"so what's missing?"

Why this article is on MacDevCenter. :P

Dan, I've been asking myself the same about a lot of what's been posted here lately...

Google should have bought been a little more patient. ThinkFree Office Online would have been a much better purchase. I bet there's a few folks at Google feeling a little ill when they look at ThinkFree. My money's on Yahoo to buy ThinkFree.

In this respect I don't know if my opinion is representative or not, but I don't want applications that don't need to be Web-based to be Web-based. It makes a degree of sense that email is a Web-based application since you can hardly send/receive email without a connection to the Internet, but for other productivity applications I see a Web-based nature to be a hinderance rather than a benefit. Sure, it'd be nice to always have the latest version of a piece of software without needing to do anything and to reclaim the disk space taken up by it, but you're completely hosed when you are either travelling without a Net connection or your existing one experiences an outage (a frequent occurance for me at the moment). I'm very happy for the likes of Google to try and improve the software that is available but I want it to work without a connection to the Internet unless I can be guaranteed to always have a connection available.

Guys Kelmon has a pretty good point.Even for email i would not like to be dependent of availability of network. I would certainly like to write my mails and once connected all outbox stuff is pushed out. Using google or even other webbased mail this is possible using a standard pop/imap MUA.

Now for the the office documents why not have a similar 'Office User Agent', when offline you can write your docs, sheets, slide shows, and when network is available you can sync it.

Let's get back to Joe User's perspective -- it seems to me that Joe User would usually be doing most of his/her work at home or at work, and it's fair to assume that both locations readily have internet available. Even if Joe is at a coffee shop, or elsewhere hanging out, at the airport, or in a hotel, it's more than fair to assume internet connectivity -- probably for free or a nominal fee. And that's now -- connectivity is only going to continue becoming more and more ubiquitous. Even if you had to pay for a slightly faster connection, the costs would be more then regained by not having to buy software licenses.

While I can understand the rationale for not necessarily wanting apps like spreadsheets to be net-centric, I have to think that Joe User doesn't really care (although you and I may care.) If the app does what it needs to, is responsive, is available at home and at work (for now anyway), and is *free*, then what could be better than that? (Ok, so I suppose we could venture into privacy concerns if you want; that's a good trump card here.)

Along similar lines, suppose you're a small business -- having a netcentric office suite that's available for free or a nominal fee would be a god send. No more maintenance, no more upgrades, no more virus/worm problems (hopefully these issues would become transparent to you) -- your IT budget would decrease significantly if you do what many offices do and technology would be less of a burden on you. Truly, and office suite would have become a commodity at this point.

But that's enough from me. And BTW, I'm not head over heels for a net-centric office suite personally -- but I do think it's up and coming, we need to be aware of it, and I do think that it'll work wonders for a lot of casual users and small businesses.

As for why this post is on Mac DevCenter -- Mac users are just as net savvy as users of any other platform, so why not post it here? Besides, may Mac users don't like MS-Office and aren't exactly crazy about NeoOffice either -- so an up and coming Google Office should sound like a pretty exciting alternative at this point.

Is the Mac connection to this that all of the information points to Google using a Mozilla product and switching to KHTML?

:)

Nice pipe dream. Won't happen.

Extending from Kelmon and Jamshed...

There are two an EXTREMELY important piecse to this that people seem to just look over without even a thought.

1 - Control of ones own data.

The notion of having globally accessible virtual hardrives is an exciting one. In fact, I am doing quite a bit of work in this area as of late (see: http://dev.extensibleforge.net/wiki/GlobalClip) so its one that I have a strong belief to be an important piece of the bigger picture. That bigger picture requires that I have at ALL times local and controlled-by-me access to all of my data, regardless of whether or not I have a connection to the internet.

The internet connection is not the important part, although it adds a layer of incredibly useful and nice features when it is. The important part is secure, reliable, data storage that I control, and applications that use this data storgage to access, edit, and save this data that I control. We have this now... Is called our hard drive. The added bit that makes this point crucial is simple: What if I want to move my storage from one hard drive to another? And what if that hard drive is controlled by [Fill in the service provider blank] and only allows me to make copies of a file to another virtual hard drive that they maintain control of? I now have lost control of that data, regardless of how secure and accessible this data is.

Of course there are a million more plausible scenarios as to why a primary storage medium that exists somewhere other than your local machine make this entire notion a bad idea all together... but I will let your imaginations take it from here.

