5 ms·
I use my Surface Pro nearly every day, so this sort of report makes me feel that: a. I am an oddball (probably true). b. Microsoft's marketing for the Surface
by bhauer 13y ago
I use my Surface Pro nearly every day, so this sort of report makes me feel that:
a. I am an oddball (probably true).
b. Microsoft's marketing for the Surface sucks (generally accepted as objectively true).
c. The Surface was a weak attempt (A subjective matter. I like mine, but there is lot of room for improvement. Ask me if you want more details).
The pundits have said their pieces, and I have written my own thoughts/rant about Microsoft's current unenviable situation [1]. But as a user of a Surface Pro, I'll offer up some of my random thoughts here as well.
Hardware-wise, the Surface Pro is one generation too early; it needs a Haswell CPU in order to provide user-friendly battery life and perhaps shed some weight. To meet their launch timing objectives, Microsoft had no choice but to use Ivy Bridge. As it was, the Pro was disturbingly late to market compared to the RT. The RT's flailing in the market muted the thunder of the Pro release. The market's response to the lackluster RT had already set in, casting a long shadow over the Pro. When the Pro finally arrived, less technically-savvy consumers were already disinterested. Even us tech folks were lukewarm despite the Pro being quite nice even if you use it as merely an ultrabook.
If Surface 2 is entirely Haswell (that is, both entry-level and Pro-level are low-TDP x86), and nothing else changes, I think the hardware will be just shy of superb. As it is, the Surface Pro packs ultrabook performance in a tablet-like form factor. Browsing the web, writing code, working with e-mails, Twitter, IM, IRC, the works--it's all as fast and smooth as the workstation I have at the office (a first-generation i7). It makes conventional tablets seem alarmingly slow and limited.
On the other hand, because the Pro is heavy and encumbered by a hungry CPU, it's not as versatile as conventional tablets.
I think it is now evident to most that Windows RT should be unshackled. Microsoft was apparently worried about cannibalism, but that worry is overshadowed by an utter dearth of sales. Microsoft should strongly shun their baked-in culture of capriciously angering consumers. They know arbitrary software limitations will upset consumers but they do it anyway. Stop that.
Personally, my computing lifestyle is not especially amenable to "computing device addition," and yet I half-reluctantly acquired a Surface Pro because I wanted portable computing capability and had never owned a high-end laptop. My tablets left me wanting more, but laptops never quite fit my desires (too large, too low-DPI). Plus, I am a bit of a Microsoft fanboy, I'll admit it.
Rather than adding another computing device, I would prefer a model where all my applications run on a compute server I own and operate (my workstation at home would suffice), and all of my devices connect to views of those applications. Adding a Surface Pro gives me yet another first-class computing device to manage. So now I have all of the following devices in my life wanting me to manage applications, manage data, manage configuration, manage presence, manage and synchronize state: my home workstation, my office workstation, my media PC, my phone, my Surface Pro, and my other tablets.
It's frustrating to add another device to the pile with no change to the underlying model of working with devices. I feel Microsoft is fairly uniquely positioned to present a unified computing model to users that Google and Apple seem reluctant to provide due to cloud stickiness (that is, a desire to solve application continuity and data continuity by moving applications to their public cloud rather than user-owned/managed applications). Microsoft is chasing the Google and Apple model and so far they are failing. As a Surface Pro owner, and Windows user on my workstation, I'd like to see Microsoft change course to provide a unified computing model focused on my devices working in a master-slave configuration over (tunneled) private networks.
Frankly, I want my Surface Pro to be a touch and keyboard-enabled, high-DPI, remote terminal to responsive (in the web-app sense of the word) application instances running on my workstation. That way, it's not a first-class computing device with all of the encumbering nags that entails. I want a subservient tablet that is just one view, one set of input and output devices, for my singular computing.
Finally, as I included in the linked rant, I believe a massive problem for Microsoft is the insanity of MSDN pricing. I strongly believe that their pricing model is actively hostile to developers, especially in this time of indie-oriented application development. They may have all sorts of special business programs that provide exceptions to the standard pricing--I honestly don't know--but the standard pricing stares you in the face and says, "Go away, Windows development is not for you." That sucks.
[1] http://tiamat.tsotech.com/microsoft http://tiamat.tsotech.com/microsoft
- pedalpete 13y agoI agree with what you said, but don't you feel that ubuntu is stepping in and providing the experience you want? One device that runs everything? I'd like to see a mobile device which can attach to an external CPU when I'm at my desktop, but aside from that, most arm chips are powerful enough for phone or tablet use.
- bhauer 13y agoThat is very close indeed. If a mobile device could augment its CPU capacity when docked, that would indeed be close to the model I seek assuming you always have your device with you (and you probably do). It may ultimately be the way things go, but if my say mattered, I'd steer in a slightly different direction. I would prefer that the computing device on my network be a hard-wired, always-on, high-bandwidth, high-compute, high-storage device (a workstation in our contemporary lingo). I want my personal applications to act as my agent when I am not present, possibly doing non-trivial computation and serving on my behalf. I want my data to be shared and federated with friends/family/colleagues on my terms, from my network or a network of trusted peers. I find it a little awkward to attempt to shoehorn all of that into a mobile device when I'd be just as happy--more happy in fact--for the compute power to be hard-wired and the mobile device to be just an input/output terminal to that compute power. As a final matter, I'd feel a little more confident with the physical security of my compute engine and data if it were in my house rather than on my person. But yes, kudos to Ubuntu for concertedly attacking our present-day malaise of multiple first-class devices!
- pedalpete 13y agoYour use case continues to intrigue me. Are you suggesting that a cloud solution where you control who has access to your data is not an option for you? A cloud that could run your 'agents'? As we go deeper and deeper into this paradigm, I'm realising how little we need more CPU. Even as a developer, I could be writing my code locally, but running it on a larger database and leveraging CPU which is not local. For some reason, I still want CPU.
- Tloewald 13y ago