Suppose for a moment that you are building a new processor for mobile devices.
The mobile device makers - Apple, Google, and Microsoft -- all have "App Stores". Side loading is possible to varying degrees, but in no case is it supported or a targeted business scenario.
These big 3 all provide their own SDKs. They specify the compilers, the libraries, etc.
Many of the posts in this thread talk about how critical it will be for the compilers to produce code well suited for this processor...
Arguably, due to the app development toolchain and software delivery monoculture attached to each of the mobile platforms, it is probably easier than ever to improve compilers and transparently update the apps being held in app-store catalogs to improve their performance for specific mobile processors.
It's not the wild west any more; with tighter constraints around the target environment, more specific optimizations become plausible.
Claim: the routing and security features on the edge devices your ISP provides as CPE are not sufficient
Claim: You want the ability to reset the shitty CPE your ISP gives you without losing LAN connectivity
Claim: Specific purpose devices are often better suited to their tasks than all-in-one devices
Solution: Treat your ISP-supplied CPE as a dumb device. Put a smarter device behind it that does routing, segmentation, translation, dhcp, etc, the way you want those things done.
Ideally, do PPPoE or something from the smarter device across the CPE, because CPE firmware is so often just terrible, but if not, double-NAT is often fine.
Critically, make your wifi APs a separate function both from your core home router and your edge device.
For a trivial amount of money, you can keep buying Ubiquiti APs and place them all over your property, as needed, and get an arbitrarily high level of speed and coverage. The configuration is completely painless, and this setup is completely independent of your edge device and edge connectivity.
Who needs weather satellites, GPS, and communication equipment anyway?
In the modern world, we all do. Which is why we should be more alarmed that all of these things are so very vulnerable to an increasingly long list of state-actors who don't like the West, and are so difficult to replace on short notice.
We've "gotten away with it" for a long time now. But any honest person knew those days were numbered.
At some point, we're going to have to really deal with the problem of space junk, and with the problem of space warfare as a prolific source of new space junk.
I think VW might contract the actual manufacturing to Chrysler.
Indeed. The VW Routan was a Chrysler Town and Country with some different skins on the inside and out. It was so much not a VW product that the VCDS system (the thing you can use to do vehicle diagnostics on any VW, Audi, Seat, or Skoda product since the early 90s) doesn't even talk to it.
In the German market, VW sells Vans of all different sizes. None of them are currently imported to the US; the Eurovan was the last rest-of-world van that was available in North America.
When we lived in town, it was great. Back then, my only serious gripe with the Odyssey is that if you are running a second set of wheels (e.g. for permanently mounted snow tires), and don't fit a 2nd set of expensive TPMS sensors to those wheels, the VSA (stability control) cannot be defeated via the console switch.
This is a problem because the VSA implementation sucks and is frankly unsafe when accelerating on surface transitions - for instance, when you are waiting on a gravel road and are about to pull onto a paved highway, the VSA system senses differing levels of wheel grip between the wheel on pavement and the wheel still on gravel, and cuts power, precisely when you need maximum power to quickly get to highway speed.
Last fall we moved to a rural area, and now poorly maintained roads (deep snow in the winters until I clear it, deep ruts whenever there are rains) has really shown me the shortcomings of the vehicle. My wife has gotten it stuck 4 times in our first winter.
The Odyssey needs 2 things to be superlative. Air suspension with adjustable ride height (it is a very low vehicle, for ease of entry/exit for small kids), and a proper AWD system.
My wife is now desperately wanting an AWD vehicle. But to get a proper AWD system (e.g. locking transfer case or at least a torsen differential), and the useful seating capacity of a minivan, you need to be looking at full-size truck based SUVs, like the Excursion or Sequoia.
I'm aware that the Sienna comes in an AWD version, but its particular AWD system and ride height doesn't inspire me that they will be foolproof enough to want to make the switch.
Sadly, my wife also refuses to drive a Mercedes G-wagen:)
As an aside, the Odyssey towing capacity isn't really sufficient. It's 3500lbs, and it requires upfitting the vehicle considerably with things that don't come factory - PS cooler, ATF cooler, hitch wiring, etc. (In addition to the actual hitch receiver).
When we were considering camping options, essentially nothing that had enough floor space for a family of 5 could be towed behind an Odyssey.
Hi there. Been an engineer at Microsoft since 2000. Have interviewed hundreds of people at all skill levels.
Why do you assume that wages at Microsoft aren't increasing?
I understand the compensation model, and how it has changed in my 14 years. The comp packages we are offering to college grads these days are astoundingly lucrative. Every few years in my career, there has been a big compensation realignment based on market realities. Everytime something at work upsets me enough that I start talking to other companies, their comp packages (especially with cost of living factored in) aren't able to match what I'm getting now from Microsoft.
Lately, high comp packages are required to compete with Google, Facebook, Amazon, etc, who all have plenty of money, and, for younger developers, are often seen as cooler places to work than old stodgy Microsoft.
I just see no evidence that H1-Bs are a mechanism for the company to save money. Dealing with HB-1 hassles involves a lot of overhead and expense that are not applicable to domestic employees.
