I was chatting with a gentleman the other day about dynamic languages and he mentioned that he was at a workshop dedicated to them. One camp focused on languages such as Perl, Python, and Ruby. Others were focused on languages such as Smalltalk and Scheme. Allegedly there's a lot of envy -- perhaps animosity -- surrounding Perl because it's ugly, has all of these weird hacks, isn't "pure," and yet is enormously successful. At the same time, the Smalltalk and Scheme guys were trying to explain to the other camp that they solved our sorts of problems decades ago, so we should really listen to them.
This begs an obvious question: if they solved these problems so long ago, why are they viewed primarily as academic languages? Yes, Disney uses Smalltalk and Yahoo! Store was originally written with Lisp, but those seem like noteworthy exceptions. The proponents of these languages at the aforementioned workshop were apparently unable to explain why (of course, this is news I heard second-hand and you're hearing third-hand.)
So why aren't they used more? I suspect that part of it has to do with how "foreign" they are to most people's thinking. I still remember when I was first learning C and wondered how you could program without line numbers. If C wasn't so widespread, I might not have bothered with it. Is this what hampers Smalltalk or is there a different mechanism coming into play? I've read that Smalltalk used to be the second most popular OO language, but was displaced by Java.
Re:Simple
malte on 2004-09-14T08:15:31
You have never heard of a Smalltalk library because a huge beautifully designed class library is so inherent to the language that it is not worth mentioning.Re:Simple
hardburn on 2004-09-14T19:20:46
I don't think any of these fully explain why Smalltalk/Scheme/whatever didn't catch on in the first place (though they would explain why it is so hard for them to catch up after other langauges took root). For instance, if Scheme was a more popular language, you would likely see a lot more Scheme books.
I also suspect the "Difficulty" and "Syntax" reasons are non-issues if functional languages were people's first exposure to programming. LOGO, for instance, has a functional syntax and has been used to teach school kids to program.
Re:Simple
ziggy on 2004-09-15T01:23:06
True enough.I don't think any of these fully explain why Smalltalk/Scheme/whatever didn't catch on in the first placeThat's a pretty steep precondition for most programmers. The vast majority of programmers today did not use a functional language when first exposed to programming. The switch from an ALGOL derivative syntax and programming model to something fundmentally different is quite unsettling to many programmers. One programmer's non-issue is another programmer's deal breaker.I also suspect the "Difficulty" and "Syntax" reasons are non-issues if functional languages were people's first exposure to programming.Re:Simple
Ovid on 2004-09-15T15:05:00
I agree about the steep precondition. I still remember coming from BASIC to C in the 80s. I struggled to understand how someone could program without line numbers. Even though I agree with the premise that languages such as SmallTalk and Scheme would not seem so alien if these were the first languages that people were exposed to, the reality is that they are not the first language and they're unlikely to become that. Hence, we have a chicken and egg problem that is essentially unsolveable from this approach.
Proponents of those languages are going to have a hard time promoting them unless they can "unscare" programmers. I suspect that Paul Graham has had a moderate amount of success attracting people to Lisp due to his efforts. I know that I looked into Lisp as a result -- but I never followed up because I had no opportunity to apply it in a real-world setting.
Re:Simple
Isaac Gouy on 2004-09-14T19:37:35
- Advertising. That's reinforcement not cause - there's a market for books because the languages are widely used.
- Difficulty.One can read "Smalltalk by Example" and have a basic grasp on the language in a very short period of time.
- Syntax. Smalltalk's syntax is simple, consistent, powerful.
- Libraries... but I've never heard of them Have you ever looked for them? Smalltalk's rich libraries have been growing since 1980.
- ROV That's why people switch to Smalltalk - get more done, faster, with fewer people.
Re:Simple
ziggy on 2004-09-15T01:33:06
Hardly. That book is obscure and very hard to find on the bookshelf or at the bookstore. It is freely available online now, but it wasn't 5, 10 or 15 years ago. And freely available materials do not make a programming language easier to use or learn. Some of the best materials about Python have been freely available since the project's inception, yet the size of the Python community is persists at about 1/10th of the size of the Perl community.One can read "Smalltalk by Example" and have a basic grasp on the language in a very short period of time.Now that this and other Smalltalk books are freely available and getting more widely advertised (Thanks, Google!) we may see Smalltalk's popularity increase.
Yes, and I think we've pretty firmly established that they are not very well advertised. Again, reinforcement, not cause.Smalltalk's rich libraries have been growing since 1980.
Perl may be ugly, but it is IMHO one of the best languages out there for interacting with other programs and foreign libraries on a Unix system. I think the Lispniks may be catching onto this, but Lisp has so many years of bad reputation to live down...