2 - Application performance

Web-browser based "AJAX" applications are neat and cool, and have pretty ways of doing pretty things. There are some key missing pieces, however, that make the notion of a web-browser based computing world a pipe dream that simply doesn't make any sense. We've got all this power sitting on our machines that implement lightning fast, compiled code, and yet we somehow believe that Javascript AJAXlets are somehow going to completely make this a moot point?

No. Don't get me wrong. The architecture of an AJAX-enabled application does make sense. And it is my STRONG belief that style of programming will find its way into mainstream, desktop applications that are fast, reliable, and take full advantage of the machine they reside on in compiled form.

However, this "idea" is nothing new... Its called Grid-Computing.

The MT engine seems to have issues with figuring out that a ")" doesnt belong at the end of a URI.

http://dev.extensibleforge.net/wiki/GlobalClip

Oh, and the reason their may be topics unrelated to Mac in the Mac section of OReillyNet is that OReilly recently moved to a MT-based system instead of their home brew solution. We all were given the chance to ask for a spot in our primary areas of focus for each of our blogs. For example, I chose XML and Windows DevCenter.

The overall result is a good thing... For example, I while back I wrote a post that had to do with Linux, but it was more focused on bringing to surface the fact that Linux is actually making quite a bit of money in the grand scheme of the computing landspace. However, because of the primary subject matter of Linux, it was posted in the Linux DevCenter, when in actual fact it really should have gone in the Windows DevCenter as it was focused on bringing information to people that didnt already know. The Linux community, of course, already knew this and didn't need to be told.

With all of this in mind, keeping our primary areas of concern as the focal point of where our blog entries show up is a good thing, but a side effect is that sometimes you'll find non-related material in each area of the site.

One last point:

Sun Microsystems tried this idea a while back, using the term "The Network Is The Computer".

That was ten years ago. Beyond Sun employees and a couple scattered customers here and there, it never took off.

The connection speeds are faster now, but thats not why this idea failed, as the boxes themselves were geared to run via a local cache of the system. While there are more reasons than this, the idea failed for one simple reason:

If the power exists, and that power is affordable, reliable, and accessible, people want it. PC's are affordable, reliable, and accessible. As such, people want them.

Regarding control of one's data: if that were a serious concern for many people, we would all be using open data formats right now (because everyone would be demanding it). Users of a certain flaky but extremely popular operating system are already used to sporadic loss of access for various reasons; they would quickly shrug off not being able to open Writely the same way they shrug off not being able to open Word now. In short, they'd go play Minesweeper for a few and try again later.

When it comes to the powerful forces that be, I don't think there are any on this planet that are stronger than the basic economics of supply and demand. Although not an economist by trade, I really do that that incentive rules all within the proper context. Again, that -- in my mind -- is the linchpin of this whole idea: that super cheap, wifi enabled PCs will soon be available and free Google-driven office apps will also be available (probably for free) on the net...and as long as "good enough" works (or even almost works), why would Joe User go out and buy a $300+ USD office suite. For the same reasons, why would many (not all -- many) small businesses choose to do otherwise?

Again, my inclination here is to consider the incentive. And I think the incentive as the future approaches is quite a bit different than it was 10 years ago. True, the network throughput has increased which makes AJAX-driven webapps plausible now, but don't forget that the very idea of a computer is so much more ubiquitous now than it was 10 years ago as well...the internet is a commodity now...before long Wifi will be a commodity...almost everyone uses office apps...etc.

To quickly comment on "control of one's own data" and "application performance" -- I think they're both good ideas that are worth considering, but I just don't see either of then outweighing the *incentive* that's otherwise offered by net-driven applications.

As for control of one's own data -- what if data were encrypted/decrypted client side (maybe in hardware, maybe not) as it moves in and out of cache before being uploaded/sync'd with the server that ultimately stores it. If this were a seamless process, the issue of control would be significantly mitigated, wouldn't it? While no encryption is unbreakable, I think that this approach would put most people's minds at rest.

And finally, getting to application performance -- the internet is just going to keep getting better and faster, so for many (not all) applications, I think the performance you're going to get most of the time will be plenty good. Now it's true that you wouldn't be doing scientific analysis and such things over an AJAX driven webapp, but that's not what we're talking about here. We're talking about Joe User, remember? Joe User generally does pretty simple things that don't demand a lot of performance.

Thanks for the dialogue, BTW. I think this is a really interesting topic to ponder.

Funny how when a high visibility company puts out free software that people find useful, it ends up being far more successful than the many ventures that die from lack of profits..

Not like there's anything to that.. or anything...

