Ideas for Windows

Where should Windows go? What improvements should Microsoft make?

I have some ideas.

First, Microsoft should drop the '10' designation of Windows. Yes, Windows 10 is different from Windows 8.1 and Windows 8 (and Windows 7, and Windows Vista, and...) but that's not important. What's important is how people talk about Windows, and how Microsoft wants people to talk about Windows.

People don't talk about "Windows 10"; they call it "Windows 1809" or "Windows 20H2". And those names are good enough to identify a version of Windows. The "10" is superfluous.

There is another benefit to dropping the '10'. It means that Windows doesn't get embroiled in marketing battles with Apple's macOS. Apple just updated macOS from version 10 to version 11. Is Apple now "ahead" in the operating system race because they have version 11? Is Microsoft behind because they have version 10? (Fans of the old movie "This is Spinal Tap" can make jokes about "our amplifiers go up to 11".)

By dropping the '10' and referring to release numbers, Microsoft avoids all of the comparisons. Or at least the meaningless comparisons of '10' and '11'.

Looking beyond version numbers, Microsoft is moving into a position which will let it reduce the number of configurations for Windows.

Today, Windows has multiple families: "Windows Server", "Windows Professional", and "Windows Home".  When Windows was sold in a box, these divisions made sense.

As Windows moves to a subscription service, the notion of different Windows versions makes less sense. The subscription model changes the idea of the purchase. Instead of "I'm buying this Windows product today and I will use it for a few years, getting free updates in that time" the idea becomes "I have subscribed to Windows, I pay a monthly fee, and this month I want this configuration of Windows". The difference is that one can change the product during the subscription (much like changing a Netflix account to allow more or fewer screens at the same time).

With a Windows subscription, one could start with Windows Home, then "dial up" the configuration to Windows Professional for a few months.

Maybe the designations of "Home" and "Professional" don't make sense anymore. Perhaps a subscription to Windows gives one access to a set of features (more granular that "Home" or "Professional"), and you can activate (or deactivate) new features as you want. Instead of choosing one of three options, one chooses from a large menu of services, enabling those that make sense.

I think we will see a number of changes as Microsoft moves from the "sale in a box" concept to the subscription concept of software.

The Demise of Apple

Future historians will look back at Apple, point to a specific moment, and say "Here, at this point, is when Apple started its decline. This event started Apple's fall.". That point will be the construction of their new spaceship-inspired headquarters.

Why do I blame their new building? I don't, actually. I think others -- those future historians -- will. They will get the time correct, but point to the wrong event.

First things first. What do I have against Apple's shiny new headquarters?

It's round.

Apple's new building is large, elegant, expensive, and ... the wrong shape. It is a giant circle, or wheel, or doughnut, and it works poorly with human psychology and perception. The human mind works better with a grid than a circle.

Not that humans can't handle circular objects. We can, when they are small or distant. We have no problem with the moon being round, for example. We're okay with clocks and watches, and old-style speedometers in cars.

We're good when we are looking at the entire circle. Watches and clocks are smaller than us, so we can view the entire circle and process it. (Clocks in towers, such as the "Big Ben" clock in London or the center of town, are also okay, since we view them from a distance and they appear small.)

The problems occur when we are inside the circle, when we are navigating along the circumference. We're not good at keeping track of gradual changes in direction. (Possibly why so many people get lost in the desert. They travel in a circle without realizing it.)

Apple's building looks nice, from above. I suspect the experience of working inside the building will be one of modest confusion and discomfort. Possibly at such a minor level that people do not realize that something is wrong. But this discomfort will be significant, and eventually people will rebel.

It's ironic that Apple, the company that designs and builds products with the emphasis on "easy to use", got the design of their building wrong.

So it may be that historians, looking at Apple's (future) history, blame the design of the new headquarters for Apple's (future) failures. They will (rightly) associate the low-level confusion and additional brain processing required for navigation of such a building as draining Apple's creativity and effectiveness.

I think that they (the historians) will be wrong.

The building is a problem, no doubt. But it won't cause Apple's demise. The true cause will be overlooked.

That true cause? It is Apple's fixation on computing devices.

