[wp-hackers] GSoC 2014 Questions

[Benedict, Keith (Digital)]

I saw this thread and thought I’d pop a note on it…

Having just worked through the development of a “Simple” app using phone gap, which is now submitted and undergoing apple review.  I have to say I would almost never do it again, and if I did I would not underestimate the numerous issues that are faced with the simplest of tasks.

For instance, something as simple as the status bar in IOS7 versus 6 initially led us to believe we would require a plugin…which didn’t work as advertised, that then led us to add pixels to push the screen down on IOS 7, which again didn’t work as well as advertised.  We ended up fussing about this tiny issue for a significant amount of time.

Phone gap makes it easier to deploy apps with no objective-c knowledge.  However it does guarantee you can make “Good” apps without this knowledge.  Whomever is building the app will need to run Xcode on their Mac OS X system, and likely will be manipulating the delegate.m file to add in some code when it comes to any sort of add-ins like analytics additions, alerts, in-app web browser, etc.. They also need to ensure classes are registered, frameworks added and modify the plist file for various settings.  This all of course is depending on your needs, I doubt most apps will make it through without at least one phone gap plugin.

Also the documentation for many of those is horribly inaccurate at times to completely wrong in others.  Be warned that developers will need to do their research.  We worked on a project that was given 2 months and it took 3 instead after having to work through the “phone gap -isms”.

The alternative to a local build is you can use the build service for phone gap, however….If you are not paying for an account you are restricted to a 10 MB project.  If you are paying, you can get up to 50mb project.  The trick here is this size is the raw size of all your generated code, media assets and application code for all platforms…so if you have a few mb of assets for each, you’ll nearly start bumping the free limit.  So the alternative is to develop the app locally on your Mac OS X machine running Xcode (For IOS builds anyway).

Lastly….just because it helps build multi platform.  Expect a lot of “Why doesn’t that work on android…” or “Why doesn’t that work on IOS…”  It’s no where near perfect, you have to hand hold both and again with Android you will want someone troubleshooting it with Java experience using the Android SDK on Eclipse…it is incredibly slow, but it is the way it is done.

Our app is completed and going through apple review, it ties into our Wordpress site nicely, and it took a lot of work. I truly think we would have had a better app if we spent the month figuring out issues and learning object-c or java as needed.

Good luck!


On Mar 3, 2014, at 1:19 PM, Marko Heijnen <mailing at markoheijnen.nl<mailto:mailing at markoheijnen.nl>> wrote:

Hey,

I disagree that the benefits outweighs the concerns but it’s not only for me to decide if PhoneGap is a viable solution for GSoC.
If the mobile team will manage most of it then it should be objective-c since they know that better. Also I’m still unsure if building it with PhoneGap is a two month project.

1. Not having a developer account would mean that you need to buy it. You do earn money with GSoC so I would see it as a investment.
2. We should have a special version for tables.
3. 3 would be solved with 1 but I guess having the ability to borrow a iPhone is probably necessary.

Marko

Op 3 mrt. 2014, om 08:57 heeft Yatin Maan <yatinmaan1 at gmail.com<mailto:yatinmaan1 at gmail.com>> het volgende geschreven:

Anybody there ?


On Sat, Mar 1, 2014 at 9:40 PM, Yatin Maan <yatinmaan1 at gmail.com<mailto:yatinmaan1 at gmail.com>> wrote:

Hey,

I am aware of that performance is not one of PhoneGap's strongest points
but the app we are developing here is simple enough to not be bottlenecks
by PhoneGap. If enough time is invested in optimising the app then it can
achieve almost native speed (Health Tap was featured as an Editor's choice
on the AppStore).
Plus, using PhoneGap we can deploy on multiple platforms ensuring maximum
availability (I am thinking iOS , Android and if time permits Windows
Phone).
And all the apps having a single code base has its own benefits , imagine
bugs are fixed/features are added to iOS versions yet Android is lagging
behind because the particular developer who added the feature doesn't know
Java. What I am trying  to say here is that yes PhoneGap Apps are a bit
slow but if enough work is put into it that lag can be overcome (eg you can
remove atleast 300ms delay per touch by using ontouch instead of
onclick)  and the benefits provided by PhoneGap far outweigh the concerns.

Now onto some questions,
1. How do compile the iOS App as I don't have an developer account ? (You
need to have a key to
3. Do we need to create special versions for tablets or would stretching
be fine ?
2. How do I debug the iOS App as  I don't have an iPhone ? (Though I have
an iPad but its not jail broken and I would need an developer account to
get the app on the iPad.)

Thanks,
Yatin