Since brian d foy pointed out that the majority of the Perl world doesn't know about some recent developements with Test::Harness, I thought Perlmonks would be a good place to mention this and solicit feedback.
As of this writing, there are over 50 bugs in the RT queue for Test::Harness (I can't tell how many because many are spam). Part of the problem is that the current Perl testing framework has evolved over almost 2 decades into Test::Harness and related modules. As anyone with a biology background knows, evolution leads to many inefficiencies and Test::Harness has become very hard to work with and extend.
Recently, due to my work with TAPx::Parser, Andy Lester (petdance) sent me an email suggesting that TAPx::Parser would be a good candidate for Test::Harness 3.0. For most of us, that simply means that the default output is more interesting and useful. It can be customized to your needs and even driven through a GUI. Basically, because the parsing, analyzing and presentation of data are all separated, the testing tools are much more useful than they once were.
I'm now working with Andy Armstrong to push this ahead further. He's been doing a lot of work making things stable on Windows and he's started the TAPx::Harness::Compatible layer. That's designed to be almost completely backwards compatible with the current Test::Harness implementation (there are some problems with Straps.pm, Point.pm, etc.).
You can read some of the discussion on the Perl-QA mailing list to get an idea of what's involved with this. As it turns out, there's quite a lot to do, but unlike with Test::Harness, it's fairly straightforward work.
This won't be happening soon as we plan to be ultra-cautious to avoid breaking anything, but this is the direction things are moving.
Cheers,
Ovid
New address of my CGI Course.
|
|---|