Dhananjay Nene » Tweets

Tweets for 2010-11-14

Dhananjay Nene — Sun, 14 Nov 2010 11:00:00 +0000

You’ve Got FMail. Interesting take on email <-> events <-> social signals http://ff.im/-tHCjU #
@sbtourist thats interesting. Seems quite fast too. #

Tweets for 2010-11-07

Dhananjay Nene — Sun, 07 Nov 2010 11:00:00 +0000

@tlockney knowing you can't get the best the company can offer "for free" is always disappointing (especially when one has been used to it) in reply to tlockney #
.@tlockney Architecture choices are based on some long term assumptions. When the assumptions become fluid, choices become harder in reply to tlockney #
.@tlockney my specific concern now is can I recommend JVM 4 a cloud app on a large no. of commodity servers? What costs should be assumed? in reply to tlockney #
@tlockney I think two things are uncertain (not changed). Specific features eg. advanced GC etc and costs for the advanced features. in reply to tlockney #
@tlockney this when I had narrowed the choice to clojure / scala because of the JVM and was about to recommend clojure bcos of dynamism in reply to tlockney #
@tlockney there are degrees of safety, and more uncertainty => more concern. Though for enterprise apps – there is still no alternative in reply to tlockney #
@tlockney One can safely assert they will change for the better technology perspective and worse from a costs perspective in reply to tlockney #

Tweets for 2010-11-06

Dhananjay Nene — Sat, 06 Nov 2010 12:00:00 +0000

SUCKS ! What I was afraid of. RT @newsycombinator Oracle cooks up free and premium JVMs http://j.mp/b4eshw #
With free and paid versions does the JVM continue to be an attractive binary target for other langs? #
@gr0kk3d to the contrary this is what I did fear in reply to gr0kk3d #
@dexterous that discussion topic doesn't make a lot of sense. OpenJDK is also oracle controlled. in reply to dexterous #
@diptanu #oracle measures profit and loss based on umm.. profit and loss. So given zero revenues today it is NOT A LOSS for oracle. in reply to diptanu #
@dexterous doesn't matter. Enterprises need a patent and IP issues free JVM – none such in the open source world exists. in reply to dexterous #
@travisennis imo it is a "lot less attractive" in reply to travisennis #
Wonder what will go in the paid version of the JVM? Advanced GC, hot deploy, non-stop kernel …? How about hotspot (imo thats less likely)? #
@dexterous jokes apart, ensuring patent non infringement would be a nightmare. in reply to dexterous #
@dexterous hotdeploy is auto bytecode reloading, hotspot is auto JIT'ing in reply to dexterous #
@dexterous extraordinarily hard, if you look at the kinds of patents oracle sued google for. in reply to dexterous #
@dexterous nope .. I meant exactly what I said. in reply to dexterous #
@dexterous hotspot is simply a technology which auto JITs frequently run code – it is not a compiler or a JVM in reply to dexterous #
.@dexterous afaik, even if wikipedia states it that way – hotspot is a technology not a JVM. in reply to dexterous #
@dexterous http://www.javaworld.com/jw-03-1998/jw-03-hotspot.html I don't know if some relabeling of the JVM happened since in reply to dexterous #
@travisennis we'll have to wait and watch Hope an upside doesn't get created by negatively influencing the free version. in reply to travisennis #
@dexterous dude you are getting into unimportant semantics – hotspot is generally understood to be a technology. I've conveyed all I could. in reply to dexterous #
@offbytwo I would cross my fingers, hold my breath and pray your expectation is met with in reply to offbytwo #

