Never entirely sure how to take some of these announcements. Prima facie it sounds great and laudable ... but I'm old, and my memory is full of stories of similar-sounding attempts to get youth mindshare with dubious subsidies of gear into schools, and the old classic of $100m to fund HIV research, $400m to fund ways to destroy GNU/Linux. [1] I know - people tell me - it's a whole new Microsoft these days.
Indeed. Unless there's a promise of migration support once funs inevitably run out or some sort of lifetime fixed price/discount, such that the nonprofits or whatever can make an informed decision about the real cost of their infrastructure once it's no longer free, this is nothing but a terrible trap.
I'd hate to see organizations suckered into this only to find that it has unsustainable costs once the freebies dry up and whatever pains come from attempting to migrate to some other platform/service and/or back onto local servers.
One can certainly design one's apps in such a way where the cloud provider doesn't necessarily matter -- docker containers running on VMs, for instance.
If you use a bunch of things like SQL Server, Document DB, etc. that is certainly much harder to migrate than PostgreSQL running on a VM.
Very good point. You could leverage Docker and an overlay networking solution like ours[1] to abstract yourself from the provider APIs and even the network itself.
They do a lot of things like most big companies, and some people seem to concentrate only the bad.
For example, they started funding OpenSSL and other core infrastructure[1] projects of the Linux foundation after the Heartbleed fiasco. But you don't find that kind of news on here.
By the way "$400m to fund ways to destroy GNU/Linux" seems to be quite hyperbolic, especially coming from 'The Register'. The link in the article goes to a 404 page, do you have any real details of the so called campaign?
TheReg link works fine here. You're right about Reg style, though I have no reason to doubt substance of the story.
Agreed on range of perception of deeds of any large company. Microsoft has long form on this front (baiting the education market), though they're certainly not the only organisation to have done this. Unlike other student discounts - say for movie tickets, public transport, etc - there's a clear and obvious future payback for providing free or heavily discounted software & services to that demographic, as opposed to just charging less to customers who can't afford full price.
Also, IIRC it's a requirement in the USA that public (and not-NFP) companies must always act to 'increase shareholder value', so the giving away of products or services could be considered a legal exposure. The obvious way to work around this concern is to label it marketing, which brings us back to where we came in.
EDIT: Sorry - misread your 404 question -- try these:
>Also, IIRC it's a requirement in the USA that public (and not-NFP) companies must always act to 'increase shareholder value', so the giving away of products or services could be considered a legal exposure. The obvious way to work around this concern is to label it marketing, which brings us back to where we came in.
>EDIT: Sorry - misread your 404 question -- try these:
I read those and I don't find anything to justify OPs "$400m to fund ways to destroy GNU/Linux" or The Register's characterization of '$421m to fight Linux".
Happy to be disabused on the shareholder value front. In AU we're obliged to consider shareholder interests, but with no constraint to prioritise short-term gains. No idea how far this has been tested by case law.
For example, they started funding OpenSSL and other core infrastructure projects of the Linux foundation after the Heartbleed fiasco. But you don't find that kind of news on here.
It all depends on what you consider to be more valuable: 3 years worth of a team's effort working with a specific stack, or 3 years worth of free cloud services. My guess is that in the average case it skews heavily in the team effort direction. So then it's a net gain by MSFT. Still, you shouldn't discard charitable efforts if they also happen to be structured to benefit the benefactor.
This is the same old strategy of giving students discounts which get them accustomed to proprietary products, then marketing to those same students when they graduate.
Now it's just adapted for remotely-hosted software.
> This is the same old strategy of giving students discounts which get them accustomed to proprietary products, then marketing to those same students when they graduate. Now it's just adapted for remotely-hosted software.
Sure. But how is Azure and Microsoft cloud services any more proprietary than AWS or Google's offerings?
They're not (other than aws being so popular/early-to-market that there are a few work-a-like open/Free olatforms emulating some of the apis/services).
The 1 billion number sounds impressive on its own - wonder how many credits Amazon has given away by now?
Lets hope MS reaches out to github or vice-versa so they can get added to the student pack:
So long as the course is designed independently of Microsoft then it's a failing of the university if that happens. Any donation of cloud services, from any given provider, could be used to teach "how to deploy to this specific cloud provider" or "how to deploy to cloud services". The university needs to go for the latter.
