Hi, I'm Riku Seppälä and this is page is my central presence on the web. I come from Finland and live in Montreal. I'm A Business Strategist And Corporate Information Systems Developer. Currently I'm Working on My Masters Thesis In Business Strategy. I'm A Firm Believer Of The Clan Of Getting Things Done. On my page you can find links to my CV and the projects I'm working on at the moment.

What I'm Thinking - My Twitter Feed

    follow me on Twitter

    Wednesday 29 July 2009

    The Shift to Utility Computing / Cloud Doesn't Erode Competitive Advantage From IT

    I sometimes see that researchers and technology analysts write that it is not possible to create competitive advantage through Information Technology. The reasoning is that as computing is becoming a utility and applications are moving to the cloud, IT becomes more of a competitive necessity that something that provides differentiation.

    The analogy is often made to electricity;
    1. first electricity was produced locally. In the same way as companies have their own data centers for computing today.
    2. electricity then shifted to grids, and was provided as a service or "utility".

    As a utility, electricity didn't provide any competitive advantage.

    Computing is now becoming a utility just as electricity did.

    But what about machines that were run on electricity?

    Companies that used machines and integrations of machines that supported their specific processes were able to create competitve advantage even after the shift to electricity as a service or "cloud" electricity.

    In the same way, companies that can use applications and integrate them in ways that support their specific processes will be able to create competitive advantage.

    It's just the infrastructure that is not in-house anymore, not the way we use it.

    Monday 13 July 2009

    Apple Time Capsule (TC) Setup Problem Amber Light, Stuck in Restart or just Flashing Amber Light, Cannot Connect through the Internet WAN port

    I just go a new Apple Time Capsule, and I had some trouble setting it up. I also couldn't find any solutions on the web, so I decided to write a short post of how I solved the issues.

    My setup is a 2008 new MacBook and the Time Capsule which came with firmware 7.4.1. I have a functioning ethernet connection which I want to turn wireless with my Time Capsule.

    When I started, Airport Utility only recognized the Time Capsule once when setting up. When the Time Capsule was supposed to restart, Airport Utility got stuck. After that, even by plugging off the power from and replugging the power the Airport Utility wasn't able to find the Time Capsule. The status light was just amber at this point, stuck in "completing its startup sequence"... Yeah, right.

    Well, I'll now describe how you can fix this problem if you find yourself in the same situation.

    I only had the functioning ethernet connection running when starting the setup, so I wasn't going to extend and already functioning wireless network, but I think it would be possible as well.

    Of course, you will need the functioning ethernet connection all the time (or an already setup wireless which you will be extending with the Time Capsule.

    First, make sure that your Airport utility is updated with the newest version (Airport Utility - Check for Updates...)

    1. Downloading the newest firmware for Time Capsule (7.4.2 for me)


    I noticed there had been problems with the firmware 7.4.1 for Time Capsule. Many people have documented that the 7.3.x versions work better, but I still decided to try the 7.4.2. It worked, but the 7.4.1 might not.

    You can download the newest firmware without connecting to the Time Capsule (which obviously doesn't work...). While on your working internet connection, go to Airport Utility (found from spotlight or applications/utilities), choose the 'File' menu and choose 'Check for updates' while holding the 'Option' (Alt on a MacBook) key. This will pop up a menu where you can choose to download new firmware for any devices. Choose the newest firmware or any you think might work if your current doesn't. (The current firmware of the TC can be found on the opening screen when your computer finds the TC after a reset of the TC... See below)

    2. Checking and Documenting your Network Values


    Make sure your internet connection works with the Ethernet cable that you're plugging into the Time Capsule Internet WAN port (or the wireless connection that you want to extend). When that connection works, check and write down the values for the IP-address, Subnet mask, Router Address and DNS servers from your Network Preferences under the 'Ethernet' Tab.