As I said earlier, I have interviewed many, many folks, for many positions. The hire rate is not as high as we would like it to be. It never feels good to have to turn someone down, and it is a waste of time for everyone when an interview doesn't go well. But the bottom line is, we talk to many more people than we can feel confident about making an offer to. There are lots of STEM graduates, foreign and domestic. But not all of them are someone we could feel comfortable hiring. I'm sure you've known people in your CS class who could get good grades but who couldn't code... those people count as "qualified STEM applicants" to people that are pushing the "H1B is evil" rhetoric, but we all know that just because someone has a degree doesn't mean they are employable in that field... and certainly not by the top organizations in that field.
I've also seen no evidence that Microsoft has a preference for hiring H1-Bs, or that there is any compensation disparity for H1-Bs. I have seen evidence that H1-Bs cost the company money that domestic employees do not. For example, the company has special lawyers and paperwork people that deal with H1-B and other immigrant-labor related problems. That's a cost. When H1-B engineers are dealing with this stuff (which is frustratingly often), they aren't writing code or analyzing tests. That's a hit to their productivity, which ultimately, is another cost.
My dad was active in Mensa when I was younger and he was newly divorced. My dad is an unapologetic anti-democrat; I think Milton Friedman and Ronald Reagan may be above Jesus in his world view.
As near as I can tell, his interest in Mensa was for social networking with people that had a chance of understanding him. He's brilliant, loyal, fair, judgmental, and not at all sentimental. He has great difficulty expressing himself emotionally. Only certain people "get" him, and that's fine with him as long as there's at least one.... He's a hardcore INTJ.
He has no desire to run the world or to run other people's lives.
I haven't bothered to apply officially for Mensa, but I think I'd be borderline for admission. I'm also NOT a technocrat and ALSO not a liberal democrat.
There is nothing worthwhile in diversity in and of itself
This is the attitude that needs to stop. Diversity may not be a value in your pantheon, but it's not social engineering to want an inclusive society. It's wisdom.
Why does it need to stop?
A huge problem -- that few people seem to speak about -- is that using gender, nationality, or, most frustratingly -- race, as a measure of "diversity" is implicitly bigoted.
The diversity that people _claim_ to want is one of perspectives, life experiences, etc.
The things that are relatively easy to bucketize - gender identity, race, socio-economic status, etc.... these things in and of themselves are not a valuable source of "diversity"
The implicit bigotry in the "diversity" argument says that, if you hire more black people, you'll get much different ideas than what you already have. Why? Because all black people are different from the white people you already have.
I've never seen a more stark illustration of _racism_ then that.
The conjecture here is that if a population distribution doesn't' look the way certain people expect it to, then there is some upstream social problem that needs tinkering with.
That conjecture is only ever true or false on a case by case basis. The real problem that needs to stop is for people to believe this conjecture in the general case; the real problem is that people don't even agree or are not willing to state what their expectations are for the "ideal" population distribution, but, are still willing to cry foul and to assert that a problem exists.
I use facebook, multiple account email, and Exchange calendar from my phone multiple times a day. Its just that, I'm usually at home or work, and both have WiFi.
Contract phone plans are absurdly expensive, and, I've been running a pre-paid SIM for over 7 years. I don't want to go back to a situation where I pay a high monthly fee for a limited selection of phones with phone company malware on them...
I am getting everything I need out of this smart phone WITHOUT a gmail account.
I'll tell you how I've landed on a Windows Phone -- one that I paid for out of pocket, and using a plan that I also pay for out of pocket.
(I mention this only because I'm an MS employee, and I want to avoid the problem of someone claiming that I am astroturfing here)
For the last year or two, I had been using a used iPhone 3G. I had to jailbreak it so I could SIM unlock it.
I never bought any apps from any appstores. Free apps, yes. Paid apps - no.
The basic problem with the iPhone series is that apple simply obsoletes its hardware too quickly. Most of the apps in the apple app store couldn't install on my phone, because my phone couldn't be updated to the newest OS. The phone was unbearably slow when browsing desktop-class pages.
I feel like apple is a premium-price for a below-average experience.
Regarding Android - every android phone I've seen has been completely different from the others. If I pick up an android phone, it always takes me a while to adjust to the quirks of that particular handset's UI. I'm attracted to the ease of "owning" an android device, but, ultimately, I want a phone that just works. I rarely want to tinker with it.
Finally, Android bothers me because I don't use gmail and I don't trust google. The people I've talked to claim that it is difficult to really make the most of an Android phone without giving your life over to your google account.
So, Microsoft finally comes out with the Lumia 521 -- a no-contract phone that is natively built for Windows Mobile 8. I really like this phone. It has a fast browser, and the 1st party apps are quite good. It is like $120 from Wal-Mart. The camera and photostitching apps are good, and it comes with a built-in Nokia mapping/navigation program that has complete offline capability. This is important for me since I don't have a data plan and I am often in places with no data service anyhow. The Nokia HERE DRIVE and HERE MAPS applications are fantastic.
The windows mobile UI is great. More consistent then Android, and better information density than iPhone.
Microsoft has a long and interesting Linux/FOSS history.
I remember in the late 90s, Microsoft actually released a Front Page Server Extensions module for Apache on Linux, so people using FP could publish sites to Linux servers.
During the early 2000s, MS shipped a bunch of GPL'd stuff via the Interix/SFU product.
Currently, System Center (enterprise management tool) can also monitor and manage Linux machines along side windows (and Mac) machines.
As noted elsewhere, Microsoft has made Linux a 1st class scenario for Hyper-V on-premise and Azure hosted uses.
