The Elusive Killer App

[hishamk]

Just thought I'd start a thread on killer apps and how vital they're for an operating system's success.

Can OS 4 succeed by having two major killer apps (MS Office and Adobe's Creative Suite [maybe just Photoshop]) ported?

Obviously porting is nonsensical from a business perspective for MS and Adobe, though perhaps the same guys that built AROS in a clean room to function exactly like AmigaOS can pull off building a Photoshop clone that functions exactly like, er, Photoshop to the t

The only hold back to people not using Linux that I often encounter is: StarOffice is not MS Office and GIMP is not Photoshop.

I think if Linus built a Unix clone by clean rooming it's API, then why not start doing the same for popular commercial software? Will there be any legal obstacles in terms of interface similarity? I'm sure there are ways around that (have it ship with a customizable interface that the user can then change/download settings from another site to make it look exactly like Photoshop for instance).

[Piru]

I think if Linus built a Unix clone by clean rooming it's API, then why not start doing the same for popular commercial software?

Apples and oranges.

What Linus did was something completely different. He wrote his own "nix" implementation loosely inspired by minix, which already was available in source code form (for eduacational purposes). He also did the early development hosted on minix. Soon after finishing an very limited early betaversion of the system he released it for the public and since then several thousand people have contributed on it. It took years and uncounted work hours to get linux anywhere close to what "real" unices (BSD et all) were at that time.

This is not what you're suggesting.

You're suggesting reverse engineering a several hundred megabyte binary applications and somehow magically producing a working clone of it. Even if you somehow could do it, you would be sued to death (trademark & patent infringement if copyright wouldn't work) instantly.

[coldfish]

I trialed Ubuntu for a while.  Openoffice and Gimp are both very good packages even if they do have a certain freeware feel to them; with little graphical glitches, hit and miss screen redraw and patchy file type compatability.

I used OSX for a while too but purchasing the same software on two different platforms (doubling our costs) meant that our already established Windows machines always got first preference.

When I have to do something in a hurry, I always fall back on Windows and its apps.  Years or practice and learning the quirks make using something else while under pressure an unnecessary chore.

This is probably the same reason people have stuck with AOS, other OSes are less familiar and can become unpleasant to use.

I think any Amiga based OS will have a hard time attracting an audience outside the Amiga scene, especially when it comes to serious (non-hobby) use.  Even with "killer apps" lined up, maybe you'll get a few hundread "once-were" users coming back but I doubt there'll be a flood of overnight converts.

[hishamk]

Apples and oranges.

You're suggesting reverse engineering a several hundred megabyte binary applications and somehow magically producing a working clone of it. Even if you somehow could do it, you would be sued to death (trademark & patent infringement if copyright wouldn't work) instantly.

To the contrary, I am not suggesting outright reverse engineering. I suggested clean room engineering. This, coupled with a good lawyer and the current precedents in the software industry, should not really cause any problems for people creating software from scratch that has the same functionality as existing software. Be it an OS's API or an end-user application.

Case in point, Compaq's 1982 clean room engineering of IBM's BIOS to produce compatibles. Or even AROS or ReactOS.

As to whether it is feasible in a free, open-source environment to create Office and Photoshop clones, I don't think so. It might best be taken on by Amiga, Inc. themselves (which in itself is highly unlikely since it's a dead corporation to begin with). Then again, every time I look at the progress made with AROS I'm amazed at the dedication and contribution of disparate individuals to create something for fun and not for profit; something that would take a major investment (and a huge argument with the number crunchers) by a corporation to pull off otherwise.

[Jope]

So.. You get right on that, and tell us when you're finished.

A killer app was important in the 80s, because back then people didn't have any hardware to run the killer app. It was easy to sell the machine running the killer app, because there was no installed base.

These days people have had computer hardware for several years, so it is going to take one hell of a killer app to get anyone to switch from what they've used for a long time.

A killer app by original definition means that people will go into the store and buy the computer by referring to it by "a microsoft word machine" or "a photoshop machine". I somehow don't see that happening in the 2000s, as the computer has evolved to become something akin to a kitchen appliance, not the magical tome of eternal knowledge it was in the 80s when this stuff was still new and amazing.

[Piru]

To the contrary, I am not suggesting outright reverse engineering. I suggested clean room engineering.

Clear room engineering doesn't protect you from patent violations.

This, coupled with a good lawyer and the current precedents in the software industry, should not really cause any problems for people creating software from scratch that has the same functionality as existing software.

Be it an OS's API or an end-user application.

Usually reverse engineering is only allowed for interoperability when there is no other way to obtain the information needed to implement it.

That's what is done with OS APIs for example.

Case in point, Compaq's 1982 clean room engineering of IBM's BIOS to produce compatibles.

That was 8k of code. Somewhat less than big application today.

As to whether it is feasible in a free, open-source environment to create Office and Photoshop clones, I don't think so. It might best be taken on by Amiga, Inc. themselves (which in itself is highly unlikely since it's a dead corporation to begin with).