    3. Connect to your Time Capsule through Airport Utility by a full reset on the TC and Update the Firmware.

    So because you'll probably be in the situation where your Airport Utility cannot find the Time Capsule, you'll first have to reset and connect to it:

    1. Connect the ethernet cable from one of your Time Machine Ethernet ports to your computer.
    2. Do a complete reset of the Time Machine by:
    1. dsconnecting the power from your Time Capsule.
    2.With a pin or something, hold the reset button while plugging in the power again.
    Hold the reset button until the Status Light on the Time Capsule starts flashing fast
    (amber).
    3. It then takes about 1 minute for the Time Capsule to start up and you'll see it in your Airport
    utilities.
    4. When you see the Time Capsule in Airport Utilities, choose 'manual setup'.
    5. In the manual setup, in the tab 'summary', click on 'version', and select to upload the version number you downloaded in step 2.

    4. Resetting and reconnecting the Time Capsule with Manual IP Settings.


    When Airport Utility shows the information about the Time Capsule, don't choose 'Continue', or you'll be in trouble. Go for 'Manual Setup'. (This is the screen where you can see the firmware version of your Time Capsule.)

    In the manual setup screen, choose 'Internet' from the options at the top of the window. Now, choose 'Setup manually' for the IPv4 settings. Fill in the IP-address, Subnet mask, Router address and DNS Servers that you wrote down in step 2. (When checking your Ethernet connection)

    (Every time you update your changes in Airport Utility, Airport Utility will ask you if it's ok to restart the Time Capsule and lose the connection for a moment etc. Press 'ok' or continue or whatever. )

    So if you now update the TC from Airport Utility by pressing 'Update', you will be warned that you will lose the connection for a while. Continue and hopefully your Time Capsule will now restart. After the TC has restarted (you can listen to the process on the Time Capsule), you'll hopefully see your Time Capsule in the Airport utility. This time the Time Capsule should show a Green Light. Now you can start feeling a bit better.

    5. Change from Manual IPv4 settings to DHCP


    The last step is to change back to DHCP and perform the other settings such as names, passwords etc. This time, when the Airport Utility has found your Time Capsule with the green light after step nr 4., you can choose the 'continue' option instead of 'Manual Setup' as before, and continue with setting up all the things you need to. You can setup the names and passwords etc. Now, choose "DHCP" instead of manual setup of IPv4.

    Follow the steps and let the Time Capsule Restart again... Now you will hopefully have a fully functional wireless network with the settings you chose.

    I hope that worked for you. When I went through this process, it was iterative and took several hours. I had to reset the TC at least 6 times with different updates to get it working. If this didn't work for you, don't lose hope:

    Extra:

    6. If it still doesn't work, go and setup the IPv4 Manually from the 'Manual Setup' again as in Step 3.

    Let the Time Machine restart again with manual IPv4 settings so you get the green light. Now choose 'Manual Setup' again, go to internet and choose 'Using DHCP' for 'Configure IPv4'. You'll have to update again and let the Time Machine restart.

    I hope it works for you. If you're having trouble with these steps feel free to comment and I'll try to guide you better.

    Shame on Apple.

    Friday 10 July 2009

    The Chrome OS and a glance of economics behind cloud computing

    Someone finally got it. Quite a wonder actually that it was Google. I will just point out two facts about the good ideas behind the Chrome OS. In addition, I'll discuss the most important two things that will drive the move towards SaaS in the enterprise. (It might not be what you think)

    Firstly, which two annoyances does the Chrome OS promise to help us with?

    1. The most annoying thing about a laptop: Slow Booting.

    Chrome OS just boots to a browser. It's fast. It skips most memory handling events which make Vista the slowest booter ever.

    2. Not being able to smoothly collaborate and access your data from a distance securely.

    Chrome OS is built around security for your web apps. And it will run your cloud applications faster, as we've seen with the Chrome browser. Chrome OS will be ready for enterprise applications to move to the cloud.

    Now we come to my second point: Why do I predict that enterprise applications will move to the cloud?