How many successful startups have actually been built on top on Microsoft software. I know Plenty of Fish is one example, but you just don't hear of that many examples of successful startups built on top of Microsoft's stack. The reason being that paying Microsoft's tax as you scale eats into your profits and your growth.
It's hard to separate out the Microsoft stack and the scale up approach because for licensing reasons they tend to go together. If you find yourself in the position of transitioning from scale up to scale out by adding dozens of cores, MS licensing will bite you.
You know what else is crazy? If you find yourself in the position of scaling your dev team because your business is successful, salaries will bite you. If you thought MS licensing was expensive, try hiring senior developers! :)
Between this and the ChakraCore news item on the front page I've been wondering whether Microsoft really has been changed its behavior since the 90s, or whether this is just how they play the game when they're not in a position of strength as they are with node.js and cloud services.
It's pretty telling from the ChakraCore thread not entirely being "They're obviously trying to embrace, extend, extinguish node.js like they did Java" that people have either forgotten their modus operandi, or that I'm too much of a cynic and they've genuinely changed.
Sometimes I think I'm being too cynical and they're now just as "nice" as some other companies which have a mostly healthy relationship with open source & open architectures.
But then again all the news where they're being more open these days have to do with areas where they don't have established lock-in, including projects like .Net & ChakraCore which formed the basis of some of the major projects where they lost the lock-in wars.
Are they just trying to lure users they've lost back into their sphere of influence so they can tighten the noose again, or are they genuinely doing business differently now?
I don't know, but it's hard to place any trust in them when you've been burned so many times, and even though they have a new CEO middle management & their business model has a lot of inertia.
Posts like this remind me of the quote in N.N.Taleb's Antifragile book, The mind fights yesterday's war, but the body fights tomorrow's war (in the context of biological systems under stress).
Some minds here are fighting a decade old war and no good deed goes unpunished in HN comments.
IMO you don't get to shed decades worth of reputation because you made a few good moves in the open source world.
For me at least, such a company would have to spend as much or more time pushing this new open idea than they did pushing the Embrace Extend Extinguish model.
Some people are more forgiving than me but I agree with the parents skepticism about how the openness seems to largely be in areas where they already lost the lockin wars.
Look they released their Javascript engine with no JIT compiler for Linux. It's the same Microsoft because they are in it for money and dominance. A leopard can't change its spots. Microsoft will always be the same, no matter what they say. They are simply changing themselves to be more like Google because they know their old strategy does not work in the new world.
If anything it's Microsoft which is fighting yesterday's war with today's strategy.
And that's all you have to do, just keep saying we're working on it, and turn to the next question. All ways of getting people to use your operating system and your eco-system while keeping the rhetoric of we're open now. And they can't be it's not in their nature. They have a whole lot of business interests which will keep them the same old, same old.
There's nothing new -- MS has been giving away free (or nearly so) Windows and Office suites to students as long as I remember. Only the product has changed, now it's cloud services. Basically, it's like a drug dealer giving you the first dose for free.
So they are giving away, to educational institutions, a limited amount of a consumable resource which is amenable to lock-in effects (i.e. addictive), and problematic for privacy (i.e. harmful to users). In effect, they are standing outside schools giving away cigarettes.
I think it's a great thing. People already use Azure or some other provider at these institutes (I know my uni does). If Microsoft wants to trade free services for market share, everyone partaking benefits from it -- unlike cigarrets at schools. If you or your institution don't want to partake, don't. I know I won't.
But there is no reason the support has to be "free" access to cloud resources at all. Why not just donate the money and let the nonprofits choose which cloud to use (or whether to use a cloud at all)?
Because if you give 1 billion dollars away, you are 1 billion less wealthy. If you give away 1 billion dollars worth of Azure credits, you are some much smaller percentage of 1 billion less wealthy, and the recipient is still receiving 1 billion in value.
Only because you declare that something has the value of 1 billion doesn't make it so.
First, you're not giving away physical goods, but eligibility for a service - but what exactly this service entails and how useful it is can still change in the future.
Second, it's a service that requires a substantial level of commitment from your clients - a commitment that causes risks for the clients in its own (lock-in) as discussed in many other places in this thread.
So without knowing more details, without some special effort on the nonprofits side, I'm guessing this will lock them into an Azure architecture. That could be OK if MS continues the donations after the initial $1B runs out, otherwise it could end up costing the nonprofits/.edu's more in the long run than if they went with a more open path now.
