Public Pad
Version 6543
Saved July 7, 2010
 
Please leave your feedback here (and don't delete anything from others!)
 
Questions to ponder about:
- What is missing today?
- What do we need?
 
Your feedback:
 
I think it's a wonderful idea that will dramatically improve quality of what's available.
 
A very good starting point would be a common infrastructure for discovering "savable" web apps to some interface (the iphone home screen, an app tab, whatnot). It could work much in the same way as discovering search engines automatically in Firefox. Every vendor of browers, devices, etc could decide themselves on how to handle this information. As long as a common web app, found through an URL would identify itself as a "proper web app", the iPhone could offer to put it as an icon on the home screen, Firefox could offer to put it in a permanent app tab, and so on.
How to find these apps then? As I just said during the talk, this works as everything works on the net; we send stuff to each other. It is really a whole other problem than the question of web app store or not. There is of course a lot of space for such stores, but they could just as well distribute URLs. If you want to chrage for your app, but a login on it and charge
    maybe, maybe not - if we recontextualize to payment-tokens-for-the-web we can reinvent the idea
    ^ this sounds like a great extension to the AppCache spec
 
http://www.toolness.com/wp/?p=821
 
What's the killer app? Why would an average user want to come to this store?
 
 
Can  an app store be a success if it isn't a monopoly distribution channel?  
 
What does a "Web App" look like- i.e: does it appear or display in a differentiated context from other web content?
 
 
If the APIs to give web apps first-class capabilities are years off, native apps will have an enormous head start. Can we drive and promote native-quality APIs for web content in a reasonable time frame?
 
'Web apps' doesn't mean 'client-side-only web-apps'. The real value in most web applications is the cloud and server side power. That can't be copy-pasted.
 
But a huge amount of those hundreds of thousands of iPhone/etc. apps are stuff like single-player puzzle games etc. Big part of the apps are actually client-only-apps, often offline apps.
 
Seriously folks, since when has piracy killed any business? It's happening in the software, film, music markets, and all of them are flourishing. Just not the big, big actors alone as back in the days and that's not what we want anyway, but rather a broad range of providers. Look at the Mac shareware market for example, works great.
 
How  should we prioritize the need for extendability (hackability) to the  need of the developers source and making "cheap copies" - should there  be a screening for ~100% alike apps?
 
I think "cheap copies" drives the internet, having a controlled environment is somehow not a very good idea in that sense.
 
Standalone vs. Backend-supported applications. How do we communicate the difference to the user? Being able to work completely offline is a critical advantage of most applications (like on the iPhone, iPad). If everything is "view source"-able there is no way that we'll have serious offline-capable applications, devs will hide logic on their servers.
    I strongly suspect that purely-offline apps will only be viable commercial products if their price is less than the "annoyance cost" of copying it.  More than $1, less than $20... I think the strong hardware-locked DRM needed to prevent offline copying is incompatible with Mozilla values.
 
So there can't be successful open source applications? :)
 
 
Discoverability: Being able to search for and find relevant applications (categories, tags, ratings, user reviews).
 
 
 
 
Easy Payment Model: Allow developers to accept variable payments - even small ones ($1, for example) without incurring fees.
 
Objective C + Xcode is a terrible platform, but it is a platform provided and maintained for developers. Nothing like that exists today in a mature form for the web
 
Could existing web apps be modified to provide premium features by way of an app store (freemium model?)
 
Using the Prism-style desktop icons? Or shown on the home-tab? (<- +1)
 
Do we need buy-in from existing app stores for an open app store to succeed?
 
As a platform, what can we (Mozilla) do to enhance the feeling of owning an app?
 
Self-contained: Web apps must be possible to use offline if they need no connectivity. Interesting thing is that a lot of iPhone apps are not really web related but small tools/games/etc of various sorts. I think this segment of apps is often ignored when speaking of web apps etc. HTML5 has the possibility to cache 100% of the content (text, images, etc.), right? So there could be ordinary tool apps in the form of "web" apps too.
 
