Is HTML5 really a good idea for mobile development?

[Fabrizio Giudici]

Some people don't think so:

http://sandofsky.com/blog/shell-apps.html


I'm discussing this privately with some people doing lots of mobile
development and they agree. I think it's interesting to extend the
discussion here.




--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."
fabrizio...@tidalwave.it
http://tidalwave.it - http://fabriziogiudici.it

[federico.y...@post.ch]

Ciao Fabrizio,

your friends agree that HTML5 is a good idea or not?

Personally I think:
- If you are living on the edge & you need always latest platform feature --> go for native
- if you are an enterprise and is more important that all the client platform are updated at the same time --> go for HTML5

Also, maintaining enterprise applications for multiple platforms (2?, 3?, 4?... how many in the future?) can become very costly...

Ciao,
Federico


-----Messaggio originale-----
Da: jugl...@googlegroups.com [mailto:jugl...@googlegroups.com] Per conto di Fabrizio Giudici
Inviato: giovedì, 10. maggio 2012 13:30
A: JUG Lugano
Oggetto: [juglugano] Is HTML5 really a good idea for mobile development?

[Fabrizio Giudici]

On Thu, 10 May 2012 15:44:41 +0200, <federico.y...@post.ch> wrote:

> Ciao Fabrizio,
>
> your friends agree that HTML5 is a good idea or not?
>
> Personally I think:
> - If you are living on the edge & you need always latest platform
> feature --> go for native
> - if you are an enterprise and is more important that all the client
> platform are updated at the same time --> go for HTML5
>
> Also, maintaining enterprise applications for multiple platforms (2?,
> 3?, 4?... how many in the future?) can become very costly...

Sorry for the late answer, but I couldn't connect yesterday.

My friends say that they found themselves in the same position as the blog
post: they happily started with non-native solutions (*), but soon
realized that they reached a (negative) break-even. Their feeling include
all the aspects that you are referring to, including deployment, etc...

(*) For this discussion I used the same subject as the original blog post,
but indeed I think that the comparison is "native vs non native" solutions
for mobile development.

[Sebastiano Cobianco]

i have mixed views on this topic, but basically my current rational
would be

-- if you are developing something plain and standard then consider
(not necessarily pick) html5
-- else go native

to support the above

first, html5 is yet to be a standard specification. in fact, by html5
we can only refer to a de facto situation whereby w3c is still in
troubled waters and months/years away from being able to release a
final specification; meanwhile sw and browser vendors are going their
way, resulting in 2 partly-overlapping major "implementations" of
html5, driven by market opportunity and corporate fights rather than
language robustness, interoperability, etc....the "safe" part of the
so-called html5 is therefore a subset of the to-be specification.
selecting html5 to build an app must take the current industry
scenario into account. the same "hrml5" code might render differently
on differing devices or browsers. also, there is no guarantee that
your working code will keep working going forward (as the spec is not
a final spec).

second, html5 doesnt provide the developer with the full api of the
device. for instance, you want to develop an app which includes the
ability to shoot pictures. you are able to capture pictures from html5
code using the device camera, but how about zoom, grid, manipulate
colours, effects, crop, flashlight, red-eye, etc...? the full camera
api is available to native apps, but html5 only plays with a subset of
it, restricting the developer's ability to cater for specific or
advanced needs.

on the other hand, i can see the allure to develop with html5,
particularly in relation to android fragmentation. taking versions and
forks into consideration my last count is 75 differing combinations in
which you could find yourself deploying your android app. quite
daunting isnt it? memories of j2me vietnam...iOS is more predictable
on that respect, but comes with strings attached (actually, ropes
attached...)

ciao


On May 12, 11:37 am, "Fabrizio Giudici"
<Fabrizio.Giud...@tidalwave.it> wrote:

> fabrizio.giud...@tidalwave.ithttp://tidalwave.it-http://fabriziogiudici.it

[The Dude]

-> Sebastiano Cobianco writes:

> i have mixed views on this topic, but basically my current rational
> would be

> -- if you are developing something plain and standard then consider
> (not necessarily pick) html5
> -- else go native

[...]

Cool post Sebastiano, thx.