    Because the fixed costs for the providers are much lower and the relationship in a SaaS model shifts the advantage to the customer:

    One of the key cost structure drivers is the need to support legacy versions of software. Because the SaaS provider doesn't have to support legacy versions, the costs of innovation will be lower, because the developers time isn't used in support. For example, if we compare the COGS (Cost of Goods Sold) to revenues of SAP and Salesforce.com:

    SAP

    48%

    Salesforce

    13%

    This is an indicator that Salesforce is able to sell its software with a margin of 87%. The rest can be put on developing the software and ensuring sales. SAP on the other hand sells the software with a margin of 52%. SAP has to rely on heavy support fees to keep its business running (54% of SAP's software related revenues come from support).

    The difference in these business models is unsustainable, which is why enterprise software will be driven to the cloud.


    In addition to the cost structure difference, the SaaS model has advantages because of the relationship that is built between the provider and customer:

    The provider relies on the customer to be satisfied enough to continue paying the monthly fee.

    In a packaged application relationship, the seller wants to extract the maximum price beforehand, and after that has a monopolistic position to demand high service fees.

    I think these differences are the key to why software in the enterprise will move towards the cloud. Security issues will just be solved, that's the beauty of technology.

    Monday 6 July 2009

    How will the business IT game play out? Legacy vs. Agility

    The amount of interesting web applications springing up each day is overwhelming. It makes me think that there are a lot of smart people out there who are allocating their scarce resource (their time) into projects that will linger a long time and then die out. I wouldn't want to be in that field, doing consumer web services. The barriers to entry are way too low, and the return on investment as well. There could be a lot of space, if we compare to tangible products for example. The problem is that the web provides zero-cost logistics. But that's a question for another post. Today I want to concentrate on Business Infromation Systems.

    Business Information Systems (or applications, whatever) have been ruled by a few companies. Everyone has Microsoft (word, excel etc.). Then there's a more directly business oriented group of giants, SAP, Oracle and so forth. These giants have been attacking the balance sheets of all the companies in the world for years, making them believe that they cannot survive without paying their ridiculous licensing fees for software that is developed by such a huge amount of developers that no one really knows where the software is going.

    More agile methods have been developing all the time. Ruby on rails, django and other rapid development methods have sprung up opportunitites to create customized applications faster.

    Service Oriented Architectures (SOA) and SOAP promised making integration a breeze, and when that didn't quite work out as expected, REST came along and is looking quite strong in enabling new, easier ways of data integration.

    What is needed now is some good ways of calculating ROI for development, fast ways to develop and easy integration. Using services such as Boomi, companies will be able to create an ecosystem that automates a lot of manual work, thus raising productivity.

    In my opinion, some web applications, such as Dabbledb are already at a point where companies should start using their services and see what they can make faster using these easy alternatives to custom application development or out-of-the-box solutions.

    It just isn't worth spending huge sums of money running an IT infrastructure that isn't agile, when it is possible to do simple collaboration with Google docs, just settle with Gmail and create fast business applications with Outsystems, Wavemaker or Hammerkit. Or just settle for Zoho and centralize your IT there.

    One exciting company that's pushing the boundaries of productivity is Mark Logic. They allow companies to retrieve unstructured data and use it for analysis. Out of the box. No code. Just deploy. Once these technologies become mainstream, and once they are easy enough to use such as any web 2.0 service, productivity will rise fast, and the dilemma of information system success will be a faint memory. (the dilemma refers to the difficulty of research in finding a connection between IT investments and productivity).

    I think the research on IT investments and productivity has been led astray. It should be built from the ground up. What can be improved by IT? How much? How much will it cost?

    Instead, IT is taken as a bulk investment, money is poured into a hole, and then we try to measure something which at the same time is influenced by another 10 meaningful factors. I think it reflects well the products of the biggest vendors. You're just buying a lot of trouble, because they make you feel like you need it to stay competitive.

    I just think companies should stay logical and start calculating what they really need.