Mad Dog 21/21: Smart Fellers
July 18, 2011 Hesh Wiener
Remember tongue twisters? You know what you want to express, but it just doesn’t come out right. One smart feller, he felt smart. Two smart fellers, they both felt smart. Three smart fellers, they all felt smart. Tongue twisters remind me of IBM‘s Lotus division. Lotus always had a little trouble getting its message across. That’s how Excel beat 1-2-3 years ago. Now Lotus’s Domino server and Notes client could go the way of 1-2-3 because people armed with mobile clients are hooking up via companies like Google and Amazon, whose cloud allure dwarfs that of IBM.
It’s not so much that IBM’s Lotus products are weak or flawed. On the contrary, IBM has made its Domino server robust and feature-rich. Domino is arguably superior to its arch-rival, Microsoft Exchange, when it comes to client security. Perhaps surprisingly, it is probably even easier for an IBM i user with fewer than a thousand seats to price and buy a Domino software stack than it is for a Windows customer to put together a viable Exchange system.
Lotus is trying to keep up. It offers mobile client apps for iOS, Android, Symbian, and Windows. Microsoft’s only mobile client is the one for Windows Mobile. There is pretty good Exchange connectivity in iOS and okay service in Android. There are also specialty apps like TouchDown for Android that do what Microsoft (or Google) should have done.
Still, when it comes to mobile messaging, business users and especially lawyers are in love with Blackberry, and not without cause. Blackberry devices provide excellent message handling, high security, and superb ergonomics. They evolved as clients of Blackberry’s messaging servers long before “cloud” became the irritating buzzword it is today and they also work well with Exchange, IMAP, POP3, SMS, and other text and communications hubs. Blackberry gizmos, particularly the newer and more elaborate models, can also record, send and receive multimedia messages by which we mean not only photos but also videos.
Nevertheless, Blackberry is struggling in a market that has gone wild for more versatile clients that run iOS or Android and the company’s managers and shareholders are understandably nervous. Blackberry, like Microsoft and IBM’s Lotus division, had the upper hand until recently only to have its prospects drastically altered by a perfect storm of technological and cultural developments.
At the heart of the storm are Apple and Google. Apple created spectacular client devices, developed a wonderful system for selling and delivering apps, and media files and attracted a huge following of affluent, lively customers. Google built vast information repositories and created superior means for people with mobile devices to find and extract information from its data banks. Google developed and diffused fantastic software tools to help people use its data in unprecedented ways and do so at no direct cost. In addition, Google opened up the client universe by developing the Android operating system and a rich and excellent set of support structures. Google also shamelessly mimicked Apple’s file marketing scheme, creating an app store and related media store and used the awesome power of its brand to boost the market profile of Android mobile clients.
Another factor accelerating the pace of change has been the phenomenal improvement in broadband networks both wired and wireless. Wireless communications technology in the United States and Europe lags that in Japan and Korea, but it is still moving forward at a good clip. Wireless speed varies from carrier to carrier and city to city (as well as from country to country). There is a good summary of the current American situation at Mobile Broadband Reviews. In the United States, for example, Verizon Wireless is rolling out LTE, which can in theory download data to a client at 100 Mbps and upload at 50 Mbps. Verizon advertises speeds of a tenth this theoretical maximum and in tests seemed to be able to deliver downlink speed of about 6.44 Mbps and upload speed of 5 Mbps. But you might not see it on any of their phones, which are constrained.