Re:Another reason
Isaac Gouy on 2004-09-15T17:22:19
"interacting with other programs, handling files, and talking to other machines"
Why don't you think Smalltalk supports "handling files"?
How do you want to interact with other programs, and machines?
Corba? Sockets? SOAP? CGI? COM? DLLs/Shared libraries? inline C-code?
Here's the VisualWorks Smalltalks documentationRe:Another reason
educated_foo on 2004-09-16T15:04:26
Alright, you RTFMed me pretty good there. Having only used two smalltalks (GNU ST and Squeak), and those only briefly, I made the same mistake many people do with Scheme: there are a bunch of languages out there called Scheme, some supporting practical extensions beyond the standard. Using the wrong "Scheme", or only reading the standard, you get the impression that the language doesn't have much built in. With Scheme, I've been fortunate enough to be exposed to both Bigloo and PLT, two implementations with many nonstandard, practical extensions.As for "handling files" -- it's not just a matter of accessing the filesystem at all, but of slinging around directories and temporary files, generating input for and parsing output from text-based programs, dealing with proto-web-interfaces, etc., and doing all of these things as painlessly as possible. In other words, scripting in the "shell scripting" sense. This involves large numbers of intermediate files and complex external programs that I don't want to rewrite or try to hook into my language via FFI.
Re:Another reason
Isaac Gouy on 2004-09-16T17:40:08
'RTFMed me'
We all seem to mix up language and language implementation;-)
'scripting in the "shell scripting" sense'
That's something Perl was designed for, and not something Smalltalk was designed for.
Having said that, once we start thinking about web apps, take a look at Seaside.Re:Another reason
educated_foo on 2004-09-16T20:47:12
We all seem to mix up language and language implementation;-)
I think we also seem to mix up implementation (the interpreter, compiler, or whatever that makes programs run on a machine) and "implementation" (the non-standard extensions an implementation supports). GCC is IMHO rightly slammed for supporting a host of weird extensions to the C language -- however useful they are, with the exception of inline asm, I don't think many people consider them a positive feature, since they lead to nonportable code. Whatever else can be said of C, at least when someone complains about its shortcomings, "well, you should use this other implementation" is not an acceptable response.Maybe it's best to think of this the way some people think of natural languages vs. creoles or dialects. There may be one version that is considered "the official language" for political reasons, but in fact each implementation's "creole" is its own separate language. R5RS Scheme is too spartan for me. The "Bigloo" language, with its optional typing, regular parsers, removable support for escaping continuations, etc., is more to my liking, but I'm not sure how meaningful it is to call it "Scheme".
The consequences that students don't learn the language in school, they can't play around with it at home, they are unlikely to use it to solve a small problem at work.
Re:Development tools
Isaac Gouy on 2004-09-14T17:47:54
"expensive, commercial software"
That you can download for free and use non-commercially as long as you like: VisualWorks 7 Non-Commercial
And powerful environments which are free for commercial use: V Smalltalk/X
Re:Smalltalk etc
Isaac Gouy on 2004-09-14T17:37:56
"So why aren't they used more?"
One reason is badly dated mis-information like this!
"Historically, lack of integration with legacy databases"
Pre-historically! Smalltalk ORM was common from '90
"actively hostile to multiple programmers and source control"
From the late '80s Envy/Developer provided fine-grained (method level versioning) source code management. All the code was in a multi-user, replicated, database.
"Wall Street had... Patching live code on the fly is scary"
Many of those systems are still around. (Working on live production systems is dumb in any language - normal work was done on development machines, sweated on staging machines, and then pushed to production servers.)Re:Smalltalk etc
ziggy on 2004-09-15T01:44:46
You've struck a cultural difference here.From the late '80s Envy/Developer provided fine-grained (method level versioning) source code management. All the code was in a multi-user, replicated, database.On this side of the fence, "source control" is an aspect of the development process independent from the platform. We generally think of source control as standalone tools like subversion, cvs, rcs, perforce, clearcase and whatnot. Source code management in Smalltalk is a different beast with the same name and similar behavior.
When we cite the lack of integration with source control in Smalltalk instances, it is the lack of interoperability with the current tool-of-choice for source code management. It means leaving our current well-worn tools behind and moving into a brave new foreign world.
Citing source code management in Envy/Developer misses the point. It's not (necessarily) a slam that Smalltalk VMs (like Envy) are inferior to, say, Java, Perl, C# or Visual Basic. It's (frequently) an observation that our favorite tools and skills aren't supported. And that's disturbing.
I won't debate whether the CVS or the Envy model is better. That's a religious war for another time.
Re:Smalltalk etc
Isaac Gouy on 2004-09-15T16:45:43
"an observation that our favorite tools and skills aren't supported. And that's disturbing"
I haven't seen people have difficulty learning the tools. I've seen people have real difficulty understanding object oriented design.Re:Smalltalk etc
Isaac Gouy on 2004-09-15T16:47:52
What makes you claim "Smalltalk VMs... are are inferior to, say, Java, Perl, C# or Visual Basic"?
