Typesy Turvy

Lambda The Ultimate feed just notified me that types are considered harmful. Ah yes, I heard this before, so what's the news? The news is that it's not Stevey vs. Cedric or anything like that, this is Benjamin C. Pierce in his own write. Benjamin C. Pierce, from "Types and Programming Languages" and "Advanced Types and Programming Languages", King of ML, Lord of the Functional Programming Commonwealth, Defender of Type Systems Faith!

Nice presentation, BTW. I think it's one of these situations when a big shot computer scientist is confronted with a real life problem. So you say the language is perfect for writing a compiler - good for you. How about a database application with a web front-end?


(click to see full-size)

Mr. Language Designer, where are you in this picture? Yes you, who designed the mousetrap which "the bug" has safely escaped from. Now it is here, so are you at least by the developer side, handing him something heavy to throw at "the bug", or are you on the bed with the rest of the crowd going "ah ah ah, what do they teach computer science graduates these days..."?

One thing that impressed me during otherwise boring (let me just read you aloud the tutorial) JRuby on Rails preso I recently attended - here is a system that tries to serve the needs of the developer. Not server vendor, not language designer, not JSR politician, not some brandthirsty marketing person or buzz-oriented architect, THE DEVELOPER. I am not used to that. So all I have to say - programmers of the world, unite! Stand for your rights! We deserve better tools, because we are the ones getting the job done.

Thank you for reading.

In whining there is truth?

What do Java developers want? Hard to say. But here is what Java bloggers don't want Sun to do:
No evolving Java syntax - no properties, no closures, etc. - fix what is already there first.
No extending Java capabilities via annotations.
No investing in a new JVM language (JavaFX, or JRuby in the past)
Ok, maybe add Groovy, but don't change JVM spec.
Hey, wait, why are key people leaving Sun? We didn't want that either!

Girl Power

I have been reading recently that women are abandoning computer science, and that percentage of women in our profession is not just low, but getting lower. I'm not sure actually that the situation here is as bad as in North America, but it's certainly true that software engineering and computer science are not very popular among women. Why? To be honest, I don't have the answer.

Here is what comes to mind:

• Geek-ness is viewed in the society as the opposite of being attractive; this is much more important for young women, than for men. For a guy - a nice high-tech salary will provide the attraction instead.
  
• Sitting in a cube by the computer all day and doing one thing, coding, is probably not very attractive for most people, but for women especially, since most of us are better at performing a variety of tasks and interacting with people - I mean there has to be an explanation why "secretary" or "teacher" are such typical women professions.
• Hi-tech jobs are very demanding, too demanding. For most women family life is at least as important as professional life, and usually family comes first. But there just aren't many positions you can find after getting computer science degree that allow you easily balance work and family life.
  
• Hi-tech is for young people - look around, how many programmers you know are over 45? Why? For the same reason you don't see many women - it's hard to compete with the smart kids when you are pregnant, or haven't slept for a week, or worried about some family matter.
• But can't we just switch roles with the husband? well, go back to first bullet, double standards of the society certainly don't make it easier on us or our partners.
  
• People we are surrounded with (nerdy young men mostly) are pretty anti-social creatures in the first place, even more so with species of the other sex, even more so with the ones that don't fit social stereotype.
  
• Now suppose you survived all the obstacles, because you passionately love science and engineering. Did you watch the knack? There's a grain of truth there... society became less tolerant of weirdness, and people who 100 years ago may have been referred to as crazy geniuses nowadays live "normal life" on prescribed medications. And for a girl it's even stranger to be a crazy genius then for a boy.
• Male domination in the field - yes, it's chicken and egg problem. The field will change only if there will be enough women in it to drive the change from within and help other women. So I will devote the rest of the post to the ones who made it.

Lady Ada (Byron) Lovelace

She was more of a technical writer really, but she was a visionary.

Admiral Grace Hopper


The inventor of COBOL and debugging.

Professor Barbara Liskov


The one from the substitution principle. I chose her among several prominent computer scientist women because she was the first female computer science Ph.D in the US.

SVP Jayshree Ullal


I originally thought to put the Google princess here, but having personally met Jayshree (she was my manager's manager's manager at one point) and impressed by her personality and professionalism (she is so clever, and yet such nice and humble person), I decided she's a better candidate to represent successful women in computer industry.

Alice, Dilbert's workaholic colleague

What did she achieve exactly? Surviving in the office should not be underestimated! So I am going to honor another ex-colleague, although we never met, for providing inspiration for a character I can identify with :-)

Keep coding girls!

Net, Goddess of Wisdom

No, it's not a joke or some kind of wordplay.