Tweets for 2010-10-22

Dhananjay Nene — Fri, 22 Oct 2010 12:00:00 +0000

Interesting counterargument to functional programming http://prezi.com/n9pfrk3vimh5/the-real-harm-of-functional-programming/ #
My talk on FP yesterday to a largely scientific / embedded / firmware kind-of audience (read C/C++) was rather nicely received – need2blog #
Interesting point from a C++ compiler writer: they've been waiting for someone to write a large app using STM, but that just hasn't happened #
Gosling's take on Apple discontinuation of java http://nighthacks.com/roller/jag/entry/steve_jobs_comments_on_apple Interesting last para. #
@psnively How about: I can count the number of enterprises that care about verified JVM on MAC OS on one hand ? @olabini in reply to psnively #
.@aloiscochard I must confess my attraction to LLVM has increased substantially ever since Oracle continued to be #disharmonic in reply to aloiscochard #
@IamSB I think so. in reply to IamSB #
@deanwampler why is it not a good way to learn idiomatic clojure? in reply to deanwampler #
@NeilRobbins hehe.. if it was overengineered you would be looking at a Ferrari and a bill for a Ferrari, the latter you didn't budget for in reply to NeilRobbins #
@debasishg @djspiewak Isn't the difference between a DSL & non-DSL existence of domain vocabulary (or overloaded operators if applicable)? in reply to debasishg #
@debasishg Sorry .. my mistake, was responding to http://twitter.com/#!/djspiewak/status/28410855157 after misinterpreting it @djspiewak in reply to debasishg #
@_sunil_ @shopsocially congrats! in reply to _sunil_ #
@debasishg @jamesiry Seems Phantom Types, could provide a vehicle to encode any state machine into Type system (not just Builder) in reply to debasishg #
@djspiewak didn't understand. An object of a non-phantom type would create 2nd object allowing original to be gc'ed. Could be nonrecursive in reply to djspiewak #
@djspiewak Still digesting. Will revert if I have any questions post understanding it in reply to djspiewak #

Tweets for 2010-10-12

Dhananjay Nene — Tue, 12 Oct 2010 12:00:00 +0000

Sometimes a blog post makes me want to respond back v strongly. Does Java Make You Less Productive Than Dynamic Langs? http://is.gd/fYI0D #
The heuristic I am comfortable for LOC between Python : Java is 1:4 and for time spent in writing properly working software is 1:3 #
@rickasaurus I am 1 of those dynamic guys On a serious note, I dont find maintaining "well written" dynamic code hard (but 4 refactoring) in reply to rickasaurus #
@rickasaurus Maintaining well written expressive but concise code is always easier than maintaining well written expressive & verbose code. in reply to rickasaurus #
@rickasaurus Agreed. Which is why refactoring python takes more time than refactoring java. Not "necessarily" true for "maintaining python" in reply to rickasaurus #
@rickasaurus Again agree. Type inferencing improves expressivity & brevity. But there's still a class of problms better expressed by dyn typ in reply to rickasaurus #

Tweets for 2010-09-28

Dhananjay Nene — Tue, 28 Sep 2010 12:00:00 +0000

@jneira the link to the actual interview on the page is returning a 404 /cc @debasishg in reply to jneira #
@radhamukkai well put. in reply to radhamukkai #
Created twitter list http://twitter.com/dnene/indian-pythonistas/members on meeting some wonderful folks@ #inpycon2010 Pls suggest additions #
@_lprasad done. Thanks in reply to _lprasad #

Tweets for 2010-09-14

Dhananjay Nene — Tue, 14 Sep 2010 12:00:00 +0000

I think I am going to ditch Powerpoint/OO Presentation. Beamer it shall be with a healthy dose of pygments. #
@aloiscochard shouldn't it rely on the built in equals() as with any other object ? in reply to aloiscochard #
@aloiscochard perhaps to work around some gotchas like http://xahlee.org/java-a-day/hashmap_equality.html ? in reply to aloiscochard #
RT @jedisct1: RT @justinvincent: Dear Twitter: Stop screwing over your developers http://bit.ly/c4FZaI #
@aloiscochard strange indeed. Not including the method would result in the same behaviour. Perhaps @cbeust may have a better insight. in reply to aloiscochard #
@aloiscochard Though the assertEquals with Object inputs does take care of the array comparison. in reply to aloiscochard #
@cbeust @aloiscochard We were wondering why static public void assertEquals(Map actual, Map expected) in Assert exists (1/2) in reply to cbeust #
@cbeust @aloiscochard Not having it would allow it to default to the "Object" parameter equivalent method which would work as well (2/2) in reply to cbeust #
@jstrachan @hseeberger even redmine is http://www.redmine.org/ quite nice in reply to jstrachan #
@aloiscochard but if the "expected" map is null, then doesn't the assertEquals(Object) actually work better ? /cc @cbeust in reply to aloiscochard #
@cbeust Am unclear. Per http://bit.ly/aDK03d apart from null params, only other diff is expected.equals(actual) vs. actual.equals(expected) in reply to cbeust #
@davetron5000 thanks. looks great, but the seeming lack of an export to PDF is a bit of a bummer. in reply to davetron5000 #
@vdichev Am trying out docutils + s5. If it works similarly may switch to rst2pdf and rst2s5 due to much cleaner syntax of Restructured Text in reply to vdichev #
@davetron5000 in my case the valid submission formats for a conference are OO/PPT/PDF Plus sharing on slideshare etc is easier for PDFs in reply to davetron5000 #
@davetron5000 Totally agreed. its superfail in reply to davetron5000 #
@rickasaurus The abstraction is deeds in the name of ThemVsUs. Have seen it apply far far beyond nationality or religion. in reply to rickasaurus #
Would generators still be required / be useful if TCO was in place / available? #
Is there a separate tool to generate such beautiful & rocking slides ? Sample presentation : http://slides.html5rocks.com/#slide1 #

