Article: Why Mobile Web Apps are Slow

[D. Hugh Redelmeier]

| From: Scott Sullivan <scott-lxSQFCZeNF4 at public.gmane.org>

| http://sealedabstract.com/rants/why-mobile-web-apps-are-slow/
| 
| It came to me from an ARM mailing list because mobile is largely ARM(*) based.
| The article covers a fair bit of ground

I don't program on mobile platforms, or in Java, but I've programmed
on generations of machines, ones with varying levels of resource
constraints.

I think he's probably pretty accurate about right now.  But a little
more nuanced position makes sense:

- resource constraints seem to always get slacker, except for periodic
  dry spells.  Of course the music has to stop some time.

- many wins have come to teams that design products that are not
  reasonable today, but are reasonable next year, when they actually
  ship.

- writing an app vs a web page has a lot of silly legal/economic
  consequences: Apple lets you provide its users with a web site
  without permission or a cut; not so with apps.  That's probably
  why they hated Flash.

  It is astonishing to me that they will not accept an app that
  provides a programming environment.  And get away with that
  restriction.  Sure seems anti-competitive.

  Javascript is the one that got away.  Somehow they've even decided
  to provide a performant implementation.

- writing an app makes it hard to support multiple platforms.  This
  lock-in is anti-competitive (but can look as if it is a technical
  requirement, so invites little scrutiny).

Most programs I write would not be different with a factor of two
change in speed or space.  A factor of 5 gets noticeable.

But a language like Java is probably fine if a suitable resource
intensive kernel is moved to some other implementation.  This kernel
is likely very small.  I don't know if JNI makes this reasonable.
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists