When Apple introduced the Siri voice assistant with the iPhone 4s in 2011, you may have felt it was a revolution. Up till then, voice recognition systems had been hit or miss, and often required intense training to become accustomed to your voice and speech patterns. You even had to use a special mic to get the best accuracy.

With Siri, Apple promised that you could ask plain English questions on an iPhone (and later an iPad) and get a response; well, some sort of response at any rate. Sure, if Samuel L. Jackson asks something, you can be sure Siri will listen and respond appropriately. I suppose the same can be said for any trained performer.

But what about regular people, asking regular things? Well, there have been class-action lawsuits against Apple claiming that Siri just can’t do the things you hear “her” doing in those popular TV ads. Apple’s response would probably be to remind you that Siri is still beta, although you have to look through a few screens on Apple’s site to find that label. I suppose that frees them of any obligation to explain Siri’s missteps.

In fact, I wonder how the public would have reacted to Apple’s new Maps service had it also bore a beta label. But it didn’t, although it does seem as if Apple is right that Maps has grown better. Maybe in a few months it’ll be good enough to really go up against Google.

Returning to Siri, I’ve discovered a number of comparisons against main competition, the enhanced Google Voice Search, with varying results. Before I cover any specifics, it’s important to realize that Google would appear to fare better with direct search queries, but cannot handle the virtual assistant requests that are Siri’s stock in trade, such as setting an alarm or otherwise reminding you of an important meeting or event.

Remember, too, that both Siri and Google aren’t self-contained. They both require access to server networks to retrieve relevant information. Because both are highly imperfect, you can easily manipulate the questions to favor one or the other. That’s the case of one article, for which I won’t supply a link, which even included an interview with someone from Google telling us how great their voice service really is. Yes, Siri failed on most of the questions posed in that article, but I still suspect the motivation behind the writer’s choices.

One highly respected tech site, Ars Technica, did a somewhat informal but fair test of the two services.  Here most comparisons rated equal, with one or the other faring a bit better.

So Google excelled with the response to the question, “Who founded the Free Software Foundation?” This seems to make sense from a political standpoint, since offers a lot of free services, including Android. Siri delivered the standard and irritating “Search the web” response, and you wonder why Siri didn’t just do it for you. Or maybe you could set that as an option if an answer is otherwise not readily available.

I did wonder, though, why Siri flunked the question, “When does ‘Les Misérables’ come out?” There have been several versions of the movie, but the context of the question would mean the version being released next month.

However Siri was guilty of one curious failure, and you have to wonder why. In response to the question “When did Tim Cook become CEO of Apple?” Siri offered to take you to Apple’s site. Really now! Google delivered such search results as Cook’s Wikipedia entry. Fair enough.

Siri also couldn’t tell you “Who painted the Mona Lisa?” Now come on!

In fairness, anything that involves a direct online search is Google’s stock and trade, which is why Siri didn’t rate as high with such requests. You’d think that she’d learn to improve accuracy over time, but I haven’t really seen any real changes, despite the enhancements in iOS 6. But those changes mostly fleshed out Siri’s features.

However, in that recent corporate shakeup, Eddy Cue, Apple’s Senior VP for Internet Software and Services, has taken control of Siri and Maps from departing executive Scott Forstall. Since both services are troubled, and Cue is considered the “Fixit” person at Apple, perhaps we can look forward to improved accuracy over time for both. Maybe Siri will earn her stripes and escape the beta label before long.

In the scheme of things, Apple needs to demonstrate that Siri can realize her promise. Yes, Google is still concentrating on search requests, but it’s only a matter of time before general voice assistant functions are added. Sure Siri has become a sort of cute cultural icon, and perhaps you can forgive the quirks and occasional slow responses. But Eddy Cue and his team really need to step up to the plate and demonstrate that Siri can become something more than a curiosity, and really become a useful service for users of Apple’s mobile gear — and perhaps Macs too if a certain rumor about OS 10.9 is true.

Supposedly Microsoft is in favor of focus group testing, the better to evaluate new products and get public reaction. Yet you have to wonder how they continue to come up with curious, ill-thought solutions that seemingly defy logic. Take the ribbon toolbar. Microsoft wanted you to believe that toolbars containing icons that were context sensitive were new and different, and superior to regular pull-down menus. But I think most of you know better.

