5 ms·

the web was a mistake. not as a concept, but as an application platform. the question i ask myself is: how do we take the Good Parts of the web and bring them t

by xj9 9y ago

the web was a mistake. not as a concept, but as an application platform. the question i ask myself is: how do we take the Good Parts of the web and bring them to the native application side? how do we create an ecosystem as free as the web, that doesn't erode user trust with spam, malware, and so on?

Andrex 9y ago
The lesson taken from email and chat platforms seems to be "centralize and control it," unfortunately. See Google's attempts with AMP as a potential solution. I know AMP isn't popular on HN, but it's a possible answer to your question. Speed, freedom, and usability - choose two, I suppose.
opportune 9y ago
I actually disagree. I think there are too many companies trying to put everything in an app when their content can be served and digested just fine via the web. And the blame there ultimately rests with the businesses that try to push their users into apps for whatever reason (gathering more data, easier to keep users within your ecosystem and not clicking away).
SirHound 9y ago
We have an ecosystem that's free, it is the web. The spam, the malware: A great deal of web apps don't have these issues. We need to continue to enhance the web as an application platform, rather than hobble it.
- xj9 9y ago
  i don't really see web applications ever being good mobile applications. to start with, jit compiling applications on a energy-constrained device is straight stupid. you should compile the application for me on some system that isn't already struggling to make it though to day on a single charge. wasm might help some, but pls don't make my mobile do more work than it has to. #p2p applications often need special capabilities or native libraries to work well (or at all). the web offers some weak apis for this, but really isn't up to par with what i would be interested in using (or building on). again wasm might help here, only time will tell. finally, web standards move very slowly and really don't exist as a disruptive force. if you need buy-in from all of the major vendors to push new/interesting features through, you aren't going to get much unless it favors them somehow.
oblio 9y ago
I keep hearing this and it seems wrong. I think the actual problem is a social one. In a free, decentralized environment, you'll get jackals (tragedy of the commons). In a centralized one, you'll get dictators. I don't think you can work around this with any kind of technology. I know that what I'm saying is basically anathema for HackerNews, but I think it's the truth.
- apatters 9y ago
  It's also worth noting that these dictators don't appear out of thin air. They come from somewhere else when they see an opportunity. The Web was successful because it was decentralized and egalitarian. Anyone could run a server. You could access any site you wanted. Everybody could link to everybody else. People loved this and the Web spread for these reasons. Modern app ecosystems spread because capital wanted a return. It noticed people were getting rich off of this tech thing, and all that money from the rest of the world came pouring in. The defining feature is control: of experience, sharing, attention, and revenue. Because controlling all these things allows capital to extract greater returns. There are pros and cons to each model, but let's be clear, one was built from the ground up by financiers looking to make another buck. If it's easier to build a bottom nav bar or something, that's incidental to their plans. Why do you want a bottom nav bar, anyway? If you're truly building for the users, the first lesson is that they want their information on their terms.
- iainmerrick 9y ago
  In a free, decentralized environment, you'll get jackals (tragedy of the commons) That phrase "tragedy of the commons" has an interesting history! It's propaganda used to justify the privatization of common land, according to George Monbiot: http://www.monbiot.com/1994/01/01/the-tragedy-of-enclosure/ http://www.monbiot.com/1994/01/01/the-tragedy-of-enclosure/
richthegeek 9y ago
I think the main problem with that, coming from someone who has been programming on the web since around 2001, is that many of the attempts to 'fix' the web as an 'application platform' seem to come in very Enterprise Java costumes. Hundreds of files of abstraction and encapsulation, layers of indirection, tonnes of new nomenclature... and all without any obvious reason other than "that's what works for desktop applications" I think? Angular 1.x is the example I'm most familiar with (primarily because it was JS and 'easy' for the first 1000 lines), where most of StackOverflow was occupied with the "factory vs service vs provider" question.
amelius 9y ago
I think the problems all start with having a small number of companies controlling the OSes.
KaoruAoiShiho 9y ago
The article addresses the Bad Parts which are all browser side. The browsers suck. The web is fine. What we need are more powerful browsers that can be programmed by websites to act like native apps. Security can be handled through curation. What Mozilla (or a new browser startup) should do: Enable a new type of "extension". This extension is pre-approved and pre-checked for malware. This extension lives on the Mozilla cdn and is only downloadable from there. This extension is automatically installed on a site-specific basis with no user visible UI. This extension is powerful enough to control the browser and make it resemble a native app. When a user goes to the website it'll automatically just work just like the native app.
dmux 9y ago
I believe this is what Rebol IOS was trying to implement: http://www.rebol.com/ios/ios-intro.html http://www.rebol.com/ios/ios-intro.html

by xj9 9y ago

Andrex 9y ago

The lesson taken from email and chat platforms seems to be "centralize and control it," unfortunately. See Google's attempts with AMP as a potential solution. I know AMP isn't popular on HN, but it's a possible answer to your question. Speed, freedom, and usability - choose two, I suppose.

