Simon’s Blog

Comments

1. smoker911 wrote:

   In the evergreen Mac vs. PC debate, this is the best / funniest string I have ever seen.

   http://www.youtube.com/watch?v=yyRTBq_tyYc&mode=related&search=

   Be sure to scroll down in related videos to see the Microsoft masked warrior respond.

   Priceless…

   Posted 14 Mar 2007 at 1:55 pm ¶

2. Asam Bashir wrote:

   All I’ll say is that who ever you spoke to has obviously not been keeping up with Leopard developments for the last two years..

   Your comment @4 is laughable, since it’s common knowledge Apple is gearing up for release of it’s major OS upgrade to revision 10.5. The Mac OS 10.4.9 update is a timely release for the Tiger series.

   Comment @5 - duhhhhhh, obviously Java SE 6 developement and deployment is coupled to Leopard release.

   For knowledgable comment on Apple Enterprise developents in Leopard, see:

   http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/C85A20BA-A56B-476C-A52A-E53B85BF6CB4.html

   http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/685B09D3-950B-4B23-8B1F-A56D448F7208.html

   Posted 15 Mar 2007 at 5:05 pm ¶

3. Asam Bashir wrote:

   Java SE 6 Release 1 Developer Preview 6 (build 1.6.0-dp-b88-34) was only released for Mac OS X 10.4.5 or later.

   I can’t comment anymore about which version will be in Leopard since that would break NDA.

   http://developer.apple.com/

   Posted 15 Mar 2007 at 5:37 pm ¶

4. simon wrote:

   @2. Why would you think they haven’t been following Leopard? Leopard isn’t out yet. And no-one seems to know for sure when it will be. Enterprise customers like roadmaps with lots of details and dates on them, provided under NDA. Apple doesn’t do that, as far as I know.

   @3. I don’t know anyone that knows whether Apple has a final version of Java SE 6 ready to ship with Leopard. As far as I know, they haven’t chosen to share that information with anyone.

   Also take a look here:

   http://viewfromthemountain.typepad.com/applepeels/

   That’s a blog by someone who worked at Apple for 20 years. He doesn’t seem to be all that positive about Apple in the Enterprise…

   Posted 15 Mar 2007 at 6:08 pm ¶

5. Asam Bashir wrote:

   You too can pay and become an Apple developer and sign NDA and have all the support necessary in the preparation for Leopard…

   Considering how Vista development was put on hold and reset (including the photocopiers) following the initial announcement of Leopard plan in 2004, can you blame Apple for it’s secrecy?

   Bhaaaaa, ex-employee with a grudge…..

   Posted 15 Mar 2007 at 6:34 pm ¶

6. Jussi wrote:

   The points you list are very vague, but perhaps that was the intention.

   I trust the company’s decision has been the right one for them but I have hard time imagining what the real problems behind the vague statements have been. For example what kind of problems with buggy kernel have they had to replace *desktops*, and why aren’t those problems seen by wider audience.

   I’m not familiar with large installations of any computers so I’m a bit curious about what the features of #4 and #6 for the “Enterprise” are exactly.

   Also #5 seems quite a weird reason to do the switch *now*. December is just 4 months away and if you really need to be on the edge with Java 6, how was this not an issue a few years back when the Java support of OS X was lagging much worse than now.

   Leopard will be out this spring, and that has been stated over and over again publicly by Apple. As mr. Bashir stated, Apple provides more information about specifics of future releases to developers who pay (relatively little) for that information and sign an NDA.

   Posted 16 Mar 2007 at 6:32 am ¶

7. simon wrote:

   Jussi,

   Re: kernel bugs. I suspect these are manifesting themselves on the server side. From an Enterprise perspective, the issue is that when these bugs are reported, the Apple support team are able to supply neither fixes nor workarounds. Answers of “Yes, it’s a bug. You’ll just have to manage.” are simply not good enough if a computer vendor is serious about the Enterprise. It’s hardly what you’d call “support” is it? Companies want Enterprise-class support so that they can pick up the phone and get things working again. I don’ have any specific examples to give I’m afraid - it was just a brief conversation. But there will be desktop issues too, I suspect.

   Problems manifest themselves in Enterprise settings where they don’t show up for consumers for a couple of reasons. First, consumers tend not to deploy complex networked server architectures e.g. clustered applications running on servers linked into SANs, that are backed up. And on the dekstop, the systems are generally more stressed. In settings where hundreds of desktop computers are connected to groups of servers - any networking problems will show up. And in an Enterprise setting, much more custom software will be running than on a consumer desktop. And if this custom software stretches different parts of the OS than those typically used by most consumer apps, then, again, bugs in the OS will manifest themselves.

   There was a definite sense in the conversation I had that there was frustration with Apple’s decision to base MacOS X on FreeBSD, which is a rather messy, incremental design e.g compared to the Linux kernel which is elegantly architected. The problem with messy designs is that it makes it hard to innovate, and hard to fix bugs.

