Re^5: canonical doc explaing need for "eval { ... } or do {...}" construct

Replies are listed 'Best First'.
Re^6: canonical doc explaing need for "eval { ... } or do {...}" construct by LanX (Saint) on May 13, 2022 at 06:20 UTC
I disagree since Try::Tiny is available for older versions, so importing will work "backwards" Cheers Rolf _{(addicted to the Perl Programming Language :) Wikisyntax for the Monastery}	[reply]
Re^7: canonical doc explaing need for "eval { ... } or do {...}" construct by ikegami (Patriarch) on May 13, 2022 at 17:31 UTC
I think you're saying that since the new keyword won't work on older versions, that it's not backwards compatible? That's not what backwards compatibility means. Backwards compatibility is the ability to continue to use existing code as-is. The new feature won't change that. Try::Tiny and its users will continue to work perfectly fine, so the feature is backwards compatible.	[reply]
Re^8: canonical doc explaing need for "eval { ... } or do {...}" construct by LanX (Saint) on May 13, 2022 at 19:59 UTC
It seems two problems were merging in my mind (which is currently overdosed with antibiotics) writing CPAN modules which work on older Perl versions (like - as an extreme case - those notoriously antique Perls on AIX ) compatibility of Try::Tiny code after the new try-catch feature leaves the experimental phase. The new try-catch has obvious and subtle incompatibilities to Try::Tiny, like the explicit `catch ($e)` argument. First case may need to called differently, probably "Old-Compatibility" ... (I'm eager to learn about a proper naming). The sheer number of modules already using Try::Tiny according to hippo make the second case a real "Backwards-Compatibility" issue. Because at some point a feature should become a default. Cheers Rolf _{(addicted to the Perl Programming Language :) Wikisyntax for the Monastery}	[reply] [d/l]