Tweets for 2010-08-31

Dhananjay Nene — Tue, 31 Aug 2010 12:00:00 +0000

Fantastic! RT @KentBeck RT @agilecoach: One of the greatest agile coaches (@dhemery) gives away some of his best stuff: http://bit.ly/9ERl5E #
Getting me a Design Concepts in Programming Languages http://is.gd/eMIBM ( Indian shoppers) http://amzn.to/9SrW8a (thx @debasishg @psnively) #
Turns out, will be speaking at two conferences at Bengaluru over the next couple of months. #
@ptrthomas One is pycon India. The other isn't announced yet. Thanks @codemonkeyism. in reply to ptrthomas #
@byjupv Cool, something for me to remember the next time. in reply to byjupv #
Changing relationship cardinalities can be so gut wrenching #
@jneira yep, but its crazy how many touch points one has to modify, when the model changes after it has been implemented in reply to jneira #
If I have one intel on-board graphics card + one more nvidia display adapter with two display ports .. can I have 3 monitors ??? #ubuntu #
@jteigen you can't have your cake and eat it too /cc @codemonkeyism *grin* in reply to jteigen #
Blog post comment spammers!! May insects crawl up the in the wrong direction of your digestive system. #
@offbytwo Ok. Great to know its at least possible. That means I can at least attempt to take a shot some time soon. in reply to offbytwo #
@jneira I prefer YAGNI except in scenarios where I believe a simpler relationship is only due to simplified/immediately useful requirements in reply to jneira #
@jneira I very often end up considering n-n or tree structures, but don't implement them unless I'm convinced they are inevitable in reply to jneira #
@jneira as happened today, every once in a while the call ends up the wrong way. But I still prefer it that way. I budget a bit of scalding in reply to jneira #
LOL! RT @arungupta @dnene So true, both in "real-life" and 4 database tables re: Changing relationship cardinalities can be so gut wrenching #
RT @sallamar: Cookie-OAuth convergence – http://bit.ly/9BCoN0 #

Tweets for 2010-08-30

Dhananjay Nene — Mon, 30 Aug 2010 12:00:00 +0000

Have concluded all my thoughts around monads and objects. Blog post in a day or 2. Some interesting thoughts /cc @Missing_Faktor @kaleidic #
Stating my conclusion b4 the analysis, "jQuery is a monad" leads to a "large subset of objects also are monads" cc @Missing_Faktor @kaleidic #
RT @etorreborre RT @fsteeg: New post: Scala, object persistence, and the original NoSQL: XML http://wp.me/pwKJk-kb #existdb #eclipse4 #java #
@Missing_Faktor Best wishes for a quick return to wellness. in reply to Missing_Faktor #
Had a phone call which reminded me of my 2 yr old blog post. Multicore for Project Managers and Junior Developers http://is.gd/eL7Iz #
@jneira it was actually a little bit more nuanced as also reflected in http://twitter.com/dnene/statuses/21911508350 in reply to jneira #
Anansi — Node.js Knockout http://ff.im/-pVI7O #
@jneira about as good as monkeys are for patching http://www.scrapbookersparadise.com/pictures/stores/09-Reminisce-Monkey-Card.jpg in reply to jneira #
Reflections on Node.js Knockout Competition http://ff.im/-pW24v #