   I think point 5, about Java, was more a sense of frustration. Not a driver behind any decision to switch to Windows now, particularly. The point being that: for years, before Mac OS X, Apple was a long way behind the other OSes - because the Mac OS 9 couldn’t support the multi-threaded features of Java. And they remained behind for years after that, because they weren’t based on Intel - they didn’t put the investment in to keep up with the advanced developments with VM technology etc. After switching to Intel, though, the hope was that Apple could finally get its act together. There are no longer any good reasons to be behind other OSes. And, as Java SE 6 (Mustang) was being developed for other platforms, it seemed that Apple really wasn’t too far behind. But then, everything stopped. No more developer previews. There was a sense of “they’re really *never* going to get their act together, are they?”

   Yes, Leopard will be out, this Spring - probably very late this Spring and it may even slip into early Summer. However, even then, it will probably only be ready to deploy on consumer desktops - not for the Enterprise.

   The bottom line was that there was a clear sense that Apple isn’t as interested in the Enterprise as it is in consumer stuff like iPod, iTV and iPhone.

   That was their perception, for right or wrong. And they’re dumping MacOS X as a platform for the desktop and on the server side.

   Now, actually, I think Apple would be/is right to focus on the consumer. My point in posting this was to try to give a flavour to people who don’t know, that it’s not an easy understaking to really deliver for the Enterprise - these are demanding customers.

   Posted 16 Mar 2007 at 7:55 am ¶

8. Mr X wrote:

   Problems, to name but a few -

   Frequent file server crashes, problems with Active Directory integration ( clients locking up ), problems getting commercial VPN solutions providers to support the Mac ( Tiger broke their products ), limitations on a user being a member of only 16 groups ( hopefully fixed in 10.4.9 ), Safari not being a supported browser for third party products ( and Firefox being a bit second class on the Mac, Camino is only just version 1), lagging, and till quite recently second class Java Support ( due to the huge effort needed to port to the Mac OS X graphics system)..

   If you just have a Mac at home you’ll never had seen these problems.

   Had Apple people on site for days without success for one of our problems.

   It seems Apple released Tiger as soon as they humanly could - and it wasn’t ready for the demands of the enterprise market - though it’s a great consumer OS.

   Leopard? Tried it, loved it, client and server - we have paid developer access and BTW Apple do _not_ share time scales with developers ( unless your perhaps, Adobe or MS - and maybe not even then ) - they have a very strong policy of not commenting on future products. Leopard client and server have big improvements - but it’s not just about features, it’s about stability - it’s my worry they will release Leopard as soon as humanly possible for a consumer release - that will mean it’s full of bugs for an enterprise customer just as Tiger was. ( Note making everything 64 bit is not a small undertaking ). Ok Enterprise customers shouldn’t upgrade straight away you say - we didn’t for Tiger - but there are some key bugs still not fixed - so what’s the choice, stay on Tiger or start the cycle again with Leopard?

   Apple appear to be running as fast as they can - you only have to look at the number of releases in the last 5 years to see that - to achieve that they have had to take some risks - the consequences of those are felt in the enterprise.

   Another example the Intel transition - overall brilliantly handled - very fast, but it has broken some key enterprise security apps - something most people won’t notice.

   Obviously any enterprise system will have problems - that’s why we have Sysadmins - but it’s more unstable, for the core services, than I’d expect.

   Doesn’t mean I don’t really like what Apple are doing with their products and I think they are set to expand their market share. I think Apple engineers are great and working their arse off - if fact I’m worried they are having to work too hard!

   Also I think they will gradually improve the stability situation overtime - as I mentioned 10.4.9 has fixed some key issues but lets not fool ourselves - their focus, and quite rightly for them, is the consumer, education and creative professional markets. Just have your eyes open - use in the enterprise is a bit raw bleeding edge.

   BTW the Windows switch isn’t driven by the above problems - it’s separate. In fact some of the problems were encountered because of the planned Windows migration ( the need to integrate with AD for example ).

   I find it funny when people, with perhaps knowledge only from owning a mac and reading rumor sites, suggest that people who have direct experience with the situation don’t know what they are talking about….

   Posted 16 Mar 2007 at 2:22 pm ¶

9. Asam Bashir wrote:

   All these critisism you lay on Apple are fair but where you are unfair is in your comparision to other OSs - XP only became a useable secure’ish system after SP2 - No one in enterprise is even thinking of deploying Vista till at least SP 1 comes out, and how long will that be? At least Apple deal with their problems faster then Microsoft. Why do you expect perfection from Apple on release on day one.

   It’s obvious that Apple has been thinking a lot about the Enterprise customer for Leopard (did you see the links I gave before?), and all I will say is that wait for the release.

   http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/C85A20BA-A56B-476C-A52A-E53B85BF6CB4.html

   http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/685B09D3-950B-4B23-8B1F-A56D448F7208.html

   The real choice is Leopard or Vista, so compare them directly, what problems will you have deploying Vista over the next year, and would they be more or less then the problems you would encounter deploying Leopard?

   Posted 16 Mar 2007 at 4:36 pm ¶

10. Jussi wrote:

    Thank you very much for clarifications. This was reason I commented, I was curious about specifics not claiming that you were complaining about nothing or incapable of choosing what is the right thing for you.

    Yes, I have just one Mac at home and have not worked with bigger installations (of any OS). But I am cursory familiar with two fairly large Mac installations because I know their IT managers. I had mostly heard good things from them. Granted that their servers mostly don’t run OS X (but Solaris and Linux). Both sites have almost 100% OS X clients, with handful of Windows and Linux. They have some OS X servers and the the bigger site also has an xSAN installation.