I suspect this move is somewhat motivated by the "first hit is free" strategy as other commentators have speculated. This is probably secondary, however, to the fact that Azure has been temporarily overbuilt (or, somewhat equivalently, over-contracted). That is, they bought all of this equipment, fewer people than expected are using it, so instead of just dropping the price for their services this seemed more interesting. It also happens to be easier to construe as "evil" than dropping the price would have been, so I'm not sure they'll come out ahead in terms of PR...
The criticisms here remind me of The Copenhagen Interpretation of Ethics[1]:
> The Copenhagen Interpretation of Ethics says that when you observe or interact with a problem in any way, you can be blamed for it. At the very least, you are to blame for not doing more. Even if you don’t make the problem worse, even if you make it slightly better, the ethical burden of the problem falls on you as soon as you observe it. In particular, if you interact with a problem and benefit from it, you are a complete monster. I don’t subscribe to this school of thought, but it seems pretty popular.
Microsoft is giving away a billion dollars worth of services, yet people are criticizing them for it. There are several reasons why that's unwarranted.
First, universities and non-profits don't have to choose Microsoft. If they think the risks of lock-in or later expenses will outweigh the donated services, they won't pick it. These organizations aren't populated by idiots, and Microsoft makes the terms of their deals clear before any contracts are entered. There's no coercion or trickery.
Second, when it comes to these sorts of discounts, Microsoft is late to the party. AWS gives free stuff to schools and non-profits. Apple has education discounts. People would be outraged if these companies ended these discounts. Yet people are outraged at Microsoft for offering similar discounts. You can't have it both ways.
Third, I seriously doubt that Microsoft is doing this as a purely cynical, profit-seeking move. I can't find any mention of a time limit on the free services. Not in Nadella's announcement. Not in responses to Q&A. Nowhere. There's also the fact that, for the past couple of years, Microsoft has been giving Office 365 away to non-profits and schools. Once they validate your organization (to make sure you're not trying to scam them), they give it away, for free, forever. What's the ulterior motive there? That people might use it, like it, and pay for it at home or at a different job? That's completely unobjectionable.
One last point: People like to accuse companies of predatory discounts, but nobody brings up the opposite. What about the countless free-loaders who use as much of a service as they can without paying, then move on to something else? I've seen people brag about using free servers from Rackspace, AWS, and Azure, with no intention of paying those companies a cent. As soon as the discount runs out, they move on. There is little condemnation of these tactics. If anything, many social circles celebrate them. This is a clear double-standard.
You make some great points, and I'm genuinely torn on how I should feel about the announcement.
I'll comment on only one of your points, though:
> Microsoft is giving away a billion dollars worth of services,
> yet people are criticizing them for it.
1. I don't think people are criticising the act of giving something away. It's the questioning of the motivations behind that.
2. The billion dollars is a retail value -- and the retail pricing is defined by Microsoft. I know these kinds of announcements (from everyone) always use retail value, but it's a disingenuous way of determining actual value.
(a) It would actually indeed be more accurate to quote the difference in assets the donation is causing for you--anything beyond that is arguably being paid by other tax payers.
(b) Your analogy isn't actually analogous. If I take a rock from my back yard and attach a price tag of "ten trillion USD" to it, and then donate it to you, would you accept my claim that I donated assets worth ten trillion USD to you? A price that has only been declared by the seller says nothing about the value of the thing the price is attached to. Only a price at which a transaction happened says anything about the value--that is, a price at which someone actually bought what you offered. Now, if they are giving something away for free, that transaction arguably indicates that the value is zero USD.
>It would actually indeed be more accurate to quote the difference in assets the donation is causing for you--anything beyond that is arguably being paid by other tax payers.
For tax purposes yes I'm not an accountant but MSFT isn't doing this for a tax break and they'll get credit for the actual financial value of their donation based most likely on their operational costs.
>Your analogy isn't actually analogous. If I take a rock from my back yard and attach a price tag of "ten trillion USD" to it, and then donate it to you, would you accept my claim that I donated assets worth ten trillion USD to you?
My analogy was pretty much correct this isn't an sole actor attributing an arbitrary price, Azure pricing is set by the market not by MSFT.
You can assign a defined discrete value for each "resource credit" for Azure, AWS, DigitalOcean, any other IAAS/PAAS service provider out there based on it's market price as these are defacto commodity resources.
