[clue-talk] Stupid OCaml tricks

[Nate Duehr]

On Aug 14, 2006, at 8:58 PM, Matt Gushee wrote:

> Dennis J Perkins wrote:
>
>>> During my presentation last week, Dennis asked something along  
>>> the lines of "what makes OCaml a better language for you?" Can't  
>>> recall the exact words, but I believe that was the gist of it. I  
>>> don't think I answered the question very well, so let me give  
>>> y'all an example that might give you a better idea.
>> My question was, "How does OCaml make something complex simple?"
>> Maybe a better question is how does it simplify somthing that is  
>> hard to
>> do in other languages?
>
> Oh, yeah, I remember that now. It's a good question, and there's  
> probably a good answer to it, but ... I don't know. I can't think  
> about it anymore just now. And whatever I say, somebody will  
> probably point out how wrong I am ... getting harder and harder to  
> believe I am right about anything, so I think I'll just shut the  
> fuck up and go die somewhere.
>
> Sorry, I know you're sincerely interested and deserve a good  
> answer. I am just fucked up and everything I try to do just makes  
> it worse.

Whoa!  Matt... that was definitely not the impression I was  
attempting to give you at all!!!  I just enjoy the discussions about  
how languages and "technology advancements" do or don't really make  
the overall big picture better or worse... but in NO way was I  
bashing on OCaml or your comments about it!

I was just taking the discussion to "somewhere different" to see if  
anyone else had some interesting thoughts about it.

After 24 years of using computers, (oh lord... has it been that  
long?) they're really not doing anything "new" since about year 5 of  
that 24.  They're (sometimes) doing it faster, and certainly more  
"pretty"... but not "different".

I keep looking around for the next "big thing" and I just don't see  
it coming -- it'll be there sooner or later, but I'm trying to not be  
surprised about it.  :-)

The advent of Linux was cool, and I've invested a lot of time and  
effort into learning it and it made a great stepping stone into a  
Unix career by some stroke of dumb luck (and I was "saved" from  
dealing with Windows desktops in the business world for almost all of  
my career!) but Linux hasn't really panned out to be as "great" as a  
lot of us zealouts (me included) thought it was going to be.  I love  
Linux, but it's gotten quite bloated and disorganized on the desktop  
stuff.  Servers... it's got it nailed... but desktop?  The problems  
with proprietary hardware have only been kinda "half-solved" over all  
these years... which seems retarded to me to some extent... and sad.

But seriously Matt -- NOTHING I said was meant to be personal or  
negative about what you're doing, or the OCaml discussion!  OCaml is  
actually neat technology -- I just wonder in general about all these  
languages... being a sysadmin I really haven't found ANY application  
on the server side that I couldn't do in a shell or in Perl... and  
it's been that way for something like 10 years now.

Python looks neat... played with it a bit... okay, whitespace zealots  
and haters aside, it had some interesting new ideas... various shells  
have been messed with over the years... and some C and C++ (I  
actually had a C++ coding job for three months... yuck... luckily I  
learned quickly that coding in C++ was not my lifelong cup of tea...  
and probably others.. (Should I admit I used a lot of BASIC years  
ago, and it's still useful to have a BASIC compiler for  
microcontrollers today?  And of course, SQL in various flavors and  
different databases maintained and used...

But...

After all that... 95% of the "stuff" that really makes companies  
money boil down to things Perl does well... pulling data from text  
files, or a database, stuff them into arrays, do some manipulation,  
and stuffing it all back into a database... the faster the better...  
that stuff makes the serious behind-the-scenes real money in the IT  
world... or so it seems from a Unix system admin's perspective.

So I get kinda jaded about "the latest and greatest language"  
stuff... even if the language is as old as OCaml.  I respect that  
it's been around a long time... but so has REXX and I've coded in  
REXX once or twice for real work (god save me, I had to admin OS/2  
boxes with Sybase SQL Anywhere databases on them once!) and it was  
kinda "just another language".

So seriously Matt -- you didn't say or do ANYTHING wrong.  And  
nothing wrong with OCaml, etc... don't take any of the comments I've  
made personally.  I think I'm bored with certain things in Unix right  
now.  Maybe it's time I learn to really code some stuff in a GUI.  I  
can DEFINITELY say you're WAY ahead of me there.  I'm kinda at a fork  
in the road... I've been enjoying programming and building little  
hardware projects on microcontrollers, and I have to decide if I like  
that enough to start playing with more low-level stuff like DSP's or  
FPGA's... or if I should go play with some of the GUI toolkits and  
find out just how hard making nice GUI code is.  :-)

So... maybe that's the real point of all my ranting here... plus that  
I love to talk about tech futures... "What's next?" for all this  
stuff?  GUI's?  Automation?  Making software simpler?  Making  
software work out of the gate?  (That's a completely different  
rant... software companies are actually NOT trying to release great  
software... their incentive is to SHIP code... even all the way down  
to how most software engineers get paid.  I once thought "open  
source" was the answer, but the reality seems to be that when someone  
releases some GREAT code in the open-source world, they get gobbled  
into a Corporation or start their own, and their contributions taper  
off.  Sure, there's a number of kernel hackers that are paid, but the  
end-user software guys don't get to stay in open-source for long...  
you see them in their college years, and maybe a few years after, and  
then they switch to maintenance mode on their code and nothing new  
comes out after a while... so I'm not convinced that the ultimate  
goal of open-source is to ever get truly "great" software, but  
sometimes it happens... Apache is a nice example of something  
generally going really "right", I suppose.

Well anyway, sorry to rant... I should probably blog some of this  
stuff instead of trying to tag it onto a useful discussion about the  
nitty gritty of OCaml or any other technology... sorry if that came  
out wrong, Matt!

--
Nate Duehr
nate at natetech.com