On Tuesday, 2 June 2015 at 11:26:06 UTC, Paulo  Pinto wrote:
> On Tuesday, 2 June 2015 at 10:18:59 UTC, ketmar wrote:
>> On Tue, 02 Jun 2015 09:44:24 +0000, Paulo  Pinto wrote:
>>
>>> Thankfully mobile OSes and desktop app stores seem to be on the right
>>> track to kill this.
>>
>> yet they pushing cromeos and firefoxos...
>
> FirefoxOS is going nowhere.
>
> http://www.cnet.com/uk/news/mozilla-overhauls-firefox-smartphone-plan-to-focus-on-quality-not-cost/
>
> As for Chromebooks, at least in Germany they are gathering dust on the few stores that bother to try to sell them.
>
> They might be on the Amazon US top charts, but I bet most of its users are GNU/Linux users, wiping ChromeOS and using GNU/Linux instead.

They're insanely popular, especially in educational environments. They do everything 98% of modern computer users do, which is generally check email, browse facebook, and use twitter.

my local public highschool ordered 800-some of them instead of upgrading their ipads(???), which would have cost far, far more. AFAIK they got the cool thinkpad versions.

On 06/02/2015 09:43 AM, "Ola Fosheim =?UTF-8?B?R3LDuHN0YWQi?= <ola.fosheim.grostad+dlang@gmail.com>" wrote:
> On Tuesday, 2 June 2015 at 13:37:47 UTC, ketmar wrote:
>> and there was a thing that allows to use real applications without
>> installing them... java web start! way too ahead of it's time, though...

And 0install. (Though I admit I haven't checked in on that in a few years. But I always wanted to see that, or something like it, succeed.)

> Java got itself a bad reputation in the 90s by providing a really poor
> implementation.
>

Over the years, I've noticed that one of the best ways to kill an idea is to get everyone onboard a really, really bad implementation of it. (Ex, see: "email" vs "exchange server and webmail clients", or "static type system" vs "C++ and Java". Or IMHO: "desktop/laptop computers" vs "every major desktop OS in existence" ;))

On 06/02/2015 05:44 AM, Paulo Pinto wrote:
>
> Thankfully mobile OSes and desktop app stores seem to be on the right
> track to kill this.
>
> http://www.quirksmode.org/blog/archives/2015/05/web_vs_native_l.html

+1billion

My god it's nice to see that finally acknowledged.

"Still, we web developers have spent the last six years in denial. Our working assumption has been that all web sites should be app-like, and therefore tooled up to the hilt."

Love it.

Thing is, one of the biggest reasons (likely even THE biggest reason) for the giant "web for apps" push was the whole no-install aspect. But then instead of actually, y'know, creating a no-install for applications, which could have been made and well-entrenched by now, the whole tech sector went and blew (probably) billions in $ and time retrofitting "rich, dynamic experience" into a document platform. And for all those blown resources, it's STILL at least as much of a broken mess as it was back when we still thought the IE/Netscape divergence was the greatest damage we could ever inflict our unfortunate web-based users.

And then on top of all that, Java/Flash/JS/basic-freaking-displaying-a-stupid-little-image demonstrated that even the whole purported "sandboxing" benefit of web apps (and VMs for that matter) was a near total bust.

On Tuesday, 2 June 2015 at 23:04:30 UTC, Nick Sabalausky wrote:
> On 06/02/2015 05:44 AM, Paulo Pinto wrote:
>>
>> Thankfully mobile OSes and desktop app stores seem to be on the right
>> track to kill this.
>>
>> http://www.quirksmode.org/blog/archives/2015/05/web_vs_native_l.html
>
> +1billion
>
> My god it's nice to see that finally acknowledged.
>
> "Still, we web developers have spent the last six years in denial. Our working assumption has been that all web sites should be app-like, and therefore tooled up to the hilt."
>
> Love it.
>
> Thing is, one of the biggest reasons (likely even THE biggest reason) for the giant "web for apps" push was the whole no-install aspect. But then instead of actually, y'know, creating a no-install for applications, which could have been made and well-entrenched by now, the whole tech sector went and blew (probably) billions in $ and time retrofitting "rich, dynamic experience" into a document platform. And for all those blown resources, it's STILL at least as much of a broken mess as it was back when we still thought the IE/Netscape divergence was the greatest damage we could ever inflict our unfortunate web-based users.
>
> And then on top of all that, Java/Flash/JS/basic-freaking-displaying-a-stupid-little-image demonstrated that even the whole purported "sandboxing" benefit of web apps (and VMs for that matter) was a near total bust.

I have to disable javascript on amazon.com to be able to use the site or else it brings my browser to a crawl.

I, for one, am in favor of scrapping javascript and replacing it with lua. At least it has a decent JIT implementation.

fun tidbit, back when the benchmark's game had luajit on it, luajit's _interpreter_ would beat V8 in every single benchmark.

On Tue, 02 Jun 2015 17:04:58 +0000, Ola Fosheim Grøstad wrote:

> On Tuesday, 2 June 2015 at 14:01:48 UTC, ketmar wrote:
>> On Tue, 02 Jun 2015 13:43:29 +0000, Ola Fosheim Grøstad wrote:
>>
>>> The situation that is emerging now is that you can integrate different browser technologies, and that browser vendors cooperate, and that makes all the difference.
>>
>> except you can't really integrate browser techs in your program, but you are required to integrate your program into browser. ;-)
> 
> https://github.com/domokit/mojo

exactly! "integrate your app into our stinkin' browser!"

