perl-modperl mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From John Dunlap <j...@lariat.co>
Subject Re: beginning and testing
Date Tue, 07 Oct 2014 22:14:08 GMT
That is heavily dependent on the architecture of your application but,
architecture permitting, this is why you need to have different types of
tests. In my case, I work with a SOFEA(Service Oriented Front End
Architecture) which is the formal way of saying that I have a
html/javascript front end which talks to HTTP/REST/JSON web services via
Ajax calls which originate from a web browser. In this kind of
architecture, for example, you can create integration tests which test the
end-to-end behavior of your system between the web service client(a web
browser in this case but could just as easily be an iphone) and all of the
server-side components all the way down to your data store(usually a
relational database) and back again. Unit tests, by contrast, are generally
for testing isolated behavior within a small subset of your code base.
Automated user interface testing, by contrast, is an entirely different(and
very complicated) topic requiring entirely different tools. This kind of
testing isn't simply concerned with the UI behaving correctly but also with
verifying that it behaves correctly in as many browsers, operating systems,
and browser versions as possible. I've previously used tools like Selenium
for this kind of testing with great success.

On Tue, Oct 7, 2014 at 5:31 PM, Vincent Veyron <vv.lists@wanadoo.fr> wrote:

> On Tue, 7 Oct 2014 17:42:56 +0000
> Mark Hedges <Mark.Hedges@Ticketmaster.com> wrote:
> >
> > I have found Apache::Test extremely useful for testing web
> applications.  Don't let anyone tell you to run manual tests, that is a
> recipe for confusion and error as you try to modify code later.
> >
>
> How do you use that with a web application that is a work in progress,
> though?
>
> I write mod_perl modules that output html forms. One has to write a test
> that sends requests and compares the responses with an expected result. But
> that result changes every time I make the slightest change in code, which
> happens quite often and in multiple places. So maintaining the test suite
> on top of the code base seems appealing in theory, but I found it
> overwhelming in practice.
>
> I can see it happening for a well matured app that changes rarely, but for
> a new one?
>
>
> --
>                                         Regards, Vincent Veyron
>
> https://libremen.com/
> Legal case, contract and insurance claim management software
>



-- 
John Dunlap
*CTO | Lariat *

*Direct:*
*john@lariat.co <john@lariat.co>*

*Customer Service:*
877.268.6667
support@lariat.co

Mime
View raw message