Stories
Slash Boxes
Comments
NOTE: use Perl; is on undef hiatus. You can read content, but you can't post it. More info will be forthcoming forthcomingly.

All the Perl that's Practical to Extract and Report

use Perl Log In

Log In

[ Create a new account ]

Adrian (66)

Adrian
  adrianh@quietstars.com
http://www.quietstars.com/

The hats I wear at work include: accessibility consultant; information architect; software developer (of the agile/XP/TDD breed - mostly in Perl); usability consultant; web site designer. I'm very dull:-) . [technorati.com]

Journal of Adrian (66)

Sunday April 18, 2004
05:28 PM

Test::Block - with added Module::Build

Yet another release of Test::Block this time. Mainly because I fouled up some of the documentation in the first release (i.e. having a vastly out of date README file).

To prevent myself making the same mistake again I switched over to the excellent Module::Build, which (if you ask it nicely) will auto-generate the README from the module POD.

(For all of you who currently dislike Module::Build I made it generate a nice Makefile.PL for you too. Since I'm not doing anything fancy it's even traditional rather than passthrough or small.)

I've fallen in love with Module::Build over the last ten or so months. It's ease of configuration and tweaking has seen it replace ExtUtils::MakeMaker for all of my "real" work, so it's about time it made it to my CPAN modules.

I also got around to adding a SIGNATURE file, just for the heck of it. My GPG fingerprint is ABDE 1C70 BFB0 1F2A 2896 3F92 22D7 1798 B194 612B for those who care about that sort of thing.

I'm really looking forward to when the plugin system for Module::Build is fully defined since I've tons of useful stuff that can be refactored out into nice generic plugins. I'm especially pleased with some subversion integration I've got in my local build system that allows me to specify module versions with properties, do automatic commits every time a test suite passes, etc. Fun stuff.

Right then... how am I doing...

Resolution-o-meter: v1.0
Week: 16, 8 weeks behind
01-02-03-04-05-06-07-08-09-10-11-12-13
14-15-16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52

Eight weeks behind - gosh it's actually going down!

Tuesday April 13, 2004
06:35 PM

Test::Block released to CPAN

Took a break from Test::Class, just for varieties sake, and finally got around to releasing Test::Block to CPAN.

Test::Block had it's genesis in a discussion on perl-qa back in May last year, and allows you to specify test plans at the level of blocks instead of the whole file, which can occasionally prove useful for a variety of reasons.

Resolution-o-meter: v1.0
Week: 16, 9 weeks behind
01-02-03-04-05-06-07-08-09-10-11-12-13
14-15-16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52

Nine weeks behind... quite a bit of catching up to do!

Sunday March 28, 2004
10:03 AM

For all the Windows folk out there

Test::Class 0.06_6 (hopefully) fixes a long standing bug in the tests that caused bogus failures on MSWin32 boxes.

I say "hopefully" because I don't have any Win boxes to hand to double check on - but it looks like a fairly simple fix. The directory separator "/" was showing up in some of the output we were testing against.

Let's see where we're at...

Resolution-o-meter: v1.0
Week: 13, 7 weeks behind
01-02-03-04-05-06-07-08-09-10-11-12-13
14 15 16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52

Oh dear, oh dear...

Friday March 05, 2004
12:36 PM

Ahhhh, it was all going so well ...

One incremental release of Test::Class every week for four weeks. Then various computers went bang in some interesting ways. The last two weeks have been spent rebuilding systems, then catching up on the work that I should have been doing while rebuilding systems. Hey ho.

Actually Test::Class 0.06_5 did hit the streets back on the 21st February - I just didn't get a chance to write about it due to said evil hardware.

Not a lot of changes.

Originally all the information about the number of methods in each class was in a big HoHoH. Since this became unwieldy The bottom layer of the hash was encapsulated in the Test::Class::MethodInfo class. This holds information on the method type and expected number of tests.

There was still some functionality that belonged in T::C::MI in T::C. Now more of it is in the right place.

As part of this clean up the add_method disappeared. Originally I was thinking of providing a mechanism to allow pre-attribute Perls to define test methods. However I've never had a need for it, and nobody has asked for it, so I've binned the idea.

Resolution-o-meter: v1.0
Week: 10, 5 weeks behind
01-02-03-04-05-06-07-08-09-10 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52
s

Hmm.... improvement needed...

Saturday February 14, 2004
07:01 PM

Default test names in Test::Class

Another week. Another Test::Class release.

A minor tweak to the new support for default test names. I also got around to documenting it :-) You can now do:

sub one_plus_one_is_two : Test { is 1+1, 2 }

and Test::Class will produce

ok 1 - one plus one is two

So as long as you name your methods well you don't have to add in a test name. Useful if you're writing in the one-test-per-method style.

That's it (apart from fixing MANIFEST so we actually include the tests for the new features from the last few releases - sigh...)

Resolution-o-meter: v1.0
Week: 7, 3 weeks behind
01-02-03-04-05-06-07 08 09 10 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52

Holding steady at three weeks behind. See y'all next time.

Saturday February 07, 2004
06:06 PM

A belated new year's resolution

A few weeks back Tony Bowden nagged me for the Nth time (sorry again Tony) to upload a new version of Test::Class to CPAN.

Various poor innocents, Tony included, have been putting up with some non-CPAN releases of Test::Class containing a mixed bag of bug fixes, experiments and new features.

I hadn't made a new CPAN release because of the long list of changes and fixes that I wanted to make to the module. With a little bit of work, so I thought, I could get everything finished! People would stand in awe of all the shiny new Test::Class goodness :-)

Unfortunately the Free Time Fairy never visited and the next release remained "ready in a couple of weeks" for over year and a half.

This is from somebody who preaches about the good things open source and agile development methods give you. Somebody who says regular incremental releases are a good thing.

Bad Adrian. No cookie.

Fortunately Tony's polite poking had some effect on this occasion. I did some tidying and Test::Class 0.06_1 hit CPAN. I knew I would quickly fall back into my old habits, so I told Tony:

My new years resolution is now to fix at least one bug/feature a week, so expect a new dev release tomorrow night.

Somehow Test::Class 0.06_2 hit CPAN the next day. Amongst other things it fixed a bug David Wheeler had reported and patched in... gosh where does the time go... January 2002. (Bad Adrian).

After one more week, through some miracle, another Test::Class release managed to escape my subversion repository. Three weeks with a new release each week! How long could this continue?

While exchanging e-mails with David on this very topic I wrote:

Maybe I should blog it to use.perl.org - that way I'd have the threat of public humiliation.

Which David thought "A very good idea. ;-)"

I think it's a jolly good idea too. However, since I don't think there's a year's worth of improvements to be made to Test::Class I'm going to modify my resolution to:

I resolve to make one incremental improvement to my particular corner of the software development world (a new module, a maintenance release, a published article or presentation, etc.) for every week in 2004. Fifty two in all. I'll blog about each one here.

Since we agile folk like public visible charts of progress I introduce you without further ado to...

Resolution-o-meter: v1.0
Week: 6, 3 weeks behind
01-02-03-04-05-06 07 08 09 10 11 12 13
14 15 16 17 18 19 20 21 22 23 24 25 26
27 28 29 30 31 32 33 34 35 36 37 38 39
40 41 42 43 44 45 46 47 48 49 50 51 52

As you can see I have some catching up to do. But considering I didn't start until 24 January - not too bad. I wonder how the rest of the year will go?

Feel free to pop back at the start of 2005 to point and laugh :-)