Tweets for 2010-08-29

Dhananjay Nene — Sun, 29 Aug 2010 12:00:00 +0000

Really unfortunate! RT @newsycombinator: IPhone App in Approval Limbo for 3 Mths, Dev Decides to Open Source It http://j.mp/c2m51z #
Some useful thoughts. Projects are evil and must be destroyed http://ff.im/-pSjcI #
An Example of Multithreaded Processing using Python's "multiprocessing" Module http://ff.im/-pSjuo #
Will we see more of these now? Paul Allen Sues Apple, Google, Others Over Patents http://ff.im/-pSjOY #
Random thoughts on Clojure Protocols http://ff.im/-pSk6M #
RT @ptrthomas: #agile RT @wakaleo: RT @evanbottcher: Holy crap. Brilliant. http://www.halfarsedagilemanifesto.org/ (via @jamesladd) #
Ladies. Look at your code. (on lambdas and JDK 7) http://ff.im/-pSn6o #
RT @jneira: skilldrick.co.uk » Static typing: the presumption of guilt http://bit.ly/a8A60K #
Evan Bottcher: DevOps and the Iteration Showcase http://ff.im/-pSYhN #
I like this notion of thinking of OODA loop http://is.gd/eJMsZ time. You perhaps cycle fastest in your favourite lang? (via @debasishg) #
@psnively I thought parts of it were a little strange. eg. "more reliable is a myth". His comment on metaprogramming was right on in reply to psnively #
@psnively I believe static typing more reliable than dynamic, but there is a flexibility tradeoff. Different folks weigh them differently. in reply to psnively #
@psnively That one is beyond my knowledge sphere. But it would be interesting to find out how metaprogramming works in that context. in reply to psnively #
@psnively I assuming you were saying that in half jest, 'cos I very firmly believe thats not true. in reply to psnively #
@psnively If I could buy "fooling myself" insurance / cover. I would always buy it regardless of my position /cc @jteigen @kaleidic in reply to psnively #
@jteigen unit testing is a strange topic that always enters the typing debates – something imo isn't a central aspect to the choice. in reply to jteigen #
@jteigen I suspect most dynamic type proponents don't actually write all the unit tests it would take to provide a iron clad assurance in reply to jteigen #
@psnively Not quite. I buy life cover since the stakes are high, but rarely if ever extended warranties based on the economic tradeoffs. in reply to psnively #
@psnively A example is people externalising metadata (to XML?) when using static langs and then feeling happy about the productivity gains in reply to psnively #
@psnively I'm going to have to read up on that one. Will revert later. in reply to psnively #
@psnively Much Ruby s/w out there where people not looking 4 that insurance. I shall continue to sit on the fence n say no 1 right way in reply to psnively #
@psnively I am not convinced of all his statements eg. Reliability, Refactoring. Some say thats 'cos I haven't seen power of smalltalk tools in reply to psnively #
@psnively As a java dev I used to feel utterly jealous of python/ruby 4 their natural expression. I guess that changed after Scala @kaleidic in reply to psnively #
@kaleidic Whoa! in my experience the darling of financial sector is Java. No one. Repeat – no one comes even remotely close /cc @psnively in reply to kaleidic #
@psnively On the face of it Camlp4 is a preprocessor. I would imagine one could implement something similar 2 implement metalinguistic (1/2) in reply to psnively #
@psnively abstractions for java. But thats going to be unnatural and the preprocessed language probably is dynamically typed (2/2) in reply to psnively #
@psnively I would still imagine python/ruby/javascript/clojure provide much more natural and integrated metaprogramming capabilities in reply to psnively #
@psnively Hypothesis: Most static langs don't treat metalinguistic abstraction because they can't implement a statically typed preprocessor. in reply to psnively #
@psnively Sorry I didn't see your camlp4/metaocaml tweet before my last tweet. Am back digging into details in reply to psnively #
@psnively imo dyn typed langs allow me to "program ahead" to a type I only partially know, or doesn't exist & will get created @ runtime in reply to psnively #
@psnively If I understood metalinguistic abstractions properly, homoiconicity would be a "higher goal", but not mandatory. in reply to psnively #
@psnively my understanding of dyn languages is they allow "internal DSLs" whereas (perhaps) camlp4 allows construction of "external DSLs" ?? in reply to psnively #
@psnively afaik, type inferencing will allow binding only if to a known interface, will not allow or bind to types on the fly at runtime in reply to psnively #
@psnively you've educated me on an interesting possibility thats going to take some time to follow up on. in reply to psnively #
@vdichev count me in as one of those in reply to vdichev #
@jteigen ahh.. understood. in reply to jteigen #
@psnively Yes and in nonhomoiconic python / ruby one can invoke methods on types which are metalinguistic abstractions(?) created @ runtime in reply to psnively #
@fogus sorry, I used it just a tad loosely. But allow me to explain. In python/ruby, one can use metaclasses and build a dsl around it (1/2) in reply to fogus #
@fogus to express new type construction, and usage of such types (internal). Play framework does it thru bytecode @ startup (external) in reply to fogus #
@fogus Other examples of external would be preprocessor macros (not the lisp kind). Scala DSLs afaik will not allow new type construction in reply to fogus #
@debasishg Yes, Lisp, PlayFramework (bytecode on fly), camlp4 are different than cpp macro model, which defy clear classification @psnively in reply to debasishg #
@debasishg I take back my statement on Lisp .. thats actually clearly understood as a read cycle macro at runtime not compiletime @psnively in reply to debasishg #
@jneira what does uau and jum stand for ? in reply to jneira #
@jneira thanks .. you had me hunting all over the internet acronym finders in reply to jneira #

