After much searching I found a graphical front end that is downloadable and free called Remember that has the memtest binary inside the package. I ripped out the os x memtest because I like command line tools better than the GUI. He made a nice gui and all, but it’s not my style.

The bad news is I still can’t get ghc for freebsd on amd64. I found some mailing list posts from mid/end of April on the topic. It looks like work is being done. It feels like it will be done quickly. But I hope I still have ambition to work in Haskell by the time that rolls around. It feels very much like a neat and powerful language with a lot of personality.

But I can actually do things in Python.

Functional programming is… well it’s neat. And different. I fought for a long time with almost every data definition and function I wrote, but as the hours wore on it started to feel more and more natural and I went back and re-factored a bunch of my code. In no time flat I had everything up and working great. One little problem. Something felt wrong. I wrote a function to recursively blast alterations into my list and when I checked it out it looked like it wasn’t freeing memory and keeping a copy of each list ever touched. Bummer.

I dropped into #haskell and asked for advice. Man. I got so many things I didn’t understand thrown at me, also a few compliments. Apparently for a first solo Haskell project, my code was far more Haskell like than expected. Anyways, I had a bunch of things pointed out to me to fix or improve and was introduced to the concept of thunks. Haskell has lazy computation which means it only evaluates things when it needs to. Thunks as I understand are partially evaluated computations which sit around taking up ram waiting to be useful again.

Pretty sure I have to go back to trying to learn monads better to fix it though.

Code: here
Executable: here

The entire body was overly technical and I spent more time beating my head against understanding the content of the examples than I did trying to grok the actual Haskell lessons. I found it very neat to be walked through building a DES cryptography system, but I don’t feel it made me understand Haskell better.

On the upside, I can now look at Haskell code and generally work out what’s going on. The syntax mostly makes sense to me, though the typing of functions still messes with my head from time to time. Some of the random code flying past on #haskell is becoming less alien. I’ve written a few small proof of concept tools and am gaining more confidence and comfort with the language.

However I don’t feel ready to set out on my own and write something real world that solves a problem yet. I don’t think I have enough of a grounding in Haskell and its workings. So I’m off to find another online tutorial to work through!

Seriously. My brain hurts from the concepts and math of the encryption, not from the programming logic. Also of annoyance is the so called helpful library they have in the appendix which you can download from their website. For non standard definitions of can that is. I could be retarded, it has happened to me before. However, I could not find it, nor could Google .

I’ve gone and formatted it all pretty like and put it on the web:
SequenceUtilities.lhs

I googled for ‘learning haskell’ and ended up at the official haskell learning page. After browsing the options I decided to go for the tutorial called, ‘Two dozen short lessons‘, and download hugs as my haskell environment. So far so good.

I’ve never done any functional programming of any kind, so I had no idea what to expect. Erik has tried to show me a few things, but I’m the type of person who learns to do by doing. Most of the things he showed me I thought were neat conceptual tricks with no real reason for existing. I’m a good little sheep, so when the tutorial told me to forget everything I know about functional and object oriented programming I did so. With great ease and joy for the object stuff as well.

I’m on page 70 of the tutorial and so far I have found one main thing that annoys me. The way haskell is set up means that I can’t just write definitions (raw code) into the interpreter like I would Python. I have to do that in a text file and then load it. From there I can run commands on the code. So I can use the functions I define, but I can’t define them in the interpreter for a quick putter. Annoying, but not nearly as annoying as the fact that I can’t declare variables in the interpreter as well. I’m learning the way and the why of haskell so it makes sense to do it this way, but I’m lazy and it’d just be easier to write it right in.

I’m understanding how things work, and I’ve taken their examples and mashed them up to do my own thing differently, but I’m not sure I grok it yet. Most other coding tutorials I’ve done have started out with a hello world program. Haskell has dived right into sets, list comprehension, reductions and infinite iterations while briefly stopping to talk about the nature of numbers. It’s not making my head hurt as much as I thought it might, but there is still some pain.

So far I’m not seeing any real advantage to haskell as a language, but I’ve only started scratching the surface and the syntax. I’m seeing it more as a tool for doing computations like statistics than as a general language useful for everything. I’m going to keep at it and see if I can’t think up a project to do in haskell that solves a real problem.