The Symbian OS (779886), страница 88
Текст из файла (страница 88)
Itwas the first product on Symbian OS v6.1, the first Symbian 2.5G (i.e.it supported GPRS) phone and, with its VGA camera, it was the firstcamera-phone anywhere outside Japan. At that time, the Nokia 9210 hadjust become the best-selling PDA, unseating Palm. Symbian OS rose forthe first time to the top of the platform chart for PDAs and smartphones.Bob Dewolf:Series 60 made us think about all the things Symbian OS had to worryabout, like distribution policies and whether components are only publishedinternally or were third-party published.
We put in a whole new layer ofcontrol to say not only that header files export but whether they get filteredout to the SDK process. Pearl became Series 60 at that point, version 0.9, andSDKs came out soon after at version 1.0.But above all, the Nokia 7650 was the first S60 phone.
The bigannouncement at 3GSM in Cannes was Symbian OS v7.0. At the sametime, Symbian announced a new UI strategy. The DFRD idea wasdropped. In fact, Symbian no longer proposed to ship a GUI implementation at all on top of its UI Framework implementation. The field wasopened to UI vendors or licensees to do their own thing. Meanwhile thenew UIQ UI was launched in place of Quartz, with UIQ spun out as aseparate, independent company, though it remained Symbian owned.
(Inlate 2006, UIQ was purchased by Sony Ericsson and became completelyindependent of Symbian.)A few weeks later, the first Symbian OS v7.0/UIQ phone, the SonyEricsson P800, was announced (it launched later that year). It was a penbased phone featuring a removable flip keypad, somewhat in the stylepioneered by the Ericsson R380, a jogdial thumbwheel for navigationand, of course, a camera.The new strategy was possibly the only strategy that made sense notjust of the unique nature of the phone market, but of Symbian’s uniqueposition in it. Perhaps, to some extent, Nokia identified this most quickly(and so platformized its own UI).
The moral, perhaps, is that whereit is impossible to have more than partial foresight of evolving marketrequirements and opportunities, multiple visions are better than one.Enabling a competitive ecosystem allows multiple different visions toemerge and have a chance to succeed.The change in strategic direction predated Colly Myers’s departure asCEO in February 2002, but its implementation spanned the interregnumand was picked up and seen to fruition by the incoming CEO, DavidLevin, from April 2002. Fittingly, Myers’s – and Nokia’s – Christmas giftto the company was a Nokia 7650 phone apiece.42016.8ONE SIZE DOES NOT FIT ALL: THE RADICAL USER INTERFACE SOLUTIONHow to Develop a World-class GUIThe most visible mobile phone trends are in conflict: devices arebecoming increasingly smaller, packing in more and more hardwareand demanding longer battery life, while moving increasingly into theconsumer end of the market as features and functions expand.
The phoneoperating system is the battleground on which these conflicts are playedout. Shrinking size and increasing complexity are not simple trade-offs.The evidence is that the successful phone needs both.Perhaps it is worth citing the definition from Pekka Ketola, a usabilityengineer on S60 [Lindholm et al.
2003, p. 172].Smartphones are an emerging product category where communication,namely voice calls and messaging, is still the main function, but wherepersonal information management is fundamentally improved compared toconventional mobile phones. Smartphones have good calendars, versatilecontact management properties, to-do lists, address books, and so on. Theyare solid platforms for imaging and gaming.In one way, phones are become increasingly personal items, statementmaking lifestyle accessories. In another direction, they have evolvedfrom being ‘expensive showoff tool[s] to an everyman’s communicationplatform’ (Nieminen-Sundell and Vaananen-Vainio-Mattila in [Lindholmet al. 2003]).
In both directions, the trend is to a normalized market, thatis, a consumer market in which sophisticated users give way to naı̈veones and complexity is not tolerated. Phones lose novelty value and areexpected to perform as easily and predictably as TV sets and stereos. Inthis battleground, the UI becomes the front line.Big User Interfaces Don’t ScaleBob Dewolf:There’s an awful lot to a UI.
It’s not just controls and their customizations,there’s a lot of interaction. That’s a surprise we got in the process of development when we did Pearl, how much more integrated you had to be. Integrationis how the applications fit together, what keys the phone is grabbing from thewindow server, and who’s managing the power key and what happens if youpress the Phone button.