If you give me a rock which is worth a trillion dollars by your account its really not the same thing unless the market agrees that this rock is worth a trillion dollars either in direct value (e.g. you can sell/buy a rock for that sum) or indirect value (e.g. the amount of resources I would need to spend to find a substitute rock).
> For tax purposes yes I'm not an accountant but MSFT isn't doing this for a tax break and they'll get credit for the actual financial value of their donation based most likely on their operational costs.
(a) You were talking abstractly about "donations", not about Microsoft, (b) no, not for tax purposes. If you give someone a million dollars and then are reimbursed, say, a quarter of that by someone else, it's simply dishonest to publicly claim that you donated a million dollars, as (ba) your assets have only decreased by 750000 USD due to the donation, and (bb) because you are taking credit for the quarter million that that other person/entity paid for.
> My analogy was pretty much correct this isn't an sole actor attributing an arbitrary price, Azure pricing is set by the market not by MSFT. You can assign a defined discrete value for each "resource credit" for Azure, AWS, DigitalOcean, any other IAAS/PAAS service provider out there based on it's market price as these are defacto commodity resources.
Except that you choose to simply ignore one giant transaction in this market that happened only once they selectively reduced the price to zero (or rather, might happen, if people don't still consider that to be too expensive). If they were giving away a hundred dollars worth of services, sure, I'd be willing to accept that as a reasonable method of valuating the donation, but if it's in the same order of magnitude as their total yearly revenue in this area, not so much.
Also, it's actually not quite as much a commodity as you make it out to be, because of proprietary interfaces--the different providers certainly are in competition to one another, but one's product is not necessarily a direct substitute for the other's.
> If you give me a rock which is worth a trillion dollars by your account its really not the same thing unless the market agrees that this rock is worth a trillion dollars either in direct value (e.g. you can sell/buy a rock for that sum) or indirect value (e.g. the amount of resources I would need to spend to find a substitute rock).
A better measure in this case would be how much you could get others to buy it from you for. If your rock massively increases the supply of rocks, that is going to considerably drop the price.
>(a) You were talking abstractly about "donations", not about Microsoft, (b) no, not for tax purposes. If you give someone a million dollars and then are reimbursed, say, a quarter of that by someone else, it's simply dishonest to publicly claim that you donated a million dollars, as (ba) your assets have only decreased by 750000 USD due to the donation, and (bb) because you are taking credit for the quarter million that that other person/entity paid for.
I'm not sure you understand how tax deductions work.
Donating to charity allows you to claim a portion of that donation as a deductible, this can vary depending circumstances but in most cases you have a limit for the deductibles for gross income and the amount it self is usually not deducted in full regardless of limits when it comes to non-cash property (this is also gets more complicated because you have carrybacks, other losses, expenses etc.).
When Microsoft donates X amount of what ever to some one no one pays MSFT that money back, they might get some tax relief for it that's all this doesn't mean that you as an individual lose money in fact direct donation in many cases is more efficient than indirect funding through taxation.
>Except that you choose to simply ignore one giant transaction in this market that happened only once they selectively reduced the price to zero (or rather, might happen, if people don't still consider that to be too expensive). If they were giving away a hundred dollars worth of services, sure, I'd be willing to accept that as a reasonable method of valuating the donation, but if it's in the same order of magnitude as their total yearly revenue in this area, not so much.
It doesn't affect the price of the service in any way regardless of what portion of the global XAAS market or MSFT's own XAAS business it is. This simply allows for organizations that would not afford access to these resources at this scale before to have the opportunity to do so.
It's also spread across 3 years and includes all MSFT cloud services so Office 365, Skype for Business/Lync, OneDrive etc. which makes is about 1/8th of their currently yearly revenue from cloud products which means overall it's about 1/24th of it's yearly cloud related revenue over the duration of the program.
>because of proprietary interfaces--the different providers certainly are in competition to one another, but one's product is not necessarily a direct substitute for the other's.
This holds true even if you build your infrastructure and platform in house it's not like MSFT can donate universal credits
>A better measure in this case would be how much you could get others to buy it from you for. If your rock massively increases the supply of rocks, that is going to considerably drop the price.
I am a non-profit or a school I don't sell rocks but I do need one, not everyone operates a tradable commodity exchange...
You don't value the price of donated food at the local homeless soup kitchen by how much you could sell it to some one on the street you value it based on how much you will have to pay to make it on your own.
And the people who actually receive the food tend to value it based on how hungry they are....