    I acknowledge the problems Mr X (hopefully no relation to Mr T) enumerates, apart from complaining about Camino “because it’s version 1″. It is 4–5 years old project and has been good and stable browser since Jaguar days.

    I am also a bit curious about the state of VPN solutions. I use the integrated VPN client of Tiger just with one site, successfully so my sample is too small to make any conclusions but it seems really weird that Apple and commercial VPN manufactures have not fixed their compatibility problems during the 2 years (hopefully in weeks or months) of Tiger. But I take your word for it.

    Posted 16 Mar 2007 at 4:45 pm ¶

11. Mr X wrote:

    Assam - I wasn’t comparing the Mac server stability to Windows - I was comparing it to what I’d expect and that comes from ‘Unix’ systems - specifically Solaris - rock solid but then they are a server company.

    I’m not expecting perfection from day one, I’m expecting fileserver software to not have key bugs unfixed after two years!

    As I said - I’ve used Leopard and it’s not features but stability that concerns me. Also I’d characterize the new features in Leopard as aimed at the small workgroup rather than at the larger enterprise per se.

    The decision to move to windows was not mine to make and you falsely assume that it will be Vista - in fact it will be XP SP2 and again it’s other factors not stability that have influenced the decision.

    The conversation wasn’t about Mac versus Pc versus Linux or Solaris - that’s a bit of a red herring - it was about whether Apple are over-stretching - taking too long to fix key bugs and deliver key updates - think about all the stuff they are doing.

    Jussi

    Maybe your right on Camino - can’t claim to be a big expert - the bottom line is the developers didn’t feel it was ready for prime time until 1.0 - but then that rather does depend on the developers standards - which in this case I think are very high - so are mine.

    The VPN situation is thus - prior to Tiger there wasn’t really a stable kernel interface and Apple discouraged people writing software that talked to the kernel as they planned to change it - but there is some software - for example that integrates right into the network stack that has to do this - Apple changed some code in Tiger and broke the VPN solution ( obviously their own still worked! ) the vendor took ages to fix it - I suspect it ws not a priority as they didn’t have many mac customers - who knows? All I know is it took them ages.

    Posted 16 Mar 2007 at 5:39 pm ¶

12. John C. Welch wrote:

    Well, to be blunt, Apple isn’t an Enterprise company, they don’t want to be an enterprise company, nor do they care about the enterprise. So saying they’re going to lose a market they fundamentally don’t care about is a non-problem.

    Having said that, they DO care a lot about the SMB market, EDU (duh), and they do understand that they make hardware that plays well in the Enterprise.

    (Some definition: By “enterprise”, i’m speaking only of companies with more than 1000 computers per location.)

    Let’s look at the points:

    The heart of Mac OS X, the kernel, is really very buggy; probably because the core of the OS is poorly architected compared to other UNIX-type OSes such as Linux and Solaris.

    Without specifics, this is somewhat meaningless. But I will say that comparing Mac OS X to Solaris borders on silly. They have wildly different design and usage requirements. Compared to things like AIX and Solaris, Linux is a bundle of half-working spaghetti code too.

    However the Mach kernel is hardly “poorly architected” or “very, very buggy”, although I’d like to see some examples here.

    The bugs are not being fixed. Reports come back from Apple saying - “Yes”, this is a bug; and “No,” we don’t have a fix.

    What bugs? Again, you make a broad generalization with no data to support it. I can say that while not every bug I’ve submitted over the years has been fixed, the majority of them have been. If you are wanting the kind of communication you get from IBM or Sun, not going to happen, Apple’s not an enterprise company.

    With new releases of MacOS X, they have been told - the OS is OK for consumers, it’s not ready for the Enterprise.

    That would be correct. Apple is not an enterprise company. So you’re mad that they’re being honest?

    Apple is overstretching itself. They haven’t delivered anything new on the Enterprise side for a while now (funny timing of this post - today Apple have released the Mac OS X 10.4.9 megapatch) ; probably because they’re so busy getting ready to ship iTV and iPhone, and all their efforts are dedicated to that.

    Define “a while”. If you mean in the last 60 days, they haven’t shipped a lot of anything. If we abandon “enterprise” as our sole criteria, and look more at “business”, the picture gets better.

    January: The new Airport base station. Not Enterprise class, but a solid choice for SMB customers.
    Certification of 750GB modules for the Xserve RAID.

    November: New Macbooks. Those are, by the way, excellent business laptops.
    The Intel Xserves shipped with (finally), dual Power supplies, LOM/IPMI support, SAS drive support.

    October: New MacBook Pros

    Everything I listed is a solid player in corporate and SMB computing.

    After a promising start, Apple has fallen behind with its support for the Java platform. Development of Java SE 6 hasn’t moved on for months (still based on build 88). Compare that to Java SE 6 on Windows, Linux and Solaris that was released in December last year.

    Is this actually causing problems, or is it a checkbox issue?

    There are key Enterprise applications that don’t run on the Mac. They can’t run their business without some support for Windows.

    Apple is not an Enterprise company, nor have they claimed to be. Keep saying that until you understand it.