Apple builds (and sells) computers. They are the sole company that has survived from the 1970s microcomputer age. (Radio Shack, Commodore, Cromemco, Sol, Northstar, and the others left the market decades ago.) In that age, microcomputers were stand-alone devices -- there was no internet, no ethernet, no communication aside from floppy disks and a few on-line bulletin board systems (BBS) that required acoustic coupler modems. Microcomputers were "centers of computing" and they had to do everything.

Today, computing is changing. The combination of fast and reliable networks, cheap servers, and easy virtual machines allows the construction of cloud computing, where processing is split across multiple processors. Google is taking advantage of this with its Chromebooks, which are low-end laptops that run a browser and little else. The "real" processing is performed not on the Chromebook but on web servers, often hosted in the cloud. (I'm typing this essay on a Chromebook.)

All of the major companies are moving to cloud technology. Google, obviously, with Chromebooks and App Engine and Android devices. Microsoft has its Azure services and versions of Word and Excel that run entirely in the cloud, and they are working on a low-end laptop that runs a browser and little else. It's called the "Cloudbook" -- at least for now.

Amazon.com has its cloud services and its Kindle and Fire tablets. IBM, Oracle, Dell, HP, and others are moving tasks to the cloud.

Except Apple. Apple has no equivalent of the Chromebook, and I don't think it can provide one. Apple's business model is to sell hardware at a premium, providing a superior user experience to justify that premium. The superior user experience is possible with local processing and excellent integration of hardware and software. Apps run on the Macs, MacBooks, and iPhones. They don't run on servers.

A browser-only Apple laptop (a "Safaribook"?) would offer little value. The Apple experience does not translate to web sites.

When Apple does use cloud technology, they use it as an accessory to the PC. The processing for Siri is done in a a big datacenter, but its all for Siri and the user experience. Apple's iCloud lets users store data and synchronize it across devices, but it is simply a big, shared disk. Siri and iCloud make the PC a better PC, but don't transform the PC.

This is the problem that Apple faces. It is stuck in the 1970s, when individual computers did everything. Apple has made the experience pleasant, but it has not changed the paradigm.

Computing is changing. Apple is not. That is what will cause Apple's downfall.

The Real Problem with the Surface RT

When Microsoft introduced the Surface RT, people responded with disdain. It was a Windows tablet with a limited version of Windows. The tablet could run specially-compiled applications and a few "real" Windows applications: Internet Explorer, Word, Excel, and Powerpoint.

Many folks, including Microsoft, believe that the problem with the Surface RT was that it was underpowered. It used an ARM chip, not Intel, for the processor. The operating system was not the standard Windows but Windows RT, compiled for the ARM processor and excluding the .NET framework.

Those decisions meant that the millions (billions?) of existing Windows applications could not run on the Surface RT. IE, Word, Excel, and Powerpoint ran because Microsoft built special versions of those applications.

The failure of the Surface RT was not in the design, but in the expectations of users. People, corporations, and Microsoft all expected the Surface RT to be another "center of computing" -- a device that provided computing services. It could have been -- Microsoft provided tools to develop applications for it -- but people were not willing to devote the time and effort to design, code, and test applications for an unproven platform.

The Surface RT didn't have to be a failure.

What made the Surface RT a failure was not that it was underpowered. What made it a failure was that it was overpowered.

Microsoft's design allowed for applications. Microsoft provided the core Office applications, and provided development tools. This lead to the expectation that the tablet would host applications and perform computations.

A better Surface RT would offer less. It would have the same physical design. It would have the same ARM processor. It might even include Windows RT. But it would not include Word, Excel, or Powerpoint.

Instead of those applications, it would include a browser, a remote desktop client, and SSH. The browser would not be Internet Explorer but Microsoft's new Edge browser, modified to allow for plug-ins (or extensions, or whatever we want to call them).

Instead of a general-purpose computing device, it would offer access to remote computing. The browser allows access to web sites. The remote desktop client allows access to virtual desktops on remote servers. SSH allows for access to terminal sessions, including those on GUI-less Windows servers.

Such a device would offer access to the new, cloud-based world of computing. The name "Surface RT" has been tainted for marketing purposes, so a new name is needed. Perhaps something like "Surface Edge" or "Edgebook" or "Slab" (given Microsoft's recent fascination with four-character names like "Edge", "Code", and "Sway").

A second version could allow for apps, much like a Windows phone or iPhone or Android tablet.