This is the obvious question but somebody has to ask it so I might as well:  Why will anyone pay for my webapp when they can hey can just go to the URL and use it?  Especially if they can discover the URL through the web app store and then just go there and use it?  Does each webapp developer have to com eup with their own way of locking out non-paying users?  And does that lockout run contrary to Mozilla principles?
  - can't we do our best to support an "advertisement or purchase" model?  doesn't need to be purchase-or-nothing.
- Well the advertisement model is easy -- we direct people to your app, you put ads there.  But that's happening already.  I'm curious how we can make the purchase model work at all.
- Seems likely that a purchase would be the method of account creation, but handled in a more generic way than it is today.
- I would propose that the app store produce a "proof of purchase" bound to some sort of identity token.  Say a signed bundle of appID + userID + some other metadata.  Need a federated ID platform to make it work, but Weave can provide that.  And the thing that appears in the home tab could be that thingy (and it's synced through Sync too)
 - Cool, that's a good start.  I think if we want to enhance the sense of "actually owning something" then I would want my proof of purchase to be transferrable if I switch machines
 - +1 to that.  It should be bound to your identity, not your machine
 
There is a basic conflict between being able to share webapp code with my friends & family and developers being able to charge money for it. The hardest technical problem that has only met with horrible solutions so far is being able to "sell" open source code. You can sell support, sure, but never the code itself. What exactly will a developer of an open webapp be charging for?
 
What can we do to make Web Apps ROCK?
 
one question i had; you mentioned objective-c a couple times hackability; what does it look like for a user to hack on a web app? pascal: i think this is the broader question of hackability in general; one of the principles we hold highly at mozila; doesn't mean every user needs to hack on apps; just that you can; if you are interested in it, you can;
 
one of the big principles of hackability is that i can learn; view source and see how other people do stuff
 
followup: i think that one thing this brings to mind is hackability is greasemonkey; jetpack is the logical successor to that; ability to make existing web app ...
 
i just wanted to mention as food for thought; was actually proposed as drumbeat project; another way to reframe question as broader question: is there a way to make it easier to give people money over the web in a way that doesn't suck as much as it does today? because right now resources required to xfer resources over web is high; just curious about drastically lowering the barrier to giving small amounts of money to people over web; if it was as easy to give 25 cents to author of good blog post as it is to toss money to street musician in subway; what would that mean for world?
 
good point; interesting because i spent the last 8 weeks thinking a lot about what web app marketplaces should look like; have background at ebay; payment component is fairly well solved by paypal; 25 cents is hard case because it's so low that transaction cost is getting high; more interesting part is discoverability; sense of ownership; but if we could have wallet to give 25 cents would be interesting.
 
app marketplace is that primarily for mobile devices or desktop browsers as well? awesome question; absolutely amazing opportunity is to build something that works cross platform; jeresig spearheading initiative; today if you build something like farmville on facebook and want to bring it to another platform you have to rewrite it; an enormous waste of resourses; if we do this on HTML 5 it works across devices; when you see integration of webapps today, on iphone it's awesome; you can turn web app into launch icon; makes use of most of html 5 features; not propagated by apple, which has incentive to bring you to app store; remember the milk is nice example; charges 25$/year; they give you full product, almost like donation; if i could pay and have access on all devices, awesome
 
does prism play a part in this story? prism for those who don't know is labs project which allows you to turn web content into a single contained desktop application that works mac/windows; i think yes, because it's a way to give you sense of ownership; you bought say calendaring application or game and you want it as icon on desktop; it should be possible; not only on desktop as one delivery mechanism, but also in launchpad style thing in browser and on website; so even if you're not on primary browser on your device, say i visit my aunt; which is interesting problem; spoke to guys at chrome store; they want you to package your app into zip file with format of crx, same format as for extensions, has manifest; what it means is you can only use them in chrome; just weird and wrong;
 
being proposed as web standard; yes, but no one wants it because it's wierld (disagreement--i like it!)
 
