Google: A Utility Company
February 12, 2009
There are several news stories bopping around the datasphere this morning (February 12, 2009). The gist of these stories is that the GOOG bought an abandoned paper mill in lovely rural Finland. The speculation about the intent of this acquisition is interesting. Here’s a synopsis of items I found interesting. If you want the kicker to this post, skip the dot points and jump to the final paragraph. Now the news and analysis that feeds naive minds:
- Reuters here reported here that the former Stora plant will be a data center.
- PaidContent.org and the Washington Post reported here that the Stora Enso facility consumed 1,000 gigawatt hours of power
- Global Markets reported here that part of the site will be transferred to the city of Hamina for other industrial uses.
The missing item is that Google’s patent document 20080209234 Water Based Data Center contains interesting language that includes systems and methods which can be applied to traditional power generation facilities. I wish I knew how to translate into Finnish Google Power & Light. Google Translate yields this phrase: “google teho ja kevyiden yritys”. I prefer GPL myself. I pronounce the acronym as “goo-pull”.
You can poke around other Google documents here. The “site:” instruction may be helpful too.
Stephen Arnold, February 12, 2009
IBM: Covering Its Bases
February 12, 2009
The love affair between IBM and Microsoft cooled years ago. After the divorce, Microsoft took the bank account and the personal computer industry. IBM entered counseling and emerged a consulting firm with some fascination for its former vocation as world’s leading computer and software company.
Jump to the present day. IBM has batted its eyes at Googzilla. IBM and Google have teamed to stimulate the flow of programmers from universities. You can refresh your memory here. In 2008, I received a copy of a letter to an intermediary that said, in part, we understand Googzilla quite well. The outfit interested in this answer was not the addled goose. The interested party was a certain government agency. That outfit was not confident IBM understood Googzilla fully. I wrote about this in a Web log story last year. You can find the article here.
IBM issued a news release here that has been picked up by various information services. The headline makes clear that IBM is not going steady with the GOOG: “IBM to Deliver Software via Cloud Computing With Amazon Web Services”. You can read the full article here. In a nutshell, IBM wants to
make available new Amazon Machine Images (AMIs) at no charge for development and test purposes, enabling software developers to quickly build pre-production applications based on IBM software within Amazon EC2. The new portfolio will over time extend to include Service Management capabilities from IBM Tivoli software for Amazon EC2 to help clients better control and automate their dynamic infrastructures in the cloud.
The idea is a good one. But the significance of this deal is that IBM is making clear to the GOOG that a certain someone is no longer numero uno. IBM is playing the field. Amazon has outpaced Google in some cloud services and by spending a fraction of the billions Google has invested. What’s IBM know that I don’t?
Stephen Arnold, February 12, 2009
Microsoft Has a PR Mountain to Climb
February 12, 2009
Fast Forward 09 zipped right on by. Not too much excitement in the blogosphere. Then Joe Wilcox, writing in the eWeek Microsoft Watch publication, dropped the hammer on Microsoft in “Windows 7 Enterprise Asks Too Much” here. The main point of his interesting article was:
Software sales are slowing with global economies, and Microsoft executives must hope that annuity license contract conversions will smooth out Windows Client division revenues and ensure they will continue. Problem with the reasoning: If businesses are buying less, there’s no reason for them to spend more for Windows’ licenses. Microsoft’s Software Assurance priority and its customers budget-constrained spending objectives are contradictory.
In short, bad economy and Microsoft is trying to take money out of organizations like it was 1999.
But the killer analysis was Preston Gralla’s “Microsoft 10,000 Patents, $9 Billion Annually in Research and We Get Vista” here. Ouch. The write up drags fingernails down the whiteboard. The issue for me was the comment section to this article. One example: anonymous (of course) wrote:
Microsoft could spend twice the amount on R&D and it would make little difference. We got Vista and Win7 because of gross management incompetence starting with Ballmer.
No comment from the goose.
Stephen Arnold, February 12, 2009
Google and Microsoft: The Complexity Spectrum
February 12, 2009
I talked with a small group yesterday about the “big” Microsoft announcement. SharePoint administrators get an opportunity to merge SharePoint and the Fast Search & Transfer technology. You can get some useful information from ChannelWeb here. Mostly vaporware and roadmaps as of February 2009, SharePoint believers will have some to brush up on their content processing integration skills. I think the learning curve will be steep, so quit reading the addled goose’s Web log and dive into the Fast documentation. Hmm. That might be a problem. We tried to locate documentation online and could not locate the information. I had to return my three ring binders when I disengaged from my poking around inside ESP for a well known organization. Sorry, my lips are sealed on which organization. Think acronyms. Think follow the rules.