Microsoft has opened some its internal projects to the external community, with acceptable licenses, and Microsoft has also contributed to existing FOSS projects where it has made sense. Internally, "should we use existing FOSS" or "should we open source this?" are questions that are coming up now where in the past, they never did, and asking them would get you some funny looks.
In the future, you're going to see Microsoft doing a better job of meeting customers in mixed/heterogenous settings. We've got a new CEO that has provided this guidance to the entire company. The market changes have certainly become too large to ignore, but the bottom line is that we're adapting.
On the business side, getting some of a customer's business is better than getting none of their business.
As always, we partner with everybody and we compete against everybody. For example, I sit in a building where most of the developers here work on Microsoft's own ERP products, yet I worked on features that let Visual Studio talk to SAP.
The party you are referring to exists - it's called the libertarian party - and it is mostly (but not entirely) ex-Republicans who think responsible adults should be treated like responsible adults -- e.g. left alone until they hurt somebody.
There _should_ be more liberals and democrats joining the libertarian cause, because the LP is much better than the dems on key issues dems claim to care about: anti-war, pro-civil liberties, anti-racism in law enforcement (especially the drug war), anti-corporatism..
So, I cannot tell you why there aren't more democrats who break ranks and join the libertarians.
One reason for that, I suspect, is that I simply cannot relate to democrats or understand how they came to be democrats in the first place. There are plenty of intelligent people who are democrats, but I've never been able to figure out how any of them "tick".
In any case, there are principled libertarians -- and that's how they've traditionally billed themselves. Principled in the sense that they think government morally/ethically should not do certain things.
Then there are pragmatic libertarians -- folks who figure government is _ineffective_ or even malicious at doing certain things, and therefore shouldn't do them. An example would be Gary Johnson.
The bottom line is that, if America were actually incredibly hungry for a fiscally conservative, socially permissive party -- that party has existed for decades. It has been getting more popular lately, but it's still basically a rounding error in most elections.
1) access to data without vendor approval/involvement.
2) interop
3) no "remote killswitch" on software
4) no strange privacy leaks
I think these are all fine requirements.
But it's not clear to me why closed software couldn't meet them.
For instance, how does Windows + Office not meet these requirements?
1) the Office XML formats are documented, open, and have reader/writer libraries on non-Microsoft platforms
2) As a result of the consent decree, and much subsequent engineering and doc work, its quite easy to interop with windows and office.
3) So far as I know, there are editions of Windows and Office that require no internet connection at all, and certainly have no provision for remote-kill.
4) Microsoft is actually pretty good about shutting off telemetry, either on a per user basis, or with centralized management tools -- because enterprise customers want this capability too.
I have one of the older e-Ink, Wi-Fi only Kindles. Still has a physical keyboard, which I rarely use. My wife has the ad-supported one with no keyboard, and she doesn't seem to miss it.
The old e-ink kindle is great. I love it. They nailed the user scenario for me -- it is actually _better_ than a physical book. I can use it anywhere I'd use a physical book, I rarely worry about battery life. It's easier to read than a real book when laying on my side in bed.
I am completely uninterested in a color e-reader until it has the battery life and contrast of the e-ink display. And I don't want music, or apps, or multi-tasking, or anything else, because history tells me that adding them will detract from the basic experience of just reading a fucking book.
Here are the improvements I want out of a new kindle.
1) some kind of magical mystery charging. Maybe there is an inductive mat. Maybe its solar. Who knows. I said I _almost_ never worry about charging it. The next step would be I _NEVER_ worry about charging it -- and, I leave the Wi-Fi enabled and continue to not worry about it.
2) bendable/flexible - within limits. If they could make the thing so that it would reliably survive on the outside of soft-sided luggage; if I could put it in a pocket and not worry about it..that would be amazing. What's interesting about this is that the basic e-ink display technology can be flexible...
3) ability to easily -- and I mean easily -- send a book I've finished to my wife's device. Like, if me and her are in the same room, with both of our devices, I ought to be able to send a book on my device to her device. For free. Without any nonsense/bullshit.
Sorry, I fell into the trap of using "right and left", but these mean different things to different people.
When I say "right", I mean "laissez-faire", "capitalist", "individualist", "deregulated"
When many people say "right", they mean "authoritarian" and "nationalist"
That's not what I mean at all. I detest authoritarianism.
There are many places that are more authoritarian than the US (but we're working hard to catch up! (grumble))
There are no places that are more pro-individual liberty than the US. There are a few places which have better pro-business environments, and more economic freedom, but they tend to have fewer civil liberties than the US.
fwiw, my ideas about individual rights may also not be what yours are. I think "hate speech" should be legal, and like any other speech, should only be prosecuted when it is threatening or slanderous. And I think individuals ought to be able to keep machine guns without any government knowledge of oversight. Finally, I think homeschooling is a critical way to pre-empt the historical evils of government indoctrination, and so support homeschooling and parental rights to an essentially unlimited degree -- not because I think all parents are good, but because I think most governments are bad:)
I take individual rights _very seriously_, and so for me, a nation that offers a high degree of individual liberty has the following characteristics: few laws restricting the content of speech, few restrictions on private gun ownership, few restrictions on how children are educated outside of state control.
The US ranks quite well on all 3 of these individually, and taken together, far and away better than anywhere else.
To be fair, all of the countries that do worse than the US _also_ have more leftist governments.
The US has the most right leaning government there is. It also has the most racially, socially, culturally, and economically diverse population there is.