@Hoby: I hear ya, but think about it: big companies like Google have a ton of discretionary income that they can pump into special efforts that aren't immediately profitable while smaller companies usually don't. i.e. Google makes tons of money from adversising related to searches and slowly ushers in cool new products like Gmail, Google Calendar, etc. (which aren't immediately profitable.) Smaller companies would have to focus on what's immediately profitable, which may or may not be some new cool and innovative effort.

I can tell what's missing: Support for Safari, Opera, and iCab (and many more browsers) in certain Google Web applications, for example. Google is behaving recently like the webmasters in the old days were websites "supported" Internet Explorer only. No, Windows-only crappy Internet Explorer and slow and not Mac-alike Firefox are no solutions. And not every Mac user wants to run Camino, though it is a fine browser.

And one has to remember that there are a lot of countries where there is Internet access not everywhere available.

ptbo,

All very interesting points to consider... let me think on it a bit and respond again once I have.

Thanks! As you say, all very interesting, thought provoking conversation. :)

FARFetched,

Hmmm... I think we're speaking to two different points:

Real control.

Perceived control.

Your point regarding open formats is a good one... But its the perception of control that I am speaking to more so than the actual control of the data contained inside. The ability to read a Word .doc file is ubiquitous... There are free readers and reader/writers available on all OS platforms. So the perception to anybody who opens, reads from, or writes to a Word .doc file is that they have full control of the contained data.

But similar to the recent (over the last 2-3 years or so) surge of customers outraged when they realized that they were unable to play a particular CD they purchased on their home CD player, or rip it onto their MP3 player, the ability to read/write from ANY hard drive, physical or virtual, no matter if that be GDrive, LiveDrive, or the physical hard drive in ones own machine is absolutely MANDATORY or the perception that they do not have control over their data will move from perception to reality and outrage will ensue.

oops... two comments above should have been addressed to "ptwo..." not "ptbo"

We could use ThunderBird as a Free outlook email software.

How about using the OpenOffice.org suite of software, for a complete free set of tools for business use.

While Google may be trying to expand its market share in the office application arena, the more it provides, the more opportunities that Joe User has to get frustrated with Google.

Steve,

Excellent point! Hadn't really looked at it from that angle, but without a doubt, you're point is spot on.

ahhh... you're = your :D

Keith,

Yup... and in doing so you are going to have a more capable, pleasurable, and overall satisfactory experience. Making the web the home-base of ALL of our applications is dangerous, at best. The merger of the two has been ongoing since Tim Berners-Lee pushed "play" back in the 89-91 time frame... Reading through his own thoughts and comments on that time era, it seems obvious to me that his vision of the web was ALL about publishing, and making it as easy as possible for anyone to do just that... Publish.

We're in the middle of (FINALLY!) converting over to TBL's original vision of the web (see: http://www.oreillynet.com/xml/blog/2006/05/overplanning_the_internet_is_i.html for more links and information)... Instead of pushing things too hard (in the wrong direction, in my personal opinion, but what I or anybody else might think doesn't necessarily equate to reality) in the hosted applications directions, we should instead focus on make the web Read/Write first, to then worry about what comes next.

In other words, we're just installing the hard drive into our new machine... Let's perfect the concept of reading and writing from and to it respectively before we concern ourselves too much with building applications that use this new read/write capability.

I agree that in a nice ordered universe, we'd all just ease into things. Databases would never need to be redesigned "live" because all of the requriements would be collected and would never change, Windows would never crash on people because MS would design higher quality software, and people's Mac Book Pros wouldn't be overheating...but since we live in such a chaotic place, these things just don't happen.

I see a pretty good parallel between these things and the web being used and abused as a platform for anything and everything, including applications, online storage, etc. Being the economist (at heart) that I am, I simply correlate all of this back to incentive: where there's a buck to be made, a buck will be made.

But I will say that it's nice to have different perspectives and ideas all over the spectrum. It keeps things somewhat stable...and maybe more importantly, interesting.

PTWO,

>> Being the economist (at heart) that I am, I simply correlate all of this back to incentive: where there's a buck to be made, a buck will be made.

Both a good point/well said.

>> But I will say that it's nice to have different perspectives and ideas all over the spectrum. It keeps things somewhat stable...and maybe more importantly, interesting.

Amen to that :D

Just reread:

>> In other words, we're just installing the hard drive into our new machine... Let's perfect the concept of reading and writing from and to it respectively before we concern ourselves too much with building applications that use this new read/write capability.

Got bit by MovableTypes inability to figure out what a URI can and can-not contain again:

TBL's mentioned post is located @ http://dig.csail.mit.edu/breadcrumbs/blog/4

Senator Sununu's quoted comments can be found @ http://www.eff.org/deeplinks/archives/004343.php#004343

Interesting comments.. :D