What surfaced in our discussion was a somewhat tired metaphor. A spectrum. The idea is that the ends are different. Red light waves at one end; blue light waves at the other. The diagram below shows the type of spectrum that elementary school teachers use to educate the kiddies about “light”.
For the purposes of this Web log post, one end is Google. Color Google blue. The opposite end is red. Color Microsoft red. Both companies have many similarities; for example, each is a software company; each competes in Web search; and each covets the enterprise market. Keep in mind. Search is complex. Within a complex task, the notion of a spectrum of complexity suggests that simple may be better. Less hassle. Lower cost. Easier to troubleshoot–sometimes.
One big difference is the positioning of each companies’ technology in the enterprise sector. I want to focus on the subject of behind the firewall search or what most trophy generation wizards call “enterprise search.” Google sales professionals run the game plan and make it clear that the Google Search Appliance or GSA is simple. The system is easy to deploy. Unpack the shipping container, plug in the gizmo, zoom through the administrative screens, and employees can search the processed content. No muss. No fuss. Let the GOOG do the work.
On the other end of the spectrum is Microsoft’s approach to “enterprise search.” For me, the Microsoft search product line is hard for me to understand. There are a couple of versions of search available for SharePoint. One if free; the other comes with an Office server. Earlier this week, Microsoft announced two or three different “versions” of Fast Search & Transfer information retrieval technology. I thought there was the ESP or enterprise search platform and the orphaned Fast Web search. I may have Microsoft’s reality clouded with what I recall Fast Search professionals telling me in the pre buy out era.
Email Alert Web Site Online
February 12, 2009
A new web site is launching Monday that will send you e-mail alerts. Crap, I Missed It! is somewhat similar to Yotify.com (The Beyond Search addled goose wrote about Yotify.com here. CrapIMissed wants to distinguish itself by focusing on “sweet spot” alerts with info unavailable elsewhere, such as Amazon.com bestsellers, new audio CDs, upcoming concerts and so on. This isn’t like Google Alerts. You don’t enter a couple keywords and get an info dump as-it-happens or daily for every search term. Crap lines up your request into a single e-mail message per day compiled from only new information. There’s no account creation, there’s a reference archive of your alerts; and a one-step unsubscribe. For us goslings who suffer daily information overload, it may be worth a try.
Jessica Bratcher, February 12, 2009
Yahoo Monetizes BOSS Soon
February 12, 2009
A happy quack to the reader who alerted me to Yahoo’s business model for BOSS (build your own search system). BOSS has been “free”. The reasons range from building traffic to expanding the reach of Yahoo. Free is generally good. Spidering the Web and indexing content can be expensive. Yahoo knows this first hand. Certain vendors have been quick to embrace BOSS because the price was right. Besides Yahoo’s own search results often leave me unsatisfied. I have written about Cluuz.com because the company uses BOSS and the firm’s technology makes much better use of the Yahoo index than Yahoo itself does in my opinion. Now Yahoo wants to charge outfits like Cluuz.com, a small Canadian company, to use BOSS. The new story in Network World here said:
Once Yahoo introduces BOSS fees towards mid-2009, it will also increase the number of search results an engine can obtain via a single API call to 1,000 from 50. The fees vary depending on the type and quantity of search result involved. Yahoo will also offer SLAs to promote the creation of more sophisticated BOSS search engines.
Yahoo assumes that some of the outfits using BOSS will monetize their services. I think that if a company using Yahoo’s BOSS could monetize its services, the company would be monetizing now. The economy is a bit shaky and among the hardest hit are small search and content processing companies. Yahoo and others of its ilk give away search systems. Presumably Yahoo perceives a revenue win.
In my opinion, BOSS users may start looking around for ways to shift from BOSS to another service. Microsoft, are you listening? The GOOG never listens but maybe the idea of tweaking Yahooligans again might spark some activity.
Maybe this is a brilliant play that will reverse Yahoo’s search fortunes? I hope so. I am uncomfortable watching the company follow the trajectory of Ask.com and America Online, among others in the information processing game.
Stephen Arnold, February 12, 2009
Exalead Accelerates
February 12, 2009
Exalead is flexing its muscles by powering search for the famous Wellcome Trust Sanger Institute (http://www.sanger.ac.uk). Sanger does genome research involving large scale sequencing and analysis like the Human and Cancer Genome projects. The institute needed a powerful search solution for its monstrous scientific data requirements, and Exalead offered a scalable, flexible program that can handle a lot of abuse: the data quantity at Sanger grows by about 120 million records annually. Now a simple text request will return results from a pool of 500+ million data–a mere puddle compared to the projection of a possible 20 billion files over time. You can read the case study here. Looks like Exalead has a huge task ahead of it. If it succeeds, it will be indexing possibly one of the biggest public databases in the world. A happy quack to the Exalead team.