I wish I knew what to do about the problem of people not caring about Math and not excelling at it.
I think you and I probably disagree on many things, although I did see elsewhere that you complained about the growth of administrative/managerial positions within the school system. On that issue, I agree with you entirely.
It's frankly not clear how much the school can do for the kids _in general_ to improve outcomes for the broadest cross section of students, but one thing that has good empirical evidence is reducing class sizes.
That means hiring more teachers.
Frankly, given how much less it costs to hire a teacher than it does to hire an administrator, this should be a move everyone should get behind -- fiscally responsible, pro-teachers, pro-students.
The fact that this doesn't seem to be happening suggests that public education is sadly serving some other set of interests...
Nothing drives engineering like a tangible end goal that is big and easy to understand. Want to make batteries better? Have an engineering target that says what size car they have to fit in...
Tesla makes a better car than a "normal" car. It is faster than almost every sport sedan. It has excellent styling, excellent practicality (seating for 7!), very good handling.
Even if it weren't a pure EV, it would be an interesting contender.
But hey! It's also an EV. It is completely quiet. It doesn't need oil changes, or coolant flushes, or gasoline. You don't have to use the brake pedal anywhere near as much.
Tesla was the first company to make an electric car in my lifetime that wasn't a total piece of shit garbage joke. Not only did they instead make a _credible_ car, they made a car that kicks the ass of nearly other car. Their car is quieter than a Mercedes, out accelerates most BMWs, and handles better any other American sedan.
Why in the world should they stop making cars? The car they've made is better than most of the competition, AND is a game changer.
And, they're selling more of them than they can build. The model S is around 2 month lead time. The Model X already has 6 months of deposits.
I don't want Tesla to make batteries. I want everyone else to figure out how to make Telsas.
Based on my experience of using SketchUp to design a part that I ultimately had ShapeWays print for me, I would say that there's plenty of room for innovation in the 3d printing toolchain.
One thing to note is that modelers like SketchUp assume a uniform internal volume. And so you use tools like slic3r to prepare the model for printing.
In effect, the 3d geometry is the source code. Slic3r would be a "compiler" -- which translates idealized geometry to something more ready for printing -- e.g. by driving an extrusion head in 2d layers (or slices)
I think there is a lot of room left for innovation in the transformation pipeline, which can backfeed requirements into the modelling stage, and which could feed requirements into the print hardware to expose capabilities information upstream as well.
For instance, how can I optimally slice a model so that I use as little print media as possible, while still retaining enough strength? Slicers and printers differ on how they fill interior volumes. Filling volumes with uniform hollow geometry speeds up printing and lowers cost. But is uniform geometry the best choice? When and where are other choices better? And should I slice the model a certain way based on the characteristics of the printer?
Suppose my model let me capture information about stressed joints, where I would want to change the how the interior volume was printed, based both on my design and on the capabilities of my printer?
These features may actually exist, but in the free tools I used, there was no indication that this was the case.
Autodesk could certainly do a lot of good things here. Printing objects that are faster to print, cheaper to print, and stronger when used.. that would be impossible for the industry to ignore.
Obviously, because I think she's neither foolish nor hypocritical.
There's a class of people who respond to Ayn Rand with ad hominem. Which is funny -- She wrote a lot of pages -- more than I care to read in one sitting. In all that, somewhere, you'd think there's be fertile soil for a response more intellectually stimulating than, "she's a crank".
Fault her for whatever reasons you've faulted her, but to me, nobody has more constancy and conviction in their writing in favor of doing the right things for the right reasons. The importance of principle is central to everything she ever wrote.
The Mozilla conversation is about principle vs. pragmatism, and I think her quotes on the topic are highly relevant.
The Hurd isn't a viable alternative because it isn't needed.
Stallman had a vision of a completely free as in speech computer system. When he started, that meant, OS, tools, and application software.
It was a radical strawman against the beginnings of an industry of for-profit software with intellectual property laws.
It turns out that Stallman and his friends created the programmable editor, the compiler suite, the tool chain, the user-space unix tooling..
and them some Finnish guy and his friends came along and made the OS kernel.
The point is that now, not only is there a free OS and development tool chain -- more successful than Stallman could have ever managed -- there is an entire philosophy around free-as-in-speech software.
Stallman has been more influential on how we think about an use computer software than arguably just about anyone. I would at least put him in the same room as a Woz or a Bill Gates.
The market share of Hurd is the wrong metric. The fact that my company -- Microsoft -- is releasing more and more of our stuff as free-as-in-speech software -- that's the metric.
Let's objectively look at what Stallman started.
Let's use this metric: how many Fortune 100 companies have capitulated to _your_ philosophical demands?
Oh look. Here's a whole _page_ of Ayn Rand quotes about compromise
In any compromise between good and evil, it is only evil that can profit
or...
Contrary to the fanatical belief of its advocates, compromise [on basic principles] does not satisfy, but dissatisfies everybody; it does not lead to general fulfillment, but to general frustration; those who try to be all things to all men, end up by not being anything to anyone. And more: the partial victory of an unjust claim, encourages the claimant to try further; the partial defeat of a just claim, discourages and paralyzes the victim.
Many folks have a go at the idea that they can somehow tame evil or compromise with it without being tainted too much. I'm not sure this has ever really worked out.
There is a lot to like about the Richard Stallmans of the world. They are clear about the what and the why, and they stick to their guns.