> When Microsoft donates X amount of what ever to some one no one pays MSFT that money back, they might get some tax relief for it that's all
I am not sure you understand how accounting works. You have a tax liability. Now, you donate a million dollars (let's assume money for the sake of simplicity). How is this donation reflected in your balance sheet? Well, your bank owes you a million dollars less, and at the same time, you owe the state 250000 dollars (or whatever it is) less. Now, if the fact that your bank owes you a million dollars less means that you paid the donation, then the fact that you owe the state 250000 dollars less also means that the state has paid you a quarter of a million dollars of your donation back. Or in other words: That part of the donation is paid for by other tax payers.
> It doesn't affect the price of the service in any way regardless of what portion of the global XAAS market or MSFT's own XAAS business it is.
You have it all backwards. This is not about whether the donation changes the price in other sales of the same service. This is about how you determine what the price of a product is. You determine the price by figuring out what people are paying for the product. Now, when you do this, you simply discard one giant transaction that happens in this market, for no good reason. This giant transaction happens to be one where the price is zero.
> This holds true even if you build your infrastructure and platform in house it's not like MSFT can donate universal credits
(a) How does that change the fact that it's not a commodity?
(b) Actually, they could. People commonly call this "universal credit" "money".
> You don't value the price of donated food at the local homeless soup kitchen by how much you could sell it to some one on the street you value it based on how much you will have to pay to make it on your own. And the people who actually receive the food tend to value it based on how hungry they are....
Which should all result in the same value, unless you are making a mistake.
>I am not sure you understand how accounting works. You have a tax liability. Now, you donate a million dollars (let's assume money for the sake of simplicity). How is this donation reflected in your balance sheet? Well, your bank owes you a million dollars less, and at the same time, you owe the state 250000 dollars (or whatever it is) less. Now, if the fact that your bank owes you a million dollars less means that you paid the donation, then the fact that you owe the state 250000 dollars less also means that the state has paid you a quarter of a million dollars of your donation back. Or in other words: That part of the donation is paid for by other tax payers.
If you donate money you donate actual money, the deductibles are for potential profits from your revenue.
Say I donated 1M to charity these 1M are gone they aren't mine any more, this means that my buying power has been reduced by at least 1M and considering credit even more, and I've just lost any potential ROI on that money.
What I can potentially get in return is the ability to deduct some of that amount from my tax future tax liability this means that if I make enough money to pay tax I could deduct a portion of the tax I would have to pay based on that donation.
No one in this scenario gave me money, I've gave out 1M, I might get some of it back however this isn't additional money that the tax payer has to pay that just money that is gone from the tax system.
By this logic a company that operates at an operational loss or an individual that their assets have resulted in a capital loss some how increase your tax liability which is simply not the case.
Now you might say that if people lose money all the time and not pay taxes eventually the Government will have to balance it out by taxing you more but this doesn't really happen that often as organization will not exist in operational loss which defers any tax liability on their account indefinitely same goes for people which their assets continuously lose value.
And while you might say but hold on a second Charity allows you to get some tax deduction and you can do that indefinitely well no no amount of charity will cancel out your tax liability, the cap ensures that the amount which can be deducted will always be considerably.
But you also need to remember that charity means that money goes to causes that in most cases would be funded by public funds (with Religious institutions being the only exception in some countries) this means that if you donate 1M and get 250K in deductibles the banks owes you 1M less, you have much less credit, you owe the state 250K less potentially but the state just got 1M in surplus because it doesn't need to allocate funding.
Charity (when people actually donate money, goods or labor) in most cases works exceptionally well when you actually donate to decent causes and not to some Kony2012 BS (and before you complain about remember that to pass the budget they US government ends up spending 100's of millions on nonsense like investigating the effects of beavers on moose mating calls and the economical implication of anti trust investigations against ice cream cartels in Guatemala on US foreign interests) to the point where some countries either do not put caps on charitable donations or are looking into removing them.
You need to understand that charity isn't a tax loop hole even with 100% deductible.
If I earned 2M paid 50% of that in taxes in a given year (n) I now have 1M$, if I donate all of that to charity the government has technically received 2M from me instead of 1M as it now doesn't have to pay out additional 1M to fund various project because I have done so in their place.
Not only that but technically depending on what I've donated too I've actually increased their tax revenue potentially even more since some of my donation could end up paying for goods (VAT/Sales tax), salaries, microloans for businesses etc, housing (deed stamp tax (not sure if you have that in the US) etc.