we know for a fact that other major browser vendors say it doesn't feel right for them; do we want packaged version of web app?
 
i wrote blog post; feeling of ownership; practical aspects; web apps can change at any time; if i pay company for service and it goes out of business, i can't use it anymore, but notion of apps is that once i bought it i can still use it even if company goes out of business; one of the things that i like about package format is that once you have that you can put it on your computer and death of uthor doesn't take it away from you; 
 
but what chrome is doing is taking it a couple steps further by forcing you into central delivery mechanism; but they need to get ...
 
do you think developers will pushback against view source? and if so, what do we do about that? interesting point; people will say proprietary information, i want DRM, yadda yadda; i follow chrome app store internal discussion; people did say i want to lock this down, this is my code, but 95% of people on mailing list responded forget it; if someone wants to steal your code, they will anyway; you can obfuscate to reduce readability; so it's a good point; we as mozilla would take the stance that at the end of the day this is a fundamental building block of the web; you can do view source on any website, it didn't hold back bringing crazy innovative shit on web sites; yes, some people will say they don't want to play in this league; most will be fine with it
 
two comments: i think flickr does a nice job showing you are a supporter with Pro badge next to your profile, i wonder if app marketplace could change perspective from paying for applications to supporting applications; going back to what atul said, i still think paying is a problem; when i want to pay on paypal, i have to log in, review purchase, receive email; account manager does logging in automatically, but microtransaction system with threshold below which i don't need to review transaction but just do it;
 
totally agree; amazon one-click experience; haven't communicated this to anyone, but one of the fundamental principles is to make it as awesome as amazon in terms of how it works; customer service like zappos; and fun as woot; the three things i want to have; payments that are easy, customer service, and fun
 
i really like the idea, but i think there are a couple major problems to deal with; first is fact that right now web developers are incredibly bad at making cross-browser applications; problem is that if we're trying to sell open web store, one of the big selling points is that it works on the web in web browser, but that means we're inclusive of ie6, blackberry, things thaht aren't good; if we can't make guarantees that they'll run everywhere, it's not really an open web store, it's modern browser store, which is like chrome (whose definition of modern browser is chrome); second problem is there is conflation of apps that you download and have or are communicating with server; and do you view source of entire application, or is logic being stored on server or not; all of this is major issue, if you're paying money, people are more apt to want to pirate, and if you can view source, it's all there; it seems ilke our ideals will conflict with what we want to create
 
yes, i agree, if we want to do this we need to work through things; we're not doing this yet, just thinking about it; to your point; first question: compatibility across browsers; one way is to go to developer and ask them to specify manifest with which browsers it works on; we need to get better at giving people right tools to build cross-browser awesomeness; jquery is one tool, but there's more missing; to your second question; in my world, there are web apps that are totally self-contained client-side, and there are a bunch that interact with web server and have server logic; that's fine, part of the sauce, i wouldn't go so far as to require showing the server code if you buy a web app; the question of preventing copying, interesting point: i know a bunch of people with iphones and a bunch who when you jailbreak your iphone it's easy to get apps that you haven't paid for; there's even a cracked site with free apps; but i haven't found a person who has actually done that; because it's too much hassle, the purchasing experience is awesome, and it's 99 cents who cares? the beauty that apple taught the world is to overcome copy protection you have to make the purchasing experience so great and pricing that it's so easy; people used to use napster but migrated to itunes store; you'll always have some pirates kids with computer games; there's a balance; but i'm not too worried if we get the experience right
 
i'm not too worried about user piracy but developer piracy; we don't want a moderation pre-step presumably and then we can delete afterwards; if it's view sourceable it's going to be a huge problem; and then they give it away free; and then it's moderated out and 50 more will pop up; that's a valid point and something we need to figure out; ultimately the answer is something like social policing; it's a really good question
 
i'm an amo editor, and to address what was just asked; we at amo have people who do what you're saying; they clone existing addon and relist it with a different branding or change the name of the author; and depending on the license, that clone will be removed; community will report it, and editors will remove it; it comes down to a policing issue; we're not going to change view source or not, but we can police what is sold or not on an app store;
 
and to a certain extent you can use technology; diff uploaded apps; haven't spent too much thought on it yet;
 
back on microtransaction problem; the fact that you have community at scale lets you solve the problem with aggregation; if you have minimum transaction purchase price like $1 you can wait until you hit threshold because charging; even the app store does that
 
yes, i agree; admittedly, tihs is a problem we have never touched, and we have no experience; how do you run a server that stores dollars?
 