I see corporations using the "Edgebook" because of its connectivity with Windows servers. I'm not sure that individual consumers would want one, but then Microsoft is focussed on the corporate market.

It just might work.

Apple and Microsoft do sometimes agree

In the computing world, Apple and Microsoft are often considered opposites. Microsoft makes software; Apple makes hardware (primarily). Microsoft sells to enterprises; Apple sells to consumers. Microsoft products are ugly and buggy; Apple products are beautiful and "it just works".

Yet they do agree on one thing: The center of the computing world.

Both Apple and Microsoft have built their empires on local, personal-size computing devices. (I would say "PCs" but then the Apple fans would shout "MacBooks are not PCs!" and we don't need that discussion here.)

Microsoft's strategy has been to enable PC users, both individual and corporate. It supplies the operating system and application programs. It supplies software for coordinating teams of computer users (ActiveDirectory, Exchange, Outlook, etc). It supplies office software (word processor, spreadsheet), development tools (Visual Studio, among others), and games. At the center of the strategy is the assumption that the PC will be a computing engine.

Apple's strategy has also been to enable users of Apple products. It designs computing products such as the MacBook, the iMac, the iPad, and the iPhone. Like Microsoft, the center of its strategy is the assumption that these devices will be computing engines.

In contrast, Google and Amazon.com take a different approach. They offer computing services in the cloud. For them, the PCs and tablets and phones are not centers of computing; they are sophisticated input-output devices that feed the computing centers.

That Microsoft's and Apple's strategies revolve around the PC is not an accident. They were born in the microcomputing revolution of the 1970s, and in those days there was no cloud, no web, no internet. (Okay, technically there *was* an internet, but it was limited to a very small number of users.)

Google and Amazon were built in the internet age, and their business strategies reflect that fact. Google provides advertising, search technology, and cloud computing. Amazon.com started by selling books (on the web) and has moved on to selling everything (still on the web) and cloud computing (its AWS offerings).

Google's approach to computing allows it to build Chromebooks, light-powered laptops that have just enough operating system to run the Chrome browser. Everything Google offers is on the web, accessible with merely a browser.

Microsoft's PC-centric view makes it difficult to build a Windows version of a Chromebook. While Google can create Chrome OS as a derivative of Linux, Microsoft is stuck with Windows. Creating a light version of Windows is not so easy -- Windows was designed as a complete entity, not as a partitioned, shrinkable thing. Thus, a Windows Cloudbook must run Windows and be a center of computing, which is quite different from a Chromebook.

Yet Microsoft is moving to cloud computing. It has built an impressive array of services under the Azure name.

Apple's progress towards cloud computing is less obvious. It offers storage services called iCloud, but their true cloud nature is undetermined. iCloud may truly be based on cloud technology, or it may simply be a lot of servers. Apple must be using data centers to support Siri, but again, those servers may be cloud-based or may simply be servers in a data center. Apple has not been transparent in this.

Notably, Microsoft sells developer tools for its cloud-based services and Apple does not. One cannot, using Apple's tools, build and deploy a cloud-based app into Apple's cloud infrastructure. Apple remains wedded to the PC (okay, MacBook, iMac, iPad, and iPhone) as the center of computing. One can build apps for Mac OS X and iOS that use other vendors' cloud infrastructures, just not Apple's.

For now, Microsoft and Apple agree on the center of the computing world. For both of them, it is the local PC (running Windows, Mac OS X, or iOS). But that agreement will not last, as Microsoft moves to the cloud and Apple remains on the PC.

Slow and steady wins the race -- or does it?

Apple and Google run at a faster pace than their predecessors. Apple introduces new products often: new iPhones, new iPad tablets, new versions of iOS; Google does the same with Nexus phones and Android.

Apple and Google's quicker pace is not limited to the introduction of new products. They also drop items from their product line.

The "old school" was IBM and Microsoft. These companies moved slowly, introduced new products and services after careful planning, and supported their customers for years. New versions of software were backwards compatible. New hardware platforms supported the software from previous platforms. When a product was discontinued, customers were offered a path forward. (For example, IBM discontinued the System/36 minicomputers and offered the AS/400 line.)

IBM and Microsoft were the safe choices for IT, in part because they supported their customers.

Apple and Google, in contrast, have dropped products and services with no alternatives. Apple dropped .Mac. Google dropped their RSS reader. (I started this rant when I learned that Google dropped their conversion services from App Engine.)

