Version 17 of Tcl bounty programme

Updated 2016-12-01 23:13:59 by AMG

FlightAware are sponsoring a programme of bounties for improving Tcl and (selected) extensions.

Full details here: https://github.com/flightaware/Tcl-bounties

Note that these bounties are not sufficiently generous to be considered pay (given how much a software engineer should be able to earn). They're more awards for things someone might be inclined to do anyway.

---

Projects

In progress

• A reasonable C API for enumerating an array
  • Discussion: Array enumeration
  • Code: [L1 ]
• Speed up clock format and clock scan (2X, 4X, 10X)
  • Discussion: speed up clock format and clock scan
• array foreach
  • Code: [L2 ]
• Support for SO_REUSEPORT on server sockets
  • TIP 456: [L3 ]
  • Code: [L4 ]
• Tcl package introspection improvements
  • Code: [L5 ]
• tclreadline improvements
  • Discussion: [L6 ]
  • Code: [L7 ]
• Support for epoll()/kqueue() to replace select() in socket handling
  • TIP 458: [L8 ]
  • Code: [L9 ]
• A first class, high-performance, non-hackish way to do named parameters
  • TIP 457: [L10 ]
  • Alternative: Proposal for proc argument handling enhancments
• Call-by-name syntactic sugar for proc definitions that would obviate most uses of upvar
  • Code: [L11 ]
  • Original patch: [L12 ]

Completed

(coming soon)

Not yet claimed

• Upgrade the scotty extension (TEA)
• Upgrade the scotty extension (UDP)
• Clean up of tcltls
• Revive the Tcl Pro debugger
• Make TclX's signal trap handlers safe to use with threaded Tcl
• Make TclX's profiler work properly with Tcl 8.6
• Stop Tcl from eating child process exit status gratuitously
• after -at
• array default arrayName value
  • Interested expressed by AMG, no code yet
• Tcl runtime performance improvements (2X)
• Tcl runtime performance improvements (10X)
  • DKF: It is not possible to get a 10x speedup without going to native code somewhere.
  • See tclquadcode for what I think is the best bet right now for achieving the 10x goal.

---