Slashdot Mirror
Suppose for a moment that you are building a new processor for mobile devices.
The mobile device makers - Apple, Google, and Microsoft -- all have "App Stores". Side loading is possible to varying degrees, but in no case is it supported or a targeted business scenario.
These big 3 all provide their own SDKs. They specify the compilers, the libraries, etc.
Many of the posts in this thread talk about how critical it will be for the compilers to produce code well suited for this processor...
Arguably, due to the app development toolchain and software delivery monoculture attached to each of the mobile platforms, it is probably easier than ever to improve compilers and transparently update the apps being held in app-store catalogs to improve their performance for specific mobile processors.
It's not the wild west any more; with tighter constraints around the target environment, more specific optimizations become plausible.
Claim: the routing and security features on the edge devices your ISP provides as CPE are not sufficient
Claim: You want the ability to reset the shitty CPE your ISP gives you without losing LAN connectivity
Claim: Specific purpose devices are often better suited to their tasks than all-in-one devices
Solution: Treat your ISP-supplied CPE as a dumb device. Put a smarter device behind it that does routing, segmentation, translation, dhcp, etc, the way you want those things done.
Ideally, do PPPoE or something from the smarter device across the CPE, because CPE firmware is so often just terrible, but if not, double-NAT is often fine.
Critically, make your wifi APs a separate function both from your core home router and your edge device.
For a trivial amount of money, you can keep buying Ubiquiti APs and place them all over your property, as needed, and get an arbitrarily high level of speed and coverage. The configuration is completely painless, and this setup is completely independent of your edge device and edge connectivity.
Your James Bond film sounds very intriguing and I would like to subscribe to your... movie.
In the modern world, we all do. Which is why we should be more alarmed that all of these things are so very vulnerable to an increasingly long list of state-actors who don't like the West, and are so difficult to replace on short notice.
We've "gotten away with it" for a long time now. But any honest person knew those days were numbered.
At some point, we're going to have to really deal with the problem of space junk, and with the problem of space warfare as a prolific source of new space junk.
Indeed. The VW Routan was a Chrysler Town and Country with some different skins on the inside and out. It was so much not a VW product that the VCDS system (the thing you can use to do vehicle diagnostics on any VW, Audi, Seat, or Skoda product since the early 90s) doesn't even talk to it.
In the German market, VW sells Vans of all different sizes. None of them are currently imported to the US; the Eurovan was the last rest-of-world van that was available in North America.
We have 3 kids in car seats, and an Odyssey.
When we lived in town, it was great. Back then, my only serious gripe with the Odyssey is that if you are running a second set of wheels (e.g. for permanently mounted snow tires), and don't fit a 2nd set of expensive TPMS sensors to those wheels, the VSA (stability control) cannot be defeated via the console switch.
This is a problem because the VSA implementation sucks and is frankly unsafe when accelerating on surface transitions - for instance, when you are waiting on a gravel road and are about to pull onto a paved highway, the VSA system senses differing levels of wheel grip between the wheel on pavement and the wheel still on gravel, and cuts power, precisely when you need maximum power to quickly get to highway speed.
Last fall we moved to a rural area, and now poorly maintained roads (deep snow in the winters until I clear it, deep ruts whenever there are rains) has really shown me the shortcomings of the vehicle. My wife has gotten it stuck 4 times in our first winter.
The Odyssey needs 2 things to be superlative. Air suspension with adjustable ride height (it is a very low vehicle, for ease of entry/exit for small kids), and a proper AWD system.
My wife is now desperately wanting an AWD vehicle. But to get a proper AWD system (e.g. locking transfer case or at least a torsen differential), and the useful seating capacity of a minivan, you need to be looking at full-size truck based SUVs, like the Excursion or Sequoia.
I'm aware that the Sienna comes in an AWD version, but its particular AWD system and ride height doesn't inspire me that they will be foolproof enough to want to make the switch.
Sadly, my wife also refuses to drive a Mercedes G-wagen :)
As an aside, the Odyssey towing capacity isn't really sufficient. It's 3500lbs, and it requires upfitting the vehicle considerably with things that don't come factory - PS cooler, ATF cooler, hitch wiring, etc. (In addition to the actual hitch receiver).
When we were considering camping options, essentially nothing that had enough floor space for a family of 5 could be towed behind an Odyssey.
Hi there. Been an engineer at Microsoft since 2000. Have interviewed hundreds of people at all skill levels.
Why do you assume that wages at Microsoft aren't increasing?
I understand the compensation model, and how it has changed in my 14 years. The comp packages we are offering to college grads these days are astoundingly lucrative. Every few years in my career, there has been a big compensation realignment based on market realities. Everytime something at work upsets me enough that I start talking to other companies, their comp packages (especially with cost of living factored in) aren't able to match what I'm getting now from Microsoft.
Lately, high comp packages are required to compete with Google, Facebook, Amazon, etc, who all have plenty of money, and, for younger developers, are often seen as cooler places to work than old stodgy Microsoft.
I just see no evidence that H1-Bs are a mechanism for the company to save money. Dealing with HB-1 hassles involves a lot of overhead and expense that are not applicable to domestic employees.