What comparison are you making?Re:Smalltalk etc
ziggy on 2004-09-17T03:12:50
Um, read what I wrote. I am not making the claim that Smalltalk is superior or inferior to anything.I am making the claim when we say "Smalltalk doesn't support source code management", it's not the same thing you hear when you respond "Envy/Developer provides SCM."
We are saying "we are happy with cvs/p4/svn/..., and cannot use our tools in your development environment." Your response that those capabilities exist in some other form does nothing to assuage our skepticism.
I do acknowledge that some less informed language bigots use the lack of CVS support to make the claim "Sma11T4lk 5UX!". I am not making that claim, nor am I defending the language bigot who makes that claim. I am stating that many who say "Smalltalk does not support [our favorite form of] SCM" are not saying that Smalltalk is inferior as a result.
Show me the money (Ugh!)
Ovid on 2004-09-15T15:18:55
I think Envy/Developer is another reason why SmallTalk has issues with acceptance. As a programmer, there are certain things I look for in a development environment, not just a language. I want to know about the language(s) I'll be using, how source control is managed, the database (if any), the test suite, the IDE (if any), etc. I like to learn about those things one piece at a time. Throw too many at me at once, or tell me I cannot use tools that I am comfortable with and I'll likely be less interested. Having tried and tested tools that the programmer already knows can inspire a lot of confidence. The barriers to entry must be low unless I'm getting paid for it.
In fact, the only reason SmallTalk really interests me right now is the "pure" OO model that I keep hearing people talk about. However, I'm brushing up on my Java because that's where the bulk of the work is. Until I hear convincing arguments about how SmallTalk can improve my professional life, I'll probably just wonder about it. I have a pretty active social life, so I don't have as much time for research as I would like. I won't argue that this is necessarily the best attitude to have, but it's a pragmatic attitude that many programmers worried about their mortgage or rent, savings, and other financial considerations have. "Show me the money" might sound rather mercenary, but I have to put food on the table.
Re:Show me the money (Ugh!)
Isaac Gouy on 2004-09-15T16:37:20
"tools that I am comfortable with"
Once in a while leave the comfort zone;-)
"but I have to put food on the table"
You and me both.Re:Smalltalk etc
dws on 2004-09-18T20:41:17
Historically, the lack of integration with legacy databases meant Smalltalk was restricted to new projects only.Visualworks 1.0 had connectors for Oracle and Sybase; later VWs had additional connectors, including one for ODBC. (Pre-merger) Digitalk Smalltalk also had database connectivity. Many of our (ParcPlace's) customers connected to legacy databases. What we didn't have was a story for connecting to "Big Iron" mainframe databases. (But then neither did Java for quite a while.)
The short answer: academic languages are eclipsed by workaday languages because the academics are focusing on solving tomorrow's problems. It's tomorrow, and we are adopting a lot of what they discovered decades ago (memory mangement, VMs, JIT, closures, objects, etc.).
A longer answer is in my journal.
FWIW, some guesses, on that:
Reputation would be the start of it, there.
Lack of publicity might be another part of it.
There may even be some cases (rare, perhaps, but perhaps possible) of "self doubt" (or "language doubt"?), among any users of said languages. I do not mean any sort of an odd joke by that, though; if it would be so, then it may be another matter hearkening back to reputation - really, of perceptions in the social sphere, which may exist around any languages or around any other tools or techniques.
So why aren't they used more? I suspect that part of it has to do with how "foreign" they are to most people's thinking.
Heh, I'll avoid jerking my knee about what might be regarded, but too-lightly, as if it was an accusation that Smalltalk or LISP would be foreign to most people's thinking.
As to the why of it, there: I could try to explain it away, myself, and not even succeed.
For the other half of my 1.9 cents, though:
Now, perhaps to smash any doubt about it: Myself, I would neither think nor would I encourage anyone else to think as if there was any "anti-{LISP, Smalltalk} smear campaign" going around, or anything like it; I doubt that anyone else would really think as so, either. "If {LISP, Smalltalk} is useful, and yet has been commonly overlooked, then that may just be a quaint accident."
Those of us who use said languages, and who derive any sense of benefit or general usefulness about them, I'd guess we're all generally content with 'em.
Now - while meaning no slight about C or else - I might wonder how many folks are quite content, themselves, with the languages that they've been sticking with. (For example, I've picked up an impression that maybe not so many people do really enjoy using C, though, evidently, very many people do use C. It is, quite so, a popular language.)
and, hey, in the end, there is a variety of languages, and "there's more than one way to do it", yah?