    The company has taken the decision to migrate all desktops in the company to Microsoft Windows, and to lose Apple on the server side (they will be a mixture of Windows and Solaris on the server side).

    Without the specific reasons for this move, the statement has no real meaning for the future, or even the present.

    Apple is NOT an enterprise company. They have products that are usable IN the enterprise. That is a critical distinction.

    Posted 16 Mar 2007 at 6:05 pm ¶

13. Asam Bashir wrote:

    Mr X, you’ve used a Leopard preview and you’re concerned about stability. That’s a valid point, but surely you ought to give Apple a chance and judge them starting from the release version and not a developer preview. Also, regarding the Leopard preview, Jobs has stated several times that there are key pieces of development that Apple has not even made available in those previews. I accept you’ve enountered problems with deploying Tiger in enterprise, but Apple never made claims that Tiger was an enterprise solution. Leopard is a very different cat, but still Apple won’t woe the enterprise market untill it’s ready, but what we’re sure about is that Apple is listening and making changes to attract first the SMB, but it does have the ability to enter the enterprise market when the time is right, and that time I belive will be when the decision for the enterprise customer is Vista or Leopard.

    Posted 16 Mar 2007 at 6:14 pm ¶

14. simon wrote:

    John,

    For details, I refer you to comment #8.

    I agree with you that Apple isn’t an Enterprise company. However, lots of people don’t agree. Those people believe Apple is going to grab lots of market share from Microsoft in the Enterprise. I don’t take that view; and this post was simply about one example where they’re not delivering, as far as a significant “Enterprise” customer is concerned.

    I don’t agree with you that Apple technology is robust up to the level of 1000 computers though. I suspect it starts to be “not good enough” or “not as good as the competition” at levels far below that. Perhaps an order of magnitude lower, in fact. Again, I refer you to comment #8 for the some of the details of the problems with Mac OS X.

    As for Apple not caring about the Enterprise - I can’t find anywhere on their web-site that says, “Apple doesn’t care about the Enterprise”. Do you have some special inside knowledge? Of course, Apple does seem to *say* they care about the Enterprise e.g. see,

    http://www.apple.com/services/consulting/

    Forgive me if I don’t address your comments about the design of OSes… but you know, the Linux kernel is actually rather well-designed. Of course, that wasn’t the case ten years ago. But it is the case today…

    In regard of the Java situation on Mac OS… well, you know it’s a pretty sensitive subject for those that have been developing and deploying true cross-platform apps for the last decade. Java support on Mac OS has been a source of much frustration for many people for many years.

    Perhaps the bigger question that I alluded to in this post, is - is Apple overstretching itself? Not just with the Mac OS X stuff… but with so much going on on the consumer side of their business. Time will tell on that one.

    The truth is: the schedule for iTV has already slipped. It remains to be seen how robust the iPhone OS turns out to be. They sure have quite a lot on their plate at the moment. Should be an interesting few months for Apple watchers…

    Posted 16 Mar 2007 at 7:04 pm ¶

15. John C. Welch wrote:

    Details? Where?

    Frequent file server crashes, problems with Active Directory integration ( clients locking up ), problems getting commercial VPN solutions providers to support the Mac ( Tiger broke their products ), limitations on a user being a member of only 16 groups ( hopefully fixed in 10.4.9 ), Safari not being a supported browser for third party products ( and Firefox being a bit second class on the Mac, Camino is only just version 1), lagging, and till quite recently second class Java Support ( due to the huge effort needed to port to the Mac OS X graphics system)..

    Those aren’t details. In fact, you cite two third party applications, Camino and Firefox, and have one obvious error, (The 16 group limit is gone in Mac OS X 10.4)

    I’ve had no lockup problems with my AD integration at all, and if a company hasn’t been able to update their code for tiger BY NOW, that’s not Apple’s fault.

    I agree with you that Apple isn’t an Enterprise company. However, lots of people don’t agree. Those people believe Apple is going to grab lots of market share from Microsoft in the Enterprise. I don’t take that view; and this post was simply about one example where they’re not delivering, as far as a significant “Enterprise” customer is concerned.

    People think that Apple is going to do a lot of things, usually centered around “what I want.” That doesn’t make them correct. If you’re running an IT shop, and you allow “What I want” to rule your decision making process, instead of “what is”, that’s not Apple’s fault.

    I don’t agree with you that Apple technology is robust up to the level of 1000 computers though. I suspect it starts to be “not good enough” or “not as good as the competition” at levels far below that. Perhaps an order of magnitude lower, in fact. Again, I refer you to comment #8 for the some of the details of the problems with Mac OS X.

    I would say, based on my knowledge of installations at the thousand machine level and higher, that you’re wrong. Based on some of the statements you made in comment 8, I’d also say you’ve not bothered to RTFM much.

    As for Apple not caring about the Enterprise - I can’t find anywhere on their web-site that says, “Apple doesn’t care about the Enterprise”. Do you have some special inside knowledge? Of course, Apple does seem to *say* they care about the Enterprise e.g. see,
    http://www.apple.com/services/consulting/

    They don’t focus on the Enterprise. They’ll work within the enterprise, but they are fundamentally not an Enterprise company, they aren’t designing for the enterprise, and I’d challenge you to find any official statement from Apple that says the ARE an Enterprise company.