As I said earlier, I have interviewed many, many folks, for many positions. The hire rate is not as high as we would like it to be. It never feels good to have to turn someone down, and it is a waste of time for everyone when an interview doesn't go well. But the bottom line is, we talk to many more people than we can feel confident about making an offer to. There are lots of STEM graduates, foreign and domestic. But not all of them are someone we could feel comfortable hiring. I'm sure you've known people in your CS class who could get good grades but who couldn't code... those people count as "qualified STEM applicants" to people that are pushing the "H1B is evil" rhetoric, but we all know that just because someone has a degree doesn't mean they are employable in that field... and certainly not by the top organizations in that field.
I've also seen no evidence that Microsoft has a preference for hiring H1-Bs, or that there is any compensation disparity for H1-Bs. I have seen evidence that H1-Bs cost the company money that domestic employees do not. For example, the company has special lawyers and paperwork people that deal with H1-B and other immigrant-labor related problems. That's a cost. When H1-B engineers are dealing with this stuff (which is frustratingly often), they aren't writing code or analyzing tests. That's a hit to their productivity, which ultimately, is another cost.
It is certainly racist.
You are using race as the determining factor to make a presumption about an individual human.
What other useful definition of racism could there be?
Counter-anecdote.
My dad was active in Mensa when I was younger and he was newly divorced. My dad is an unapologetic anti-democrat; I think Milton Friedman and Ronald Reagan may be above Jesus in his world view.
As near as I can tell, his interest in Mensa was for social networking with people that had a chance of understanding him. He's brilliant, loyal, fair, judgmental, and not at all sentimental. He has great difficulty expressing himself emotionally. Only certain people "get" him, and that's fine with him as long as there's at least one.... He's a hardcore INTJ.
He has no desire to run the world or to run other people's lives.
I haven't bothered to apply officially for Mensa, but I think I'd be borderline for admission. I'm also NOT a technocrat and ALSO not a liberal democrat.
Why does it need to stop?
A huge problem -- that few people seem to speak about -- is that using gender, nationality, or, most frustratingly -- race, as a measure of "diversity" is implicitly bigoted.
The diversity that people _claim_ to want is one of perspectives, life experiences, etc.
The things that are relatively easy to bucketize - gender identity, race, socio-economic status, etc.... these things in and of themselves are not a valuable source of "diversity"
The implicit bigotry in the "diversity" argument says that, if you hire more black people, you'll get much different ideas than what you already have. Why? Because all black people are different from the white people you already have.
I've never seen a more stark illustration of _racism_ then that.
The conjecture here is that if a population distribution doesn't' look the way certain people expect it to, then there is some upstream social problem that needs tinkering with.
That conjecture is only ever true or false on a case by case basis. The real problem that needs to stop is for people to believe this conjecture in the general case; the real problem is that people don't even agree or are not willing to state what their expectations are for the "ideal" population distribution, but, are still willing to cry foul and to assert that a problem exists.
FIRST POST
(however, the apparent local time when you see this post may differ based on the apparently non-constant nature of c )
Well, not exactly a feature phone.
I use facebook, multiple account email, and Exchange calendar from my phone multiple times a day. Its just that, I'm usually at home or work, and both have WiFi.
Contract phone plans are absurdly expensive, and, I've been running a pre-paid SIM for over 7 years. I don't want to go back to a situation where I pay a high monthly fee for a limited selection of phones with phone company malware on them...
I am getting everything I need out of this smart phone WITHOUT a gmail account.
I'll tell you how I've landed on a Windows Phone -- one that I paid for out of pocket, and using a plan that I also pay for out of pocket.
(I mention this only because I'm an MS employee, and I want to avoid the problem of someone claiming that I am astroturfing here)
For the last year or two, I had been using a used iPhone 3G. I had to jailbreak it so I could SIM unlock it.
I never bought any apps from any appstores. Free apps, yes. Paid apps - no.
The basic problem with the iPhone series is that apple simply obsoletes its hardware too quickly. Most of the apps in the apple app store couldn't install on my phone, because my phone couldn't be updated to the newest OS. The phone was unbearably slow when browsing desktop-class pages.
I feel like apple is a premium-price for a below-average experience.
Regarding Android - every android phone I've seen has been completely different from the others. If I pick up an android phone, it always takes me a while to adjust to the quirks of that particular handset's UI. I'm attracted to the ease of "owning" an android device, but, ultimately, I want a phone that just works. I rarely want to tinker with it.
Finally, Android bothers me because I don't use gmail and I don't trust google. The people I've talked to claim that it is difficult to really make the most of an Android phone without giving your life over to your google account.
So, Microsoft finally comes out with the Lumia 521 -- a no-contract phone that is natively built for Windows Mobile 8. I really like this phone. It has a fast browser, and the 1st party apps are quite good. It is like $120 from Wal-Mart. The camera and photostitching apps are good, and it comes with a built-in Nokia mapping/navigation program that has complete offline capability. This is important for me since I don't have a data plan and I am often in places with no data service anyhow. The Nokia HERE DRIVE and HERE MAPS applications are fantastic.
The windows mobile UI is great. More consistent then Android, and better information density than iPhone.
Microsoft has a long and interesting Linux/FOSS history.
I remember in the late 90s, Microsoft actually released a Front Page Server Extensions module for Apache on Linux, so people using FP could publish sites to Linux servers.
During the early 2000s, MS shipped a bunch of GPL'd stuff via the Interix/SFU product.
Currently, System Center (enterprise management tool) can also monitor and manage Linux machines along side windows (and Mac) machines.
As noted elsewhere, Microsoft has made Linux a 1st class scenario for Hyper-V on-premise and Azure hosted uses.