opportune 9y ago

I actually disagree. I think there are too many companies trying to put everything in an app when their content can be served and digested just fine via the web. And the blame there ultimately rests with the businesses that try to push their users into apps for whatever reason (gathering more data, easier to keep users within your ecosystem and not clicking away).

SirHound 9y ago

We have an ecosystem that's free, it is the web. The spam, the malware: A great deal of web apps don't have these issues. We need to continue to enhance the web as an application platform, rather than hobble it.

xj9 9y ago
i don't really see web applications ever being good mobile applications. to start with, jit compiling applications on a energy-constrained device is straight stupid. you should compile the application for me on some system that isn't already struggling to make it though to day on a single charge. wasm might help some, but pls don't make my mobile do more work than it has to. #p2p applications often need special capabilities or native libraries to work well (or at all). the web offers some weak apis for this, but really isn't up to par with what i would be interested in using (or building on). again wasm might help here, only time will tell. finally, web standards move very slowly and really don't exist as a disruptive force. if you need buy-in from all of the major vendors to push new/interesting features through, you aren't going to get much unless it favors them somehow.

oblio 9y ago

I keep hearing this and it seems wrong. I think the actual problem is a social one. In a free, decentralized environment, you'll get jackals (tragedy of the commons). In a centralized one, you'll get dictators. I don't think you can work around this with any kind of technology. I know that what I'm saying is basically anathema for HackerNews, but I think it's the truth.

apatters 9y ago
It's also worth noting that these dictators don't appear out of thin air. They come from somewhere else when they see an opportunity. The Web was successful because it was decentralized and egalitarian. Anyone could run a server. You could access any site you wanted. Everybody could link to everybody else. People loved this and the Web spread for these reasons. Modern app ecosystems spread because capital wanted a return. It noticed people were getting rich off of this tech thing, and all that money from the rest of the world came pouring in. The defining feature is control: of experience, sharing, attention, and revenue. Because controlling all these things allows capital to extract greater returns. There are pros and cons to each model, but let's be clear, one was built from the ground up by financiers looking to make another buck. If it's easier to build a bottom nav bar or something, that's incidental to their plans. Why do you want a bottom nav bar, anyway? If you're truly building for the users, the first lesson is that they want their information on their terms.
iainmerrick 9y ago
In a free, decentralized environment, you'll get jackals (tragedy of the commons) That phrase "tragedy of the commons" has an interesting history! It's propaganda used to justify the privatization of common land, according to George Monbiot: http://www.monbiot.com/1994/01/01/the-tragedy-of-enclosure/ http://www.monbiot.com/1994/01/01/the-tragedy-of-enclosure/

richthegeek 9y ago

I think the main problem with that, coming from someone who has been programming on the web since around 2001, is that many of the attempts to 'fix' the web as an 'application platform' seem to come in very Enterprise Java costumes. Hundreds of files of abstraction and encapsulation, layers of indirection, tonnes of new nomenclature... and all without any obvious reason other than "that's what works for desktop applications" I think? Angular 1.x is the example I'm most familiar with (primarily because it was JS and 'easy' for the first 1000 lines), where most of StackOverflow was occupied with the "factory vs service vs provider" question.

amelius 9y ago

I think the problems all start with having a small number of companies controlling the OSes.

KaoruAoiShiho 9y ago

The article addresses the Bad Parts which are all browser side. The browsers suck. The web is fine. What we need are more powerful browsers that can be programmed by websites to act like native apps. Security can be handled through curation. What Mozilla (or a new browser startup) should do: Enable a new type of "extension". This extension is pre-approved and pre-checked for malware. This extension lives on the Mozilla cdn and is only downloadable from there. This extension is automatically installed on a site-specific basis with no user visible UI. This extension is powerful enough to control the browser and make it resemble a native app. When a user goes to the website it'll automatically just work just like the native app.

dmux 9y ago

I believe this is what Rebol IOS was trying to implement: http://www.rebol.com/ios/ios-intro.html http://www.rebol.com/ios/ios-intro.html