    Forgive me if I don’t address your comments about the design of OSes… but you know, the Linux kernel is actually rather well-designed. Of course, that wasn’t the case ten years ago. But it is the case today…

    The quality or lack thereof of the Linux kernel has precisely nothing to do with the quality or lack therof of the Mac OS X kernel. That is what we would call a strawman.

    In regard of the Java situation on Mac OS… well, you know it’s a pretty sensitive subject for those that have been developing and deploying true cross-platform apps for the last decade. Java support on Mac OS has been a source of much frustration for many people for many years.

    That’s nice, but it’s not specific issues.

    Perhaps the bigger question that I alluded to in this post, is - is Apple overstretching itself? Not just with the Mac OS X stuff… but with so much going on on the consumer side of their business. Time will tell on that one.

    Well, the consumer side does make them a nice bit of change, far more than the business computing side ever has. Which would you say is the better one to service for Apple?

    The truth is: the schedule for iTV has already slipped. It remains to be seen how robust the iPhone OS turns out to be. They sure have quite a lot on their plate at the moment. Should be an interesting few months for Apple watchers…

    That’s drawing a conclusion from a coincidence. You’ve not provided a lot of real detail, and what little you have, is either incorrect, or talking about products that aren’t Apple’s for the most part.

    Posted 16 Mar 2007 at 9:39 pm ¶

16. Asam Bashir wrote:

    Cheers for the input John Whilst we wait for details of those specific ‘kernel bugs’, I was reminded of a recent article by Siracusa speculating about a change in the kernel for Leopard,

    http://arstechnica.com/staff/fatbits.ars/2006/6/19/4376

    Posted 16 Mar 2007 at 11:19 pm ¶

17. simon wrote:

    @15…

    “Based on some of the statements you made in comment 8, I’d also say you’ve not bothered to RTFM much.”

    Those statements weren’t made by me - so, perhaps you should errr.. RTFB!

    In fact, it’s clear you mis-RTFB in a number of areas. I’ll let everyone draw their own conclusions from that.

    Now…

    “I’ve had no lockup problems with my AD integration at all,”

    You know, just because *you* haven’t seen a particular problem doesn’t mean it doesn’t exist. I’d have thought someone that claims as much IT experience as you do would realise that.

    “and if a company hasn’t been able to update their code for tiger BY NOW, that’s not Apple’s fault.”

    Eh? If a software vendor hasn’t updated their code for Tiger “BY NOW”, it’s rather unlikely to because they *can’t*. It’s much more likely to be because they don’t *want* to; presumably because they don’t believe they would derive any commercial benefit from supporting Mac OS X.

    “The quality or lack thereof of the Linux kernel…That is what we would call a strawman.”

    Well, you brought it up in your comment #12. Did you forget?

    “They don’t focus on the Enterprise. They’ll work within the enterprise, but they are fundamentally not an Enterprise company, they aren’t designing for the enterprise”

    Yes, I quite agree. Apple aren’t focussed on the Enterprise. They fundamentally aren’t an Enterprise company. And, more than that, I don’t see much evidence that they want to be.

    So let’s be clear - does the above mean we *agree* that Mac OS X is unlikely to make significant in-roads into the Enterprise? Or do you believe (as Asam does) that, despite Apple not being focussed on the Enterprise, there will be a veritable flood of companies dumping Microsoft for Apple?

    As for your and Asam’s requests for *more* details, perhaps Mr X might be able to provide you with some of these. I have to wonder though, given the way you’ve constructed your arguments, whether *any* level detail will be enough for you…

    Posted 17 Mar 2007 at 8:27 am ¶

18. Asam Bashir wrote:

    I was making some confusions between enterprise and SMB, and also EDU has to be considered a seperate market. So my position is that Apple is interested in increasing SMB and has included specific services in Leopard to attract that market. These services are also needed by the enterprise market, but Apple isn’t interested in developing for that market, nor is it likely to provide enterprise level support. BUT - I still hold on to the position that Apple could enter the enterprice market, since it does produce hardware and software that will be of interest to enterprise. I think with some support mechanisms and with virtualization technology it becomes easier for enterprise to deploy on Mac hardware, but of course, Apple would not want to get involved in supporting other OSs running as VMs. Maybe, there is room here for a third-party to step in and provide the support for enterprise using Apple hardware and software, but then, the profit margins are too small. As far as I can judge, enterprise is very interested in Apple hardware, and also very interested in having non-microsoft solutions.

    I can understand Apples reasons for not wanting to support enterprise, but there is a huge gap in the market here for someone to fill, and apart from Apple there doesn’t seem to be any other contenders.

    Posted 17 Mar 2007 at 3:39 pm ¶

19. John C. Welch wrote:

    You know, just because *you* haven’t seen a particular problem doesn’t mean it doesn’t exist. I’d have thought someone that claims as much IT experience as you do would realise that.

    I can point at some rather large setups that haven’t had problems either. The only inherent problem with directories and Macs is not Active Directory specific. If you have laptops, then you will have issues, due to limitations in lookupd. The fix for it, unfortunately, isn’t happening in the Mac OS X 10.4 timeframe. There are workarounds that are reliable and relatively painless.

    But that’s not an AD or even a DS issue, that’s a lookupd issue, and is easily dealt with.