Microsoft has opened some its internal projects to the external community, with acceptable licenses, and Microsoft has also contributed to existing FOSS projects where it has made sense. Internally, "should we use existing FOSS" or "should we open source this?" are questions that are coming up now where in the past, they never did, and asking them would get you some funny looks.
In the future, you're going to see Microsoft doing a better job of meeting customers in mixed/heterogenous settings. We've got a new CEO that has provided this guidance to the entire company. The market changes have certainly become too large to ignore, but the bottom line is that we're adapting.
On the business side, getting some of a customer's business is better than getting none of their business.
As always, we partner with everybody and we compete against everybody. For example, I sit in a building where most of the developers here work on Microsoft's own ERP products, yet I worked on features that let Visual Studio talk to SAP.
Bugs weren't missed in mainline openSSL. Bugs were logged, sat around for years, and didn't get fixed.
The project management and software engineering practices for openSSL were/are simply not acceptable.
The code is salvageable. The people and processes that allowed the code to get that way are not.
"This code under new management"
The party you are referring to exists - it's called the libertarian party - and it is mostly (but not entirely) ex-Republicans who think responsible adults should be treated like responsible adults -- e.g. left alone until they hurt somebody.
There _should_ be more liberals and democrats joining the libertarian cause, because the LP is much better than the dems on key issues dems claim to care about: anti-war, pro-civil liberties, anti-racism in law enforcement (especially the drug war), anti-corporatism..
So, I cannot tell you why there aren't more democrats who break ranks and join the libertarians.
One reason for that, I suspect, is that I simply cannot relate to democrats or understand how they came to be democrats in the first place. There are plenty of intelligent people who are democrats, but I've never been able to figure out how any of them "tick".
In any case, there are principled libertarians -- and that's how they've traditionally billed themselves. Principled in the sense that they think government morally/ethically should not do certain things.
Then there are pragmatic libertarians -- folks who figure government is _ineffective_ or even malicious at doing certain things, and therefore shouldn't do them. An example would be Gary Johnson.
The bottom line is that, if America were actually incredibly hungry for a fiscally conservative, socially permissive party -- that party has existed for decades. It has been getting more popular lately, but it's still basically a rounding error in most elections.
There are some great points in there
1) access to data without vendor approval/involvement.
2) interop
3) no "remote killswitch" on software
4) no strange privacy leaks
I think these are all fine requirements.
But it's not clear to me why closed software couldn't meet them.
For instance, how does Windows + Office not meet these requirements?
1) the Office XML formats are documented, open, and have reader/writer libraries on non-Microsoft platforms
2) As a result of the consent decree, and much subsequent engineering and doc work, its quite easy to interop with windows and office.
3) So far as I know, there are editions of Windows and Office that require no internet connection at all, and certainly have no provision for remote-kill.
4) Microsoft is actually pretty good about shutting off telemetry, either on a per user basis, or with centralized management tools -- because enterprise customers want this capability too.
I have one of the older e-Ink, Wi-Fi only Kindles. Still has a physical keyboard, which I rarely use. My wife has the ad-supported one with no keyboard, and she doesn't seem to miss it.
The old e-ink kindle is great. I love it. They nailed the user scenario for me -- it is actually _better_ than a physical book. I can use it anywhere I'd use a physical book, I rarely worry about battery life. It's easier to read than a real book when laying on my side in bed.
I am completely uninterested in a color e-reader until it has the battery life and contrast of the e-ink display. And I don't want music, or apps, or multi-tasking, or anything else, because history tells me that adding them will detract from the basic experience of just reading a fucking book.
Here are the improvements I want out of a new kindle.
1) some kind of magical mystery charging. Maybe there is an inductive mat. Maybe its solar. Who knows. I said I _almost_ never worry about charging it. The next step would be I _NEVER_ worry about charging it -- and, I leave the Wi-Fi enabled and continue to not worry about it.
2) bendable/flexible - within limits. If they could make the thing so that it would reliably survive on the outside of soft-sided luggage; if I could put it in a pocket and not worry about it..that would be amazing. What's interesting about this is that the basic e-ink display technology can be flexible...
3) ability to easily -- and I mean easily -- send a book I've finished to my wife's device. Like, if me and her are in the same room, with both of our devices, I ought to be able to send a book on my device to her device. For free. Without any nonsense/bullshit.
Sorry, I fell into the trap of using "right and left", but these mean different things to different people.
When I say "right", I mean "laissez-faire", "capitalist", "individualist", "deregulated"
When many people say "right", they mean "authoritarian" and "nationalist"
That's not what I mean at all. I detest authoritarianism.
There are many places that are more authoritarian than the US (but we're working hard to catch up! (grumble))
There are no places that are more pro-individual liberty than the US. There are a few places which have better pro-business environments, and more economic freedom, but they tend to have fewer civil liberties than the US.
fwiw, my ideas about individual rights may also not be what yours are. I think "hate speech" should be legal, and like any other speech, should only be prosecuted when it is threatening or slanderous. And I think individuals ought to be able to keep machine guns without any government knowledge of oversight. Finally, I think homeschooling is a critical way to pre-empt the historical evils of government indoctrination, and so support homeschooling and parental rights to an essentially unlimited degree -- not because I think all parents are good, but because I think most governments are bad :)
I take individual rights _very seriously_, and so for me, a nation that offers a high degree of individual liberty has the following characteristics: few laws restricting the content of speech, few restrictions on private gun ownership, few restrictions on how children are educated outside of state control.
