What, exactly, constitutes a desktop, and for that matter, what exactly constitutes an operating system? I beg a bit of an indulgence here in order to talk about a couple of cool apps and a couple of very disturbing developments, then would like to come back and reconsider these questions in light of this.
I am writing this particular post in a little Google applet called Google Notebook which is nothing exceptional (basically just a rich-text-edit control with a small piece of back end storage) that’s nonetheless quite exceptional in what it implies.
I am a writer, so spend a great deal of time, well, writing … putting words on a page, though the definition of that last term has changed considerably over the years. For some tasks - writing books, for instance - I need to have a heavy duty word processor capable of advanced stylesheet management, though I have to admit that in recent years most of that is accomplished (for me at least) using docbook in Oxygen. Yet for a great number of other tasks, I find that opening up a full bore editor is a time-waster and overkill … I want to put my thoughts down, that’s all, and by the time I wrestle MS Word (or more accurately Open Office, which is the only word processor that I have on my system at the moment) open, those thoughts are generally scattered hither and yon.
More to the point, those thoughts, even once transcribed, end up in a file in a folder among literally tens of thousands of folders, and not surprisingly it is all too easy to misplace that critical word document. Something as simple as Google Notebook presents information in order of when you wrote it. usually with the first dozen words or so appearing as the “title”. You can add headings to the notebooks, making it possible to create interrelated pieces of content. Oh, and as a little bonus, if viewing this in Mozilla Firefox 2.0 or above, it also has automatic spell-checking thanks to the new spell-check features of FF2. It’s not perfect - you still have to manually save content, rather than having it auto-save (something which forced me to lose a bit of content before I realized this), though I suspect that this may very well change once the system moves out of the lab and into prime time, and you can easily switch from a weblet view to a full screen view and back with remarkable ease.
Expanding into a full screen mode also serves to bring up a somewhat fuller set of rich text functions, as well as to give you access to even MORE functionality - you can rename a notebook or add a new notebook, you can print the entry or the full notebook, and you can set the various sharing options. This is when you begin to realize that you are no longer in Kansas anymore, Toto. In terms of collaboration, I can send a private access link to those people who I wish to collaborate with so that they can both read it and, if appropriate, edit it or I can set the option to make the page “published” on the web with a distinct URL. All it needs to shift into the mode of being a blogging tool is the infrastructure support for RSS feeds (which I have no doubt are in the works) and the distinction between writing notes to yourself and blogging becomes largely moot). While it appears to generate HTML content, I suspect that if I dig around long enough, I can probably find an XML option, which makes XIncludes in particular an interesting proposition.
I have to admit that I’ve become something of a convert with regard to online storage and editing of documents. The biggest problem that exists for such online content is the fact that you are in fact trusting a third party to hold your content. This is not an insignificant issue, though there is also something of a nuclear strike option going on here. Is it worthwhile for Google (or any other company) to sell the contents of your notes to the highest bidder? No … in general it isn’t. Should such a thing happen (or even if such a system was hacked), the reaction would be swift and merciless, with people basically retrieving their documents as quickly as possible and then deleting them from the system, before going back to their own machines and their own (perhaps marginally more secure) sandboxes. In short, no matter what Google’s market cap is, such a breach of trust would very quickly spell the end of the company. (Indeed, I suspect that the biggest challenge that Google faces with such a service is in keeping the sharks from creating bogus reports of such breaches in order to blackmail the company). So there is an issue there.
The second key issue with regard to online documents is the question of what to do when people are offline. The gamble that Google is taking here is that bandwidth (especially wireless bandwidth), has reached a stage where in most cases people can get online to update or access their content. I don’t think its an unreasonable gamble. I find that I’m in general no more than ten minutes away from most wireless connections, and any time I do a presentation to a client I usually have an online connection available as a matter of course, and I suspect I’m far from atypical in that regard. Indeed, it doesn’t take much to envision a Google based offline extension for applications such as Firefox that would make such text services at least partially usable in an offline mode - I could write the interface in XUL in about two hours - so most of the key work comes down to the integration aspect (which I suspect, given their reliance on Atom elsewhere, is probably using the Atom Publishing Protocols (APP) so even this should be a fairly simple proposition).
Stepping back from the Google Notebook for a second, I think its worth highlighting the corresponding Google Home Page concept, because I see in it for the first time the concept of content dashboard done correctly (in nearly a decade of looking at similar portals). Dashboards are certainly nothing new, but all too often the reason that dashboards failed stemmed not from the concept - which when implemented correctly is quite useful - but from the mistaken notion that the only agencies who should be able to put up dashboard components are paying corporate customers. The irony is that in most cases, these corporate customers are the ones I’m least likely to wish to get content from. I’m sorry but I’m not interested in the latest ad campaign from GM or Coca-Cola or The Gap. Very, very occasionally, I’d like to be able to get coupons and sales from my local supermarket in a weblet (and THERE is an untapped market if I’ve ever seen one), but for the most part what I’m looking for falls into a few basic categories - I want usable applications that make my day easier or make it possible for me to do specific actions, I want news (whether image, audio, or video) from sources I specifically request (and some of them are very obscure), I want reference materials - the periodic table of the elements (or better, the entire CRC reference), APIs at my fingertips, Wikipedia access, and I want communication tools - messengers, email and RSS readers and editors and so on. I need to be able to prioritize the display of this information in a rapid and intuitive manner, and I need it to be there for me whenever I log in, just as I left it.
