Valan said, in another thread that was getting a bit long,
We all want the Amiga to succeed in one form or another. For years it has seemed the obvious way to go and other systems are slowly getting there.
But for all the knowledge in the Amiga community we find it hard to identify what the Amiga actually is.
Sentences like 'the AmigaOne/Pegasus runs hundreds of Amiga program' are humerous bordering on sad since it defines the new Amigas as allowing users to step back at least 5 years and run old software.
Then people point to the 'killer app' as if it must appear. Well it never appeared for Atari or Segas new hardware.
We cannot take the success for the Amiga for granted no matter how well intentioned.
I think we should be talking about how we now want to use Amigas, I mean what would be your the 'killer app'?
...Seems like this is a discussion worth having again, so I split it off here.
Killer apps I've been waiting for:
Peercast - Done, but weak video support and so on; room for improvement there.
---
Post-blog, post-wiki "convergence" - This is a tough one to describe.
Chandler is so close yet so far. Basically, right now we segregate our information into "personal" and "public;" 'blogs' are just online diaries (what used to be called "web pages" before Geocities lost its cachet), 'Wikis' are great, but weren't envisioned with security models, and they're heavy on the 'hypertext' model - users explicitly define relationships. 'Forum sites' require more Apache monkeying than should really be necessary, and the backends are usually of debatable quality. Meanwhile, your "address book," "calendar," or "PIM" probably don't integrate with any of this junk, even though you conduct a good portion of your life through it. (Bandaids like present usage of RSS and whatever the iCal protocol is notwithstanding.) Meanwhile, basic concepts like file sharing in the non-pirate domain are either hopelessly broken or haven't advanced much since 1990. (Sure, they *work*...)
I'd like to see something pull all that together. I'd like to be able to log in - however I log in - and see all my *stuff* sorted by date of creation, date of last access, creator (is it mine, or is it something useful I cribbed off someone else's site?), type, whatever. Okay, all these relational storage experiments are doing that, but only when I'm *at* (or otherwise VNC'd, ssh'd, whatever) the machine. I'd like everything to be able to have a discussion thread if I want... I'd like every message and discussion thread to be a 'thing.' I'd like to see any visitor able to drop in and contribute, without having to worry too much about security or the like... I'd like to give my pals space with quotas without that having anything to do with privilege to run executable code on my machine... I'd like to be able to flag data as 'private' and only share it with a group (UNIX/POSIX does it, but again, those implementations approach from the concept of program execution; I want a *data* management system), I'd like the whole thing to offer an interface that works from any SSL-capable browser, since that's the lowest common denominator right now... I'd like the possibility of other interfaces when I use it from the console or 'desktop' ... I'd like it to define open standards and protocols for anything it needs new ones for... I'd like to see an 'economy' of protocols and solutions grow around it in almost the exact way the browser market now operates... I'd like to not have to roll it myself, but I'll probably have to try, someday.
Things like a 'raw' *NIX box with Webmin, WebDAV (I've never used WebDAV, maybe it's great if you have it integrated in your file browser, but it seems like it was designed from the perspective of making life easier for 'the web' - that is, to improve the lot of conventional site designers and managers - than to make 'the web' work for people), Chandler with its focus set on existing applications, "easy" programmability (Hm, wouldn't it be easier to
not have to program in the first place? Is including a featureset that can be equally well-represented in 'language' and 'controls' really that hard? If you *need* to program something, shouldn't it just.. call an actual program?) and that huge Zope dependency anyway... are close, but seem "wrong" in one respect or another.
In particular, when all is said and done, Chandler looks like it'll end up focusing more on benefit to corporate monstrosities (yay, unified address books and calendars, exciting...) than actual users and their interaction. ("Hey, snag this file from me." "Did you see that interview with Jay Miner from 10 years ago? Here's my copy..." "Wow, I just took a global view of public commits... Someone came up with an awesome way to simplify and finish that project I gave up on last year.")
Of course, if I could explain the minor differences in approach and realization succinctly, I would, and someone smarter than me would probably have already done it by now. But this has been on my mind for a long time, and I've taken it nowhere, so maybe someone else will at least be inspired. Instead of relying on 'services' for all this junk, it'd be the equivalent of solar power; your "blog," your "work" and so on could all stay on your machine conveniently; in turn, you'd keep using a group/corporate/project one when it made sense. (Amiga.org would still serve a purpose, for instance, but what we need are better ways to empower everyone to be their own mini-Amiga.orgs.)
And yes, REBOL could be involved, but REBOL in the 'client' concept is explicitly bass-ackwards (more power to 'em) from the perspective of what I'd like to see here. It's about pushing *code* to users (a-la Java-as-plugin, a-la DE) - handy, indeed, if you have a fast machine and, say, don't want to keep a satellite tracking program hogging your disk but do want a pretty, realtime-updated display when you need it... I want the equivalent for *data,* and the conventional webserver / user-accounts / CMSs-that-don't-really-manage-much / databases but-each-site-probably-has-to-roll-its-own-tables model doesn't do it for me. Works for Slashdot, works for A.org, works for IBM.com, but those sites *know* what they want to be. I want a 'portal' to manage, share and organize my
own cr*p.
---
Well, that became a nice rant. Other thoughts:
Home automation. From discussion with friends, the first person to come up with a Bluetooth/ZigBee X-10 socket or light bulb will probably be a hero. I sort of roll this into the whole concept of "omiscience" and "omnipotence;" for instance, all these "spy toys" are getting dirt cheap and wireless - cameras, phones (themselves with cameras and sensors on them), wireless weather stations... all sorts of junk. Windows will probably get all the drivers first, but finding smart or just plain fun things to *do* with all this data... I can already go out, buy a camera, and watch my front stoop from my desktop, but there's got to be more interesting things we can do along those lines.
In turn, conveniences are nice, and become selling points as well... Everyone remembers the keyboard garage; what if Eyetech or one of the dealers were the first guys to bundle a machine with a motion sensor, so it could, for instance, automatically mute the sound 10 minutes after you leave the room? (This is more of an amusement when you run a *NIX box with a locking screensaver and your machine has internal speakers with no hardware volume control...) Or unsubscribe from your camera's stream (if we're still in a unicast world) and save you some local segment bandwidth? Or refresh your mailbox the moment you walk in?
Heck, how about games that take advantage of that sort of 'awareness?' There's already one for the GBA that forces kids to go stand in the sun...
