 1) 337

The keyboard and trackpad on the new Type Cover are actually quite nice. I certainly prefer the buckling spring keyboard on my desk for extended typing, but it easily matches or exceeds the majority of mobile keyboards I've used.

A bluetooth keyboard and mouse is one solution, but not necessarily an idea one. Most options right now (including the Apple keyboard and mouse) haven't been updated to use the new "Bluetooth Smart" low energy mode, so you're going to lower the effective battery life of your tablet or phone by using it. The devices themselves need power as well, which means keeping those devices charged or dealing with removable batteries.

And of course you have additional parts to carry around, unless you're using a keyboard case. For everything but the iPad Pro, that's a third party accessory (which isn't necessarily a problem). And given the lack of native iOS support for pointing devices, that will still be a separate device with third party support.

I actually think the relatively constrained storage is the biggest mistake in the iPad Pro. Cloud storage is fine for some applications, but local storage is still necessary for offline access (like keeping the kiddies occupied with movies on a road trip), application storage (they just upped the limit for an app to 4GB in the Apple Store) and production. That last point is of particular import as the iPad moves away from being a consumption first device. In the era of 4K video, 20+ megapixel cameras, and multichannel high resolution audio, 128GB doesn't go very far (and 32GB is a bad joke.)

I'm frankly surprised they didn't go at least 64GB on the base model and 256GB (if not more) on the top end. I'm less surprised they didn't include an SD card, though the inclusion would have been as useful for loading content as expanding storage (think digital photographers, who can easily fill a high capacity SD card in a single shoot).

But then for editing tasks the storage may not matter. Even if the A9X is surprisingly fast, the relatively meager 4GB will become an issue. (Amazing that's considered meager these days, but that is absolute entry level for notebook from Apple. Microsoft does offer a Surface with only 2GB, but that's on entry level Surface 3, which is $499 (or $599 with LTE).

Assuming that ISVs actually supports Apple's bid to expand the scope of iOS as a platform, I fully expect future iterations bump both the base and top end specs, but I have to wonder if Apple made a mistake not aiming higher in the first place. If people buying it are just buying it as a "big iPad", they will fail to fundamentally change the app ecosystem. On the other hand, the number of people who actually WANT just a big iPad may render that moot (and it will keep producers buying the more expensive Mac OS X based products).

Comment Re:He's got his talking points (Score 1) 478

Yeah, back in the day even running contemporary applications on different distributions was a challenge. Maybe you were running Slackware and couldn't run the latest MySQL because it was linked against glibc instead of libc5. Or maybe you were on Caldera and couldn't run a commercial app because it was compiled for Red Hat's controversial GCC fork (2.96). Or maybe you were just still stuck on a 486 and were trying to run a binary from Mandrake after they started compiling with an i686 target.

I can't bring myself to miss those days. :)

Comment Re:He's got his talking points (Score 1) 478

A stable kernel ABI is only one aspect of backwards compatibility. It's great that Alan Cox can run a static binary from 1992, but the majority of applications are dynamically linked and therefor depend on having the appropriate libraries available. To make matters worse, GCC didn't provide a forward compatible ABI until 3.4 (released in 2006), so for older applications not only do you need the appropriate version of the library, they would need to be compiled with a compatible version of the compiler suite.

Individual Linux distributions provide their own backwards compatibility guarantees, but only for a limited pool of libraries, and only across a limited number of releases. For example, Red Hat supports the following libraries across three releases: alsa-lib, krb5-lib, libtbbmalloc, mesa-libGL, elfutils-libelf, libgcc, libtbb, mesa-libGLU, glibc, libgfortran, libusb, motif, glibc-utils, libgomp, libvirt-client, pam, gtk2, libstdc++, libxml2, SDL, hesiod, libtbbmalloc_proxy, libxslt. So RHEL7 can run applications dynamically linked against those libraries for applications compiled on RHEL5, which was released in 2007.

The next tier of libraries is only supported within one major release. In other words, it if was compiled for RHEL 7.0 it will work in all subsequent RHEL 7 releases, but there is no guarantee it will work in RHEL 8.0.

And sure, you can still run apps written for Motif - or any other toolkit - but with the same caveats mentioned about. Any and all dynamically linked libraries need to be available, compiled with a compatible compiler. And applications may fail regardless. For example, a GNOME application which depended on bonobo for intra-application purposes (the way Evolution did) would simply fail when it couldn't contact an object broker.

TL;DR: Linux mostly sucks for backwards compatibility when it comes to desktop applications. Part of this is due to a fair amount of experimentation by the competing desktop environments. Part of it is due to a focus on source compatibility rather binary compatibility, which is fine. The silver lining is that the applications are largely open source, and most of them evolved along with the platform.

Comment Re:He's got his talking points (Score 4, Insightful) 478

How well does MacOS run applications from 1996 (like Civilization II)? Not at all. Apple was still on System 7 back then. Support for classic apps was dropped in 10.5 (2007) and for PowerPC apps in 10.7 (2011).

How well does Linux run applications from 1996? Largely a moot point, since there were relatively few compiled applications in the first place. But the Linux world was only just transitioning off a.out binaries and libc5. Anything written in C++ would be a non-started since we're talking GCC 2.5 or 2.7. Newer applications are potentially even worse, as they might depend on abandoned pieces of the nascent desktop frameworks (e.g. Bonobo, ORBit, DCOP, ARTS, ESound, etc).