I was about to chide Google and Apple for their inappropriate behavior when I thought of something.

Maybe I am wrong.

Maybe this new model of business (fast change, short product life) is the future?

What are the consequences of this business model?

For starters, businesses that rely on these products and services will have to change. These businesses can no longer rely on long product lifetimes. They can no longer rely on a guarantee of "a path forward" -- at least not with Apple and Google.

Yet IBM and Microsoft are not the safe havens of the past. IBM is out of the PC business, and getting out of the server business. Microsoft is increasing the frequency of operating system releases. (Windows 9 is expected to arrive in 2015. The two years of Windows 8's life are much shorter than the decade of Windows XP.) The "old school" suppliers of PC technology are gone.

Companies no longer have the comfort of selecting technology and using it for decades. Technology will "rev" faster, and the new versions will not always be backwards compatible.

Organizations with large IT infrastructures will find that their technologies are less homogeneous. Companies can no longer select a "standard PC" and purchase it over a period of years. Instead, every few months will see new hardware.

Organizations will see software change too. New versions of operating systems. New versions of applications. New versions of online services (software as a service, platform as a service, infrastructure as a service, web services) will occur -- and not always on a convenient schedule.

More frequent changes to the base upon which companies build their infrastructure will mean that companies spend more time responding to those changes. More frequent changes to the hardware will mean that companies have more variations of hardware (or they spend more time and money keeping everyone equipped with the latest).

IT support groups will be stressed as they must learn the new hardware and software, and more frequently. Roll-outs of internal systems will become more complex, as the target base will be more diverse.

Development groups must deliver new versions of their products on a faster schedule, and to a broader set of hardware (and software). It's no longer acceptable to deliver an application for "Windows only". One must include MacOS, the web, tablets, and phones. (And maybe Kindle tablets, too.)

Large organizations (corporations, governments, and others) have developed procedures to control the technology within (and to minimize costs). Those procedures often include standards, centralized procurement, and change review boards (in other words, bureaucracy). The outside world (suppliers and competitors) cares not one whit about a company's internal bureaucracy and is changing.

The slow, sedate pace of application development is a thing of the past. We live in faster times.

"Slow and steady" used to win. The tortoise would, in the long run, win over the hare. Today, I think the hare has the advantage.

Centralizing or decentralizing? What's happening now?

One might be confused with the direction of today's PC technologies. Are they becoming more distributed or are they becoming more centralized?

There are two large-scale changes in today's technologies. Depending on which you look at, you will see the trend for distributed control or the trend for centralized control. (Kind of like those optical illusions of stairs that go up, or down, depending on your visual perception.)

The distribution of PC applications are becoming centralized: Apple's business model is centralized. Applications for iPhones, iPods, and iPads are distributed through iTunes and through iTunes only. This is very different from the approach used by Microsoft for PC-DOS and Windows applications, in which anyone could write and distribute an application, and anyone could purchase and install an application (provided they had administrator privileges) without any involvement or supervision from Microsoft. Apple has complete control over iTunes and one can distribute an iPad/iPhone/iPod app only with Apple's permission.

Apple has declared intentions to move the Mac world from the "open distributor" model to an "Apple centric" model with the "App Store". Indeed, applications that use iCloud must be distributed within the App Store. Applications distributed via the "open distributor" model cannot use iCloud services.

Microsoft is considering to closed distribution model with Windows 8 and the Metro environment.

The selection and ownership of PCs are becoming decentralized: The "bring your own device" fad (for now, let us call it a fad) shifts the ownership of PCs from employers to employees. The previous model saw employers specifying, providing, and provisioning PCs for workers. Often a company would have a standard configuration of hardware and software, issued to all employees. A standard configuration reduced support costs, since there was one (or a limited number) of hardware and software combinations.

With the "bring your own device" fad, employees pick the device, employees own the device, and employees provision and maintain the device. One person may pick a Windows laptop PC, another may pick a MacBook, and a third may pick a Linux tablet. This is clearly a decentralization of decisions -- although the employer retains decisions for the development of company-specific applications. An employer may develop custom software for their employees and build it for one or a limited number of platforms. (Such as custom software for insurance adjusters that runs only on iPads. You can be sure that insurance adjusters at that company will select iPads.)