Jessica West Bratcher, February 12, 2009
Google and Compression
February 11, 2009
Google’s technical papers are often useful. The Reddit post by Biogeek is interesting. The item is brief: An often undersold component of the Google technology stack is their sophisticated library of compression algorithms. Here’s one they presented at vldb in 2007. [pdf] (vldb.org). More information on Google’s compression inventions can be viewed by clicking this link. Google’s patent documents provide useful context information and somewhat more detailed explanations of the company’s methods. The 2007 paper is interesting, and it shows how Google engineers work on incremental improvements. Biogeek is spot on. Compression is a core competency at the Google. One note: it’s not a recent interest.
Stephen Arnold, February 11, 2009
Business Intelligence: An Useful and Intelligent Review
February 11, 2009
A happy quack to the reader who alerted me to Stephen Swoyer’s “How BI Habits Are Changing” here. In general, I agree with his points. What impressed me was the jargon free analysis of the changes in business intelligence. For me, the most interesting comment was:
SAP has a lot of account control in SAP shops that makes up-selling to Business Objects possible. With IBM, it seems the WebSphere and consulting services bring influence.” she [Cindi Howson — founder of BIScorecard.com] says. “Customers that are only using DB2 — so that’s all that IBM is in the account — don’t seem to be looking at Cognos in a new light. So again, it really depends who has strong account control… [emphasis added]
You may find other nuggest in Mr. Swoyer’s write up. I don’t get too excited about business intelligence, but I think one can discern a movement in this market sector that may provide some hint of what lies in the future for enterprise search and content processing. Worth reading.
Stephen Arnold, February 11, 2009
Microsoft Fast: An Integration Roadmap without Google Maps
February 11, 2009
Fast Forward 09’s “big” announcement is history. I have had a couple of calls and a few emails about the bundling of Fast Search & Transfer’s Enterprise Search Platform with SharePoint. You can read Todd Bishop’s review of some of the history or deal and some azure chip consultant market projections here. His article has attracted some comments. I flicked through these and found two of interest.
The first is the article in All about Microsoft here. Mary-Jo Foley’s “Microsoft Updates Its Enterprise Search Roadmap” is well named. The “big” news is not a product that one can use right now. The “big” news is a roadmap. For me the most interesting part of her column was this comment:
Fast’s technology soups up the enterprise search capabilities that are part of SharePoint Server. Fast adds more sophisticated user-interface elements, like thumbnail and preview views; cluster support and more compute-intensive tasks like entity abstraction and the creation of relationships between concepts
In my experience, anyone who asserts that Fast ESP “soups up” enterprise search has not performed two hands on Fast ESP tasks. [1] Installing, tuning, and updating the system with Fast ESP hot fixes. And [2] building out an infrastructure to give Fast ESP sufficient room to breathe. Ms. Foley does a good job of tracking Microsoft, but I am not convinced that her inclusion of the “soups up” reference matches reality. Fast ESP is a collection of components. Some of those identified in the comment above come from third parties and will stretch the expertise of the average SharePoint wizard. The article does not point to several of the well known issues associated with Fast ESP, and those issues help explain how the company ended up in a bit of a financial jam which evolved into the police action on October 16, 2009.
The second is Search Engine Watch’s article “Microsoft Integrates Fast Search with Existing Enterprise Search Offerings” here. The write up is based on this comment from a Microsoft professional:
FAST Search for SharePoint will combine high-end search with the broad portal, collaboration, content management and business intelligence capabilities of SharePoint. And FAST Search for Internet Business will deliver search capabilities tuned to drive more revenue through Web sites.
I don’t know what this statement means. I find myself reluctant to believe that the Fast ESP system will “drive more revenue through Web sites.” Maybe this is a reference to the Fast Search publishing technology. Again I think the writer is feeding back marketing lingo without providing any detail.
Let me come at this announcement in a different way:
First, Fast Search was designed a year or two before the Google stomped into the picture. The Fast Search plumbing was Linux and the system did a very good job of indexing Web content quickly. Even after the sale of Fast Search’s Web index (AllTheWeb.com), Fast Search’s indexing of news was more efficient and timely than Google’s news service. (Google has now bypassed Fast Search in news in my opinion.)
Coming to a SharePoint installation if I understand the news announcement at Fast Forward 09 on February 10, 2009. © Fast Search & Transfer 2007. From Enterprise Search Report, 3rd Edition. Used in that study with the permission of Fast Search & Transfer.
Hooking the Fast ESP into SharePoint with or without Dot Net is going to be a bit of an effort.
My take on this “big” announcement. The competitors who offer alternatives to SharePoint search will have many opportunities to make new sales. My hunch is that Google is greatly encouraged by this “big” announcement.
Stephen Arnold, February 11, 2009
-
- Subscribe to Beyond Search
Feature archive
News archive
-
