I think that all of this is true and that it presents a problem for those
of us that are entrepreneurially minded. Just as the angel and vc community
is getting used to the new pace of things and realizing that you can start
something big by throwing less than 20K at some kids, it's no longer true.

Building highly modularized app that supports several integrations to
disperate services and that supports several clients and building those
clients in their own native languages is way harder than building the
original twitter, and the tools to pull off this major feat aren't really
that streamlined because suddenly your not in the "80% of web apps are just
the same stuff with a different design" mode anymore. They are really
different and use different technologies and have different areas that need
scale and manage data in different ways. To write apps like this, you need
experts and capital.

Historically speaking, I think this is more the normal mode. You can't make
something really new and useful without some serious input. But it's too
bad, because for a while between about 2003 and 2008 you could.

--Jon

On Wed, Nov 30, 2011 at 3:08 PM, John Lynch <[email protected]> wrote:

> I think architecture is just a lot more important, especially when
> building larger systems. Lots of smaller, testable black boxes with
> well-defined interfaces. This lets you scale easier, experiment with
> different data stores/languages/frameworks for different problem domains,
> etc. However, in our recent experience building a Rails app for a new
> start-up, as much as we wanted to use a SoA style from the beginning, we
> just didnt know enough about the app.  It was easier to prototype/sketch
> out the app as a standard monolithic Rails app, and then once the natural
> fault lines started to emerge, extract out functionality into a service.
> Rinse, and repeat.
>
> Regards,
>
> John Lynch
> [email protected]
>
>
>
>
> On Wed, Nov 30, 2011 at 10:50 AM, Chris McCann <[email protected]>wrote:
>
>> An interesting post about where Rails fits in with the current web-
>> enabled application landscape.
>>
>> http://broadcastingadam.com/2011/11/moving_on_from_rails
>>
>> Thoughts?
>>
>> Chris
>>
>> --
>> SD Ruby mailing list
>> [email protected]
>> http://groups.google.com/group/sdruby
>
>
>  --
> SD Ruby mailing list
> [email protected]
> http://groups.google.com/group/sdruby
>

-- 
SD Ruby mailing list
[email protected]
http://groups.google.com/group/sdruby

Reply via email to