I think that the Google Home Page has succeeded in this to a degree that people are only beginning to realize. They have created minimal bandwidth versions of two key pieces of functionality that are accessible nearly instantaneously: word processing and spreadsheets. More to the point, they have opened the API so that others can also create additional pieces of functionality - some of it banal (Homeland Security Warning icons combined with pictures of the Power Puff Girls), some of it surprisingly sophisticated (Flash based gaming will likely do quite well under this setup), and all of it easy to add at no cost to me. This of course is the ultimate paradox of the “information age” - there is a zero point energy level that creates an energetic false vacuum in cyberspace (I’ve often though that you could do an interesting paper on the similarity between Bose-Einstein particles (bosons) and information - may write it one of these days) - that sustains such desktops as Google even in the absence of (or perhaps especially in the absence of) money changing hands. The challenge that the Yahoos and Microsofts of the world have is that they are still beholden to the older corporate model of the world, and tend to denigrate their user generated content as being so much fluff. Thus when the Web 2.0 explosion occurred, these companies were frankly caught flat-footed … they were too busy servicing their corporate clients to realize that the real activity was going on in the apparent fluff - My Space, YouTube, Flickr, and so forth. There’s a certain level of Zen that one needs when dealing with the web, and I suspect strongly that at least for the moment Google has achieved that state far more readily than Microsoft has.
This is perhaps most recently been illustrated in a recent open letter by a security analyst named Peter Gutmann that I have no doubt is causing more than a little nervousness at Microsoft. Entitled A Cost Analysis of Windows Vista Content Protection ,it examines the complete rearchitecture of Vista’s hardware and software driver specifications that are designed to specifically provide system level digital rights management to content providers. It’s a damning piece, especially so given that it was not written with the usual Microsoft bashing rhetoric but instead focuses on many “improvements” to Windows that may have the effect of destroying the PC market altogether. Among the charges that Gutmann levels is the fact that video and audio chip manufacturers must deliberately downgrade their processing in the face of any protected content to the extent of potentially making the system unusable - the system would revert to a much lower resolution if DRM content is attempted to be played without the proper money being paid, hardware would need to provide 30ms interrupts to their devices just to insure that no one is attempting to hack such content, device manufacturers would be largely prohibited from building chipset templates that create a common architecture base that would just have specific ports or connections disabled for lower end devices because of the potential that they could be hacked to steal DRM content (adding significantly to the cost of the manufacture of these devices), and that fault tolerances in such devices would be disabled because of the potential for hacking, meaning that such things as a small spike in power could effectively shut down an entire building of PCs.
Read the article. At first, his executive summary statement:
The Vista Content Protection specification could very well constitute the
longest suicide note in history.
seemed to me to be so much hyperbole, but after reaching the end of the article I began to realize just how accurate he may have been. Microsoft has chosen here to kowtow to the media providers in order to get the “desirable content” from them and become the de facto premium distribution channel because such content would be unplayable anywhere else. My suspicion is that this may very well spell the death knell not only for Microsoft but for these same content providers, because it will make out of band (i.e., hacked) content far more desirable - it won’t cause your computer to come crashing to a halt every time you try to play a TV show, and it will probably end up working with the nucleus of the film community that’s now establishing itself over at YouTube to accelerate this next wave of entrepeneurial video studios. While the vast majority of YouTube content providers are producing marginal content, I’ve noticed of late that there are an increasing number of professional production houses that are going straight to YouTube … when the audience goes there, so do the advertisers - and all of a sudden Microsoft and the DRM Content producers have created the exact nightmare that they were attempting to thwart.
So what, exactly is a desktop, or an operating system? If I can get to most of my workaday applications from a browser, do I really need to have all of the bells and whistles of a Vista? If my resources are available with a click from the web do I need to have an oft promised, never delivered universal file system? When HTML or XML (and yes, the occasional flash or PDF) is good enough for almost all of my application needs, do I need to have the rich capabilities of a WPF? Especially when the bill comes in - handing to the manufacturer of that operating system the sole monopoly for distributing content?
Microsoft hasn’t learned its lesson from Google - the power in the web does not derive from those who hold the existing content, but rather derives from the ease by which those who may create new content can do so in a manner that is secure, trustworthy, obvious, and ultimately in the best interests of the new content creators. Those new content creators are far more likely to be twelve year old kids putting up the latest “anime mix” or may be studios that realize they won’t get their shot in the established channels and so work into the new ones. They won’t staunch the flow of piracy that has the DRM studios so much in arms - they will only make such piracy more rampant, because people who are treated like criminals when they aren’t will see less and less value in remaining virtuous in the face of such efforts. Moreover, even advertisers are beginning to realize that the Battlestar Galactica phenomenon is very real - people don’t watch BG on television, they download it or view it in YouTube in numbers that by conventional standards far overwhelm the anemic numbers from TV or the movies, and it is far easier to actually observe that particular metric than it is to rely on the increasingly archaic Nielsen ratings. Advertising, not subscription, has always been the lifeblood of the video and audio industry, and so the metrics go, so too do the advertisers.
So, in the countdown to 2007, I can now see the shape of the new real operating system … one that doesn’t have big fanfares and major release numbers and significant rollouts but instead simply is there, is organic, and evolves with time. It is pretty much ubiquitous, doesn’t require a $200 license to play, and for the most part works not by pushing the edges of what’s possible with the hardware at hand but by working with established technologies and tools, albeit sometimes in very new ways. It isn’t produced by any one vendor, and it doesn’t penalize a person because he doesn’t have the latest and greatest hardware and software and the money to spend on all that premium content. It recognizes the value of advertising - advertising is, at its sole, the nervous system of capitalism, and in its place is valuable - but it doesn’t bend to the point where the advertisers can corrupt the system. The new operating system places the emphasis instead upon the individual participant, not as a captive user or potential customer but as a creator in his or her own rights, no matter how simple the creations. In an anemic PC market, heading into the spectre of a major economic downturn, I think I’d place my bets on the simple desktop of Google rather than the late, bloated and hostile operating system that’s being foisted off to the world in the next couple of weeks. KISS - Keep It Simple, Stupid - is a principle that Google fundamentally gets and Microsoft … doesn’t.


I found Gutman's article yesterday, and laughed out loud at how badly Microsoft misses the point of the Internet still ... and don't even get me started on the RSS patent!
you're on crack.
And the content providers keep taking it up the snoot while the technologists take the money. This bit is funny, Kurt:
"Microsoft hasn't learned its lesson from Google - the power in the web does not derive from those who hold the existing content, but rather derives from the ease by which those who may create new content can do so in a manner that is secure, trustworthy, obvious, and ultimately in the best interests of the new content creators."
That is so but it is so far off the mark from what happens. Ultimately, it is simply the shift of power from owning the software operating system to owning and securing a server farm, so from software to hardware. The only thing it relies on is the willingness of the content authors to submit to the rules and dictates of the server-farm owner. Ask the SL graphics artists about Copybot. So far it hasn't slowed Linden Labs down a bit. Your mythical uprising doesn't happen as long as the power stays on the other side of the pitchfork. Make as much noise as you like; without a means to retaliate, it's just noise and the vaunted 'we'll walk out' just doesn't work without the power of the law behind it. Union busting pretty much did away with that option.
I won't even bother to quote the Who this time.
No way I'm giving up the desktop. Not a chance in hell.
len
Well, one vote for crack and one for server farm control.
Not on crack - I have a hard enough time dealing with caffeine - so on to Len ...
I'd differ with you on this one (and I don't differ with most of what you say) though not because I think Google is warm and cuddly (I think the company is still relatively young, but like most large companies the imperative for corporate survival eventually outweighs the mandate to "do good".
However, I think that Google is nowhere near so secure that they can afford to lose the good graces of their customer base. Momentum CAN shift quickly on the web - far more quickly than it does in the brick and mortar world, even in areas such as politics. Microsoft is largely immune to this effect in part because their distribution channel is usually at one remove from the customer - they make very specific deals with the OEM, and that OEM knows that they would be hard pressed to sell any PCs at all if all they offered were Linux boxes (Apple gets away with it by being a lifestyle choice, not a technology choice, and even so Apple computer sales have been sluggish even in the face of high demand for the IPod.
Google exists on quicksand, and they know it. Their firmament is shored up somewhat with useful/helpful apps that can also bind users to their services - gmail, notebook, spreadsheets, et al., but overall there is an implicit agreement there about the relative security and privacy of the data being tied into the agreement to use the service in the first place. What does Google get out of the relationship? It basically becomes the portal of choice for the user, which means that ad revenues will always start from this point whenever the user logs on. Ditto for Gmail.
