Hi, * Kyle Wheeler wrote: > On Sunday, April 19 at 02:26 PM, quoth Rocco Rutte: > > The problem with hcache is (currently, without the patch attached to > > ticket #2942) that it's written only once when initially parsing the > > message. Thus, if the date is Epoch upon first parse, it'll be that > > forever in hcache and in all reads afterwards. Yes, this is broken > > (though I'm running the patch since months without a problem and > > have an updateable hcache).
> Huh - I thought that patch had been integrated with mainline a long > time ago... What's been holding it up? I have no feedback from others about it whether it works. Given the current design of hcache, the patch does some dangerous things (i.e. it might crash mutt, return wrong information, etc.), so that I have no final "go" from Brendan to include it. Feel free to attach further info to the ticket :) though I have seen no trouble since using it. Rocco
