NCI 't' parameter type removed; no replacement?

Patrick R. Michaud pmichaud at pobox.com
Sat May 14 19:50:52 UTC 2011


On Sat, May 14, 2011 at 01:37:48PM -0400, Peter Lobsinger wrote:
> It is clear you want a level of stability not afforded by following
> HEAD rather than the supported releases. Perhaps Rakudo would be best
> to revise its decision to do so (and implicit associated agreement to
> increased risk of breakage as a result).

I firmly believe that having Rakudo follow the Parrot supported
releases simply means that we'll encounter these same problems in
July instead of May.  And, unfortunately, it will then take until
October to have them "fixed" such that we can use them, or require
additional supported releases in July where the broken features
get backported/backpatched into Parrot.

It has been generally agreed in *many* previous Parrot discussions that 
Rakudo is one of Parrot's key canaries in the coal mine for finding these 
sorts of issues, and that it's also in Parrot's vital interest for
Rakudo to report problems early rather than later.  If you believe 
that Rakudo should not be serving in this capacity, and if the 
rest of the Parrot leadership agrees, then Rakudo will figure out 
how to deal with that.  

Pm


More information about the parrot-dev mailing list