The two shifts are symptoms of larger changes: Software is becoming a commodity, with the important packages running on multiple platforms (or equivalents running on different platforms). Second, power is shifting from PC customers (large user corporations) to PC platform manufacturers (Apple, Google, Microsoft).

Software is a commodity, and the different packages offer no compelling advantages. For word processors, Microsoft Office is just as good as Libre Office. And Libre Office is just as good as Microsoft Office. In the past, Microsoft Office did offer compelling advantages: it ran on Windows, it ran efficiently and reliably, and it used proprietary formats that demanded that new uses have the same software. Those advantages have disappeared, for various reasons.

Software is a commodity, and current products are "good enough". There is little to be gained by adding new features to a word processor, to a spreadsheet, to an e-mail/calendar application. I may sound a little like the "we don't need a patent office because we have invented everything" argument, but bear with me.

The core packages used to run offices (word processors, spreadsheets, e-mail, calendars, presentation, etc.) are good enough, the data is interchangeable (or convertable), and the user interfaces are easy enough to understand.

If we need additional functionality in an office, a company will get it (either by building it or buying it). But they will do so with extra software, not through extensions to the core packages. (The one possible exception might be spreadsheet macros.) The core office software are commodities.

Apple knows this. It spends no effort building its own version of office tools. I suspect that Microsoft understands this too, and is preparing for the day when lots of customers move away from Microsoft Office.

Apple and Microsoft are building new mechanisms to extract value from their customers: walled gardens in which they are the gatekeepers. The application software will be less important and the walls around the garden will be more important. Apple uses its iTunes as a tollbooth, extracting a percentage of every sale. I expect Microsoft to do the same.

What this means for "the rest of us" (individuals, user companies, developers, etc.) remains to be seen.

Why the Microsoft tablet for Windows 8 is significant

This past week Microsoft  announced a tablet for Windows 8.

Microsoft has a checkered history of hardware. Their successful devices include the Kinect, the XBOX, the "melted" keyboard, and the Microsoft Mouse. Failures include the Zune and the Kin, and possibly the Surface. Many people want to know: will the tablet be a success or a failure?

This is the wrong question.

Microsoft had to release a tablet. The market is changing, and the rules of the new market dictate that vendors provide hardware and software.

Microsoft must succeed with the tablet. Perhaps not this particular tablet, but they must succeed with *a* tablet, and probably several tablets.

The old PC market had a model that separated hardware, operating systems, and applications. Various vendors built PCs, Microsoft supplied the operating system, and multiple (non-hardware) vendors supplied application programs. In the early days, one purchased an IBM PC and PC-DOS (supplied to IBM by Microsoft), and Lotus 1-2-3 and WordPerfect. Once manufacturers figured out ways of (legally) building PC clones, one could buy a PC made by IBM, or Compaq, or Dell, or a host of other companies, but the operating system was supplied by Microsoft.

People who are knowledgeable of the history of computing will recognize that the separation of hardware and software originates in an anti-trust lawsuit against IBM, for mainframe software of all things. Yet that decision influenced the marketing arrangements for the original IBM PC, and those arrangements (software separate from hardware) influenced the entire PC market. (To be fair, the pre-IBM PC market had similar arrangements: Radio Shack TRS-80 computers, Apple II computers, and others let you add software -- any compatible software -- to the computer.)

That model endures with today's desktop and laptop PCs. We buy the PC, an operating system (usually Windows) is included, and we add our separately-acquired software. That software may come from any source, be it another company or our internal development shop. We are responsible for configuration and upkeep, and for problems due to incompatible software.

Apple, with the iPhone and iPad, uses a different model: they supply the hardware and the operating system, and while other vendors supply the applications, Apple limits the freedom of users to select application providers. With iTunes, only those apps approved by Apple can get onto an iPhone or iPad. Users cannot select any application, or any provider, or even have them custom-written. They must go through iTunes (and therefore Apple).

The phrase for this arrangement is "walled garden". The environment is a pretty place, but one cannot leave easily. The vendor has erected a wall around the garden, and occupants must remain within the constructed garden.

Walled gardens, especially in tech, have a downside. When leaving one garden for another, one often loses content. You can buy a dozen books for your Kindle. Buy a replacement Kindle and your books are available. Buy instead a Nook, and your books are not available. Barnes and Noble knows nothing about your purchases with Amazon.com, and Amazon.com has no incentive to make it easy (or even possible) to transfer your purchases to another device. While we can leave the garden, we do so only by  leaving things behind.