Tweets for 2010-08-27

Dhananjay Nene — Fri, 27 Aug 2010 12:00:00 +0000

Hmm… Duck typing allows imprecise communication. Python Sequence VS Iterable protocols http://ff.im/-pMMXP #
Ksplice » Six things I wish Mom told me (about ssh) http://ff.im/-pMNuM #
Scala is for VB programmers. @ErikEngbrecht on the risk of dumbing down leading to black boxing http://ff.im/-pMNXV #
RT @newsycombinator: Self-Improving Bayesian Sentiment Analysis for Twitter http://j.mp/99sTIM #
@irohan great. Looking forward to it. in reply to irohan #
@rsdighe is this your tech account? in reply to rsdighe #
Sparse Matrices: infer, Clojure and the jvm http://ff.im/-pNhKU #
@answerguy I would imagine it wouldn't be worth holding your breath. Many s/w books don't make it to Indian editions. Better buy the e-book. in reply to answerguy #
@mamund @wmartinez I realised @tweetbffs had surreptitiously made me follow it. Probably did the same to you too ? #
Lego analogy is getting overused. The key to lego is not what you can build with it. It is that a 4 yr child can start putting it together. #
@tweetbffs Did I judge you unfairly or was the option recently added ? in reply to tweetbffs #
@jteigen Perhaps. I've generally ended up understanding it as one of (depending on occasion) composability, modularity, plug-n-play etc. in reply to jteigen #
So totally agree! RT @etorreborre RT @raganwald: The Monad terminology problem http://bit.ly/cL4nnx #
ROFLMAO!! RT @sbtourist: Brilliant: "MongoDB is Web Scale", http://nosql.mypopescu.com/post/1016320617/mongodb-is-web-scale #
@Areai51 its good for a particular class of usage. Detailing many of the tradeoffs is beyond the ability of 140 chars. in reply to Areai51 #
16 digit int as a unique id? Now I need to create a table to create/persist the id. API Writers : Please use UUID when wanting uniqueness #
RT @debasishg: Scalar – Scala-based DSL for cloud computing running on top of GridGain runtime .. http://bit.ly/dc8I3h #
RT @newsycombinator: Coding Horror: Vampires (Programmers) versus Werewolves (Sysadmins) http://j.mp/94e2k6 #
File uploads and MVC Controllers http://ff.im/-pONbd #
OpenStack – The Answer to: How do We Compete with Amazon? http://ff.im/-pONKN #
The History of Python: Why Python's Integer Division Floors http://ff.im/-pONUU #
Multi-core, Threads & Message Passing – igvita.com http://ff.im/-pOPkp #
@inphina Thanks, thats encouraging. in reply to inphina #