    Eh? If a software vendor hasn’t updated their code for Tiger “BY NOW”, it’s rather unlikely to because they *can’t*. It’s much more likely to be because they don’t *want* to; presumably because they don’t believe they would derive any commercial benefit from supporting Mac OS X.

    That’s still not Apple’s fault. Apple has over 20 million active users. If someone doesn’t think they can make money in that pond, then that’s a personal choice.

    “The quality or lack thereof of the Linux kernel…That is what we would call a strawman.”
    Well, you brought it up in your comment #12. Did you forget?

    In response to a (still) unsupported statement that the Mac OS X kernel is “very buggy”.

    So let’s be clear - does the above mean we *agree* that Mac OS X is unlikely to make significant in-roads into the Enterprise? Or do you believe (as Asam does) that, despite Apple not being focussed on the Enterprise, there will be a veritable flood of companies dumping Microsoft for Apple?

    It really depends on what you mean by “significant”. Do I see Apple taking over the Enterprise? No. It’s not something they’re interested in, or currently capable of doing. But that’s not new, that’s status quo. Do I see Apple as having a solid place within the Enterprise? Sure, they have a lot of products and features that fit well into any enterprise setup. So it really depends on what you mean by significant.

    I DO however see them doing really well in the SMB market, and doing well at the expense of Microsoft.

    Posted 17 Mar 2007 at 8:35 pm ¶

20. simon wrote:

    “In response to a (still) unsupported statement that the Mac OS X kernel is “very buggy”.”

    Yes - I was reporting what was said to me, by someone who is pretty knowledgeable about Mac OS X; and more than that, by someone that knows *a lot* about software engineering. I didn’t ask them to enumerate a list of kernel bugs. So, I don’t have any more information to give you I’m afraid.

    However, it’s not like this is the first time anyone has ever suggested that the Mac OS X kernel might not be as well put together as other OSes, is it?

    For example,
    
    From the technical perspective, OS X security is rather poor, at least when it comes to kernel-land code…. obviously when you take code from many different places and stick it together, it’s prone to problems.
    [From:
    http://www.macrumors.com/2006/12/01/month-of-kernel-bugs-ends-first-adware-for-mac-os-x/ ]

    or CodeWeavers saying,
    
    We gets a lot of help from Apple, they listen to us, but we don’t get a lot of fixes. Their (Mac OS OX) kernel is pretty buggy…
    [From: http://www.winehq.com/?issue=320 ]

    You said,

    “I DO however see them doing really well in the SMB market, and doing well at the expense of Microsoft.”

    OK. Thank you for this. That is interesting. Can you quantify that? We all have different ideas of what success is - what’s your definition of “doing really well” here? That is, how much market share do you estimate Apple will take from Microsoft in the SMB market, and over what time-frame? What do you think the Mac OS X %age share of the SMB desktop will be at the end of 2007, and the end of 2008?

    Posted 18 Mar 2007 at 8:05 am ¶

21. Mr X wrote:

    Perhaps Mr Welch will believe Apples own release notes for 10.4.9 client and server

    http://docs.info.apple.com/article.html?artnum=304821
    http://www.apple.com/support/downloads/macosxserver1049updateppc.html

    Note the top three on the server release notes

    - avoiding AFP connection errors when authenticating Active Directory users who are members of lots of groups
    - preventing SMB/CIFS server deadlocks and service interruptions
    - authenticating Windows clients when bound to Active Directory

    Also in the client release notes:
    - Resolves an issue when using kerberos authentication with Active Directory if the user is a member of many groups.

    Perhaps you think Apple releases all these updates for fun and the bugs they try and fix have never affected anyone?

    And just because it makes it into a release not doesn’t mean it is actually *fixed* - for example
    byte range locking - introduced in 10.2
    ( http://developer.apple.com/technotes/tn2002/tn2053.html )
    A search on search.info.apple.com
    for “byte range locking”
    shows it’s been mentioned in release notes for
    10.4.3
    10.4.6
    10.4.7
    10.4.8

    There are improvements in 10.4.9 as well - though I’m still not sure it’s fixed - the symptom is a contended document on a file server gets into a permanently locked state - believe me it’s a real problem - obviously you may choose not to believe me - *shrug*

    I don’t think there is anyone here who thinks Apple doesn’t do some great stuff - but not everything they produce is insanely great - it’s takes a long time do develop and debug enterprise class systems for things like fileserving - it’s extremely complex to get the right balance of performance and stability ( look what happened to Microsoft’s attempt at a new file system - 10 years and who knows how many programmers failed ).

    The original thrust of the blog as I saw it was - with all that’s going on at Apple will the resources to work on stability get even more squeezed?

    Ultimately success with the iPhone may be good for Mac OS X as Apple will be able to afford more developers - right now it looks stretched.

    Posted 18 Mar 2007 at 3:29 pm ¶

22. Asam Bashir wrote:

    Simon, I simply don’t agree with you that Mac OS X kernel is not as well put together as others, they have fundamentally different objectives and Mach provides XNU with a competitive advantage in its unique design.

    http://www.roughlydrafted.com/0506.linuxmyth4.4.html

    Most of this attempt at rubbishing Mach stems from an article written by Cringely which has been proved wrong and misleading,

    http://weblog.infoworld.com/enterprisemac/archives/2006/05/os_x_leopard_he.html

    This was all happening at a time when Apple had not released open source code for x86 and the community wrongly assumed the reasons behind this. The only reason Apple delayed releasing the code was because it need to stip off some proprioty code relating to the Rosetta translation layer.

