Forgot your password?
typodupeerror
Open Source Microsoft Software Linux

Microsoft Dilutes Open Source, Coins 'Open Surface' 191

Posted by timothy
from the scientology-favors-sanity-don't-you dept.
sfcrazy writes "Now, Microsoft is coining yet another term to further confuse users — 'Open Surface.' Senior Director for Open Source Communities at Microsoft, Gianugo Rabellino, said at Oscon 2011 that customers don't care about the underlying platform as long as the APIs, protocols and standards for the cloud are open. That's when he threw the term 'open surface.'" This seems to have more than a grain of truth to it — after all, programmers have been creating open-source software with closed-source programming languages for many years, and I'm certainly more impressed by Google's willingness to let me export my data than I am turned off by the fact that they use a mix of open and closed source software to run the Google circus.
This discussion has been archived. No new comments can be posted.

Microsoft Dilutes Open Source, Coins 'Open Surface'

Comments Filter:
  • by Gr33nJ3ll0 (1367543) on Thursday July 28, 2011 @10:32PM (#36916966)
    I believe this is step 2
    • by starfishsystems (834319) on Friday July 29, 2011 @01:32AM (#36918002) Homepage
      Indeed. I'm glad you got first posting for this.

      It's not that specifications and standards aren't important. Of course they are. But Microsoft is more than a bit disingenuous in pretending to advocate them when it has been so egregiously, perennially active in undermining them. This hypocrisy is all too familiar.

      Thanks, Microsoft, for reminding me why I loathe you.
    • Re: (Score:2, Interesting)

      by mwvdlee (775178)

      Microsoft would be correct in assuming that most users wouldn't care about the openness of the underlying platform as long as API's, protocols and standards are open.

      The problem is that API's, protocols and standards pretty much never are open without releasing a reference implementation. The API's for Windows are not open simply because they do not offer access to every aspect of the underlying platform. To make a completely open API, you have to release the sourcecode to whatever the API is for; it's the

      • by TheRaven64 (641858) on Friday July 29, 2011 @07:08AM (#36919436) Journal

        The problem is that API's, protocols and standards pretty much never are open without releasing a reference implementation.

        Not true. In fact, having a reference implementation is a good way of making a kludgy standard. The IETF's requirement, for example, does not require a reference implementation. It requires two independent implementations. If the standard can't be implemented by two groups who aren't sharing code then it's not a standard it's just documentation.

        • I don't think that works for "open surface" with MS.

          It is, by your standard, designed to be broken.

        • The IETF's requirement, for example, does not require a reference implementation. It requires two independent implementations.

          The use of that rule (by people other than IETF) is why we can't have WebDB as a part of the HTML standard: there were several implementations, but all of them left the heavy lifting to the same (public domain) backend (SQLite) rather than reinventing the wheel (or, rather, the lightweight embedded relational database), and so they weren't deemed to be sufficiently independent.

          • And that was a good reason. The WebDB specification said 'accept the SQL that SQLite accepts'. It was a horrible specification. SQLite is okay, but creating a specification that requires SQLite to implement it correctly is horrible. A client-side SQL API for the web should define the exact subset of SQL that it accepts and it should be possible to implement it on top of any SQL database with a query validator, or on top of a your own storage mechanism with a SQL parser.
            • And that was a good reason.

              I never said it wasn't.

              The WebDB specification said 'accept the SQL that SQLite accepts'. It was a horrible specification.

              That wasn't actually the problem, the absence of the required degree of implementation independence was (at least, in the discussions of the issue, incorporating specifications of the required behavior directly into the standard was, IIRC, specifically called out as an inadequate fix, the problem wasn't the spec per se [that was a problem, but correctable] the

      • Or so said Alan Kay, more or less, arguing for open reference implementations.

    • by tchdab1 (164848)

      Perfect MS architecture: you think it's open, we own what's under the surface.
      Perfect in that it fits their historical culture.