Now I'm out of 1M$ and i get it as a deductible and I carryback this deductible to the next year to get my money back I need to make just as much as I did last year while paying no taxes, so if I make 2M again I will have 2M which is exactly the amount of money I would've have if i donated 0 to charity and paid the full tax for each year (if i do not donate to charity I could potentially have more money from capital gains).
This is pretty much the definition of a zero sum game or well to more accurately put it equivalency - Charity is pretty much paying more "taxes" if your deductibles are capped (especially at a percentage rather than a flat sum) or extending credit / giving a tax free loan to society if you can deduct the full amount (given you can make sufficient revenue or carryback the deductible over subsequent years).
There is no case in which when Money Inc LLC or Richie Rich give 1M to charity Johnny Bluecollar ends up paying more tax to in order to cover for them.
> Microsoft is giving away a billion dollars worth of services, yet people are criticizing them for it. There are several reasons why that's unwarranted.
The correctness of this statement depends on the correctness of the valuation you are quoting. If you interpret the critics correctly, they are saying that the long-term costs might outweigh the short-term benefits of this "donation". If that is the case, the donation's rational valuation wouldn't be a billion dollars, as claimed by Microsoft, but actually negative. If you accordingly rephrase your argument using some negative valuation, so as to mirror the position that you are apparently trying to argue against, you might notice that it's not a sensible argument at all.
>> The Copenhagen Interpretation of Ethics says that when you observe or interact with a problem in any way, you can be blamed for it. At the very least, you are to blame for not doing more. Even if you don’t make the problem worse, even if you make it slightly better, the ethical burden of the problem falls on you as soon as you observe it. In particular, if you interact with a problem and benefit from it, you are a complete monster. I don’t subscribe to this school of thought, but it seems pretty popular.
That's well put and explains the furor over Bill Gates' charity work while other immensely rich folks get zero heat for essentially hoarding their wealth like Scrooge McDuck instead.
What? I've literally never seen anyone say anything negative about Bill's charity work. Can you cite some sources for that?
Microsoft "charity" however will allways be suspicious because how ridiculosly backhanded they were during their era of monopoly. Whether its justified or not remains to be seen.
You can find a single person on some forum on the internet who is against anything. That's hardly a furor.
I don't think there is any substantial negativity associated with Gate's charity work. The closest you'll find is some in the educational space who don't agree with the metrics his foundation uses to measure progress.
There seems to be a lot of concern these days about platform lock in with Microsoft's new found philanthropic initiatives.
If one were to stick to IaaS, using only Azure hosted Linux VMs and not get involved in any special Azure PaaS offerings, then any project hosted under this model could easily avoid platform lock in and be transported to another service provider at any time.
As a student, I prefer Google.
The integration or multi-platform services of Google is better.
I got free Microsoft services from my school but I never use it.
Cloud services alone don't make what is said possible. Perhaps they'll be able to afford specialists to build what you are thinking of with the money saved from the Microsoft granted credits? Won't the university or non-profit need to hire a 'big data engineer' to build something using the services? Also analysts to know what to look for? The statement makes it sound as if the cloud can unlock it, but it's really people trained to know how to use the cloud to unlock it that make it possible, in my opinion.
Would you feel the way if Apple donated $1B worth of iPads and iPhones? I feel like most people wouldn't, yet it also lock-in into their proprietary technology.
Seeing what Microsoft has done recently with privacy on their newest operating system, I would not have high hopes for their cloud services. Of course the organizations should be weighing all the pros and cons but I share the sentiment that once the freebie well dries up, terms might be re-evalutaed or services held hostage. Also, and I know the OneDrive plan shakeup fiasco is not apples to apples[0], but I'm sure these free plans could suffer a similar fate. With this, it's not necessarily as easy as moving mom's photo album back to your desktop.
I know the Univeristy of Wisconsin used to roll its own email, but recently started using Outlook acounts for everything and I'm definitely not a fan. Migration was painful or it at least seemed so. If orgs are able to isolate their architecture from the cloud and services, it might work out. Unfortunately, a lot of companies and organizations treat aggregate privacy like a preference instead of a requirement. Bartering chips.
My cynicism may be unfounded, but a company by the same name helped write the book on fostering lock-in not too long ago.
[1] http://www.theregister.co.uk/2002/11/13/gates_gives_100m_to_...