Let's be polemic and have a real 'mud fight'! Who has some contra
arguments? Come on guys, don't be shy;-)

ciao
--
-dude

Too brief? Here's why! http://emailcharter.org

[Fabrizio Giudici]

On Tue, 15 May 2012 09:04:38 +0200, The Dude <rafal.k...@mac.com>
wrote:

> -> Sebastiano Cobianco writes:
>
>> i have mixed views on this topic, but basically my current rational
>> would be
>
>> -- if you are developing something plain and standard then consider
>> (not necessarily pick) html5
>> -- else go native
>
> [...]
>
> Cool post Sebastiano, thx.
>
> Let's be polemic and have a real 'mud fight'! Who has some contra
> arguments? Come on guys, don't be shy;-)

Exactly, this is the typical topic where the more opinions, the better,
because it's hard to define that boundary where you make the decision.
What are the points that HTML 5 can't properly handle? Gestures? Speed?

Personally, I don't dislike HTML 5 as a rendering technology. What I
really don't like is the normal way it's used (AFAIU) in Android, that is
with JS and mimicking a client-server application. I think it would be
nice if I could use HTML 5 to lay out the user interface (e.g. in place of
the XML layout facility provided by Android), but natively binding to Java
(probably I'd like to do that even in plain Java desktop apps).

--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."

[federico.y...@post.ch]

I don't like Sencha Touch, but I think this video has some good arguments about why to go for HTML5 instead of native:
http://www.infoq.com/presentations/A-Snapshot-of-the-Mobile-HTML5-Revolution

I'm currently developing a small prototype using AngularJS & jQueryMobile, and I think the complexity is now manageable (much less than JavaScript without frameworks).

My next question is: how many experts in the HTML5+JS+CSS3 there are in Ticino? I mean, how much hard is to find people able to properly implement complex & rich web applications?

-----Messaggio originale-----
Da: jugl...@googlegroups.com [mailto:jugl...@googlegroups.com] Per conto di Fabrizio Giudici

Inviato: martedì, 15. maggio 2012 09:16
A: jugl...@googlegroups.com; The Dude
Oggetto: Re: R: [juglugano] Is HTML5 really a good idea for mobile development?

[mas...@massimocorner.com]

> My next question is: how many experts in the HTML5+JS+CSS3 there are in
> Ticino? I mean, how much hard is to find people able to properly implement
> complex & rich web applications?

It's time to pay more money for senior front-end developers :-)

Massimo

[Sylvain Jermini]

On 15 May 2012 15:43, <federico.y...@post.ch> wrote:
> I don't like Sencha Touch, but I think this video has some good arguments about why to go for HTML5 instead of native

And others arguments against it:
http://blog.mobtest.com/2012/05/heres-why-the-facebook-ios-app-is-so-bad-uiwebviews-and-no-nitro/
. Talk about gimping the webview in IOS.

[Davide]

Dear all,
 I would like share with you two white-paper about HTML5 and open web technology to develop mobile app; I hope it can contribute to this converation-thread.

Bye

--
Davide Carnevali

[Sebastiano Cobianco]

hmmm more on Android fragmentation, this time taking differing devices
(ROM) into account
http://arstechnica.com/gadgets/2012/05/android-fragmentation-one-developer-encounters-3997-devices/

it seems like almost impossible to deploy native app for "the
majority" of the android market!!....intimidating for developers, and
for business sponsors too.
will this drive android native development down in the long run?

ciao


On May 13, 3:48 pm, Sebastiano Cobianco

[Fabrizio Giudici]

On Tue, 22 May 2012 18:12:18 +0200, Sebastiano Cobianco
<sebastian...@googlemail.com> wrote:

> hmmm more on Android fragmentation, this time taking differing devices
> (ROM) into account
> http://arstechnica.com/gadgets/2012/05/android-fragmentation-one-developer-encounters-3997-devices/
>
> it seems like almost impossible to deploy native app for "the
> majority" of the android market!!....intimidating for developers, and
> for business sponsors too.
> will this drive android native development down in the long run?

Well, honestly I find a bit misleading to count the number of devices, as
the important points are the os version, screen size, capabilities, etc.
It's probably not a very small number, but not so high. Does really HTML 5
solve anything in this field? After all, if the os version is different,
even the WebView component is different and might render HTML in with
different bugs. If the screen size changes, HTML 5 doesn't solve
automatically the problem. You can design a resizable UI with Android code
or HTML as well. The relevant point is that you have to *test* with the
higher number of devices and testing is the part of the workflow where I
see a cost increment because of so many devices. But the need for testing
is independent from the technology used for the development.

--
Fabrizio Giudici - Java Architect, Project Manager
Tidalwave s.a.s. - "We make Java work. Everywhere."