Philip Monk

Opened #5903 against this branch to resolve the slowness in prettyprinting. I will say that subjectively it feels like it takes significantly longer to compile desks with these changes.

What do you think about merging this just as a generator plus the psq jets? That way we can easily use either compiler, and it's guaranteed to stay working because...

We're not the sender, though. The sender is some unrelated ship. On Tue, Apr 7, 2020, 14:39 Ted Blackman wrote: > To clarify, the fix I described would happen in...

Yes, under some circumstances. For now, I recommend not using the same port for different ships.

You should never see those printfs from anything other than a galaxy

I believe this is ready to go. I made the following changes: - Adapted the wip from jon/doccords-vere to a working version here: urbit/vere#114 - Virtualize calls to dprint in...

This works on my machine as well. We still need to glob and update the kelvins.

Yes, thank you for the work here, but this specific issue should be resolved now by the Kelvin 417 release. Also, chalk this bug up as another reason to prioritize...

Some assorted measurements on ~nibset: In the 560MB snapshot, substantially all of the pages are actually different than what's on disk (around 1% of pages happen to be identical). When...

In general yes, it's a really good cleanup. It is not complete, and it's not clear when it will become a priority.