Net (also Neith) was an ancient Egyptian goddess, not very well known perhaps, unlike her Hellenistic reincarnation - Athena. Net was a warrior goddess, just like Athena, and the goddess of weaving - "weaver" is actually the translation of her name.

Athena was a weaver too, according to the Arachne myth, and of course the goddess of wisdom.

So why weaving, or net, and wisdom? Well, for several reasons. What our ancient ancestors have probably recognized is that wisdom is acquired via communication. Today we know in addition that human thinking is powered by neural net in our brains. Net, yes net. We even use the same word. Internet, collective intelligence... "there's nothing new under the sun" said another wise ancient man.

Synthetic bridge method annotations

Update regarding "walls and bridges": Sun has kindly opened an enhancement request on my behalf for the issue. Please vote if you are affected by it too.

Java properties links

Since I am trying to keep up to date with what's going on in "Java Properties" land, I started with Alex Miller's collection and have been adding things that I find either relevant or influential. Another idea I nicked from Alex was to put them on Tumblog.

A disclaimer: this is a personal collection, so if something looks irrelevant, it means that I think it is influential :-)

I'll be more than glad to hear suggestions for more links or feeds.

My OO My

No, I don't mean the song, but it's to do with Scandinavia. Mads Torgersen mentioned this during interview with Joe Armstrong (highly recommended, there is also a part II), and later I saw another reference in a paper: apparently there is "Scandinavian school of object orientation", sometimes compared to "American school".

Here are some insights on the difference between the two. The point is that the former is about concepts and philosophy, and the latter is about pragmatic aspects like software reuse and organization. I am not gonna pretend that I'm smart - all I know about it comes from searching Google, so you can go there and find out for yourself.

* On a side note, since I am linking to MSDN here, I want to say that even though my last post wasn't very favorable to Microsoft, to be fair - the company plays an important role in renaissance and democratization of functional programming (LINQ, F#, etc.), with guys like Mads Torgersen and Erik Meijer largely responsible.

Open the source, sesame

About a month ago a big shot from Microsoft came to speak at the university - looks like he's on some kind of a tour, and his mission is to convince universities to teach Windows operating system aside Unix/Linux. He waved at the audience with a free license and tried very hard to prove that some aspects of Windows design are better than Unix, mainly because it is "newer" design and better suited for nowadays computers.

Long story short, one of the things he said, was that open source is bad, because engineer who looks at the source of a library will design his application assuming particular implementation, which is now tightly coupled to library internals - bad. Obviously a discussion erupted - design by contract and Eiffel were thrown in the air, relation between programming language and operating systems..., but frankly all I could think of is my distant past as a VB developer and those 1000 pages "Windows *** Unleashed" books. They were written by hackers who used trial-and-error against Windows DLLs trying to make some sense of the APIs - the formal documentation was either sloppy or intentionally incorrect to misguide us, Windows application creators outside Microsoft, since we were potential competitors. Things were so different when I moved to Java. (Though my first Swing experience made me - believe it or not - miss VB, but that's another story.)

Anyway, statement like this coming out of the mouth of a Microsoft employee, even if he has an impressive Unix record, was easy for me to dismiss. But here is a respected Smalltalker saying

"I think people should have the source ...not to get miss quoted saying I'm against open source, but I think it's important class libraries should be viewed like caves."

(Ah, wait, he is from IBM - he can be ignored too... just kidding)

I mean, seriously look at this puzzler for example - it's pretty cool, but the solution heavily relies on a particular implementation, I would even say this is actually a hack. And sometimes a hack is necessary... or is it? Is there or will there ever be a perfect environment where we can program without hacking?

P.S. The reference to Ali Baba tale in the title is intentional. Open source is good, very good, it's a treasure and in the world dominated by large corporations - almost a miracle. But IMO we should resist the temptation to exploit it to the last bit, because we might find ourselves locked in the cave.

Static methods as function objects

In one of the earlier posts I described a policy object created by "functional style programming" in Java. The basic idea is to make methods into Function and Predicate objects and then composite the logic from these building blocks. The problem though was that the policy implementation looked alien to Java. To re-cap: I wanted to invoke a different function based on the class of the object - classical multi-dispatch with a single argument.

So inspired by extension methods idea I added a new mechanism to define properties (and functions in general) - via static methods. You define a bunch of static methods with the same name (or appropriately annotated) like this:

static int size(Collection c) {
 return c.size();
}
static int size(Map map) {
 return map.size();
}
static int size(Object obj) {
 return (obj.getClass().isArray() ?
   Array.getLength(obj) : 1);
}

Now assuming I have a property declaration

Property<Object,Integer> size =
 new Property<Object,Integer>(0) {};

The underlying implementation of the size.of(object) would be a "policy" (composite function) dispatching to one of the static methods according to run-time type of the object. So that size.of("foo") is 1, size.of(Arrays.asList("a","b","c")) is 3 and size.of(null) is 0.

