PL/Parrot San Francisco Perl Mongers 2010/05/25
•Télécharger en tant que KEY, PDF•
0 j'aime•727 vues
Recommandé
Contenu connexe
Tendances
Tendances (18)
En vedette
En vedette (20)
Similaire à PL/Parrot San Francisco Perl Mongers 2010/05/25
Similaire à PL/Parrot San Francisco Perl Mongers 2010/05/25 (20)
Dernier
Dernier (20)
PL/Parrot San Francisco Perl Mongers 2010/05/25
- 1. PL/Parrot San Francisco Perl Mongers May 25, 2010 Copyright © 2010 David Fetter david.fetter@pgexperts.com All Rights Reserved
- 3. Um, WTF?!?
- 5. Parrot
- 6. Parrot • A Virtual Machine for Dynamic Languages
- 7. Parrot • A Virtual Machine for Dynamic Languages • Register-based
- 8. Parrot • A Virtual Machine for Dynamic Languages • Register-based • Really, really #&^%#@%^ cool
- 9. Parrot • A Virtual Machine for Dynamic Languages • Register-based • Really, really #&^%#@%^ cool • Pluggable!
- 10. Parrot • A Virtual Machine for Dynamic Languages • Register-based • Really, really #&^%#@%^ cool • Pluggable! • Interoperable!
- 11. Parrot • A Virtual Machine for Dynamic Languages • Register-based • Really, really #&^%#@%^ cool • Pluggable! • Interoperable! • Dynamic!
- 12. PostgreSQL
- 13. PostgreSQL • If you gotta ask,
- 14. PostgreSQL • If you gotta ask, • you ain't never gonna know.
- 15. PostgreSQL • If you gotta ask, • you ain't never gonna know. • (apologies to Louis Armstrong)
- 16. Why?
- 17. Why? • Writing, Debugging and Maintaing PLs is a
- 18. Why? • Writing, Debugging and Maintaing PLs is a • Pain
- 19. Why? • Writing, Debugging and Maintaing PLs is a • Pain • In
- 20. Why? • Writing, Debugging and Maintaing PLs is a • Pain • In • The
- 21. Why? • Writing, Debugging and Maintaing PLs is a • Pain • In • The • As^Wneck
- 22. Why?
- 23. Why? • PL
- 25. Why?
- 26. Why? • Don't
- 27. Why? • Don't • Repeat
- 28. Why? • Don't • Repeat • Yourself!
- 29. Why? • Write in PL/Perl6 • Call from PL/Python
- 30. What • First things first: • PL/PIR
- 32. <EMBED>
- 33. Parrot Embed In Flux
- 34. Parrot Embed In Flux • Parrot trunk!
- 35. Parrot Embed In Flux • Parrot trunk! • PostgreSQL master!
- 36. </EMBED>
- 37. <HLL>
- 39. • HLL API not implemented yet, so:
- 40. • HLL API not implemented yet, so: • PL/PIR (mostly ok)
- 41. • HLL API not implemented yet, so: • PL/PIR (mostly ok) • PL/PIRU (rough edges)
- 42. • HLL API not implemented yet, so: • PL/PIR (mostly ok) • PL/PIRU (rough edges) • OBTW
- 43. • HLL API not implemented yet, so: • PL/PIR (mostly ok) • PL/PIRU (rough edges) • OBTW • PL/Specification
- 44. </HLL>
- 45. <MARSHAL>
- 47. PostgreSQL Data Types • We support basic data types!
- 48. PostgreSQL Data Types • We support basic data types! • Working on time types.
- 49. PostgreSQL Data Types • We support basic data types! • Working on time types. • Parrot needs bytea/time object...
- 50. </MARSHAL>
- 51. What It Looks Like CREATE FUNCTION test_float_add(float) RETURNS float LANGUAGE plparrot AS $$ .param num x x += 5 .return(x) $$;
- 52. What It Looks Like CREATE FUNCTION text_concat(text) RETURNS text LANGUAGE plparrot AS $$ $S1 = 'PL/' $S2 = 'Parrot' $S3 = $S1 . $S2 .return($S3) $$;
- 53. Access Control
- 54. Access Control • HAVE:
- 55. Access Control •HAVE: • Filesystem access: DENIED!
