Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
rodion_89
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
9 ms
·
31.
▲
by
rodion_89
14y ago
We made a similar tool for making apps like this to make our mobile Reddit app. It works really well on iOS and Android, blending into their UI design. http://reddit.kik.com Framework: https://github.com/kikinteractive/app
32.
▲
by
rodion_89
14y ago
FYI, it took Facebook 2.5 years to release the first version of their API.
33.
▲
by
rodion_89
14y ago
A couple of months ago I mentioned these same notions about the Facebook app. http://jairaj.org/2012/08/27/facebook-for-ios.html Although I have to admit that Sencha's implementation isn't all that great. If you're going to offer an alter
34.
▲
by
rodion_89
14y ago
I wrote a similar library for some mobile web apps I'm working on. https://github.com/kikinteractive/clickable It has support for: * "fast" clicking * managing downstate in CSS (like :active) * using the native disabled property * sticky
35.
▲
by
rodion_89
14y ago
> and ooh, they can't tell you anything about the idea because you might steal it In my experience the fact that someone needs to hide it usually means that they have little to bring to the table other than that idea.
36.
▲
by
rodion_89
14y ago
I'm curious, how many clicks is it to replace Bing as the default in IE? Isn't it impossible for DDG to become the default in Safari? Not saying they're wrong to point it out, but rather that they should be consistent in pointing fingers if
37.
▲
by
rodion_89
14y ago
> whereas jQuery UI datepicker is 42kb minified For anyone who cares, the Closure Compiler gets jQuery UI Datepicker down to 30kb (11kb gzipped). Not quite as drastic as he claims but still significantly larger compared to his 6.7kb (2.
38.
▲
by
rodion_89
14y ago
> HTML5 on the desktop I wasn't talking about mobile. Try iCloud.com on Mac Safari, it's quite nice. Not sure what browsers Apple is targeting for this. My assumption is that the 5 major desktop browsers can host web apps of this qualit
39.
▲
by
rodion_89
14y ago
I actually like the way Web Workers can't touch the DOM. It's nice to finally give the DOM it's personal space for UI operations, which often leads to much more responsive apps especially at a large scale. The pain points for me come in the
40.
▲
by
rodion_89
14y ago
I guess my question then is, why does that hold more true for the App Store than for any of their other native apps?
41.
▲
by
rodion_89
14y ago
I'm pretty sure I made no comment as to which is better, jQuery or CSS3 animations. On mobile there is no question whatsoever that CSS3 animations are much better. The big issues tend to be regarding 3d stacking contexts. You can end up wit
42.
▲
by
rodion_89
14y ago
JavaScript execution speed is no longer a major limiting factor on mobile. In terms of UI JavaScript execution speed will help you if you're using jQuery (or any naive animator) to animate 2d CSS properties. The DOM, 3d transforms, etc are
43.
▲
by
rodion_89
14y ago
> There is no question that native apps provide better UI's This is was me talking about mobile. Then I went on to explain why. So I don't think I need to provide any examples there. As for HTML5 on the desktop you can peruse the Chrome
44.
▲
by
rodion_89
14y ago
UI rendering and a sane threading model are two major steps that could end these premature HTML5 vs native debate. I wait for the day when these debates are equivalent to arguments about programming language preference.
45.
▲
by
rodion_89
14y ago
I've always wondered why the App Store was a web app. It seems really out of place considering its job is to distribute native apps. On the other hand it has probably been the driving force to make mobile safari better.
46.
▲
by
rodion_89
14y ago
The argument that HTML5 apps cannot provide sufficient quality is growing increasing weak. There is no question that native apps provide better UI's, especially in the case of iOS. As for Facebook, they built a terrible iOS app and blamed t
47.
▲
Show HN: I enhanced Google's new sidebar [open on mobile]
(google.jairaj.org)
3 points
by
rodion_89
14y ago
|
0 comments
48.
▲
Zerver - A new kind of webapp server
(zerver.jairaj.org)
1 points
by
rodion_89
14y ago
|
0 comments
49.
▲
by
rodion_89
14y ago
What browsers support this? Chrome 23.0.1271.22 on OS X and it doesn't seem to be supported.
50.
▲
by
rodion_89
14y ago
These number are very different than that shown by StatCounter [1] and Wikipedia [2]. If you look at Wikipedia's comparisons, they compare data from many sources. NetApplications (the dataset used by this article) is by far the one source
51.
▲
by
rodion_89
14y ago
According to Wikipedia, that version was released last month. http://en.m.wikipedia.org/wiki/Firefox#section_6
52.
▲
by
rodion_89
14y ago
> people for some reason buy into it It's not just "some" reason. They have one of the best marketing departments in the world who do an excellent job. > I don't understand why so many people blame Apple for this. I'm
53.
▲
by
rodion_89
14y ago
> Why does every new phone have to be a fucking revolution? I completely agree. But Apple has a habit of making it seem like everything they do is revolutionary. It's very good marketing. > Apple reinvents the phone [1]
54.
▲
by
rodion_89
14y ago
According to comScore, Yahoo Mail is the third largest mail provider with over 300 million users. Kicker: > in the U.S., Yahoo is No. 1, with 96.6 million active users http://news.cnet.com/8301-13772_3-20114975-52/microsoft-aimi...
55.
▲
by
rodion_89
14y ago
If I remember correctly, Messages were already native. The news feed was the only major change. But yes, I'm not saying that their app is fully HTML5. I'm saying that HTML5 still has a major role in their app.
56.
▲
by
rodion_89
14y ago
That's kinda the point. > No app store? No problem.
57.
▲
by
rodion_89
14y ago
Facebook's new iOS app is not fully native. Not even a mostly. http://jairaj.org/2012/08/27/facebook-for-ios.html
58.
▲
Facebook's iOS update is not about HTML5
(jairaj.org)
3 points
by
rodion_89
14y ago
|
0 comments
59.
▲
by
rodion_89
14y ago
That's just a discovery page for apps that use the Kik API.
60.
▲
by
rodion_89
14y ago
470K bits == 58.75K bytes (assuming 8-bit bytes)
More ›