There are couple of issues though.

Issue #1: automatically ordering the rules. Now when rules are not added manually, I should be careful to test for more specific class first. The solution is to calculate a distance between two classes and always look for a best match - basically same algorithm javac uses when choosing methods at compile-time. This isn't fully unambiguous though: if class C implements 2 interfaces A and B, having different implementations of size(A) and size(B) would put me into a dilemma which one to choose for size(C). Javac in such a case reports an error - thanks to Java being statically typed. But once I make it a run-time decision the only logical solution is to report an error at run-time. To make the problem less acute, I could add an annotation that controls ordering of the methods.

Issue #2: this is less generic than a general predicate-based policy - it is limited to "instanceof" predicate. One of the policies I use internally in the properties project is reflectively invoke getFoo() if the class has a "getFoo" method. This isn't possible with static methods.

Issue #3: reflection. It's not really a problem, but rather an enhancement I want to make: instead of using reflection I could generate the functions at compile-time with APT, like Bruce Chapman does here. This is one of my next adventures, which I will hopefully describe in a future post.

Digg a pony

Here is another post with no helpful information, sorry, I meant to write a meaningful piece but I am on vacation and in fun spirits :-)

Anyway, I just got reminded of something that crossed my mind before and is now more relevant than ever - the term John Lennon invented almost 40 years ago: "bagism". It's amazing that at the time everybody laughed - a person giving an interview in a bag. You can look this up, or even better watch the video (fast-forward about 2 and a half minutes). Here's a short re-cap:

"if everybody had to go in a bag for a job (interview) there would be no prejudice, you see, you'd have to judge people on their quality within, you know, we call it total communication..."

I think this is really the web: blogging, mash-ups, syndications, and social networks, and why all this is a lot of fun - you really know nothing about the one you are talking to, except what he (she? it?) tells you.

Dig a pony lyrics mention

"penetrate any place you go" (...internet?...)
"radiate everything you are" (...wireless?...)
"imitate everyone you know" (...blogging? syndication?...)
"indicate everything you see" (...search? google?...)
"syndicate any boat you row" (...social networks?...)

Cheers to total communication!

P.S. A tribute to Passover holiday (the reason for my vacation): there were once four boys (ארבעה בנים) - the wise and wicked one, the cute one, the quite one, and the one that can't sing... and they were fab.

Java and closures

My reaction to the likely postponing of closures till Java 8.

Mercurial adventure on Widows

With a friend at work we decided to check out Mercurial as source control system. The repository was set up on a Linux box and zillion lines of code and libraries were sucked in from ClearCase, now I just had to clone it to my lap-top running Windows XP.

Installed Mercurial 0.9.5 via TortoiseHg (I am familiar with Tortoise with SVN so it was a natural choice) - very easy and straightforward so far. Issued clone command - hm, filename too long. Switching to cygwin - doesn't help.

So Ok, we have pretty deep directories structure and Mercurial makes it worse because it translates everything to lower case by "escaping" capital letters - adding underscore before the letter, so "A" becomes "_a" and a single underscore becomes a double one. Why? This is to avoid problems when working with case-insensitive operating systems like Windows. Fair enough, but what do I do with these long filenames now... Python bug? Oh, great. In the meantime the error became "file not found", but I (correctly) suspected the cause is the same.

I found this bug report (and this one) - at least it happenned to other people, it says I can use a 'usestore' workaround. Cool, found this - says change one of the scripts, but hey, they are compiled?! Ok, downloaded ActivePython (I admit, I never touched Python in my life) also downloaded the sources. But apparently I need a C compiler too for the full build from source - that's too much, I decided to just run Mercurial with the interpreter.

Found the right script, changed it. Boom, of course extra space in indentation - until now I only heard jokes about this (besides similar experience with Fortran half a life ago). Nevermind, got it now. Still not working. I take another look at the script - the if statement I just added was already there before! Oops, what's going on? Trying to understand the Python code - it is looking for some 'usestore' configuration option, which BTW is supposed to disable the "escaping". All I need now is to set it. Hm, how? Eventually I find out that I need to edit <TortoiseHg-install-dir>/Mercurial.ini adding

[format]
usestore=False

(If you use plain Mercurial, not Tortoise, create Mercurial.ini under <mercurial-install-dir>)

I do it and... voilà, the repository cloned!!! So bottom line - all I needed was ini file change. It is somewhat dangerous 'coz I still have the mixed case, but at least I have the code on my laptop. Why didn't I find it in an FAQ or tips'n'tricks or something? No idea. Hope this post helps the next person who googles for it.