    The Mac OS X kernel is a rapidly evolving beast that saw specific changes in Tiger which provided oficially supported kernel programming interfaces (KPIs).

    http://arstechnica.com/reviews/os/macosx-10.4.ars/4

    What we still don’t know from Apple how it will deal with hypervisor technology, this maybe part of the top secret plans in Leopard. It’s just not like Apple or Jobs to say there are ultra-top secret features in Leopard, then pull out something lame, we expect something huge from Apple.

    Bootcamp and the ability to dual boot is already proving to be a valuable tool in the desession making process for large IT deployments, this recent case example of a US univeristy moving from 1,700 mixed computers (Mac + PC) to 1,450 Macs only, and saving £150,000 directly,

    http://www.macworld.co.uk/mac/news/index.cfm?newsid=17512

    It’s not only me that thinks Macs are enterprice-worthy, just check out this recent Computerworld article,

    http://www.computerworld.com/action/article.do?command=viewArticleBasic&articleId=9012644&source=rss_topic123

    Posted 18 Mar 2007 at 5:27 pm ¶

23. simon wrote:

    Asam,

    I don’t have particularly strong opinions on the Mac OS X kernel. I was relaying something that someone said to me.

    Re: Macs in the Enterprise - I think we’ve done this one to death now. It’s now time to wait and see if there is, or isn’t, a flood of large corporations switching from Windows to Mac OS X.

    I am looking forward, though, to seeing John Welch’s predictions for Mac OS X future market share gains among SMBs. Do you have some predictions there - along the lines of the ones I asked John for?

    Posted 18 Mar 2007 at 6:46 pm ¶

24. Asam Bashir wrote:

    I’m sorry Simon, if you choose to dis Mach and XNU from here say, without researching the facts yourself, then I will continue to point out where you are misleading the wider audience here.

    I’m not falling into any trap by even attempting to quote numbers about market share gains predicited for the future. What I will say is that the IT mindset is talking about an aggresive attempt by Apple to play the SMB market and it’s not Apple talking about the quality of its hardware and software in an enterprise environment, it’s the rest of the industry.

    I’ll leave you to put a figure on how much that mindshare is worth in real-terms…..

    Posted 18 Mar 2007 at 7:28 pm ¶

25. Asam Bashir wrote:

    Will you now withdraw this statement and clearly state it is misleading and inacurate:

    1. The heart of Mac OS X, the kernel, is really very buggy; probably because the core of the OS is poorly architected compared to other UNIX-type OSes such as Linux and Solaris.

    Posted 18 Mar 2007 at 7:43 pm ¶

26. simon wrote:

    Asam - the kernel issue is simply the opinion of one person (not mine - I don’t have an opinion on this topic). There is nothing to withdraw - read what I wrote - it wasn’t written as a “statement of fact” - it was reporting on one individual’s opinion.

    So… You have your opinion. He has his. You’re both entitled to them. However, Asam, I have to say that this person knows a *lot* about Mac OS X. It was because of that, I thought his opinion worthy of mention in the first place.

    You’ve been free to express your opinions and make your arguments on this topic. And John Welch has supported you. Mr X has posted some opposing points of view.

    In the discussions, I find that Mr X has been long on relevant detail - contrary to John’s and your assertions. And, I have to say that I found both you and John less convincing in your arguments - both of you simply saying that problems with Mac OS X don’t exist… even in cases where said problems have been documented by Apple.

    People will take from all that what they will from. Just as they will take what they will from the fact that you regard a simple request to quantify what “success” looks like as some kind of “trap”.

    Don’t assume you lost the argument, by the way. You and John may well have convinced everyone that reads the blog. I fully accept that Mr X is less credible than he might otherwise be because he is posting anonymously. That is his choice though. As I say - people will take from this discussion what they will…

    Posted 18 Mar 2007 at 8:37 pm ¶

27. Asam Bashir wrote:

    I don’t think I’ve ever said that Mac OS X does not have its own problems, indeed part of my participation in this discussion is to focus on where those problems are and what is being done by Apple to address them. In this regard I’ve listened to Mr X state the nature of some problems and also state that at least some of them have been addressed in the 10.4.9 update, that to me at least sound like Apple listening and fixing in a timely manner (not overstreaching itself at all). I’ve also listened to John and understood that the problems were not as big as they seemed and indeed work-rounds where available prior to the software update.

    If the kernel at the heart does have problems then it is right to bring about awareness and discussion so that Apple listens and fixes those problems. However it seems to me that Mach and XNU have certain features that will give Apple a competitive advantage with room to accomodate both multiple processing units and hypervisor technology that is ahead of the curve for application of Intel virtualization technology.

    I find this whole discussion enlightening, in the past the arguments against Apple ranged from, oh they use slow processors, oh the OS isn’t very modern or forward thinking, to now when the last resort seems to be, oh, the kernel isn’t as well constructed as Linux.

