Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 13:02:37 UTC, Adam D. Ruppe wrote: On Tuesday, 14 June 2016 at 09:30:02 UTC, Chris wrote: So, what I ideally want to do isn't actually a scripting language, but rather a really easy to use IPC and D build library. The end user thinks it is a script, it works the same way, they write a little program and load it up, but the application actually compiles it as D into an independent exe and spins it up as a new process, communication over a pipe and/or shared memory or something. The user script can then crash independently! And they can use the whole D language. The whole interface of interop can be automatically generated. Of course, the problem with the build thing is the program you write might need to be distributed with a D compiler though I don't think that's a dealbreaker, D compilers aren't that big and could be downloaded on demand by your program. (Well, dmd isn't, ldc/gdc is 10x bigger.) Could we build on this: https://github.com/Hackerpilot/libdparse ?
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 13:17:46 UTC, Chris wrote: In my experience, statically typed languages prevent a lot of silly and time consuming bugs by simply checking the type. Yes, but I would put it more generally. Matching a program against a specification of constraints prevents a set of runtime errors and bugs. But providing the specification is also tedious. You can have much stronger static verification of constraints than in C++/D. For instance check the various legal states and what transitions you can have between them. I am certainly in favour of more static checks, but I am pleasantly surprised at what can be achieved with optional static analysis.
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 12:14:45 UTC, Ola Fosheim Grøstad wrote: On Wednesday, 15 June 2016 at 11:33:23 UTC, Chris wrote: But Python for example doesn't care. Python is fairly dynamic and static analysis tools such as PyCharm helps a lot when you write larger Python projects. What you describe is basically trying to mimic static typing. No, you can have something like this: var x = f(42); if(x>100) freeze x; // whether x is immutable or not is not statically known at this point But 'freezing' ain't got nothing to do with the type. It's basically D's `immutable`. In my experience, statically typed languages prevent a lot of silly and time consuming bugs by simply checking the type. If, after a few changes, function arguments don't match, the compiler (or interpreter) complains immediately. In Python the code may still work: ` def loophere(variable): for i in variable: print (i) loopy = "Ola, mundo!" loophere(loopy) # Developer decides that `loopy` should be a list of strings loopy = ["Ola", "mundo", "!"] loophere(loopy) `
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 11:33:23 UTC, Chris wrote: But Python for example doesn't care. Python is fairly dynamic and static analysis tools such as PyCharm helps a lot when you write larger Python projects. What you describe is basically trying to mimic static typing. No, you can have something like this: var x = f(42); if(x>100) freeze x; // whether x is immutable or not is not statically known at this point
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 10:27:14 UTC, Ola Fosheim Grøstad wrote: Sorry, I missed this question. No, some languages distinguish by requiring you to use special syntax like "let x = 4" for introducing the name "x" and so on. Some languages also allow you to freeze a variable so it becomes constant. There is a pretty large space of different semantics to choose from. But Python for example doesn't care. What you describe is basically trying to mimic static typing.
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 10:08:41 UTC, Ola Fosheim Grøstad
wrote:
That's just because your example isn't realistic. A realistic
case in Python etc is where you accidentally assign when you
wanted to introduce a new symbol. That is not a typing issue.
A realistic D/C++ scenario:
import std.stdio;
// imported from libraries:
auto create_name(string n){
return ["PersonName",n];
}
auto create_name(const(char)* n){
import core.stdc.string: strlen;
auto slice = n[0 .. strlen(n)];
return slice.dup;
}
void main(){
auto myname = create_name("Ola");
writeln("Letter count: ", myname.length);
}
It's much harder to shoot yourself in the foot, though:
`
auto create_name(string n)
{
return ["PersonName",n];
}
auto create_name(const(char)* n)
{
import core.stdc.string: strlen;
auto slice = n[0 .. strlen(n)];
return slice.dup;
}
void main()
{
auto myname = create_name("Ola");
writeln("Letter count: ", myname.length);
auto p = Person();
p.firstName = myname;
writeln(p.firstName);
}
struct Person
{
char[] firstName;
char[] secondName;
}
`
Error: cannot implicitly convert expression (myname) of type
string[] to char[]
If you have
`
void main()
{
import std.conv : to;
auto myname = create_name("Ola");
writeln("Letter count: ", myname.length);
auto p = Person();
p.firstName = to!string(myname);
writeln(p.firstName);
}
struct Person
{
string firstName;
string secondName;
}
`
Then you will convert `string[]` into the string `["PersonName",
"Ola"]`, and you have a bug. However, factually, I hardly ever
encounter bugs like this, whereas in Python this happens quite a
lot once you deal with more than one module.
