On Tue, May 12, 2015 at 4:30 PM, Tim Guan-tin Chien <timdr...@mozilla.com> wrote: > Agree with the proposition too but I don't see the link between this > goal and the New Security Model.
I personally think that the "New Security Model" and the UX proposed in this thread are fairly independent. We could probably do each without the other. But I do think that the "no apps" UX proposed in this thread is easiest to do if we also implement the "no installs" aspect of the new security model. > If trusted contents are still > delivered in packages, which is distinctively different from the rest > of the web? I'm not sure that I fully understand the question here. But one important aspect of the new security model is the fact that we no longer use multiple cookie jars. Even for packaged apps. This means that if a packaged app redirects the user to facebook in order to enable the user to long in, facebook will be loaded using the users normal cookies which means that the user won't have to type the facebook username/password. And the referrer URL that facebook receives will be a http URL which it can then redirect the user back to after login. And if the packaged app does an XHR request to google.com, that request will be loaded with the user's normal google cookies. And if the packaged app creates an <img> pointing to flickr.com, that image will be loaded with the users normal flickr cookies. So while the content is downloaded as a package, all other aspects will still be like normal http content. The *biggest* difference that signed content (which is what I assume you mean by "trusted"?) has, is that it's signed. That dramatically changes the deployment model for the developer. Compared to that, packaged vs. unpackaged is a pretty small difference. / Jonas > On Tue, May 12, 2015 at 6:32 PM, Benjamin Francis <bfran...@mozilla.com> > wrote: >> On 12 May 2015 at 00:10, Jonas Sicking <jo...@sicking.cc> wrote: >>> >>> We've debated various forms of making FirefoxOS more webby and less >>> app focused. The new security model is one example of that. The >>> pinning proposals from Ben is another. >> >> >> In case you're wondering what these things are... >> >> New Security Model >> https://wiki.mozilla.org/FirefoxOS/New_security_model >> >> "Pinning the Web" design concept >> https://docs.google.com/presentation/d/1JTRM1qr25tW_GLe5zeDV-P2p0x6mwefMy5TcGkzAK9c/pub?start=false&loop=false&delayms=3000#slide=id.p >> >>> >>> We can create something >>> which cause people to not just count the number of apps that we have, >>> but instead look at other differences. >> >> >> Definitely on board with this. >>> >>> >>> Second, I think the web model provides substantial benefits. All >>> modern operating systems, desktop and mobile alike, use an application >>> model. But this model came about due to engineering reasons, not for >>> UX reasons. On desktop the web model has proven remarkably popular and >>> successful, and I see no reason why it couldn't on mobile too. >> >> >> I'm not sure it's true that the app model only exists for engineering >> reasons, but it's certainly not the only possible model, or necessarily the >> best model for the web. >>> >>> >>> Third, by creating something more webby, I think we'll make better use >>> of the content that we have. I.e. it'll make better use of the web as >>> it is and as it will be for the near term future. >> >> >> I agree. The web has a billion web sites, content is the web's biggest >> strength. Let's make the most of that and unlock more of the value of the >> web for users. >> >>> >>> So, in summary, the idea is to: >>> * Don't use separate windows for separate "apps" >>> * Enable platform-wide "tabs" which can be navigated to any URL, >>> including gaia URLs. >>> * The same URL can be open in multiple tabs at once, including gaia >>> URLs. >> >> >> I think this is all consistent with the Pinning the Web design concept. Any >> window can be navigated to any URL, no app windows vs. browser windows, the >> browser chrome collapses and expands as you navigate into and out of pinned >> sites/pages. It also suggests dynamically grouping open windows by site in >> the task/session manager as an optimisation. >> >>> Even including the homescreen. >>> * Icons on the homescreen are just like links and just navigates the >>> current tab. >>> * The homebutton simply navigates the current tab to the homescreen URL. >> >> >> It could make sense for the homescreen to be like about:home, the browser >> start page to be like about:newtab and settings to be like >> about:preferences. Also things like the dialer which depend on APIs which >> are never likely to become web standards are essentially chrome. I think it >> also makes sense to be able to modify this chrome with Firefox OS Addons. >> >> It would be shame for apps like Contacts, Gallery, Music, Video etc. not to >> have real URLs, and a unique URL for each contact/photo/song/video as is >> being worked on with the new architecture, so that you can deep link to a >> particular item. Maybe some parts of Gaia would become chrome, and some >> would become web apps? >> >> I think these ideas make sense architecturally, but need exploring further >> from a UX point of view to figure out what problems are actually being >> solved for the user. Being different isn't enough on its own, we should be >> aiming to provide real value and a compelling experience for users. >> >> Long live the web :) >> >> Ben >> >> >> _______________________________________________ >> dev-b2g mailing list >> dev-b2g@lists.mozilla.org >> https://lists.mozilla.org/listinfo/dev-b2g >> _______________________________________________ dev-b2g mailing list dev-b2g@lists.mozilla.org https://lists.mozilla.org/listinfo/dev-b2g
