Tweets for 2010-03-05
- @talios the authors are doing it voluntarily and if their choice is GPL then so be it in reply to talios #
- @_navin @harshadoak @indicq – when requirements change, its time to drop some of them *not* extend sprint date. in reply to _navin #
- FWIW A flow chart rendition of a popular song http://schnouki.net/2009/11/01/hey-jude/ #
- Interesting conversation especially the comments about how HR screening can put off some of the good guys http://is.gd/9JsuL via @vdichev #
- @harshadoak @indicq Canonical agile keeps deadlines fixed. "Overall project deadlines" gives a distinct non scrum smell. /cc @_navin in reply to harshadoak #
- @debasishg @sbtourist imo the #nosql writings were not nuanced enough 2 explain applicability. good to see recent writings restoring balance in reply to debasishg #
- @debasishg @sbtourist I meant recent pro RDBMS writings in reply to debasishg #
- @codemonkeyism I remember your post was one of the earlier ones which did point out some difficulties in reply to codemonkeyism #
- @sbtourist Most focused on what NoSQL giveth, but fewer on what it taketh away. eg. How can I run CrystalReports (or equiv) on nosql ? in reply to sbtourist #
- @sbtourist I think the thresholds for CAP to be applied are often too high to be of any significance to most devs .. I didn't see that much in reply to sbtourist #
- @debasishg +1 on your horses for courses argument. I suspect we've focused on horses, imo need more focus on courses
/cc @sbtourist in reply to debasishg # - Hypothesis/Assertion: Currying/Partial Functions primarily help in temporal separation/decoupling. Is it the primary usecase? #
- @Mendelt Is that how/why currying is more frequently used than say temporal coupling ? That was the essence of my question. in reply to Mendelt #
- @codemonkeyism I suspect "use together" suggestion is much stronger today than say after 5 yrs (maturity/tooling/reliability etc) @debasishg in reply to codemonkeyism #
- @codemonkeyism speaking strictly for myself I attempt to use them together more out of fear of losing data /cc @debasishg in reply to codemonkeyism #
- On SQL/NoSQL +1 all your thoughts @codemonkeyism @sbtourist @debasishg I am perhaps circumspect of nosql not having taken one to production #
- Never been so deep into metaprogramming .. not only using metaclasses .. now manufacturing metaclasses on the fly. Not yet sure #wtf or #ftw #
- I just did (was waiting for the promo
) RT: @fogus
The Joy of Clojure 50% off today (code: dotd0305tw) http://joyofclojure.com/buy # - When you meet that rare use case asking for metaprogramming, static typing is soooo passé #flamebait
# - @KentBeck LOL! with the amount of sins I've committed in the past, I think I shall need a MetaSweatLodgeFactoryGenerator in reply to KentBeck #
- I know why Java APIs have increasingly become crappy from painful. They color the CRAPPY ones GREEN http://bit.ly/9p3MYS
via @arungupta # - @psnively deal. I'm almost done. Should be blogging soon. in reply to psnively #
- Pro Git – Smart HTTP Transport: http://bit.ly/dtzxUQ #
- Cool Javascript Cloth Simulation: http://www.andrew-hoyer.com/experiments/cloth #
- If Nerds Can Learn Linux, Why Can't They Learn Not To Interrupt People?: http://bit.ly/ahUIbp #
- Better title to last tweet could've been : If Nerds Can Learn Linux, Why Can't They Learn How To Interpret People #
- @vdichev I was just being intentionally mischievous imo there is never going to be an obvious winner in the static/dynamic debate .. ever in reply to vdichev #
- @psnively Your breadth of awareness quickly moves the debate beyond the realms of mine. Now I have to go figure out what you meant
in reply to psnively # - @psnively wrt my metaprog usecase I create metaclasses on the fly for each ResultSet for high perf, low mem but simple db access in reply to psnively #
- @psnively ..contd.. Use a different __slots__ per class. but my blog post on the topic shall wait since I now suffer from scope creep. in reply to psnively #
- @psnively Thanks .. shall follow up and read the stuff. in reply to psnively #
- @psnively my current benchmarking makes it as fast as any other mechanism, with much lower mem and using x.y instead of x[idx] column access in reply to psnively #
- @vdichev same here .. but my refusal to be defensive does not extend into my refusal to be mischievous
in reply to vdichev #