Re: WWDC2k5 and ROI

FROM : Scott Ellsworth
DATE : Tue Apr 12 02:59:01 2005

On Apr 11, 2005, at 10:33 AM, Joseph Graham wrote:

I am only going to respond to the actual Cocoa development component 
of this question.  I choose to interpret it as "what is the business 
case for Cocoa?"

Well, I am going to WWDC to stump for my favorite technologies, to 
ask the engineers a bucket of questions, and to learn new stuff.

We do biotech work, primarily, and thus we primarily do Java and 
Perl.  Both of those are well supported on MacOS X, especially if you 
are running Tiger such that you have Java 1.5.  <http://
developer.apple.com/java/faq/#anchor2>.  The Mac is a keen platform 
for these, but I have some requests of the engineers.  This justifies 
the WWDC trip.  (This is, of course, tangential to Cocoa.)

Some of my projects need the fastest turnaround I can get, and Cocoa 
solves that problem elegantly.  I can write a UI in Java, and make it 
look pretty good.  I can write it in Cocoa, and have it look pretty 
good with less effort.  If my desired end result is a gui app to run 
on a Linux or Windows box, then I rarely start with Cocoa.  On the 
other hand, if the purpose is to wrap a gui around a command line 
tool that might be MATLAB, FORTRAN, Java, Perl, or whatever, Cocoa is 
not a bad way to do it.

So, for those projects, Cocoa is pretty keen.  It is also a good way 
to write productivity enhancers for my workflow.

> But I think that Cocoa is a steep learning curve to many veteran IT 
> developers who are not versed in ObjC.


True.  On the other hand, it lets you push out apps very fast.  This 
was one of the key bennies for WO in years past, and is one of the 
cookies of Cocoa today.

> I know [PHP, J2SE, Mono, Ruby, ...] are mostly available
> for OSX and they are not mutually exclusive.


In point of fact, we develop a lot in Java, my last contract was in 
PHP, and a prior one was in Perl.  In the main, I get more paying 
work in Java than anything else, but things do change with time. 
Further, if I can deliver a proof of concept through the magic of IB 
in a day, then I can turn it into a real app later.  Cocoa excels at 
this.

> I think that by adopting these technologies for development 
> projects have relatively massive ROI because of deployment 
> scenarios available.


Frankly, you pick the technology based on what the client needs and 
what they expect to use it for.  If they want more than just the Mac, 
then Cocoa is a non starter, but Java might not be.  If they are a 
windows only shop, then C# might be the language of choice.  Given 
that Cocoa gets functional apps out the door very quickly, and with 
full access to the latest Apple features, it has substantial ROI.

So, if your clients do not want the extra deployment scenarios, you 
are not getting any extra ROI.

> So my main question is what is Apple's ROI plan for Cocoa?  Why 
> would anyone put forth a huge investment in Cocoa such as massive 
> Cocoa-Native projects?


Speed of development and power of resulting tools.  Cocoa lets you 
write useful apps very quickly, and upcoming Tiger technologies make 
that happen faster.  Frankly, if Core Data also had Extra EOF 
Sprinkles, then I would have everything I need to prototype in Cocoa, 
curate with custom Cocoa apps, then implement web and desktop front 
ends in Java or PHP.  Even as it is, I can do a prototype fast, then 
get the client's feedback on where they want to take that prototype.

Neat.

Scott