Microsoft also has an annoying tendency to take a minor refinement or change in the OS or an app and pretend it’s something unique. So I recall two consecutive releases of Office for the Mac where modest interface changes were said to make the suite more discoverable. They pulled that a third time with Office 2011, which, aside from a Mac version of the ribbon (thankfully tamed compared to the Windows version), didn’t sport a whole lot of new features. In fact, the replacement for the Entourage email client, Outlook, actually lost a few that users liked. Yes, that’s different.

In all fairness, Apple is known to remove features, but they sometimes do it in the interests of making OS X easier to use. Or they say nothing at all about the reasons behind the change.

With Windows 8, it’s a sure thing that Microsoft isn’t feeling the love. From the launch of the very first public beta, reviewers and Windows lovers alike wondered what the developers at Microsoft were drinking, or smoking, for coming up with this travesty.

It all comes down to Microsoft’s version of the PC+ era, where Windows must appear everywhere. Rather than follow the playbooks of Apple and Google and develop a simple, highly optimized OS for mobile gear that didn’t necessary borrow much if anything from a desktop OS, Microsoft decided you and I really wanted to have it all.

So they took the flat tile look and feel from the failed Zune player and the failed Windows Phone OS and combined it with the traditional Windows desktop, more or less. This schizophrenic approach, where you are living in two different, separated environments depending on which app you want to run, was just guaranteed to confuse people.

There are already reports from a noted Windows advocate claiming that Windows 8 sales are far lower than Microsoft expected, according to his sources within the company. Microsoft continues to claim great sales, except for the Surface tablet, where the results are merely “modest.” At the same time, the fellow who was in charge of Windows, Steve Sinofsky, is now looking for another job. Despite all the silly reports that he really needed a rest, or that he worked badly with others, it really seems he is being held accountable for the Windows 8/Surface debacle. That makes sense.

But it’s not just tech reviewers and the public who have qualms about the look and the feel of Windows 8. Consider this analysis from world-famous usability guru, Jakob Nielsen. In this detailed and damning report, Nielsen recruited “12 experienced PC users to test Windows 8 on both regular computers and Microsoft’s new Surface RT tablets.”

It’s a very complete study, pointing out that most elements of Windows 8 are poorly designed, difficult to use, and, in fact, would appear to present a severe impediment to actually getting real work done. The OS, particularly the interface formerly known as Metro, is not just in your face, but severely detracting. Actually mastering the complicated touch system involves practice and more practice, and it’s not always clear what the actual hot spot is to accept a tap.

One particular shortcoming is the removal of support for multiple document windows in the tiled interface. Supposedly there is an option that will allow you to reveal another window in a small part of the screen, “but none of our test users were able to make this work.” This is a reason why Nielsen suggested that “the product ought to be renamed ‘Microsoft Window.'”

I can understand their pain as I write this article on a Mac running Mountain Lion. One Safari browser window has the standard WordPress text editing page, while a second browser window carries Nielsen’s report. And that is a typical use of multiple windows. When I’m recording a segment for one of my radio shows, I usually need to work in three apps at the same time, something that the Windows 8’s touch interface would pretty much render impossible.

This single — or mostly single — window scheme is fine for a smartphone, and it works reasonably well for many tablets. But there are reports suggesting that iOS 7 may, at least on an iPad, let you work in two apps, or two document windows, at the same time. Microsoft has clearly forgotten what made the traditional versions of Windows highly successful, and they’ve attempted, with Windows 8, to throw everything they learned out of the window unless you find a way to exist all the time in the desktop layer.

Except for Office, which still exists on the Windows 8 desktop, and only pays occasional lip service to a touch-based computer. Even Microsoft couldn’t figure a way to make Office operate under the native interface of Windows 8, so why expect any other major developer to want get involved with this mess? Can you actually imagine a Modern UI version of Adobe Photoshop?

But I don’t feel pity for Microsoft. They jumped into this quicksand with eyes open.

Over the past couple of weeks, some Web logging sources have detected what may be the first indications of the existence of an OS 10.9 testing program within Apple. This shouldn’t be surprising. If Apple stays on the path to annual OS X upgrades, it makes perfect sense that very preliminary versions of the next release are already being evaluated as feature additions and changes are considered.

This week there’s a published report suggesting that Apple may be adding Siri voice recognition and Maps integration. While it would seem curious to expect turn-by-turn navigation on a MacBook Air or MacBook Pro, it’s possible that technology would be used for integration with other apps. Siri would simply expand upon the existing Dictation capability in OS 10.8 Mountain Lion. But all of this comes from a single site that’s being widely quoted, even though there’s no evidence the story is true.