  • by ToasterMonkey (467067) on Thursday July 28, 2011 @10:40PM (#36917038) Homepage

    This is so backwards from Slashdot norm. A summary with a tidbit of "news" in it and intelligently written opinion, no FA to read.

    Am I missing something, did Microsoft not really coin this term or is there some biased, slanderous opinion that was unintentionally left out of the summary?

    • by Sulphur (1548251) on Thursday July 28, 2011 @11:57PM (#36917546)

      This is so backwards from Slashdot norm. A summary with a tidbit of "news" in it and intelligently written opinion, no FA to read.

      Am I missing something, did Microsoft not really coin this term or is there some biased, slanderous opinion that was unintentionally left out of the summary?

      In other words, an Open Surface post.

    • Re: (Score:3, Informative)

      by Anonymous Coward

      Submitter is an idiot.

      Anyway, for a coherent link, zdnet [zdnet.com], which has more than three sentences. What it boils down to is that the Microsoft Azure platform is not open source - but how to interact with it is well known and open. You can then run Open Source programs on top of a closed-source platform.

      To be honest, I think it's a complemntary idea to Open Source; and I'm not sure that he explicitly set out to 'dilute' the term open source.

      • by hedwards (940851)

        Yes and no, it's a medium ground between fully opening the source code and keeping the APIs shut, but I'm not really sure how this differs from typical APIs where the programmer knows what the API is but doesn't necessarily know how any of it is accomplished internally.

        • by GumphMaster (772693) on Friday July 29, 2011 @12:40AM (#36917758)

          I'm not really sure how this differs from typical APIs where the programmer...

          One is a marketing term that Microsoft can slap a trademark application in for, the other is not? Then Microsoft can claim to be the only vendor with Open Surface(tm) systems, or OSS.

        • by smash (1351)
          These days, there are enough levels of abstraction for the typical web/cloud app that the programmer doesn't generally need or want to *CARE* about the low level implementation details. So long as the API is open and it is portable from platform to platform, the low level source of the host platform is not of concern.
        • by rtb61 (674572)

          It is all about thinking for today versus thinking for yesterday, today and tomorrow. You might know the current APIs, protocols and standards but when the underlying platforms are closed off from input and control, you have no idea what you will have to deal with tomorrow and how that will affect what was done yesterday.

          When it comes to M$ they have been unreliable, manipulative, insensitive and arrogant. If it saves them money and enables them to make more whether it be saving costs by not fixing bugs

        • Yes and no, it's a medium ground between fully opening the source code and keeping the APIs shut, but I'm not really sure how this differs from typical APIs where the programmer knows what the API is but doesn't necessarily know how any of it is accomplished internally.

          Its hard to tell a lot of detail (and its possible that the concept is fluff that doesn't have a strong concept behind it) from anything I've seen about it, but the "open surface" description includes not just APIs but also "protocols and sta

    • You mean libelous, since it's in print. But otherwise I agree with you: the article is poorly written garbage, and the summary is decently insightful.
    • "Now, Microsoft is coining yet another term to further confuse users ... [emphasis added]"

      You missed the first half of the first sentence apparently, but fear not because Microsoft does suck and almost everyone here knows it. The fact that you haven't figured it out yourself yet is beside the point.

  • "Published API" (Score:3, Insightful)

    by Alex Belits (437) * on Thursday July 28, 2011 @10:46PM (#36917094) Homepage

    It's called "published API".

    Microsoft, as usual, is trying to conflate "published protocol" (an interface that can be used by independently developed software that may share no components with software providing interface) and "published API" (an interface that requires direct use of software providing the interface within common framework such as libraries, plugins, compilers' handling of interface definitions, etc.)

    Shut up, Microsoft. Nothing short of published, open protocol is going to suffice. And none of your products will survive if you won't hide and obfuscate protocols used by them. You know that and we know that, so don't pretend that you are not our enemies.

