Sunday, December 26, 2010

ycombinator and things I read

Recently a friend of mine send me a link to ycombinator - great resource for someone like me. So, if you like to read things that are more pragmatic then going deep but stay focused on one technology, methodology or technic, then this is something like /. but more techie. Simply I love it, usually, really short articles behind, on various subjects. Some time ago I spend tremendous amount of time looking for some good resources of techie knowledge, and so far /. is the only thing I am using on daily basis, but ycombinator seems to be great as well.

In terms of books, recently I read:
Knuth's The Art Of Programming (first 3 tomes)
Avodey's Category Theory (Oxford Logic Guides)

I enjoy reading few books at this same time (well, almost this same). When a part of book, or some topic is not really interesting, I switch to second book, and give my brain a fresh topic.

In terms of Knuth's books, I enjoy the way how they are written, easy to read, with sense of humor. Excellent exercises at the end of each topic. I don't like the MIX language. It is too old fashion to me. Sometimes, dealing with the language is harder then the algorithm itself, and a problem is more about how to write the code well (or even, how to write it), and not how to deeply understand, why some operations take place, and why they are done now, and that way. But still I recommend this book.

I spend some time deciding between Category Theory (Oxford) and Basic Category Theory for Computer Scientists (MIT). The O book is newer and bigger than M book, but also more expensive. I made a decision, base on fact that I am more Oxford then MIT boy (sorry chaps:)). So far I love it, easy to understand, lot's of new stuff to learn. I don't see to many exercises that are good for me, and It makes it hard to test new knowledge.

Sunday, December 19, 2010

Enabling grep in windows emacs installation

New windows, means new emacs and firefox configuration.

First step is to install dired plus, so it would be easy to search using grep directly from dired.

To enable grep in windows installation of emacs one needs to install grep, it is a good idea to install also find utilities. I tend to change the find.exe file name to efind, because windows already defines find utility. Then add bin directory to windows path. And copy following libraries:


to emacs installation bin directory:


In addition in my last post I forgot to add following line in .emacs

(setq make-backup-files nil)

In terms of Firefox some additional plugins, that I tend to use recently:

Tamper data
Visual Event

Haskell Platform

Long, long time ago, when I started to learn haskell I spend many days trying to choose between ghc and hugs. Some people used ghc and other hugs, and there was no right choice. The problem for me was pretty big, because I come to haskell from the world of scheme, and in scheme sooner or later one finds out that there are 'some' differences between implementations, and later on when you want to use some public library you need to switch to other implementation, and code that you already have is not 100% compatible with it. And some migration errors are really hard to find. For example if an element of a code is not used usually, but will fire up when some weird criteria are met. Sad thing, isnt it?

Today, I created a new Windows box, and I needed an haskell environment. I tend to write my haskell code in Emacs ( edition), and I love it, but today I went to, and I was surprised to see that there are dramatical changes in the contend of this site (I don't use this site at all). And finally somebody have written some cool looking haskell tutorial. I love the look and feel of it (guys, great job there). Also I was surprised to see that they are promoting a Haskell Platform. Witch is an environment and set of tools on top of ghc. I gave it a shot, and so far so good. Good to see that the haskell is still improving. Let's keep it that way:)