And even if Apple is testing those features, that doesn’t necessarily mean you’ll see them in the next version of OS X, although it would be in keeping with the enhanced integration with the iOS. The larger question would be how Apple might flesh out the feature set to tout 100 to 200 new features for Mountain Lion’s successor.

If the timetable is similar to 10.8, you’ll learn the first meager details of 10.9 in February, and a full feature set and final beta will be released at the WWDC next June. That assumes the timetable remains annual. But predicting Apple’s intentions can be a losing game. Just as you begin to assume you have the pattern down pat, it’s changed.

However, the real issue is how Apple might flesh out OS X, and to the extent to which iOS integration will continue. Already some Mac users have chafed over such changes, and it may well be that Snow Leopard, 10.6, has become Apple’s equivalent of Windows XP. It’s a perfectly useful OS, fast, reliable, and, most important, free of the irritating (to some) excesses of 10.7 and 10.8.

Now I happen to think that this perceived irritation is misplaced. Yes, there are some changes that are designed to smooth the path to iOS and back again. Having apps with the same names, such as Calendar and Contacts, helps reduce confusion. But the objections are more in the way the interfaces have been changed.

So consider the loss of the traditional scroll bars in OS X, and not displaying them (unless the setting is changed) until the cursor hovers over them. Reversing the scroll direction can also be confusing. Aping the iOS scheme, Apple calls it “natural,” and some of you might just call it a useless change. But there is that issue about consistency, and natural scrolling is easily turned off.

I’m sure some of you are concerned about the way a potentially valuable feature, Auto Save, was implemented. Yes, having your document saved periodically in the background can prevent trouble if your app or even your Mac crashes for any reason. But when you just want to noodle with a document and not save any of the alterations, things get a tad complicated. The return of Save As in Mountain Lion was implemented in a questionable fashion, because the original, in addition to the document saved under a new name, would receive the changes made before Save As was invoked. But that questionable “feature” can now be unchecked in the Save As dialog, so the original is left unaltered. Apple does listen, at least sometimes.

The real problem with Auto Save and Version, the feature that stores different sets of a document using Time Machine technology, is that few apps have received the updates required to support these features. Key productivity apps, such as the components of the Adobe Creative Suite and Microsoft Office, only pay lip service to 10.7 and 10.8. This forces you to jump back and forth in save strategies, and such situations are ripe for confusion.

Maybe Apple tried too hard. Auto Save features are already baked into some apps, and the handy utilities that add this function are global. They don’t require changes to an application, which means your workflow pattern doesn’t have to change.

The other question about OS 10.9, which may be called Lynx (or maybe not) is human interface. Now that Sir Jonathan Ive, of minimalist intentions, has control over the look and feel, perhaps the OS X developer team will tame some of the interfaces excesses (such as in Contacts and Calendar) and the sometimes inconsistent look and feel. That should be a really positive development. The best advantage of OS X is that it can get out of the way and let you concentrate on your apps. In contrast, Windows 8 is the most “in your face” OS I’ve ever seen. I suppose Microsoft hasn’t begun to realize that you might use your PC for something other than running Office, or checking your Face-book and Twitter status.

But perhaps the best thing Apple could do with OS 10.9, beyond smoothing the interface, is to go back to all those OS 8 and OS 9 features that were dropped over the years and seek a different sort of inspiration for the future.

Some interesting news tidbits have appeared this week, and, despite the sharp drop in Apple’s stock price, there are clear indications that situation won’t last for long. Now I’m not about to try to get involved in Wall Street matters; I’d do better predicting the odds at making money with the slot machines in Las Vegas, which means there’s no chance whatever.

But here’s what I’m talking about: According to recent published reports, sales of the iPad and iPad mini are actually doing better than previously expected. One of the recent estimates comes from industry analyst Katy Huberty of Morgan Stanley, who surveyed so-called supply chain sources in Asia, and concludes that,”Key suppliers into iPhone and iPad noted above seasonal March quarter order trends, stronger than expected December quarter revenue, and the potential to further upside before year-end.”

She is estimating sales of 46 million iPhones and 23 million iPads for the current quarter. Don’t forget that Apple sold five million iPhone 5s the first weekend in September, and the device has been seriously backordered since then, though the situation is reportedly settling down. In turn, some three million iPads were sold on the weekend they first went on sale, though Apple doesn’t break down the actual numbers for each model.

Continue Reading…