one thing you can look at is paypal and the apple store; you tel paypal that woot is allowed to take my money, and then every 5$ or so it happens
 
one meta-question i have, and again i would really love to get your feedback and ideas over the next couple of days: just as temperature check: does this all make sense? does the world need an open web app store? or is it yeah nice but who cares? second question: if the world needs one, is this something mozilla should take a stance in given who we are and our principles?
 
yeah, i think so; the reason i say that is when we sat down a year ago to talk labs mission, one was to promote web as first class platform; if so, you need to make the leap from selling bookmarks to selling something that is double-clickable, sense of ownership, promoting that for web is important part of making that first-class platform
 
can you think of any organization other than web that would make one truly open?
 
i think yes on both counts; i like it
 
one of the biggest challenges is presenting these in frist-class status; making sure they have functionaliyt that can rival native app or just something that feels like native software; if they're inferior it will cripple adoption;
 
having seen demos this morning, i have no doubt in my mind that web and browser as delivery mechanism of it will become unbelievably awesome; if we encourage people to build apps on this technology stack we'll see unbeleievely cool innovation
 
quick question: have you seen NYT article death of the open web? make distinction between tap and bottled watter, packaging into nice package makes big difference; people might appreciate apps so much, because they want the little shiny thing; defensive move to protect the web
 
i thought about this a lot; this is a good example; on iphone there's function in safari where you put page as icon on home screen; my internet bank didn't have iphone app, but i put it on my home screen and it loads safari with mobile version of app, and it worked great; then they released their iphone app which was just their mobile view in safari; good example about combining ownersihp, icon, but it's still just web page, which leads to opinion that open web app store wouldn't be necessary; i don't think it's good idea; if you could just go to web page and there's a good interface for "put this on my screen", it would be sufficient
 
you make an interesting point that i would throw back as counterargument; it's easy on iphone, it's not an android; on desktop browsers it's impossible; i think the URL is dead. telling my parents to go to super long URL is dead; if i tell them get app;
 
on the other hand, links are the thing; what if there was common system for saving links like app tabs? what i fall organizations vendors agreed on discoverable system?
 
yes, but you still didn't solve the question, what if your bank wanted to sell it to you? and how do you find it? with bank it's easy, but what about sudoku application?
 
the same way i find funny pictures of cats; people send them to me;
 
apps on iphone? people tell me which ones to get; i think social dimension to it is really important;
 
where's the incompatibility? i don't see it
 
i really ilke this idea; i think it should exist; i don't think mozilla should do it; i think we should work to front a new non-profit with other vendors; we know our ideals; we know we're unbiased, but people at large don't know that; they see firefox as another vendor that wants to win the browser wars; we don',t need to confuse people with that; 
 
but you agree it should be non-profict? yes
 
any way to have distributed model with payments through certain vendors, we looked at this with addon tags... one way to assuage concerns that guy had; you can find apps through common mechanisms; maybe that's something that can help
 
my time is up; thank you all for the stuff happening in etherpad; continue hacking on it if you want to; find me and other people in crew; hanson has done a bunch of work thinking about how it would work technically; i encourage you to go to session tomorrow about apps marketplace, a different but related; we have AMO; check out that session; thanks again; your input is super-valuable; come to us, talk to us; talk to your friends; spread the idea; let's get as much feedback as we possibly can to determine if we want to do something, how should it look like, how to make it awesome; amazon, zappos, and woot combined for the open web