    • Re: (Score:3, Informative)

      "published API" (an interface that requires direct use of software providing the interface within common framework such as libraries, plugins, compilers' handling of interface definitions, etc.)

      It doesn't. You can take a published API, and provide your own clean-room implementation of the same - see .NET/Mono.

      Nothing short of published, open protocol is going to suffice.

      http://www.microsoft.com/openspecifications/en/us/technical-specifications/default.aspx [microsoft.com]

      • by Anonymous Coward on Thursday July 28, 2011 @11:10PM (#36917258)

        I believe the following statement on the front page says it all:
        "rights under Microsoft patents covering such specifications are available separately"

      • Re: (Score:2, Troll)

        by Alex Belits (437) *

        It doesn't. You can take a published API, and provide your own clean-room implementation of the same -

        I am describing normal use, not reimplementation. Reimplementation of a product made by a large company is usually a massive, rarely successful effort, prompted by the company being negligent or outright malicious with the product development, use or licensing.

        see .NET/Mono.

        Mono is a failed reimplementation of a useless product. The only successful reimplementation of a proprietary API (as opposed to protocols/formats) that I have ever seen is Lesstif, and arguably more good was done by destroying Motif dominance by supe

        • I am describing normal use, not reimplementation.

          In normal use, there's no difference between e.g. a network protocol and a shared library interface. Both are ultimately about pushing some bytes in, and getting some bytes (and side effects out). That in one case your bytes are packaged in packets and sent between processes and computers, and in another they are pushed onto the execution stack, is immaterial.

          Mono is a failed reimplementation of a useless product.

          Your expert and well-referenced opinion on this will doubtlessly be extremely valuable to people who mistakenly use either or both in their products w

        • by unity (1740)

          Mono is a failed reimplementation of a useless product.

          That is funny, I coulda sworn that my customers put through hundreds of millions of dollars of sales on .net software every year. Hell of a useless product.

          • by jimicus (737525)

            Mono is a failed reimplementation of a useless product.

            That is funny, I coulda sworn that my customers put through hundreds of millions of dollars of sales on .net software every year. Hell of a useless product.

            The "useless product" bit may be untrue, but I'd be inclined to agree with the "failed reimplementation" bit, if only because I have never yet seen a commercial software firm officially support their product under both .NET and Mono. Usually, the best you'll get is someone saying "It might work - we have no idea. If you get it working, good for you. If you have any trouble - that's your problem."

            • by Alex Belits (437) *

              I have never yet seen a commercial software firm officially support their product under both .NET and Mono.

              Fogbugz does that. Not to say that Fogbugz is a good product or that Joel is not an idiot for choosing .NET as the platform for it.

      • by jabjoe (1042100)
        Yer, only if the specification is perfect. Which it never is. There is always things not documented that projects like Wine, Mono, Samba, etc must work out. Most of the time I don't think it's a deliberate MS policy (though I bet some of the time it is), it's just the nature of software. Really you want not just a published specification but a open reference implementation. MS, intentionally or otherwise, use this in both directions.
        http://tuxdeluxe.org/node/296 [tuxdeluxe.org]
    • by unity (1740)

      You know that and we know that, so don't pretend that you are not our enemies.

      Speak for yourself; MS is my sugardaddy and I like it that way.

      • by nschubach (922175)

        Money buys all kinds of willingness to ignore one's morals and law. Just look to our Congress-people. ;)

  • Take the word as it is - open as in open, and forever to be open.

    That requires whatever that is supposed to be open, not to be owned by any private party. Because, that only means 'open for now'. until the owner decides to close it.

    Open means open - not 'open for now'.

    So, microsoft, if you own anything, it cannot be open. now go shove your intentional concept obfuscation up your ass. the people who are technically apt enough to be working on these matters, are not as clueless as your customers to b
    • Um... what? That's not what the word "open" means. If a door is "open" is might at some point in the future be "closed". The word open describes the current state of something.