You need a wireless modem to get the fastest service Verizon offers. Sprint, using the WiMax technology pioneered by Clear (which it has acquired) is rolling out service that in textbooks runs at 128 Mbps down, 56 Mbps up. Independent tests indicate Sprint’s average delivery rate is 2.15 Mbps down and 0.85 Mbps up. AT&T, the biggest American provider of world-standard GSM technology, is not the leader when it comes to bandwidth, but nevertheless offers HSPA+ (and probably LTE one of these days) with theoretical speeds of 84 Mbps down and 22 Mbps up and advertised speeds of a quarter or half this max; mileage varies with local market. In practice, customers can get 2.45 Mbps down and 1 Mbps up. T-Mobile, which AT&T is trying to absorb, also has implemented HSPA+ (and has plans for LTE) with the same theoretical bandwidth as ATT’s offering. T-Mobile seems to be delivering a peppier connection right now, with testers clocking the service at 2.83 Mbps down and 0.85 Mbps up.
I did a tiny test of my own from my outer boroughs New York office. My phone loaded with the Ookla Speed Test app showed T-Mobile broadband service (from a server in Manhattan) at 5.76 Mbps down, 1.65 Mbps up. At the same location, the same Android testing device linked to my local network showed Roadrunner cable service and my ancient wireless router providing speeds of 10.76 Mbps down and 0.50 Mbps up. I also ran the desktop version of the test, which reported my cable Internet services ran at 18.39 Mbps down and 0.49 Mbps up. Verizon DSL service from a DSLAM exchange that’s a couple miles away worked very poorly for me; recent logs at DSL Reports show my speeds of 1 to 1.5 Mbps down, about 0.35 Mbps up. (Customers located close to this exchange apparently get very good services.)
I’ve been promised an opportunity to get a fiber connection and the potential of FiOS is impressive, but in practice in New York City, Time Warner Cable generally provides better value but worse upload speed. In any event, in New York City it’s possible to get wireless broadband that is about as fast as fiber, coax, or DSL broadband . . . and New York is far from the best place in the USA to be wireless. Well-regarded testing by PC World indicates that plenty of other cities have superior service. Baltimore, Chicago, San Jose and San Francisco generally rank higher than New York, but performance varies (quite a bit) by carrier.
Even in places with so-so wireless speed, a smartphone or tablet app can chit chat with Google’s servers so fast it’s hard to believe the software is strictly some code in a thin client using server side apps and database engines to do the actual work. You tap a smartphone and put it into talk-to-me mode, say “find the nearest pizza parlor” and right away you see a map of the place you happen to be with signals showing where you can get that hot slice. This is not the world Lotus lives in, nor the one where Exchange thrives. But it is the world where I am going to use a smartphone to find my lunch. At lunch, if I happen to spot an interesting book in the hands of another diner I can fire up Google Goggles, an image-based search engine, point the phone’s camera at the book cover and in a fraction of a second later I’m on the Amazon.com website where, using that company’s servers, as fast and powerful as those of Google, I can buy that book and instantly get it delivered to my Kindle or to the Kindle app on my phone. If I’m shopping I can similarly point my phone at the barcode on some merchandise and get instant client/server comparison shopping from Amazon and other merchants.
Where’s the IBM i toolkit that makes it easy for newbies to whip up the server side of these smartphone apps (and to create the client apps themselves)? Okay, IBM, do it under Linux if you want, so it’s cross-platform, and the i guys will adjust just fine. Same goes for Microsoft, which is knocking itself out trying to put its Office dreck on the Internet instead of making Windows a place a hundred thousand app developers want to work. And Oracle, which should know boatloads or maybe it’s yachtloads about making databases do useful things, has tragically put Sun, its most creative holding, where the sun don’t shine.
The business of the Internet isn’t email or slide shows. The business of the Internet is business.
IBM ought to understand that, but something has gone wrong. Lou Gerstner famously boasted that when he was running IBM he got an elephant to dance. Well, he seems to have done this with the same technique that’s used to get a bronco to buck. The result isn’t anything like Apple’s genius at gadget invention, Google’s brilliance with data, or Amazon’s nearly sexual gifts for shopper gratification. No, after watching for years as phones got smart and its executives didn’t, IBM acts like the technically excellent but conceptually dim accomplishments of Lotus are just wonderful.
Yep, it’s IBM all right. Give the company a lemon and it will make Lehman.