- 56. Access Control
- 58. Access Control • WANT • Configurable Controls
- 59. Access Control • WANT • Configurable Controls • Network Access Control
- 60. Access Control • WANT • Configurable Controls • Network Access Control • TESTS
- 61. What We Have Now
- 62. What We Have Now • An actual PL, PIRU
- 63. What We Have Now • An actual PL, PIRU • TAP tests via pgTap
- 64. What We Have Now • An actual PL, PIRU • TAP tests via pgTap • A Git Repo
- 65. What We Have Now • An actual PL, PIRU • TAP tests via pgTap • A Git Repo • An Issue Tracker
- 66. What We Have Now • An actual PL, PIRU • TAP tests via pgTap • A Git Repo • An Issue Tracker • #plparrot on Freenode
- 67. What We Have Now • An actual PL, PIRU • TAP tests via pgTap • A Git Repo • An Issue Tracker • #plparrot on Freenode • Enthusiasm!
- 68. What We'll Do Soon
- 69. What We'll Do Soon • BETTAR Argument Passing
- 70. What We'll Do Soon • BETTAR Argument Passing • MOAR Data Type Marshalling
- 71. What We'll Do Soon • BETTAR Argument Passing • MOAR Data Type Marshalling • TESTS. LOTS OF TESTS
- 72. What We Want
- 73. What We Want • Sandboxing
- 74. What We Want • Sandboxing • HLLs
- 75. What We Want • Sandboxing • HLLs • MOAR DEVELOPERZ
- 76. What We Want • Sandboxing • HLLs • MOAR DEVELOPERZ • Users!
- 77. Next: http://groups.google.com/group/plparrot http://github.com/leto/plparrot http://pl.parrot.org/
- 78. Thanks! San Francisco Perl Mongers May 25, 2010 Copyright © 2010 David Fetter david.fetter@pgexperts.com All Rights Reserved
Notes de l'éditeur
- We'd like to thank EnterpriseDB for dulling your wits this morning
- Thanks, Sarah! :)
- Normally, one doesn't think of bleeding-edge features on an RDBMS ...except that This. Is. Postgres! we'll figure out where we're dining tonight
- and Daniel Arbelo Arrocha, whose picture we were too lazy to get in time for this talk
- This is where we're going. PL Neutrality. Original Parrot Project
- Parrot Intermediate Representation (like an assembly language, but way cooler)
- Parrot trunk (yep. it has a trunk) PDD10
- Parrot trunk (yep. it has a trunk) PDD10
- Say what they are, name a few
- PL/Squaak first. PL/LOLCODE may be next (outside the Parrot tree). HAAAAAAALLLLLLLLLLLLLLP!
- PL/Squaak first. PL/LOLCODE may be next (outside the Parrot tree). HAAAAAAALLLLLLLLLLLLLLP!
- PL/Squaak first. PL/LOLCODE may be next (outside the Parrot tree). HAAAAAAALLLLLLLLLLLLLLP!
- PL/Squaak first. PL/LOLCODE may be next (outside the Parrot tree). HAAAAAAALLLLLLLLLLLLLLP!
- PL/Squaak first. PL/LOLCODE may be next (outside the Parrot tree). HAAAAAAALLLLLLLLLLLLLLP!
- Beg for help Current plan is to fall through to bytea.
- Beg for help Current plan is to fall through to bytea.
- Beg for help Current plan is to fall through to bytea.
- UNTRUSTED for now. Need more trust. PL/ParrotU will be a synonym for PL/PIRU
- Bet you didn't think assembler code could manipulate strings this way :)
- PG needs a PL access control document
- PG needs a PL access control document
- PG needs a PL access control document
- PG needs a PL access control document
- PG needs a PL access control document
- PG needs a PL access control document
- those would include performance
- those would include performance
- those would include performance
- We want opcode removal just like PL/Perl did. Have you upgraded yet? Naughty, naughty
- We want opcode removal just like PL/Perl did. Have you upgraded yet? Naughty, naughty
- We want opcode removal just like PL/Perl did. Have you upgraded yet? Naughty, naughty
- We want opcode removal just like PL/Perl did. Have you upgraded yet? Naughty, naughty