Showing posts with label google. Show all posts
Showing posts with label google. Show all posts

Sunday, May 27, 2012

The Software Adoption Spectrum

I'm sure many have seen so-called "political spectrum diagrams" like this:

Libertarians in particular have used such diagrams to make clear their distinction in the American political spectrum between liberalism and classic conservatism.  A particularly commonly cited example is the Nolan Chart, devised by American Libertarian Party founding member David Nolan.  another good example was created early in his academic career by science fiction author (and, yes, libertarian...it sort of goes w/the territory) Jerry Pournelle.

These diagrams are useful in that they serve to dispel the myth that political orientation and issues map solely to a linear spectrum of "right vs. left".  They also look cool.


It occurred to me recently that a similar spectrum diagram could be used to map the relative advantages/disadvantages of various desktop operating system solutions.  After giving the matter some thought, this is what I came up with:

My "Software Adoption Spectrum" chart maps a little differently, in that it takes four specific criteria or metrics for evaluating desktop solutions, and ranks them from 1 to 5.  A "perfect" solution that ranked at the top of the scale for all four criteria would map out as a circle as large as the spectrum grid itself.  A completely flawed solution that failed all four criteria would be a dimensionless point in the middle of the grid.

Where this is potentially useful is determining whether or not a particular solution makes sense for a particular user.  Even though the effort required to switch, for example, from Windows to Linux is greatly exaggerated there's still some effort involved.  This chart is a tool to determine if the effort is warranted or not.  I'll come back to the chart in a bit, but first I'd like to look at the four criteria/metrics and rank each of the main Desktop OS solutions--Windows, Mac OS, & Linux-- in each case.

Control

Control is just that--the degree to which your computer can be made to do what you want it to do, the way you want to do it.  One of the principal attractions of desktop Linux, particularly to the technically inclined, is that control over your personal computer is limited solely by your own knowledge and expertise.  Given time, expertise, and the willingness to do so, you can essentially build a working Linux desktop system from the kernel up.

But even lacking that level of commitment, you can make fundamental decisions about how your system operates and how you interface with it that neither Windows or Mac OS permit.  A Linux desktop user can pick between file system managers, desktop shells, office packages, and much more.  Windows has, at various times in the past, offered some options for customizing individual systems.  Mac OS started out very rigidly controlled and has since become flexible on minor issues. My personal ranking of Windows, Mac, & Linux on the issue of "control" is as follows:

OSRank
Linux5
Windows2
Mac OS1

Security

Security is the measure of how likely your system is to be hacked or compromised.  Although Windows has improved in this area, it continues to be the most problematic of the three main desktop OS solutions.  Mac OS continues to be more secure, although the recent "Flashback" exploit has proven this to be a relative measure.

Desktop Linux continues to be, by a large margin, the most secure desktop OS solution available.  While both current versions of Mac OS and Linux share a UNIX-derived secured architecture heritage, the open nature of Linux ensures that security issues become widespread knowledge as soon as they occur, and thus wind up being fixed.  The Flashback exploit that recently afflicted Mac OS was, to a certain degree, a self-inflicted wound brought about by Apple's insistence on centralized corporate control over patches to community-maintained software (in this case, Java).  On the issue of security, I rank the major OS products like this:

OSRank
Linux5
Mac OS4
Windows2

Convenience

Convenience is the measure of how easily you can use a given Desktop OS in the first place, as well as a measure of how easily you can collaborate and share data with other users.  Windows ranks at the top for this metric, by virtue of its sheer ubiquity.  If you bought a personal computer with the desktop pre-loaded, the odds are better than 90% it came out of the shipping crate running some variation or other of Windows.  If you need to share documents with other users, the odds are probably closer to 95-99% that their first preference is going to be a document created in Microsoft Office--and again, your computer came out of the crate with an evaluation copy of Office pre-loaded, just like the evaluation copy of anti-virus software, both of which you or your employer wound up shelling out additional cash to keep running.