There were a lot of issues like that and it was importantto get them right.I remember very fundamental decisions about what happens when youpress the menu key. For instance, does it just put the menu on top? In the wedecided on something very simple, but there were lots of different proposals,HOW TO DEVELOP A WORLD-CLASS GUI421which led us to decide that simplest is the best and we just leave everythingstacked where it is and go with the window-server policy. But it’s amazinghow long the debates took.
In that particular case, it ended up very simple,but not in all cases.An absolute rule is that small interfaces are different from big ones.Direct manipulation (the familiar Windows or Macintosh model), forexample, does not scale down. You do not expect to drag a mousecursor across your phone screen. The parallel model of the desktop, withmultiple open windows between which you task, does not scale either.On a small phone screen parallel is out, sequential is in.Just as even the best UIs do not scale, nor do they move easily fromone device class to the next. In effect, this is exactly what Symbian’s UIstrategy reflects and what the architecture has been evolved to support.One size does not fit all, and there is no single right model.
Whatworks on a flip phone probably does not work on a keyboard-centricCommunicator; what works on a pen-driven tablet probably does notwork on a phone designed for the shirt-pocket.The first generation Quartz devices (that is, before the evolution ofQuartz into UIQ) were strongly tablet-based designs. There have beentablet phones (for example, the original XDA based on Microsoft Windows CE and, perhaps, the phone-enabled iPaq counts as tablet-like)but they have not been huge successes in the mobile phone market.The evidence suggests that consumers want phones with productivity functions, rather than PDAs (which are really productivity devices)with phone functions. There is little to suggest that the UI is a critical factor in the decline of the PDA market; neither Palm PDAs norMicrosoft PDAs have fared much differently.
(Palm has now gone all theway and adopted Windows Mobile, so the point becomes moot goingforward).Given the differences between Windows on a desktop and WindowsMobile on a phone, there is not much case to be made that usersare seeking the desktop-style behavior on their phones; they won’t getit. What they will get is familiarity at some level, even if the behavior is different. Perhaps what they are really seeking is the unstatedpromise of compatibility; but in that case the value of having Microsoftrunning on their phones is only incidentally about the UI itself andhas more to do with the promise of the brand.
But as the Palm andMacintosh combination demonstrated quite ably years ago, smooth interoperation has not much to do with sharing a UI on different platforms,it’s just a matter of plain old-fashioned good, careful, user-centereddesign.422ONE SIZE DOES NOT FIT ALL: THE RADICAL USER INTERFACE SOLUTIONThe most interesting point about phones is that while the design pointhas in a sense been fixed and is the same across manufacturers, UIdifferentiation makes different phones quite radically different from eachother (to the point that picking up someone else’s phone and trying tomake a call can be surprisingly difficult). Kiljander and Jarnstrom (in[Lindholm et al.
2003]) argue that it is because phones have no standardUI and, therefore, everyone is using their own solutions that the UI hasbecome such a significant competitive asset. This is as true at the low endof proprietary operating systems and custom UIs as it is at the high end ofSymbian OS and its competitors.If this is the case, Nokia’s commanding market share (at the time ofwriting anyway, but with little sign of it waning) is a testament to its UIdesigners and to its strategy of evolving a family of UIs, each tuned to thedesign point of one of the categories in the famous Nokia segmentationmatrix.9Usability ValuesAs well as becoming cameras, music players and diaries too, phones havebecome open platforms for third-party software of all kinds, includinggames.
It is hard to devise a UI that is as fit for playing fast-paced shoot’em-ups as it is for managing your daily, weekly and monthly meetingsand appointments, while still allowing users to answer (and initiate) callswith a single button press. While phones have become the archetypeof the omnifunctional, converged device, Symbian OS is still designedto be flexible enough to support a wide range of possible device typesincluding more narrowly specialized categories – digital cameras, in-carnavigation systems and set-top boxes have all been rumored at one timeor another as possible target devices.Arguably, however, the particular, and unique, strength of SymbianOS is its power (as well as its compactness) as an open, thoroughlyGUI-centric, standards-driven application platform.