      Unless you mean "Take the word as it is given the specific context I will now thrust upon it, thus defeating the point of my own argument that the word has a specific universal meaning"

      • by unity100 (970058)
        considerations and importance of open technologies comes from the concerns of longevity, modifiable-ness and accessibility of the technology.

        so, if a technology may suddenly become closed and inaccessible in future, that means those considerations have no been fulfilled.

        therefore, 'open' does not mean open in that context. because context was different from the start :

        longevity, modifiable-ness and accessibility
  • by Stradenko (160417) on Thursday July 28, 2011 @10:53PM (#36917140) Homepage

    I'll be the first to coin "free (as in freedom) surface."

  • by petes_PoV (912422) on Thursday July 28, 2011 @10:59PM (#36917182)

    as long as the APIs, protocols and standards for the cloud are open

    The key thing is to ensure that the APIs cannot be controlled, or changed or withdrawn or have conditions of use imposed on them. Open means more than just having them documented.

    The only way to ensure that the APIs remain usable is to have the ability to rebuild the underlying software, rather than simply have a third party provide us with it - where the way the API is still under their sole control. To do that requires unencumbered access to the source code, and the entitlement to copy it and make other things that use it.

    Without those abilities, there will always be the possibility that the original owner could arbitrarily change it, refuse to support it, add private functions and features or prevent certain classes of users from benefitting from it. These are the attributes that make free software valuable.

    • Sure, but if the APIs are standardized, if a vendor like MS arbitrarily change theirs, you could just move to other vendor.

      Standard (open) APIs remove the possibility of lock-in.

      • by nschubach (922175)

        But EEE has a history of disqualifying that statement. Microsoft Embraces an open API, they start adding their own Extensions to that API and eventually they Extinguish parts of it they no longer want to support. Being in the position of market leader, those wishing to continue making money (more customers) will use the Microsoft methodologies invalidating the whole point of standards.

  • by jenningsthecat (1525947) on Thursday July 28, 2011 @11:04PM (#36917216)

    Whether or not this is a move to co-opt FOSS, I can't say, although I have my suspicions. But from a security standpoint, it sucks. Security breaches are becoming more and more common; with the underlying code being closed, there can be no independent confirmation of the quality of security measures, patches, etc. So when a vulnerability is found and 'patched', we still won't have any assurance, beyond Microsoft's say-so, that the patch fixes the problem and doesn't introduce any new ones.

    This announcement doesn't really change anything, and on the face of it it's non-news. But as propaganda, it stands a good chance of getting more people to drink the MS Kool-aid. And remember when MS used to use undocumented OS calls to give their own applications an edge over competitors? I think we can expect such abuses to increase greatly - the appearance of openness will hide what's really going on.. The 'surface' may be 'open', but the underlying code, and the underlying politics, are murkier and more closed than ever.

    Besides, 'Open Surface' sounds rather shallow, doesn't it?

    • by devent (1627873)

      I think you missed the point of the whole "cloud" thing. If you can't get your own "cloud" then it doesn't matter whether the software on the server is open source or not. Even if all the software is open source, you don't have access to the servers and can't verify that the software is properly patched and have to rely on the vendor's say-so.

      That is why, in my opinion, it's so dangerous if governments rely on cloud vendors for their I.T. needs. It's the same as to rely on a proprietary vendor, whether the

      • I think you missed the point of the whole "cloud" thing. If you can't get your own "cloud" then it doesn't matter whether the software on the server is open source or not.

        If the software is open source, then (presuming it isn't open source software that only runs on very peculiar hardware that isn't generally available on the open market) you can get your own cloud.