Tweets for 2010-08-26

Dhananjay Nene — Thu, 26 Aug 2010 12:00:00 +0000

@kaleidic That leaves me impressed, no, much much more than that. in reply to kaleidic #
A product is a collection of use cases and their utility. Intrinsic properties support the use cases, don't define the product. #
RT @ijuma: Is Scala Simple or Complicated? @odersky gives his thoughts on the subject with illustrations and everything http://bit.ly/aUStli #
Now installing bluetile tiling manager on ubuntu to see if it makes me more productive. Would you recommend something else for ubuntu ? #
@brijwhiz http://en.wikipedia.org/wiki/Tiling_window_manager in reply to brijwhiz #
@vdichev Its always the implementation which leaks, and thus increases complexity beyond the abstraction. #being_pedantic in reply to vdichev #
@vdichev iPhone is a good example which demonstrated that the abstraction was always simple in reply to vdichev #
Git – SVN Crash course (for subversion users) http://ff.im/-pKxXZ #
Spring Python 1.1 book is published http://ff.im/-pKz24 #
All Testing is (not) Confirmatory http://ff.im/-pKA0F #
Introduction to the Reactive Extensions for JavaScript – Buffering http://ff.im/-pKC3t #
You know you're a Python programmer when… #humour http://ff.im/-pKCiK #
Intro to Monads via reddit http://ff.im/-pKLtA #
My first experience with tiling window managers is actually quite nice. Seems like I will stick to it. #bluetile on #ubuntu #
Google realtime search http://www.google.com/realtime is awesome. http://bit.ly/bypVnX (via @shrikant) #
Interesting to take a recent hot topic and step back in time to see its evolution on google realtime search http://www.google.com/realtime #
@offbytwo I read bluetile is based on xmonad http://is.gd/eFgub in reply to offbytwo #
@debasishg I think so. I've found either separate implementations or splitting the abstraction quite helpful in situations of leakiness. in reply to debasishg #
This "I just want to copy .. word for word .." meme is spreading virally. Google realtime search tells me it started about 7 hrs ago. #

Tweets for 2010-08-25

Dhananjay Nene — Wed, 25 Aug 2010 12:00:00 +0000

Reduction of complexity is not the same thing as boilerplate avoidance. #
@codemonkeyism Correct. Brevity reduces one dimension of accidental complexity. That still may influence total complexity either way. (1/2) in reply to codemonkeyism #
@codemonkeyism I've been alleged to make my code too complex in situations when I've attempted to make it too brief (2/2) in reply to codemonkeyism #
@codemonkeyism Thats a great post. Wish I had seen it earlier. in reply to codemonkeyism #
A reason to use OO+Functional than only OO languages. Easier to write pluggable functions than functors for handling most corner cases. #
@mrkkrj I find so. eg. just a function reference instead of "new" of a functor, boilerplate of the functor etc. in reply to mrkkrj #
@mrkkrj If your question was is it worth changing the language used ? The answer would be no. in reply to mrkkrj #
@mrkkrj Hmm.. I was suggesting it as a reason to use one language over another, not change one already in use. in reply to mrkkrj #
RT @markhneedham: RT @MarkDalgarno: I Lehman's 'laws' of software evolution: http://is.gd/eClVJ http://is.gd/eCm2m #
@vdichev Unfamiliarity introduces learning complexity needing to be amortised over reduced coding/operational complexity in reply to vdichev #
@vdichev Yes, but sometimes library/tool authors handle far bigger feature set than users leverage which can increase complexity. eg. Maven in reply to vdichev #
@vdichev the hard part is never the actual amortisation, its selling/buying the initial fixed cost in reply to vdichev #
@jneira Python Answer (period = 4, data = list_of_vals) is
tuple((sum(data[i:i+period])/period for i in range(len(data)-period+1))) in reply to jneira #
@jneira so I'm a little confused what the fuss is all about. in reply to jneira #
@jneira The haskell solution seems to be the neatest (only if I could grok it better ) in reply to jneira #
@jneira couldn't resist playing some code golf myself http://is.gd/eDCNY in reply to jneira #