This being said, Windows does not rank as highly in advance of its competitors as you might think.  Office is available for Mac OS, of course, albeit w/stripped down functionality.  If you happen to have legally licensed copies of Windows and Office, you can easily set up a stripped down virtual PC running office on your vastly more secure Linux desktop machine.  You can also produce files that can be read in MS Office in any of several Open Source office suites (again, w/stripped down functionality).  You can also just share files via Google Drive.  For the moment, we will leave aside the supposed (and utterly inflated) inconvenience of installing an alternate operating system.  Looking only at interoperability, the convenience rankings work out about like this:

OSRank
Windows5
Mac OS3
Linux2

Privacy

At first glance, "Privacy" would seem to be virtually the same concern as "Security"--it isn't.  Security is the measure of how secure your system is from having either data or applications compromised by hostile third parties.  Privacy is a measure of how secure your pesonal information is from either internal or external prying, either by on-line third parties (including government agencies), other users of the same computer....or the manufacturer of your computer's operating system.

In terms of this metric, Desktop Linux is the clear and indisputable winner.  Both Windows and Mac OS "phone home" varying degrees of information about what you and your computer are doing.  Linux does not.  Every major Desktop Linux distribution offers the opportunity to encrypt the "home" directory containing all of your personal information, protecting it from any incursions, even local ones from other users of the same computer.  All large corporations are susceptible to government coercion (I'm looking at you, Google), but if your means of accessing personal information stored "in the cloud" is potentially compromised in the first place, the odds of losing your privacy is vastly lessened in the first place.  Why this matters is beyond the scope of this article.  Anyone who thinks it doesn't matter is probably wasting their time reading any of this.  In any case, the ranking looks like this:

OSRank
Linux5
Mac OS2
Windows1

Results

If we use these ranking metrics to create elliptical domains on the software adoption spectrum chart, it looks something like this:
I've removed the scale number and replaced them w/arrows to make the chart easier to read, but the results otherwise pretty much speaks for themselves.  Microsoft Windows has the smallest of the three domains, weighed heavily toward convenience but falling short in all other measures.  Apple's Mac OS has a larger footprint, almost matching Windows in terms of convenience, but almost completely sharing its drawbacks in terms of security and privacy, and suffering from even greater shortcomings in terms of control.

Unsurprisingly, Desktop Linux has a larger domain on the chart than the two leading commercial OS's put together.  To some extent this reflects my own personal software preferences, but I think that even more it indicates what a poor metric mere market share is for decisions on software adoption and information technology decisions in general.  A great many people use MS Windows on their computers.  A great many people watch "American Idol"... in neither case is the possibility of consuming rubbish precluded.

This chart is no more going to convince anyone to adopt Desktop Linux than the Nolan chart is going to convince someone to join the Libertarian Party.  But, hopefully, either Nolan's chart or mine might provide some useful perspective outside the usual propaganda for anyone weighing their options.  I'm not trying to make converts in any case...I just like writing about this crap.

A word about Tools

Of course I can hardly fail to mention that I created the two "Software Adoption Spectrum" images in the very excellent Open Source drawing program Inkscape, performing final edits in the ever-reliable OSS Photoshop equivalent GIMP.  Both tools are available free of charge for both Linux and Mac OS. Inkscape is also available for Windows... if you're into that sort of thing.

Saturday, May 26, 2012

Thoughts on Google/Motorola

Even though this blog is mainly about Desktop Linux, its impossible to ignore the impact of the ongoing migration of information users to mobile devices-- or the potential impact on the entire computing world of Google going into the hardware business.  Also, there is possible model for Google/Motorola going forward from the history of desktop computing that bears consideration.

It is, of course, impossible for an outside party to guess the direction Google intends to take with their new division--but anyone with any familiarity with the history of desktop computing can see an obvious parallel--and potentially beneficial business model--by looking at the relationship between Microsoft and Compaq over most of the 80s and 90s.