        That is why, in my opinion, it's so dangerous if governments rely on cloud vendors for their I.T. needs. It's the same as to rely on a proprieta

    • Sort of like the "Clear Skies Initiative".
  • by thegarbz (1787294) on Thursday July 28, 2011 @11:09PM (#36917252)

    His revelation is right on the mark. I constantly see proponents of Open Source say things such as "It's auditable because the source code is free". Well yes it is, but no one cares. I think even from the Slashdot crowd the number of people who bothered to build Firefox from source is a small minority compared to those who downloaded it. Those who actually look at the code are an even smaller subset of those bothered to build from source.

    People talk about open source as if users give a damn. Users are only interested in 2 things, how much it costs, and if it works. Open APIs are part of the ability for something to work if your idea of working is interoperability.

    • by Jah-Wren Ryel (80510) on Thursday July 28, 2011 @11:42PM (#36917452)

      "It's auditable because the source code is free". Well yes it is, but no one cares. I think even from the Slashdot crowd the number of people who bothered to build Firefox from source is a small minority compared to those who downloaded it

      That exact same logic condemns "open surface" too - the vast majority of customers don't give a damn about documented interfaces, they just want to use the product.

      It's only a very tiny minority that need to get their hands dirty. And of that tiny minority, only a minute fraction are good with "open surface" but not actual open source. The minute that one of those API's turns out to be only partially documented, or the code behind it buggy, or disabled for marketing reasons, then those same people now need real open source and not a sucker's stand-in like "open surface."

    • by izomiac (815208)
      It's fairly unusual for me to examine a project's source code, but if you're not willing to show others what you've written, then I assume it's of very poor quality. If you're worried about code theft then that indicates you have few plans to further improve your software. These are rules of thumb, obviously, but I find them to be true more often than not. How else shall I determine which program to try first? It's nonsense to pick the program with better marketing, or to wade through pages of clueless
    • His revelation is right on the mark. I constantly see proponents of Open Source say things such as "It's auditable because the source code is free". Well yes it is, but no one cares. I think even from the Slashdot crowd the number of people who bothered to build Firefox from source is a small minority compared to those who downloaded it. Those who actually look at the code are an even smaller subset of those bothered to build from source.

      It isn't necessary for a user to personally view, modify, or even compile the source to benefit from open source. At some point the copyright holder may add shovelware, spyware or just plain bugs. They may choose not to port to other platforms. They may just abandon the product. In these cases. a user of a closed source app can do little but continue to use the old version until it no longer runs on current platforms or until advanced security threats make it unsafe.

      But as long as one person has the wi

    • There's a difference between "having the right to vote", "actually voting" and "STFU I'm in charge!".
    • You are wrong, people care. Maybe most end users do not care but Azure is not marketed to end users. It will be marketed to businesses and they care very much about which technology they stake their entire existence on.

      And no you do not have to compile from source or even look at the source to get the benefit of open source. The main benefit of open source from a business point of view is that other people can look at the source. Thus, if the technology you are using is popular you can be sure that there wi

    • by devent (1627873)

      It's only the "normal" people who doesn't care. I don't care if Photoshop is open source, because if I buy it I expect it to work and the source code is of no use to me. Neither have I the time nor the knowledge to dig into the code of Photoshop.

      But there are clients that clearly benefit from the availability of source code. Like firms that have special needs and can adapt the code of the product they buy; governments that can adapt, certificate and check the code; other firms like banks that have special n

    • by gbjbaanb (229885)

      the number of people who bothered to build Firefox from source is a small minority compared to those who downloaded it.

      true, but so what. That 1 or 2 security researchers (who obviously have different ideals to the rest of us) can and will analyse the code means that you and I get the benefit of their expertise.

      Just because I don't look at the source doesn't mean there's no value in having it available. And I think you'll find that users do care that someone else has this opportunity.

    • by b4dc0d3r (1268512)

      Firefox is not a good example. Or alternatively it's the best counter-example. The source tree I looked at, in the 2.x series, had so many instances of duplication and cruft I couldn't tell what would actually run once compiled. And then most of it is written on top of XUL, so if you're a C++ expert you'll have a learning curve to understand even how to read it, how to search.

