Now you can buy SuSE Linux from Microsoft: am I dreaming?

The Microsoft - Novell saga continues Maybe Microsoft will stop selling Operating Systems??? I don't know if you found it a shock announcement, but it certainly confused me. Released on 2nd November is the announcement that Novell and Microsoft have agreed a set of broad business and technical collaboration agreements that will help their customers realize unprecedented choice and flexibility through improved interoperability and manageability between Windows and Linux. There you have it. There is even a picture of Novell Inc. President and CEO Ronald W. Hovsepian and Microsoft CEO Steve Ballmer shaking hands an all smiles. The agreement basically means that you can ask your Microsoft salesman to quote for X copies of SuSE under a reseller agreement. What is happening? Best of enemies / best of friends? There is a saying that you should keep your friends close to you... and your enemies even closer. Well, Microsoft and Novell certainly come into that category. Don't forget that Novell could be credited with creating one of the first practical file / print server, a workstation authentication mechanism, a small systems directory, a ... until these were trumped by Windows NT, and things like Active Directory. Even today, there are people who prefer Netware to AD and think that we should all be running token ring instead of NetBIOS (or whatever it is now called. Don't forget Word Perfect, one of the really good early word processors (under MS-DOS), until Word began to become the de-facto standard. I think the case has been made - Microsoft and Novell have a history of competition. So what does Microsoft do with competition? They either kill them off, or buy them out. So has Steve Ballmer gone soft? Some of the small print... On the technical side, the two companies will set up a facility where engineers will work on enabling co-location of Windows and Linux, using virtualisation technology. In addition, there will be common standards on web services management, interoperability between AD and the Novell Directory, and translators for MS Office XML file format and the Linux OpenDocument format. But examine some of the detail, and you begin to see what is happening. Firstly, the decision to sell SuSE via Microsoft is just a concession. I doubt if MS salesmen will be given priority commission rates on Linux sales. And Mr. Ballmer himself has been quoted as saying " If you've got a new application that you want to instance, I'm going to tell you the right answer is Windows, Windows, Windows." Pretty conclusive. So it appears that Novell will get precious little sales from this agreement. What they do get is time. They have effectively bought off the giant by feeding him some scraps. In return for a percentage of the revenue from every SuSE license they sell, Microsoft has dropped any intellectual property legal actions they may have against Linux users. So Microsoft gets some cash from SuSE sales (whether they contributed to the sale or not, they still get the money), and Novell gets some breathing space. I think Microsoft is the winner here. Novell needs to take advantage of this calm before the storm. They have until 2012 (when the agreement may run out) to build sufficient impetus to be able to stand on their own two feet. Of course, that assumes that the agreement goes full term. Some companies have been known to exit long-term agreements ahead of time...

Does SOA spell the end of SAP

Is this the end of the SAP Consultant's gravy train?. Want a job as a SAP Consultant? SAP Consultants have always had an interesting life. The challenge centres around the fact that you cannot ever have a generic solution to a specific set of customer requirements. It is all very well persuading people in the Infrastructure world that a generic solution like Tivoli or Unicenter or Patrol will address their needs. Crudely speaking, Infrastructure is Infrastructure. But when you get into Enterprise Resource Planning you are touching key differentiators. Each company's balance sheet and profit and loss accounting is different. Touch that, and you must have a specific solution tailored to each customer. So what does this have to do with SOA? Everything. In the past, this customization has been done by highly skilled (and highly paid) consultants who would crawl all over the organization and require significant amounts business analysis before they could work their magic in customizing SAP. As a result, deployments of SAP could sometimes be measured in years and months, rather than the months and days which other applications may have required. I am not suggesting that there was anything wrong with this. It was simply a by-product of having to customize each deployment of SAP to address the specific business needs of the customer. But now things may begin to change. SOA - the next generation But re-writing SAP to conform to a Service Oriented Architecture (SOA) could potentially (and I use the work "potentially" deliberately) solve this challenge and reduce the time to deployment of SAP solutions. It might also reduce the complexity and deployment challenges, so that SAP could deploy upgrades and enhancements quicker. The SAP approach employs NetWeaver and Web Services, so that all SAP software will be SOA-enabled by 2007 This should be good news for Systems Integrators or Customers. In theory, with a common services interface, it should be easier to customize SAP, and/or integrate it with other Office or Enterprise software products. After all, there are many developers these days that know about Web Services. The recent announcement of SPA Discovery will also help smooth this transition. SAP could become part of a so-called "composite application" containing Business and Financial workflows. According to one Research Group, SAP's decision to use a model-based approach will make it easier to tailor the application to fit the business, not vice-versa. In short, SAP's pragmatic approach may reduce time to deployment and thereby increase the attractiveness of SAP to small and medium size businesses. To do this, SAP may well have to look at it's pricing model. SOA Footnote On a general note, there is also some disillusionment arising about the use of SOA these days. Even David Chappell is quoted as saying that software re-use (one of the key justifications for introducing a SOA) has failed because of the cultural and business barriers. So, for many organizations like SAP, the key benefit of SOA-enabling their software may be to open it up to the customers. This may mean that the age of the specialized SAP customization consultant may be numbered. On the other hand, I seem to remember them saying a similar thing when COBOL was introduced...

