Posted: 26th Oct 2006 02:45
Who knows if this will be successful or not...yet another stab at that "write once run anywhere" ethos.
So a web application that does not run in a browser (that thing we all have installed right now). Not sure that's the greatest tag line in the world, as most of the time, the major appeal of web apps is that they DO run in a browser, the browser (kinda) becomes the OS, the platform that all this different software is written on!
Instead this Appollo runs in a "VM" that is 9 or 10MB or so to download (and trust) and install.
Then, when/if you run a program written to this VM, it will have local storage access and be able to access the internet.
So kinda like Java then. Or .NET even (to a lesser degree).
But on paper at least, this looks weak to me. Ajax has been with us for a while now and can deliver most of what is promised here but without trust, without the client having to install a "Richer Plugin".
Also, doesn't this "water down" the Adobe offering still further? They have so many products that appear to overlap on functionality (due to them buying up devco after devco) and they seem to be responding with knee jerk driven "glue" to try to create some synergy beween their (otherwise unrelated in anything but name) products.
I mean, Flash can do exactly what they say this does. Can't it? It has application development ability, local storage (limited but better than cookies), Video, Audio, Bitmaps, vectors, an RPC Web Server offering (using propriatory binary formats) SOAP based Web Service support....yadda yadda yadda. A fairly rich client and a long way from the web vector animation package it started as...as its grown its become more like its big brother.....
....Director....similar to Flash but more power, access to HW graphics OGL & D3D, Havok Physics etc etc....a very rich client indeed.
They are popular (especially Flash) but mostly because they are browser embedded, and can at least give the impression that you've not left the "comfort" of the web.
And Flash especially comes bundled with most all OEMs these days. Its pre-installed, works well in a browser so of course it's popular! If anything, its THE browser plugin, the one that's grown up with the web and we welcome it to augment the things that web browser based apps can't do very well, namely high quality/speed animation and audio / video.
So Apollo? Like we need another Java? Why re-invent Java? M$ already has with .NET (in a windows only kinda way). Isn't the 10 year lifespan and maturity of Java good enough? And yet Java, for web apps, is still not as popular as it should be. Its a good language to develop in, it can run standalone, as an applet in a browser, with Web Start, is truly multi-platform (and prominent of mobile devices) can do high speed 2D & 3D graphics, audio, applications, web services, JSP servlets....on and on, it ticks all the boxes and yet its never really got where it wanted to get. And I can only presume that it didn't because it required a VM that M$ stopped shipping with Windows. As soon as you get a client who *may* not or *probably has not* got your required "Richer Plugin" installed, you;ve just shot yourself in the foot and raised the entry requirements for your app. This means that ou are always leaning towards the lighter, browser based app if at all possible. And these days with Ajax design patterns becoming popular and well documented, the ays of the "clunky" refreshing web app are behind us.
The bottom line is, as with .NET, "what can this give me as a web developer that I can't currently do with my current toolset?"
With ASP.NET I found very little, if anything (if you compare it with COM rather than asp classic). The original (pre-Atlas) ASP.NET was a bad joke IMHO. The Treeview on MDSN2 was a great example of how backward a step it all was, 1.5MB+ to view a page with 2 paragraphs on it (due to the unneccessary postback required by ASP.NET forthe tree).
From the "on paper" description of this Appollo thing it just sounds like corporate point scoring, Adobe saying "me too" to Sun & M$ but not really offering anything major to the web dev community, another rehash and marketing excercise to try to get developers on board, when in reality we can probably do all these things already without paying out for their new tools.
Just my tuppence...