Tweets for 2010-08-24

Dhananjay Nene — Tue, 24 Aug 2010 12:00:00 +0000

@kaleidic I myself am professionally schizophrenic. The straining to be heard alphageek engineer agrees with you and bruce in reply to kaleidic #
@kaleidic But I believe a large ecosystem eg in the enterprise s/w markets in US, especially in India is not better served by these in reply to kaleidic #
@kaleidic Apart from collaboration, its also a function of the ecosystem/mileu. My ideal team size is 1, current is 3 & in the past was 200+ in reply to kaleidic #
@kaleidic I tend to view situations both from the perspective of engineering and economics. And i believe economics is not given its due. in reply to kaleidic #
RT @codemonkeyism: Automatic testing is the new build #
RT @KentBeck: OH: "if you don't crave feedback, iteration is just annoying" #
RT @rickasaurus: Seems to depend on frequency of use. “@JohnDCook: Software that tries too hard to help can harm http://bit.ly/cjK80s” #

Tweets for 2010-08-24

Dhananjay Nene — Tue, 24 Aug 2010 12:00:00 +0000

@kaleidic I myself am professionally schizophrenic. The straining to be heard alphageek engineer agrees with you and bruce in reply to kaleidic #
@kaleidic But I believe a large ecosystem eg in the enterprise s/w markets in US, especially in India is not better served by these in reply to kaleidic #
@kaleidic Apart from collaboration, its also a function of the ecosystem/mileu. My ideal team size is 1, current is 3 & in the past was 200+ in reply to kaleidic #
@kaleidic I tend to view situations both from the perspective of engineering and economics. And i believe economics is not given its due. in reply to kaleidic #
RT @codemonkeyism: Automatic testing is the new build #
RT @KentBeck: OH: "if you don't crave feedback, iteration is just annoying" #
RT @rickasaurus: Seems to depend on frequency of use. “@JohnDCook: Software that tries too hard to help can harm http://bit.ly/cjK80s” #

Tweets for 2010-08-24

Dhananjay Nene — Tue, 24 Aug 2010 12:00:00 +0000

@kaleidic I myself am professionally schizophrenic. The straining to be heard alphageek engineer agrees with you and bruce in reply to kaleidic #
@kaleidic But I believe a large ecosystem eg in the enterprise s/w markets in US, especially in India is not better served by these in reply to kaleidic #
@kaleidic Apart from collaboration, its also a function of the ecosystem/mileu. My ideal team size is 1, current is 3 & in the past was 200+ in reply to kaleidic #
@kaleidic I tend to view situations both from the perspective of engineering and economics. And i believe economics is not given its due. in reply to kaleidic #
RT @codemonkeyism: Automatic testing is the new build #
RT @KentBeck: OH: "if you don't crave feedback, iteration is just annoying" #
RT @rickasaurus: Seems to depend on frequency of use. “@JohnDCook: Software that tries too hard to help can harm http://bit.ly/cjK80s” #

Tweets for 2010-08-23

Dhananjay Nene — Mon, 23 Aug 2010 12:00:00 +0000

"When I wrote this, only God and I understood what I was doing. Now, God only knows." A code comment. http://goo.gl/EXD9 via @Missing_Faktor #
RT @andreisavu: 10 Common Mistakes Made by API Providers http://bit.ly/9oWkoh #
From a blog post: The most platform independent technology you can use is either Flash or Silverlight. <- me : Silverlight ??? #
@IamSB I was just quoting someone else in reply to IamSB #
New codeblog post: Clojure style multi methods in python http://is.gd/eylkx #
@codemonkeyism was that in the context of my codeblog post ? in reply to codemonkeyism #
@codemonkeyism If y, its trivially easy in python. Lambdas can represent conditions (probably half an hr more work) in reply to codemonkeyism #
Updated codeblog post: Clojure style multi methods in python http://is.gd/eylkx Now no switch func, but conditions per function. #
@codemonkeyism Ahh, Wasn't aware of the non clojure posts. But your tweet inspired me to extend my post to include a non switched version in reply to codemonkeyism #
Very begrudgingly with a lot of study coming around to the thought that jQuery could be termed a monad @Missing_Faktor @vdichev @kaleidic #
@kaleidic Certainly. If for no other reason than that it was a little too intense journey for a OO guy. in reply to kaleidic #
@fogus I just saw ur comment n approved it. Had received a similar comment on a mailing list earlier. (though it took about 5-7 mins 2 impl) in reply to fogus #
@fogus Yes. Let me take a look at it. While prima facie feasible to impl it, I suspect its very unidiomatic for python which has no symbols. in reply to fogus #
@fogus ie. allowing "data" to derive from each other seems extremely uncharacteristic for python. in reply to fogus #
@fogus Completely agreed. Most langs cannot compete with the minimalistic elegance of lisp PS: Responded to your comment with sample code in reply to fogus #