Admittedly, the monopoly of the server farm is always a possibility, and I think for this reason you need to temper utility with a healthy dollop of paranoia. If I send an email with a manuscript or some other item of value, that manuscript will ultimately end up residing, at least for a time, on a server somewhere. If I am concerned about the trustworthiness of the server, then I put my manuscript on a USB key and send that, and if I am concerned with the trustworthiness of the mail or express service, then I get into a car and drive to the recipient, and hand the key to them directly. I am trading trustworthiness for time and money. I see Google as being trustable for an arbitrary value of trust, just as there are those who see Microsoft in the same light.
The shift to cloud based services is happening now, largely because people have chosen to trust certain entities to handle their data (or are simply ignorant of the process, which to me entails a trust of a different sort). I think that people will likely end up setting up for themselves the balance between what can be safely kept on the cloud and what should be kept closer to the vest - the difference is largely who owns the data clouds.
I WAS oversimplifying things somewhat in the discussion concerning operating systems - the desktop is of course only one program running on top of this OS. Realistically, however, what seems to be emerging is that by abstracting many of the common features of the desktop (through the agency of the browser) companies such as Google or Yahoo or Microsoft are able to function across systems. In many ways, the decision by Microsoft not to create an Internet Explorer for Linux (and to kill support for it on the Mac) may very well prove to be disastrous in the long run, because what this ends up doing is making cross platform browsers such as Opera and Firefox far more attractive in the long run.
Whether I think that storage on Google (or Amazon, or anywhere else) is preferable to that of keeping it on your own drive, I suspect in the long run the advantages of online access (the ease of sharing, or publishing, of replication, of collaboration, and of upgrading) will end up making this migration pretty much inevitable ... for good and bad.
I'm still holding on to my desktop until they pry it from my cold dead hands. :-)
We agree on the matter of trust. I still don't want to use in-house tools to build 3D worlds. Here I need the performance I get from being local and I don't have broadband. OTOH, I've no problem with using the in-house tools such as this text edit box I'm typing in now for that which once posted, I'm unconcerned. Further if an employer was relying on the cloud, I'd be happy to join the choir there. But the cloud isn't really that; it is a server farm and with those come problems that I won't suffer to risk high value content. I suspect the rising costs for energy and green legislation will begin to impact those businesses soon, and I don't want to be unable to work when the lights on the West Coast go dim because the next Enron decides to make a few dollars more just at the oil companies do.
I think Google is more secure in their position than you do but you are right that any pure services play is more vulnerable to the zeitgeist than an operating systems vendor. MS doesn't have that many worries although some of us keep making up more. I expect Vista to have a rougher take off than XP though. Simply, it comes down to a lot of people not needing a new operating system today. It gets tougher to push them off the old one now that the web offers other options and what they want to do and need to do they can do with what they have. Features fatique sets in.
RE: caffeine. THere is never enough the older I get and the really nice caffeine venue in the local mall was just replaced with YetAnotherStarbucks. That sucks.
Len,
I'm not saying that I think you can get by with a pure server side solution - at this stage that would be suicide - broadband or no broadband. I look upon network access as being another pipe, and one that has far higher latency than any local pipes. You distribute your latencies appropriately.
I agree with that, but think it all the way through. Why are the tools simple vs complex? Is complexity a side effect of something more important?
The idea that all data lives in the cloud on the net is sort of dumb. The idea that the tools that do best are those that are in-world (in the 3D parlance, in the server-side world illusion) makes sense in the same way a MAC does: close it, keep it simple for the simple people, and control all avenues. Eventually, you control all of the content too. That's the idea for Second Life too and it is successful in the same way Google is and for the same reasons. Ease. On the other hand, security isn't provided and griefers raise hell at will.
The social issues ARE the issues that make the web a hard problem.
The Apple commercials make the PC user look like an overstuffed untalented and ultimately uncool user. The truth is, the PC user is the person who places a high value on their own freedom to control their environment and is willing to be responsible for that. Complexity doesn't faze them because they actually are smarter than the Mac user.
That's why I think where you were headed in that article is wrong. I'll take the cost of the complexity to keep the market open. You are blindly headed down the path where all your data is in the clouds but your head ends up elsewhere unable to see it.
The operating system of the future will make more use of the local disk and peer-to-peer computing. The Google model works for some applications but not all. There we agree.
http://www.siliconvalley.com/mld/siliconvalley/16313374.htm
That's and example of how Google loses. Note the comment about minimalist platforms. Note also that Anshe Chung took her 3D content business to Wuhan. It isn't about simplicity anymore. The T-model era is passing fast. It is about features and cost: the GM is God era of automotive manufacturing. The impact on that market is energy, the devil in the server farm works.
Sun is replying to that one. It will be interesting to see what has to go to hit the sweet spot.
@len, kurt,
Interesting thread -- plan to extend some comments upon my return tomorrow.
Hope your holidays have been fund :D
uhhh... make that "fun", though fund wouldn't be such a bad thing either ;)
An OS controls hardware. What is being talked about here is a supplement to Operating Environment. It feels like Plan 9 all over again...