It's certainly something I would like to add to counterclockwise. If only paredit's code was written in clojure, it could have been more easily reused by enclojure, La Clojure and counterclockwise ! :-(
2009/12/19 Avital Oliver <avi...@thewe.net> > It seems that noone has brought up the amazing paredit mode for emacs which > gives you keyboard commands for s-expression structure manipulation. It also > makes sure you never get your close parenthesis wrong. I use it and can't > imagine writing any medium+-complexity code without it. > > On Dec 19, 2009 6:11 AM, "ajay gopalakrishnan" <ajgop...@gmail.com> wrote: > > Hi, > > I've a different take on the entire thing. Based on my experience I have > realized the following: > > 1. A lot of software engineers do not have a CS background. Out of > them, many joined this field for the fun of coding and are willing to learn > new things. I know many like that. However, I feel a large majority of them > are those for whom Java is probably what Software development basically is. > (This has been a boon and bane of Computer Science. It reduces the barrier > to entry, but also prevents them from appreciating the true beauty. Many > can > code a web application, thanks to Rails, Hibernate and the like without > even > having an idea of what B-tree is) Everything begins and ends with Java. > They > don't even know anything about C/C++. > And I think it would be foolish to assume that Clojure acceptance will > reach critical mass unless it is accepted by this group of people. > 2. Second issue is Bosses (higher-ups) and convincing them is hard. > Language features and the beauty of Lisp is not going to convince them. The > biggest problem is I think when they realize that they will have to loose > all the Imperative programming skills they have acquired all over these > years and now they will be on equal footing with the rest when it comes to > Functional programming experience. > 3. Sad as it is, most of the work gets done by Copy-paste work and > unless there is a online cookbook type thing, people will find it hard. > > Unless Clojure (not just Clojure, any functional programming language) > reduces this barrier to entry for both the non-core developers and the > Bosses, I think any FP language will find it hard. > Now, if we want Clojure to get accepted fast, I think we must be eager to > reduce this barrier to entry. Just saying that "If you can program in LISP" > then you can code Clojure is not a very smart answer. Similarly saying "get > used to parenthesis" is not a smart answer either. > > *It's really wrong to think that if they do not accept Clojure they are > doomed. The users have a higher hand and if not Clojure, new languages will > keep on being built until some language gets accepted by all. * > > I think what needs to be done is the following: > > 1. In LISP like languages, Parenthesis is a non-issue as long as code > is properly indented. And hence *strictly following indenting rules*while > defining functions/Macros etc. is really really crucial. Hence my next > point. > 2. Excellent IDE indenting support for code Indentation. I have tried > Enclojure and it is good. But what I find is difficult for the LISP newbie > is deciding when to put code on the new line for function arguments and all > that. I know that there are rules for that (most Scheme like), but frankly > speaking, people have become very lazy and wouldn't waste time reading the > indenting rules, especially when the other Haskell,F#, Scala camp boasts to > address the same issues and say that we do it all without a lot of > parenthesis. *So in short, Can we do something in the IDE to make > Clojure code indentation a totally no-brainer?* > Perhaps Enclojure must show some *placeholders* properly indented based > on whether it is defmacro, defn or fn etc and the developer just fills it > up. This would be an excellent aid in promoting LISP indenting standards. > Overtime, once Clojure is widespread people will automatically start > doing it correctly. > 3. A nice and good article written by Clojure folks that highlights how > to transition from the OOPS thinking to Functional thinking. > 1. Just saying "Practice coding recursively" does not help at all. > Even core CS people do not use recursion for every thing these days. > Students have OS labs, Networking Labs and projects but everyone knows > that > it does not involve much of recursion and all - just a plain sequence of > operations, that's all. At most 2-3 courses on Algorithms is where they > use > Recursion heavily. > 2. I think the keyword "let" and the pipeline is very understated. > "let" is the single most important thing that makes this transition > easy. > 4. Have an article that has some decent functions in it and each should > have an explanation of how an experienced Clojurian/LISPian would read this > code mentally. Just write it as a transcript. Or may be a Screen cast. > 5. Have a cookbook like article for all common tasks. > 6. Explain how idiomatic code looks like in Clojure. I feel experienced > Lispers can help in this. It feels very scary when an experienced OOPS > developer knowing all idiomatic code in OOPS does not know anything in > LISP. > I'm afraid to say that this feeling of shame must be prevented. After all, > writing idiomatic LISP code is not rocket science. It is just lack of > experience > > *I personally feel that Clojure has enough of nice features.* All that is > left to do is to make the transition easier. Although it may seem less > technical and less interesting to the geeks, it is important to realize that > all coders are not geeks and not everybody behaves rationally. Looks, > presentation and ability to lure people also matters. After all, looks and > ease in *every* way is an integral part of a Programming Language. > > Finally, comments like "doing XYZ will encourage them to write sloppy code" > is not good. Nobody likes to write sloppy code. They do it because writing > good code is very hard in the language or the IDE does not support it. > > A language need not be very good to be a success. A lot depends on the IDE > support. It is known to everybody that most of the Java code is written in > IDE like Eclipse only. And Eclipse goes a long way in reducing the > boilerplate code, thus reducing a flaw in the Java language. Similarly > Enclojure and Counterclockwise support be just splendid. Just saying "use > Slime/Vimclojure" does not help again. > > My 5 cents. it is too long for 2 cents. > > Thanks, > Ajay G > > On Fri, Dec 18, 2009 at 9:28 PM, Sean Devlin <francoisdev...@gmail.com> > wrote: > > Look, Clojure... > > -- > You received this message because you are subscribed to the Google > Groups "Clojure" group. > To post to this group, send email to clojure@googlegroups.com > Note that posts from new members are moderated - please be patient with > your first post. > To unsubscribe from this group, send email to > clojure+unsubscr...@googlegroups.com<clojure%2bunsubscr...@googlegroups.com> > For more options, visit this group at > http://groups.google.com/group/clojure?hl=en > -- You received this message because you are subscribed to the Google Groups "Clojure" group. To post to this group, send email to clojure@googlegroups.com Note that posts from new members are moderated - please be patient with your first post. To unsubscribe from this group, send email to clojure+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/clojure?hl=en