    We’ll have to wait to see if John has any more valuable insights, but I will certainly say that Apple will sell more desktops and laptops to SMB and enterprise and that would easily be quantified as success in my books..

    Posted 18 Mar 2007 at 9:45 pm ¶

28. Asam Bashir wrote:

    Well ok, if you want figures then here’s an easy one,

    “Based on Gartner’s estimated sales forecasts for 2006-2010, Apple’s percentage of the overall installed base will jump by 50%.”

    And some interesting articles about “market share” and what it really means,

    http://slashdot.org/articles/07/03/18/1314230.shtml

    Posted 18 Mar 2007 at 10:20 pm ¶

29. simon wrote:

    OK - so I’ll buy those numbers (maybe). It might be optimistic, but Leopard looks like a strong product, so it’s not unreasonable to be optimistic.

    Based on Gartner’s estimates, then, I think it would follow that 1% or possibly 2% of small businesses would switch from Windows to Mac OS X by 2010.

    Now - this might represent significant revenue and profit growth for Apple, depending on how its mobile, music and video businesses perform over the next 5 years. But equally, it might not.

    However, one thing is for sure: this kind of growth would make a negligible difference to Microsoft as far as Windows is concerned. In my book, that’s not even close to “doing well at the expense of Microsoft”…

    Posted 19 Mar 2007 at 8:07 am ¶

30. Asam Bashir wrote:

    That’s the ‘trap’ I was talking about, market shares and what they really mean. For example, in the US market share of installed base is 10%, by 2010 that would at minimum be 20%, but SMB growth might be even higher.

    Did you see those articles,

    http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/FFE4A8E2-9816-4344-9FB0-61BED246674C.html

    http://www.roughlydrafted.com/RD/RDM.Tech.Q1.07/9E601E8E-2ACC-4866-A91B-3371D1688E00.html

    I found them informative,

    “In the final quarter of 2007, Apple earned $7.1 billion in revenue, compared to Microsoft’s $12.5 billion in total revenue. Yes, that’s right, Apple brought in more than half as much money as Microsoft, despite Windows owning 98% of the PC market.

    Even stripping Apple of its iPod revenues, which PC pundits love to do, the company still earned $4.4 billion on its Macintosh business, over a third as much Microsoft brought in from its entire Windows, Office, and server operations combined. Apple’s 2% of the PC market doesn’t seem so small anymore.”

    So, a very strong Apple, colossal failure of Zune and Xbox 360, now I wonder if Microsoft is feeling the pressure….

    Posted 19 Mar 2007 at 3:22 pm ¶

31. Asam Bashir wrote:

    And that’s not even considering the halo effect of the iPhone…..

    Posted 19 Mar 2007 at 3:26 pm ¶

32. Asam Bashir wrote:

    or Apple TV….

    Posted 19 Mar 2007 at 3:27 pm ¶

33. simon wrote:

    Asam, I think you need to read these articles a bit more critically.

    Any sentence that begins, “In the final quarter of 2007″ should make you suspect a few things… You do *know* we’re in the still in the first half of 2007, I take it?

    Anyways, the point I want to make here is that revenues are not the same thing as profits. If you want to look at companies like Apple and Microsoft, you really need to look at them in terms of the their profits, rather than their revenues.

    So, if we look at 2006, then we see that Apple made profits of $2B (on revenues of $19B); compared to Microsoft that made profits of $12B (on revenues of $44). So, Microsoft made more than six times as much money as Apple did in 2006. And, critically, that’s even with huge amount of money Microsoft is spending (some would say wasting) on super-expensive, loss-making ventures. I don’t know the exact numbers, but they probably spend about $20B-$30B on these loss-making activities each year (no idea how they manage to do that, but they do!).

    The bottom line, then is that it’s Microsoft’s 98% market share of the OS market that enables them to spend tens of billions a year on R&D and speculative business ventures, and even after that still make $12B a year in profit.

    So, yes, I did read the articles you posted about market share. I thought they were badly flawed in a number of places, and showed a lack of understanding of even the basics of how financials “work”… which brings us back to the Enterprise stuff…

    If you look at the financials of Apple’s business, you can clearly see that it’s the consumer stuff where they’re going to make almost all their money in the future… which is why that’s what they’re focussing on… and probably why they changed their name from “Apple Computer” to “Apple”.

    Posted 19 Mar 2007 at 4:01 pm ¶

34. Asam Bashir wrote:

    I did notice that, thought it might be something to do with when a financial tax year ends, like in UK it’s April, so that would be the last Q of that financial year - then again it might have been a typo and meant last Q of 2006…

    Of-course, I don’t have the time to go through each and every statisic quoted in that article, but I found it informative in how market shares are calculated by the rest of the industry…

    Posted 19 Mar 2007 at 4:29 pm ¶

35. Asam Bashir wrote:

    Was gonna ask the author, but it’s already been mentioned,

    Is this correct: “In the final quarter of 2007, Apple earned $7.1 billion…”? Shouldn’t that be for 2006, especially since Apple’s financial year ends 9/30?

    [Yes that's an error: Q1 2007 = final quarter of 2006 - Dan]
    R Brown | 03.18.07 - 1:11 pm | #

    Posted 19 Mar 2007 at 4:59 pm ¶

Post a Comment