No corporation can do it or they'll be ground to dust with patent lawsuits.

Then again, every time I look at the progress made with AROS I'm amazed at the dedication and contribution of disparate individuals to create something for fun and not for profit that would take a major investment by a corporation to pull off otherwise.

What AROS is doing is trivial in comparison to even one modern application. The KS ROM is only 512KB and the Workbench applications easily less than a megabyte. Almost everything in the OS is well documented (autodocs & includes) and the applications are very simple.

[hishamk]

Clear room engineering doesn't protect you from patent violations.

...

No corporation can do it or they'll be ground to dust with patent lawsuits.

I guess I wasn't concerned about patent infringement since there are apps out there that incorporate features that were once only part of Photoshop. Though I suppose the gazillion patent numbers that come up when you load up PS are there to be called on by Adobe in case someone really pisses them off.

I see your point about AROS being a clone of a rudimentary system of an age past.

Thanks

[persia]

First step, create the universe...

Seriously, even if you could write a clone of Photoshop why would you release it only for the Amiga?  The APIs are primitive in Amiga OS 4 compared to OS X, Windows or Linux so it would be even harder than writing for them.  And then it wouldn't be a killer app, people could still use Photoshop on a modern OS.

Let's face it, there are no probably killer apps in 2009, and certainly none that would convince people to plunk down €1000 on a computer than is roughly as powerful as an iPhone.  

Look at what people do with their computers:

1) Write letters and share documents, there's a rudimentary word processor that won't open current MS Word Documents.  

2) Sync their phone/iphone - MP3s, contacts, calendar, etc.  No can do.  Forgot my cable, guess I'll use blue tooth - forget about it...

3) Edit pictures - 12 MP raw?  Remove red eye?  

4) Make a video for youtube, edit video.  Nothing on OS 4, video toaster, slow & primitive on old hardware.  Doesn't handle modern digital input well.

5) Surf the web, sort of, in a half arsed way, yes.

6) Email, possible, if you happen to like the one program available (YAM).

7) That printer you bought in this century, wouldn't it be nice to use it's features?  

That autofill pdf

Just a few things to start.  There's more, an OS that doesn't remember where you put icons, that doesn't have a decent search facility, that doesn't have memory protection.

A decent, modern IDE would help, but even then the programmers aren't there.

[ElPolloDiabl]

If you make a portable or cheap amiga like natami you might have a foothold at least.

[Zac67]

Before thinking of the really BIG stuff (Office + Photoshop clones would probably require 5000+ man years to develop), you should take care of the basic apps - a decent browser up to current standards, email, ... Nobody would want an expensive piece of hardware to run a somewhat decent Photoshop and nothing else.

Include sufficient processing power, drivers for current mass hardware, make it cheap, stir lightly and you're a step closer to thinking about 'killer apps'.

This is more ludicrous than elusive - sorry.

[Ilwrath]

Can OS 4 succeed by having two major killer apps (MS Office and Adobe's Creative Suite [maybe just Photoshop]) ported?

No.  A lack of quality applications can hurt sales, but having those applications won't spur sales.  

I already have Photoshop and Office on cheap, powerful, and plentiful hardware.  What do I need them on expensive, underpowered, and low volume hardware for?

And that's not to mention that there is no way you'd get these things, anyhow.  Heck, Firefox has taken how many years, so far?  It's far simplier than either of those, and we have the source code to it!

The only way a 'killer app' approach that might work is to come up with something no one has done before and no one realized they were missing until they see it.  And even then, it would only take about 2 days from the first time a linux user sees it until there is then a Linux clone of it underway.

[spihunter]

I dont think that there really are "killer apps" anymore. These days it seems to be more about a "killer device" like the Iphone or a "killer service" Like Twitter/facebook.

[Flashlab]

I dont think that there really are "killer apps" anymore. These days it seems to be more about a "killer device" like the Iphone or a "killer service" Like Twitter/facebook.

I agree. You need a killer concept. Something that is unique and hard to copy correctly. That's exactly what Apple did with the iPhone. It's a killer combo of hardware, software and App Store service that just works. No ICT product since the Amiga has put a smile on my face like my iPhone does.

Porting any existing application or copying anything else won't set ...(fill in any Amiga flavour your like here...) apart from the competition, hence will not bring more users to the platform.

[persia]

So all Amiga, Inc, distributors of the evil snowman app, have to do is come up with TNGT (the next great thing), make it before the competition and market it like hell...

I agree. You need a killer concept. Something that is unique and hard to copy correctly. That's exactly what Apple did with the iPhone. It's a killer combo of hardware, software and App Store service that just works. No ICT product since the Amiga has put a smile on my face like my iPhone does.

Porting any existing application or copying anything else won't set ...(fill in any Amiga flavour your like here...) apart from the competition, hence will not bring more users to the platform.

[Flashlab]

Yep, quite simple, eh !