This walled garden model is used by game consoles. Games are made specific to consoles; the XBOX version of "Diablo 3" will run only on XBOX systems, not on Playstations. If you purchased an XBOX and lots of games, and then you decide to switch to the Playstation console, you don't get all the games available in their Playstation form.

We are entering an age of walled gardens. Apple has their iPhone/iPad garden, Amazon.com has theirs, Barnes and Noble is building theirs. With the introduction of the Microsoft tablet, we can see that Microsoft is building theirs. Google has built some infrastructure with Google Docs and the Chromebook laptop/browser.

This new age of walled gardens, of separate kingdoms, requires developers, users, and companies to make choices.

Developers must select the platforms to support. They can focus their efforts onto a limited number of platforms, or they can develop for all of them. But development for each platform requires tools, skills, and time. A large company can invest in multi-platform efforts; a small company with limited resources must choose a subset, possibly forgoing revenue from the omitted market segment.

Users must select their platforms carefully. The cost of changing is high, much higher than changing from a Dell PC to an Asus PC, or changing from a Ford to a Chevy.

Companies must select their platform. They have done so in the past, usually picking Microsoft, the safe choice. (How many times have you heard the phrase "we're a Microsoft shop"?) But now the choice is riskier; there is no safe choice, no one dominant provider (and no one provider that appears ready to become dominant). A company's success will depend not only on its talent and ability to execute business plans, but also upon the success of its platform. A company may do well in its market, yet fail when its technology provider fails.

These are difficult decisions, and they must be made. One cannot defer the selection of platforms; others in the world are moving to new platforms.

To wait is to be left behind.

Discoveries about discovery

Recent developments in tech have created an automated process to handle "discovery", the process of reviewing materials for a legal case.

One might think that law firms will adopt the technology, as a way to reduce costs. Or one might think that law firms will *not* adopt the tech, believing that they are traditional and unwilling to change. Or perhaps one might think that law firms are risk-averse, and do not want to try new tech that could miss something and cause the loss of a case.

I have a different outlook. I think law firms will avoid the tech of automated discovery, for economic reasons.

Law firms use the time-and-materials business model. They bill by the hour: the more hours, the higher the bill. Law firms have constructed their hourly rate to cover their costs (including labor) and to provide profit. Thus, each billable hour generates profit (not just revenue) for the firm. A reduction in labor (billable hours) equates to a reduction in profit.

Industries that have adopted technology (specifically to reduce labor hours) are industries that sell products with prices fixed by the market. Automobiles, hair dryers, books, computers... these are all sold at the market price. Profit is revenue less the cost of goods and manufacturing. The costs of labor and goods does not drive the price, and a manufacturer must manage costs to live within the market price. A reduction in labor hours equates to an increase in profit.

Thus, we can expect that businesses selling goods (or services) at prices dictated by the market will adopt cost-reducing techniques and technologies. They will computerize accounting systems, automate assembly lines with robots, and outsource software development.

We can also expect that businesses selling services on a time-and-materials basis will *not* adopt technologies that reduce labor hours. (They may adopt techniques that reduce labor costs, replacing highly paid workers with lower-wage workers. But a reduction in billable hours is unlikely.)

What of this can we apply to software development?

The software development industry is a complex one. Some software is created and sold on a time-and-materials basis. Some is sold in the market. Some is given away. Some software that is sold is not sold in a free market; Microsoft enjoys a monopoly position with Windows and Office. (A monopoly that is perhaps less strong now than ten years ago, yet still part of the complexities of the market.)

Companies that build software with the time-and-materials business model have little incentive to reduce the workload in their projects. These projects benefit from high labor efforts, and therefore we can expect them to use little in the way of effort-reducing techniques. Companies that build software for sale on the open market have strong incentives to minimize their costs. Cost reduction methods include:

• outsourcing
• agile development (pair programming, automated tests)
• modern languages (Python, Ruby, Scala, Lua, Haskell)

Companies in the time-and-materials business model do not need such cost-reducing measures. If you're not working with the above techniques, then you're probably on a time-and-materials project. Or you're in a company that will soon be out of business.