That isn't to say it wouldn't be nice to have every older application work out of the box, but Microsoft has still maintained a laudable level of backwards compatibility in their products.

I've actually moved to Windows on my personal machine for the first time after running various flavors of Linux for twenty years. Why? Obviously not for backwards compatibility. Rather, the advent of web applications have largely rendered my desktop needs down to a web browser and a terminal. I can get that anywhere, but right now Windows offers competent HiDPI support, working trackpad gestures, and mature touchscreen support.

I still run Linux on my main work machine, but new releases continue to be plagued by a host of petty annoyances, like the secondary displays on my docking station not being recognized until I open a new window. Or corruption in the text rendering in my window title bars. Given tho problems I see in conventional hardware that is several years old, even on a days old version of Linux, there is no way I will be wasting my time trying to coerce it onto a brand new Surface Pro 4.

Comment Re:I remember a time... (Score 1) 478

Of course a mobile processor isn't going to match the speed of a desktop processor. The question is not whether it is the fastest machine, but whether it is "underpowered and worthless for 'work'".

Processor speed has long since stopped being the bottleneck for the majority of workloads. Likewise memory bandwidth, which is why trading off lower speeds for lower latency makes a lot of sense. (This was often the case even when memory bandwidth was drastically lower. Look at the failure of RDRAM to capture market despite offering 1600MB/s when SDRAM was still pushing only 1066MB/s. Price was a factor, as well, but in enterprise applications the drastically higher latency made the technology less suitable regardless of nominal bandwidth).

In the end, design always involves trade-offs. For someone graphics rendering, this would be an awful choice. They would probably be better off with a desktop, or at least a "mobile workstation" class machine that trade weight and portability to cram in a 45W processor and a higher end GPU. For my wife, who largely uses her machine for typical office applications or remote access to work, the portability and the ability to use the stylus for signing are far more important than raw performance. For myself, a lightweight machine is perfect for taking to meetings or while travelling. Our development resources live in racks in a data center, not in laptop bags.

Comment Re:I gotta ask while I'm here (Score 2) 478

They are certainly moving into that space. They actually acquired the pen and digitizer tech from N-Trig, rather than simply licensing it, and have been hinting that the capabilities will continue to evolve. (The latest firmware update included an update to the Pen driver that "adds support for future functionality".)

Comment Re:Maybe (Score 1) 478

Yeah, I looked at the SP3, since they still have them on display in the store, but the difference in price with the current discounts isn't nearly large enough to justify foregoing the new model. The display improvements are very apparent and the cooling improvements keep the SP4 comfortable to hold even when the fans kick in.

Comment Re:I remember a time... (Score 1) 478

The Surface Pro 4 configured with an i7 and 16GB is $1799. And you can get the Surface Book with an i7 and 16GB (plus a dGPU w/ 1GB GDDR5) for $2499. That's well within the ballpark people pay for a well-equipped MacBook.

For more modest budgets, the midrange models provide plenty of performance. I'm typing this on the $1299 model, which has "only" the i5, 8GB and 256GB SSD.

Comment Re:Maybe (Score 5, Informative) 478

That was the Surface Pro 3, not the new Surface Pro 4, and Microsoft largely addressed his issues in their firmware update last October: http://www.penny-arcade.com/news/post/2014/11/01/surface-3-update

The new model is significantly more powerful, with no noticeable parallax or lag, and a greatly improved display: http://gizmodo.com/the-surface-pro-4-has-the-most-accurate-tablet-display-1738801322

Comment Re:Too costly (Score 1) 152

Sure, the 1920x1080 resolution is so much better than 2736 x 1824 on the Pro 4 or 3000 x 2000 on the Book.

And carrying around a 5.68# laptop is so much easier than carrying a 1.73# Pro 4 (2.41# including the keyboard) or 3.34# Book.

That is a different class of product aimed at a different market segment. Some people are willing to pay extra for portability and versatility. Others just want the best bang for their buck. And neither is Right or Wrong.

Comment Some perspective... (Score 1) 822

Consider that in that same week an average of 948 fires were accidentally set by children playing with fire (mostly lighters and matches) resulting in at least 1 death, 16 injuries and $4-5 million in property damage.

In that same week an average of 50 children (mostly toddlers) will be backed over by a car.

In that same week 94 children will end up in an emergency room due to an accident in a pool or spa. About 7-8 will die. Most of them will be pre-schoolers.

In that same week over 5000 people will call poison control because of a child accidentally ingesting medication, and at least 1000 children under the age of 5 will land in the emergency room. Most weeks one will die.

In that same week over 200 children will accidentally inject household cleaners. Of those about 12 the poisoning is life-threatening or results in a long-term disability.

In that same week about 30 children will die as the result of neglect or abuse. In most case the parents will be directly at fault.

In that same week about 10 children will be murdered, 6 of them by one of their parents.

None of this detracts from the tragedy of accidental shootings, of course, or exonerates gun owners who do fail to properly secure their weapons. But the relative rarity of accidental shootings compared to other accidental injuries would seem to suggest that the majority of the 300+ million guns in the country /are/ properly secured.

If you're not careful, you're going to catch something.