The US ranks quite well on all 3 of these individually, and taken together, far and away better than anywhere else.
To be fair, all of the countries that do worse than the US _also_ have more leftist governments.
The US has the most right leaning government there is. It also has the most racially, socially, culturally, and economically diverse population there is.
I wish I knew what to do about the problem of people not caring about Math and not excelling at it.
I think you and I probably disagree on many things, although I did see elsewhere that you complained about the growth of administrative/managerial positions within the school system. On that issue, I agree with you entirely.
It's frankly not clear how much the school can do for the kids _in general_ to improve outcomes for the broadest cross section of students, but one thing that has good empirical evidence is reducing class sizes.
That means hiring more teachers.
Frankly, given how much less it costs to hire a teacher than it does to hire an administrator, this should be a move everyone should get behind -- fiscally responsible, pro-teachers, pro-students.
The fact that this doesn't seem to be happening suggests that public education is sadly serving some other set of interests...
Nothing drives engineering like a tangible end goal that is big and easy to understand. Want to make batteries better? Have an engineering target that says what size car they have to fit in...
Tesla makes a better car than a "normal" car. It is faster than almost every sport sedan. It has excellent styling, excellent practicality (seating for 7!), very good handling.
Even if it weren't a pure EV, it would be an interesting contender.
But hey! It's also an EV. It is completely quiet. It doesn't need oil changes, or coolant flushes, or gasoline. You don't have to use the brake pedal anywhere near as much.
Tesla was the first company to make an electric car in my lifetime that wasn't a total piece of shit garbage joke. Not only did they instead make a _credible_ car, they made a car that kicks the ass of nearly other car. Their car is quieter than a Mercedes, out accelerates most BMWs, and handles better any other American sedan.
Why in the world should they stop making cars? The car they've made is better than most of the competition, AND is a game changer.
And, they're selling more of them than they can build. The model S is around 2 month lead time. The Model X already has 6 months of deposits.
I don't want Tesla to make batteries. I want everyone else to figure out how to make Telsas.
Based on my experience of using SketchUp to design a part that I ultimately had ShapeWays print for me, I would say that there's plenty of room for innovation in the 3d printing toolchain.
One thing to note is that modelers like SketchUp assume a uniform internal volume. And so you use tools like slic3r to prepare the model for printing.
In effect, the 3d geometry is the source code. Slic3r would be a "compiler" -- which translates idealized geometry to something more ready for printing -- e.g. by driving an extrusion head in 2d layers (or slices)
I think there is a lot of room left for innovation in the transformation pipeline, which can backfeed requirements into the modelling stage, and which could feed requirements into the print hardware to expose capabilities information upstream as well.
For instance, how can I optimally slice a model so that I use as little print media as possible, while still retaining enough strength? Slicers and printers differ on how they fill interior volumes. Filling volumes with uniform hollow geometry speeds up printing and lowers cost. But is uniform geometry the best choice? When and where are other choices better? And should I slice the model a certain way based on the characteristics of the printer?
Suppose my model let me capture information about stressed joints, where I would want to change the how the interior volume was printed, based both on my design and on the capabilities of my printer?
These features may actually exist, but in the free tools I used, there was no indication that this was the case.
Autodesk could certainly do a lot of good things here. Printing objects that are faster to print, cheaper to print, and stronger when used.. that would be impossible for the industry to ignore.
Obviously, because I think she's neither foolish nor hypocritical.
There's a class of people who respond to Ayn Rand with ad hominem. Which is funny -- She wrote a lot of pages -- more than I care to read in one sitting. In all that, somewhere, you'd think there's be fertile soil for a response more intellectually stimulating than, "she's a crank".
Fault her for whatever reasons you've faulted her, but to me, nobody has more constancy and conviction in their writing in favor of doing the right things for the right reasons. The importance of principle is central to everything she ever wrote.
The Mozilla conversation is about principle vs. pragmatism, and I think her quotes on the topic are highly relevant.
The Hurd isn't a viable alternative because it isn't needed.
Stallman had a vision of a completely free as in speech computer system. When he started, that meant, OS, tools, and application software.
It was a radical strawman against the beginnings of an industry of for-profit software with intellectual property laws.
It turns out that Stallman and his friends created the programmable editor, the compiler suite, the tool chain, the user-space unix tooling..
and them some Finnish guy and his friends came along and made the OS kernel.
The point is that now, not only is there a free OS and development tool chain -- more successful than Stallman could have ever managed -- there is an entire philosophy around free-as-in-speech software.
Stallman has been more influential on how we think about an use computer software than arguably just about anyone. I would at least put him in the same room as a Woz or a Bill Gates.
The market share of Hurd is the wrong metric. The fact that my company -- Microsoft -- is releasing more and more of our stuff as free-as-in-speech software -- that's the metric.
Let's objectively look at what Stallman started.
Let's use this metric: how many Fortune 100 companies have capitulated to _your_ philosophical demands?
Oh look. Here's a whole _page_ of Ayn Rand quotes about compromise
or...
http://aynrandlexicon.com/lexi...
Many folks have a go at the idea that they can somehow tame evil or compromise with it without being tainted too much. I'm not sure this has ever really worked out.
There is a lot to like about the Richard Stallmans of the world. They are clear about the what and the why, and they stick to their guns.