      I gave up bug-hunting because it's so terrible, and I'm close to completely not using it as a result. It may be better now, but I

  • Strawman attack (Score:4, Interesting)

    by enoz (1181117) on Thursday July 28, 2011 @11:14PM (#36917288)

    Published APIs are important, nay, necessary for "cloud" applications and services to be useful to developers to build upon. Open source is necessary for community based development of the underlying applications or services.

    Open source software is completely irrelevant in this instance and this appears to be a simple strawman attack from Microsoft against the open source movement.

  • customers don't care about the underlying platform as long as the APIs, protocols and standards for the cloud are open.

    customers don't care about the underlying platform as long as stuff works. Problem with Microsoft is that their stuff does not work and their users don't have any option for fixing it.

    And this open shit coming from MS which has history of closed APIs, protocols and standards.

    • by Anonymous Coward

      Yeah, their stuff is "not working" on 95% of personal computers, keep that up, dork, but know that not even the FOSS community likes you.

  • yeah, just wait till you use the APIs and get hit with suit for violating patents and copyrights.

    Kaching!

  • Assuming you had this open surface* then in the context of cloud computing* i'd say you probably wouldn't care what the back end platform is since you could be running on any platform that implements the open surface*, in the same way that most people don't care about what OS the server that their website is hosted on is running. Then again if you want vendor lock-in then i don't see how this would be beneficial.

    *yeah i don't think we need yet another term for an existing concept
  • This is, as stated, a ploy to dilute what open source means. To say customers don't care about the platform is silly considering how serious platform intrusions are, how serious platform reliability is and how platform updates are handled. Can Microsoft really say that Windows gets out of developers way and out of the way of maintenance admins? No, and for decades they designed Windows so it wasn't out of the way. This is just another Microsoft ploy in their attempt to Embrace, Extend, and Extinguish open
  • by Bob9113 (14996) on Friday July 29, 2011 @12:15AM (#36917636) Homepage

    > customers don't care about the underlying platform as long as the APIs, protocols and standards for the cloud are open.

    That seems true. Customers want openness in the part that they deal with. Since the customer does not deal with system maintenance and development, he does not care whether the underlying platform is open. The provider of the cloud service, on the other hand, has a deeply vested interest in the openness of the platform. Maintenance, repair, and extension requirements all strongly favor an open platform from the cloud service provider's perspective.

    Pointing out that the customer does not care about platform openness as long as the protocol is open is a bit like saying that automobile drivers do not care if the paving crew uses horse-drawn paving machines as long as they get the job done in a timely manner. It does not necessarily follow that horse-drawn paving machines make sense.

    • There you go, talking about things making sense again...

      And in a M$ thread no less.

      Get with the program!
  • by Anonymous Coward

    Microsoft has taken some big hits and seems like they are ready to play ball with the community. They are totally OK with modifications to their phone platform where Apple and Android Handset makers are not. They have reversed their stance on linux and are actively using it, contributing to it, and offering it as part of their cloud services. Is Ballmer a nutjob? Yes. Do you really think after how badly he tanked the company since Gates left he's calling the shots anymore, and will not be "stepping dow

  • by PPH (736903) on Friday July 29, 2011 @12:39AM (#36917754)

    ... any style house you can. As long as its built in our company town, we own your ass.

  • besides English, or is Microsoft marketing just losing it? Anything that's on the 'surface' in English is bad. 'Scratch the surface' and you haven't done enough. 'On the surface' means superficial, shallow. Maybe it's different in British English, but in American English this just sounds awful.

    Either that or it's April 1st again...
  • They had their chance to brand it to the point of Nintendo, Nike, and Kleenex

    and ?

  • With a majority of 'customers' moving towards open source technologies such as Linux,

    ...what? What majority is that?