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 09:40:39 UTC, Chris wrote: On Wednesday, 15 June 2016 at 09:09:42 UTC, Ola Fosheim Grøstad wrote: This isn't related to dynamic typing. It is related to variable assignment with implicit declaration and initialization. But this is part of the definition of a dynamic language, isn't it? Sorry, I missed this question. No, some languages distinguish by requiring you to use special syntax like "let x = 4" for introducing the name "x" and so on. Some languages also allow you to freeze a variable so it becomes constant. There is a pretty large space of different semantics to choose from.
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 09:40:39 UTC, Chris wrote:
But factually I don't:
That's just because your example isn't realistic. A realistic
case in Python etc is where you accidentally assign when you
wanted to introduce a new symbol. That is not a typing issue.
A realistic D/C++ scenario:
import std.stdio;
// imported from libraries:
auto create_name(string n){
return ["PersonName",n];
}
auto create_name(const(char)* n){
import core.stdc.string: strlen;
auto slice = n[0 .. strlen(n)];
return slice.dup;
}
void main(){
auto myname = create_name("Ola");
writeln("Letter count: ", myname.length);
}
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 09:09:42 UTC, Ola Fosheim Grøstad
wrote:
This isn't related to dynamic typing. It is related to variable
assignment with implicit declaration and initialization.
But this is part of the definition of a dynamic language, isn't
it?
Conceptually you would have the same problem in C++ and D since
they both use duck-typing (e.g. both overloading and templates
are essentially providing duck-typing).
But factually I don't:
`
import std.stdio;
void main()
{
auto name = "Walter";
writeln(name);
name = ["Walter"];
}
`
Error: cannot implicitly convert expression (["Walter"]) of type
string[] to string
Re: I'd love to see DScript one day ...
On Wednesday, 15 June 2016 at 08:48:19 UTC, Chris wrote:
On Tuesday, 14 June 2016 at 21:23:38 UTC, Walter Bright wrote:
On 6/14/2016 11:55 AM, Dicebot wrote:
I find a typeless language convenient when it's less than one
screen in size. Their advantages fall away when things get
larger. I don't know how people cope with a large project in a
dynamic language.
In the long run the disadvantages of dynamic languages outweigh
the advantages. There is the issue of redefining variables and
values. It can introduce subtle bugs that are hard to find. You
spend a lot of time debugging stuff that would have easily been
caught in a static language.
Stuff like this is not uncommon:
`# example.py
name = "Walter"
print(name)
name = ["Walter"]
if len(name) == 1:
print("Your name has length 1")
`
This isn't related to dynamic typing. It is related to variable
assignment with implicit declaration and initialization.
Conceptually you would have the same problem in C++ and D since
they both use duck-typing (e.g. both overloading and templates
are essentially providing duck-typing).
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 21:23:38 UTC, Walter Bright wrote:
On 6/14/2016 11:55 AM, Dicebot wrote:
I find a typeless language convenient when it's less than one
screen in size. Their advantages fall away when things get
larger. I don't know how people cope with a large project in a
dynamic language.
In the long run the disadvantages of dynamic languages outweigh
the advantages. There is the issue of redefining variables and
values. It can introduce subtle bugs that are hard to find. You
spend a lot of time debugging stuff that would have easily been
caught in a static language.
Stuff like this is not uncommon:
`# example.py
name = "Walter"
print(name)
name = ["Walter"]
if len(name) == 1:
print("Your name has length 1")
`
It prints
`Walter
Your name has length 1
`
Horrible!
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 21:23:38 UTC, Walter Bright wrote: I find a typeless language convenient when it's less than one screen in size. Their advantages fall away when things get larger. I don't know how people cope with a large project in a dynamic language. By adding asserts, testing, logging and if available: static analysis. :-) Just adding an assert can help a lot on static analysis. But I'd say that gradual typing is a pretty nice tradeoff. It is faster to iterate without all the typing syntax, and then you can add the types when you are happy with the result. Although the best solution would probably be to throw a "switch" on the static analysis before release so that all undecided types (not able to deduce it by static analysis) are caught before the program is shipped.
Re: I'd love to see DScript one day ...
On 6/14/2016 11:55 AM, Dicebot wrote: Because for some people static typing and compile-time checking is not a burden but a great help in designing bug-free programs. And if you stand by those values, JS is absolutely terrible language no matter the domain (it drives me mad every time I have to touch it). I find a typeless language convenient when it's less than one screen in size. Their advantages fall away when things get larger. I don't know how people cope with a large project in a dynamic language.