Tweets for 2010-08-22

Dhananjay Nene — Sun, 22 Aug 2010 12:00:00 +0000

RT @softmodeling: RT @rodrigosalinas: If you think that #LinesOfCode has to do with project duration, you must also believe in Santa #
Cool! but, wheres da code? Clojure in Python. http://ff.im/-pxnRH #
GNOME’s SeedKit aims to bring web tech to desktop apps (I wish mobile apps too) http://ff.im/-pxrtf #
Gaming foursquare with 9 lines of perl http://ff.im/-pxrI7 #
Some lesser-known truths about programming (Nice, but a little repetitive) http://ff.im/-pxrLM #
What ever happened to common courtesy (or attribution) ? http://ff.im/-pxsez #
RT @umairh: awesome post by @doriantaylor on the fail that is management as we know it. http://bit.ly/bZNP6T #

Tweets for 2010-08-21

Dhananjay Nene — Sat, 21 Aug 2010 12:00:00 +0000

.@Srini_T Thanks. Thats v. encouraging. re: @dnene Your explanation on dicts in bangpypers is awesome. in reply to Srini_T #
RT @dalmaer: Opensource and Javascript: Polymaps Used To Make PrettyMaps http://j.mp/d5JF3O #
Now live tweeting Indicthreads cloud computing conference from @d7yliven #IndicU Other livetweets from @punetechlive #
Correction. Am livetweeting from @d7ylive #IndicU Other livetweets from @punetechlive #

Tweets for 2010-08-20

Dhananjay Nene — Fri, 20 Aug 2010 12:00:00 +0000

RT @bobleah: Are you a student in India? Then check out The Great Mind Challenge 2010 Contest http://bit.ly/8ZW8uT #tgmc2010 /cc @punetech #
@arungupta Hey, I'm delayed due to some urgent work. Hope to see you in the afternoon. Very likely may miss your session in reply to arungupta #
@djspiewak heading towards a StackOverflowError in reply to djspiewak #
Am delayed from reaching the IndicThreads cloud computing conference. Thankfully @punetechlive is live tweeting it. #
Shall be talking about "Functional Programming with Python" http://is.gd/eqal6 at Pycon India 2010. #
@_sunil_ Thanks. Here's the topic http://is.gd/eqal6 in reply to _sunil_ #
@punetechlive whats the liveblog URL ? in reply to punetechlive #
@Missing_Faktor Thats only during object construction. In a monadic implementation returning a selector string would also work @kaleidic in reply to Missing_Faktor #
@Missing_Faktor The object wrapping in monads happens on the values returned by the return statement. in reply to Missing_Faktor #
@Missing_Faktor Just posted my response on StackOverflow. in reply to Missing_Faktor #
My opinion posted on StackOverflow. It may walk and talk like a Monad, but a Fluent Interface is not a monad. http://is.gd/eqiX5 #
@Missing_Faktor FluentInterface is direct talking . Monad is a translator. Can't say Talking is Translation. In Fl Iface, no role 4 a monad in reply to Missing_Faktor #
@Missing_Faktor maybe you could try entering a comment its not possible. in reply to Missing_Faktor #
@Missing_Faktor Will collate and compose all my thoughts and put a blog post out on Monday. in reply to Missing_Faktor #
Hmm McAfee was acquired for a bigger amount than Sun. Moral: Writing anti-bad software worth more than writing good software. #
@vdichev I thought a monad had the ability to type wrap after every return. In case of jQuery its only on construction. in reply to vdichev #
@Missing_Faktor Looked at it and understood. Though will think about it tomorrow .. want to sleep over the topic. in reply to Missing_Faktor #
@debasishg I guess the functional programming way should be currying (favours) in reply to debasishg #