Vista delays could open up Linux on the desktop

Will corporate users finally lose patience?. One more delay too much? The news has broken (not unexpectedly), that the latest copy of Windows - now to be called "Vista" has been delayed, at least until the start of 2007. Will this be a delay that would finally encourage people to move to Linux or other desktop technologies? If such a revolution was going to happen, now is probably one of the better times for it. After all, Linux is (almost) able to be deployed by a non-technical user, and it is backed up by OpenOffice, Thunderbird, FireFox, Samba and other compatible products that enable it to co-exist in the Microsoft world. Will people jump ship following yet another Microsoft delay? Reasons for the revolution It's not just the delays that users are upset about, it's the lack of functional benefits that they will get when / if they deploy Vista. Firstly, what is there in Vista for the average commercial user? The new File System, one of the new features which were touted some time ago, will not make it at all. Basically, it is just a new interface, which is a bit closer to the Apple "glass" interface, with the ability to have semi-transparent "3-D" appearances. Neat, but not exactly revolutionary. And not much, if people are being encouraged to move away from Windows XP. Of course, there is a new programming interface, and the wide use of XML to configure applications and define the interface. Useful for programmers CVs, but not exactly ground-breaking. In fact, when all said and done, there is not a lot of stuff in Vista which the average user will start writing home about. And that is the problem. What will users get for their money? In short, why should they abandon XP? The only real reason is that, one day, Microsoft will stop supporting it, just as they no longer support Windows NT workstations. So, to coin a phrase from Star Wars, "Fear will keep the local systems in check". Will it happen? My view is that the take-up of Vista by Corporates (who, let's face it, are the people who pay Microsoft the big bucks) is likely to be slow. Corporate customers want to see a return on their investment, and a new glossy front-end does not do it for them. There is also the fear of security holes. Just as XP has been locked-down by Service Packs and other patches, the last thing that corporates want is another string of security risks with a new architecture. In fact, Microsoft might get a better take-up if they promoted Vista as just a new shell to XP (just as Windows 2000 was promoted as "Built on NT Technology"). In short, I predict a longer future for XP. Corporates will wait and see. If there are security scares, or performance issues, then you could see an exit from Windows on the Desktop. Who knows?

Will Intel-inside-Apple become a corporate standard?

It will be software packages and interoperability that carry the corporation. Apple Inside? The decision by Steve Jobs to ditch Power chips for Intel might have made sense if he had taken it a year ago, but somehow I think he may have missed the boat. Intel, on the other hand, have found another outlet for their entry- and medium- level chips and given a sharp jolt to the anti-Intel camp (which, from what I have seen, appears to be growing daily, with the rise of AMD). So why did Apple suddenly decide to change camps? It's puzzling in some respects. There was an argument that the Intel 32-bit architecture with multiple core chips had a lot more power than Power. Certainly the new Powerbook G4 with Intel Inside has been reported to have better performance. But was the Power so bad? Not really, since you will notice that the battery consumption has dropped off with the new chip. Swings and roundabouts with any architectural design. Winners and Loosers Of course, there are downsides. One of these, which has barely been hinted at, is battery life. The Apple notebooks had a deserved reputation for long battery life. I know of one person who claims to regularly get 5 (five!) hours life from his G4 Powerbook. Not any more... The Intel chips requires lots of juice. So battery life will, like as not, be down in the medium of PC-type notebook standards. Can't have everything, I'm afraid. So what of the future? Some analysts have said that Apple's move to Intel technology is the beginning of a process towards opening the Apple MACOS X operating system to other hardware. How about purchasing a Toshiba or Compaq, and having MACOS X loaded instead of Windows XP? Sounds very tempting. After all, I can run Microsoft Office on MACOS, can still use email, can be authenticated with an LDAP environment, share Folders using Samba. Sounds promising to me. But will it happen? I don't think so. And the reason is to do with Drivers. One of the things that makes Windows XP so pervasive, but which also can lead to instability, is the fact that it works with pretty much any hardware technology you care to name, Dell, HP, IBM, Toshiba, Loveno, Tiny, Sony, ... the list goes on. In order to do this, Microsoft have had to invest in (or persuade vendors to create) device drivers that will work with these technologies. But therein lies the problem. The more drivers, the more complexity, the more likelihood that they will not easily co-exist. What happens when a NIC from one manufacturer has an IRQ conflict with a 17-inch display driver from another manufacturer? Most times, nothing. But the introduction of signed device drivers in Windows 2000 Server was one indication of the extent of the potential problem, at least in Microsoft's mind. Apple, on the other hand, does not have these sort of problems. They have one set of hardware, and that is all the MAC operating system has to work with. Any problems, Apple make both the hardware, firmware and the software. Co-existance is easy. If Apple introduced new hardware support, they would fall foul of the device driver issues that have bedeviled Microsoft these many years (ever since Compaq cloned the IBM PC BIOS, in fact). I don't think Apple want to go there. Whatever it's faults, an Apple is still a single-supplier solution. Incompatibility problems simply don't exist. Long may it continue