    I think this guy is a bit delusional when it comes to how he perceives the broad majority of users. Most users don't give a shit if the product is open source/free software and especially what that means at the source level. They just want something that works for them. If it happens to be free, great, but it often isn't, and it more than often isn't Linux. Vendor lock is meaningless to the broad majority of users.

  • by AdamWill (604569)

    "I'm certainly more impressed by Google's willingness to let me export their data"

    Once you give it to Google, it ain't yours any more.

  • by joeaguy (884004) on Friday July 29, 2011 @02:06AM (#36918170)

    All I am going to say to those of you who think "open source" does not matter is read Richard Stallman's paper "Who Does That Server Really Serve?"

    http://www.gnu.org/philosophy/who-does-that-server-really-serve.html [gnu.org]

    Having open and honestly published API's and protocols is important and certainly better than nothing, but there are so many other reasons why access to source code is important for trust and freedom in computing.

  • I think it's a good idea to differentiate between a truely Open Source thing and something that operates with the outside world in an open way.

    For example, a standard USB port verses an ipad connector.

    Although it might dilute attitudes, it's also a useful term. You shouldn't turn ones back on that with emotion. Sure, avoid the term but recognise it's usefulness, if only to coin another term because otherwise we're blinding ourselves with emotion... and if having a holy war this is where the enemy creeps in.

  • .. to just avoid, at every conceivable cost and in any way possible, the term Open Standards.

    It hurts, doesn't it? I mean, having to work with others?

  • Would have been more fitting... Followed by Open Wallet.

  • by qxcv (2422318) on Friday July 29, 2011 @06:08AM (#36919174)
    Is it just me, or are open* and cloud* becoming the new cyber*?

    Middle manager: "Hey, did you try the new CyberCloud 2.0 Open-Surface ® operating system by OpenMicrosoft ®?"

    MS patent goon: "You bet your Zune ® I did! It allows me to innovate my value-added cyberdata to enhance availability in scalable cloud based enterprise architectures channeling an enterprise virtualisation solution, thus leveraging existing ROI and increasing key delivarables as per OpenMicrosoft ® BestPractise ®!"
    • by cheros (223479)

      That.. is brutally awesome. Can I hire you the next time I need some total bullshit?

      It beats 7 types of crap out of the Internet BS generators. Wow.

      I'm in absolute awe - I just hope you don't speak like that in daily life :-)

  • I'm fairly hard-core, so I apply [what I call] the 'walk away' test to everything. That is: 'can I take my data and set up another instance of x on infrastructure that I control?'.

    As far as I'm concerned both protocols and APIs, even if published, are potential lock in, I prefer to have the whole stack and my data -available-, even if I'm not going to move every month.

    I deal with a certain amount of non-tech-savvy non-profits in the UK who end up glued to apparently 'free' stuff, because they don't un
  • Docs aren't perfect. We all know this. To me, if there isn't a open reference implementation I'm not sure it's a open. Only an implementation covers everything required. Yes, in theory, the docs should be an implementation written in English, but that fails as it can't be run and tested, so it's always an incomplete implementation. Also, personally, I often find it easier to dig out exact details for code from other code, rather than from written English.
  • I'm certainly more impressed by Google's willingness to let me export my data.

    I hope you are aware of the fact that they can still hang on to your data, even if you leave? See their Terms of Service [google.com], chapter 11. Yes, it really says "perpetual".

    Just so you go into this with your eyes open - few read this stuff.

  • I'm getting a vision of a solid block of cement with clear windows glued on to the outside.
  • Anyone who is too naive or too young should now be able to see what those of us who know better have been trying to tell them for years: Microsoft is our enemy, not because *we* choose them as our enemy, but because they are making themselves our enemy. Thank goodness Microsoft is being marginalized; I only wish there were something better than Google taking their place, although Google's a heck of a lot better than Apple.

10 to the 12th power microphones = 1 Megaphone

Working...