On Tue, 02 Jun 2015 18:41:20 -0400, Nick Sabalausky wrote:

> On 06/02/2015 09:43 AM, "Ola Fosheim =?UTF-8?B?R3LDuHN0YWQi?= <ola.fosheim.grostad+dlang@gmail.com>" wrote:
>> On Tuesday, 2 June 2015 at 13:37:47 UTC, ketmar wrote:
>>> and there was a thing that allows to use real applications without installing them... java web start! way too ahead of it's time, though...
> 
> And 0install. (Though I admit I haven't checked in on that in a few years. But I always wanted to see that, or something like it, succeed.)

that's a different story. 0install is a self-containing bundle with all libs and so on. you got it, you run it, you use it.

and javaws is a something completely different. it doesn't even download the whole app at the start -- it can download only the classes it requires to run right now, downloading the rest on background or when user activates a feature for the first time.

besides, application is not a bundle, it's more like a browser cache. javaws can check for updates transparently and update only parts of application (think about fixing some bug in some classes -- whoa, only that classes need to be downloaded). and if user didn't use a feature that requres updated class yet, he doesn't even need to restart the app.

that said, javaws was a great idea -- killed by java and by being too advanced for it's time.

On Tue, 02 Jun 2015 23:51:13 +0000, weaselcat wrote:

> I, for one, am in favor of scrapping javascript and replacing it with lua. At least it has a decent JIT implementation.

it doesn't really matter. be it js, Lua, basic -- it's all equally bad. websites are not applications, and they doesn't need any scripting at all.

On Tuesday, 2 June 2015 at 22:38:47 UTC, weaselcat wrote:
> They're insanely popular, especially in educational environments. They do everything 98% of modern computer users do, which is generally check email, browse facebook, and use twitter.

Not really.  While they do sell some in education, they were 1.8% of the PC market last year, much less than even Macs despite being much cheaper:

https://www.petri.com/chromebook-continues-to-be-a-tiny-slice-of-the-pc-market

Compare that 5.7 million in sales to a billion Android devices sold last year, native is definitely winning.

On 06/02/2015 07:51 PM, weaselcat wrote:
>
> I have to disable javascript on amazon.com to be able to use the site or
> else it brings my browser to a crawl.
>

When I finally figured out how to hack (by that I mean "load it down with an endless list of add-ons" and THEN waste an evening configuring) a modern Firefox to be tolerable, I thought I was finally in pretty decent good shape. (Well, at least until I was subjected to Mozilla's next couple rounds of UI blunders anyway...)

But now that I've finally accepted defeat on "screw cellphones" and been in the android habit for over a year, I've found myself in a new web hell: Now, no matter what mobile browser I use (Chrome, Firefox, Dolphin, or the amazingly-incorrectly named "Internet"), I have a choice:

A. Use the goofy, straightjacketed, sluggish "mobile" versions of websites (Wikipedia's is particularly bad, what with the auto-collapsing of every section on the entire page *while* you're scrolling and reading and then again every time you navigate "back" to a section you'd already re-expanded manually. Gee, thanks for closing my book, Wikipedia, never mind that I was reading it.)

Or B. Switch on "request desktop site" mode and get *improved* mobile UX but can easily take 30 seconds to a full minute (per page) before becoming responsive enough to accept clicking on a link. And that's for pages that aren't even dynamic beyond the initial flurry of JS onLoad() nonsense.

(Seriously, nearly anything that can be run during onLoad(), BELONGS on the server. Why would ANYONE in their right mind EVER make every single client do several Ajax/REST/whatever requests, THEN render the EXACT SAME page every time, for every single incoming request? Instead of, oh, I dunno, rendering the EXACT SAME page ONCE when content ACTUALLY changes and having the server spit THAT out to every request? Not enterprisey enough, I guess. Really, how often does a blog or news site actually post or modify an article? That exact same pages REALLY need to get completely regenerated on every hit even though NOTHING has changed since the last 500 hits?)

What I find very interesting is that it's consistently big businesses that have the most impossible-to-use sites. Ex: Just look at any site by SCE. You'd almost think they *don't* want any viewers and customers.

On Wednesday, 3 June 2015 at 03:41:39 UTC, Joakim wrote:
> On Tuesday, 2 June 2015 at 22:38:47 UTC, weaselcat wrote:
>> They're insanely popular, especially in educational environments. They do everything 98% of modern computer users do, which is generally check email, browse facebook, and use twitter.
>
> Not really.  While they do sell some in education, they were 1.8% of the PC market last year, much less than even Macs despite being much cheaper:
>
> https://www.petri.com/chromebook-continues-to-be-a-tiny-slice-of-the-pc-market
>
> Compare that 5.7 million in sales to a billion Android devices sold last year, native is definitely winning.

chromebooks weren't even really usable until the latter half of 2013/start of 2014 when Acer/HP/Dell/Toshiba/etc all got on board and it stopped being just Samsung making them. 2% is huge for less than 2 years. That was the chromebook revision that featured the ultra low power Haswell CPUs(2955U,) before that they were incredibly slow and suffered from general netbook issues.

And they're not even comparable to an android /phone/. Compare them to tablet sales.