A little history...

pretty much where it all started--Apple II
A recap would probably be of use to those of you joining this game in progress.  In the early 80's, Microsoft's fortunes were essentially made by IBM's decision to develop a personal computing platform to compete with the Apple II--but not develop their own operating system.

That opportunity went to a small startup called Microsoft.  Whether or not the MS-DOS operating system was developed or merely stolen  is at this point a matter of little more than academic interest.  Of considerably greater import:  IBM's decision to make their licensing agreement with Microsoft non-exclusive.

As a consequence of this decision, Microsoft was as free to distribute their desktop operating system to hardware vendors then as Google is free to distribute Android to Mobile Phone manufacturers now (although the software itself was hardly free). 

the great-granddaddy...
This encouraged another small start-up venture to create a product that has essentially changed the world.  The Compaq Portable was the first so-called "PC Clone", although it hardly looked much like an IBM PC.  Designed as the first mobile computing device, this suitcase-sized device was basically the great-granddaddy of that smartphone in your pocket.

There were soon many other PC Compatibles on the market, with Compaq itself rapidly following up with it's DeskPro non-portable product.  Soon enough, microprocessor-based computers also diversified into laptops and servers as well.



...and why it matters--the Compaq/Microsoft Model

Where this serves as a model for the future development of Android-powered devices is this:  until their acquisition by Hewlett-Packard in 2002, Compaq continued to produce the "best in class" of DOS/Windows powered desktop computing devices, collaborating closely w/Microsoft in the process.  Dell may have managed to out-compete Compaq in sales volume and price, but there was never any real doubt among knowledgeable users as to which company produced the better computers.

...of this--in more ways than one
OK, back to the current millennium.  Albeit as an owned subsidiary, Motorola can be Android's "Compaq"--one of many companies that produce hardware running the same operating system, but the one routinely acknowledged as the standard setter.  As an Android user, this is the model I would personally like to see.

In order for this to occur, Google has to be serious about maintaining a firewall between their hardware and software divisions.  If mobile device manufacturers come to believe that Motorola has an unfair advantage with Android, they will start looking for another OS solution--possibly the now open-sourced Web OS.  Fear of such anti-competitive behavior has much to do with China's stipulation in agreeing to the merger that Android must remain an open platform for a minimum of five years.

An example that Google should not follow is the fairly porous "firewall" that continues to  exist between Microsoft's Windows and Office divisions, which only exists because (a) Microsoft's de facto monopoly status in the corporate/enterprise software market lets them get away with it, and (b) you can pretty much do anything you want with closed-source commercial software.

this sucks...stop doing it
Motorola as a best of breed Android device manufacturer can also serve as the remedy to the single biggest problem Android faces as an OS platform.  While the problem of "fragmentation" is greatly overstated by Apple fan-boys (who, let's face it, like being told what to do w/their hardware), it's still an issue, and an inevitable one given Android's openness.  In order to differentiate themselves in the market, Android device manufacturers "improve" the Android experience w/irremovable desktop shells and social media widgets.  Motorola has itself been one of the worst offenders.  Motorola Android devices should run Android...period.


almost as bad as Win8, IMHO
How this plays out is going to matter to Desktop Linux users in any case--let's not forget that Android is just a specialized shell on top of Linux.  The vast number of Linux boxes quietly and reliably chugging away in server rooms around the planet ensure that the innovations and improvements will continue to occur at the kernel level, but do nothing to improve the Desktop Linux experience.  Android (and iPhone) mobile devices have already had an enormous impact on the development of Desktop Linux.  Whether you love it or loathe (count me among the latter), it's hard to imagine Ubuntu's Unity even existing, absent the influence of handheld devices on user interfaces over the last few years.

Google has just been handed an enormous opportunity to make good on the years of goodwill and somewhat naive trust they've received from the Open Source software community.  There is at least one time-tested and well-documented way they can do this right.  We are all about to find out how "non-evil"--and smart-- Google really is.