Bug #50402 for Perlanet: CHI should be required, not recommended

Sun Oct 11 17:57:29 2009 stennie [...] cpan.org - Ticket created

Subject:

CHI should be required, not recommended

Perlanet 0.32 fails to load because it wants to use CHI. CHI should be "requires" not "recommends" in Build.PL. Cheers, Stephen

Mon Oct 12 03:36:35 2009 DAVECROSS [...] cpan.org - Correspondence added

On Sun Oct 11 17:57:29 2009, stennie wrote: Show quoted text

> Perlanet 0.32 fails to load because it wants to use CHI. > > CHI should be "requires" not "recommends" in Build.PL.

Oops! Yes you're absolutely right, of course. But I don't want to force people to use CHI as it's quite a heavyweight module. So I think I'll do something like this: * Don't load CHI as the module loads * When the config has been loaded, check to see if caching is required * If caching is required then attempt to load CHI * If CHI doesn't load, warn user but proceed without caching How does that sound? Look for a new release later today. Cheers, Dave...

Mon Oct 12 03:36:36 2009 The RT System itself - Status changed from 'new' to 'open'

Mon Oct 12 03:36:42 2009 DAVECROSS [...] cpan.org - Taken

Mon Oct 12 04:13:34 2009 stennie [...] cpan.org - Correspondence added

Show quoted text

> But I don't want to force people to use CHI as it's quite a heavyweight > module. So I think I'll do something like this: > > * Don't load CHI as the module loads > * When the config has been loaded, check to see if caching is required > * If caching is required then attempt to load CHI > * If CHI doesn't load, warn user but proceed without caching > > How does that sound?

Hi Dave, Sounds great, thanks. I had problems installing CHI on OS X 10.5 and had to follow the twisty, windy path of dependencies to an apparently broken version of Test::Tester (which was causing Test::Log::Dispatch to fail). Making CHI optional definitely seems more sensible :). Cheers, Stephen

Mon Oct 12 04:16:50 2009 DAVECROSS [...] cpan.org - Correspondence added

On Mon Oct 12 04:13:34 2009, stennie wrote: Show quoted text

> > But I don't want to force people to use CHI as it's quite a

> heavyweight

> > module. So I think I'll do something like this: > > > > * Don't load CHI as the module loads > > * When the config has been loaded, check to see if caching is

> required

> > * If caching is required then attempt to load CHI > > * If CHI doesn't load, warn user but proceed without caching > > > > How does that sound?

> > Hi Dave, > > Sounds great, thanks. I had problems installing CHI on OS X 10.5 and > had to follow the twisty, windy path of > dependencies to an apparently broken version of Test::Tester (which > was causing Test::Log::Dispatch to fail). > > Making CHI optional definitely seems more sensible :).

Version 0.34 on the way to CPAN now. If you can't wait, you can get it from http://code.mag-sol.com/Perlanet/. Cheers, Dave... p.s. Now I'm thinking I should do the same thing for XML::OPML::SimpleGen.

Mon Oct 12 06:28:05 2009 DAVECROSS [...] cpan.org - Status changed from 'open' to 'resolved'

Mon Oct 12 06:28:06 2009 DAVECROSS [...] cpan.org - Broken in 0.33 added

Mon Oct 12 06:28:06 2009 DAVECROSS [...] cpan.org - Fixed in 0.34 added