Re: I'd love to see DScript one day ...
On 06/14/2016 07:09 AM, dewitt wrote: > I know its a dream but really whats wrong with JS in the browser? Now > with no need for JQuery and ES6 its not really that bad. Better to use > it for a scripting lang and other things. My opinion only.. If were > dreamin Id rather see the next Kafka or Elasticsearch big idea written > in D... Because for some people static typing and compile-time checking is not a burden but a great help in designing bug-free programs. And if you stand by those values, JS is absolutely terrible language no matter the domain (it drives me mad every time I have to touch it).
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 13:04:32 UTC, Adam D. Ruppe wrote:
that will literally compiler in D today if you stick an "import
arsd.jsvar;" in there too :)
You can also do
import std.variant;
alias var = Variant;
void main()
{
var name = "Herbert";
name = "Brown";
name = 5;
}
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 12:57:01 UTC, Chris wrote:
Dynamic typing should only be allowed to the extend `auto`
works:
var name = "Herbert"
// ...
name = "Brown" // ok
name = 5 // error
this is still error-prone. why don't just do static typing
instead, and introduce `anything` type, which is "variant", and
can hold literally anything? with `auto` you'll barely need to
think about types at all.
and yep, compiler can turn runtime typechecks in compile-time
specialization — kinda like templates. so we'll have something
like this:
// this does runtime type checks
auto add (auto a, auto b) { return a+b; }
// this does compile-time specialization, but
// can also be executed by very straightforward
// interpreter
auto add2 (auto a, auto b) {
if (typeof(a) is int && typeof(b) is int) {
// here compiler knows actual types of `a` and `b`
return a+b;
} else {
throw new Error("wut?!");
}
}
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 12:57:01 UTC, Chris wrote: Dynamic typing should only be allowed to the extend `auto` works: var name = "Herbert" // ... name = "Brown" // ok name = 5 // error So var name = "Herbert"; name = "Brown"; name = 5; that will literally compiler in D today if you stick an "import arsd.jsvar;" in there too :) That's where my script.d came from - I wanted to see how closely I could simulate a "var" type in D itself, then got so close I decided to add a thin dynamic script around it too. But since the script and D both use the very same var type, the interop is easy.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 09:30:02 UTC, Chris wrote: Rather than pursuing JS endeavors, I was thinking of creating a light weight scripting language on top of D, something better than JS and not tied to the Web. Something people could use for day to day scripting tasks (like Python or Lua). Maybe we can build on the DMDScript engine, I don't know. So, what I ideally want to do isn't actually a scripting language, but rather a really easy to use IPC and D build library. The end user thinks it is a script, it works the same way, they write a little program and load it up, but the application actually compiles it as D into an independent exe and spins it up as a new process, communication over a pipe and/or shared memory or something. The user script can then crash independently! And they can use the whole D language. The whole interface of interop can be automatically generated. Of course, the problem with the build thing is the program you write might need to be distributed with a D compiler though I don't think that's a dealbreaker, D compilers aren't that big and could be downloaded on demand by your program. (Well, dmd isn't, ldc/gdc is 10x bigger.)
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 12:49:03 UTC, ketmar wrote: On Tuesday, 14 June 2016 at 06:52:38 UTC, cym13 wrote: browser. It's main forces (compile-time stuff and speed) would have no impact at all and the static typing would just make it harder for no good reason. The web is a very dynamic place so I'd rather have a very dynamic language. ah, that's why we have dart and typescript with static typing bolted on top of js... 'cause web needs more dynamics! Dynamic typing should only be allowed to the extend `auto` works: var name = "Herbert" // ... name = "Brown" // ok name = 5 // error
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 06:52:38 UTC, cym13 wrote: browser. It's main forces (compile-time stuff and speed) would have no impact at all and the static typing would just make it harder for no good reason. The web is a very dynamic place so I'd rather have a very dynamic language. ah, that's why we have dart and typescript with static typing bolted on top of js... 'cause web needs more dynamics!
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 06:52:38 UTC, cym13 wrote: The web is a very dynamic place so I'd rather have a very dynamic language. Maybe not javascript but really no D. Isn't the idea of Typescript that there is a demand for static typing, or at least type annotations and compile-time type checking?
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 10:46:46 UTC, Chris wrote: I'd say let's start with as few features as possible start with clean language design, or you will end up with javascript anyway. ;-)
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 10:16:33 UTC, ketmar wrote: you can start with "language report" (i.e. short semi-formal description of language). then people will have at least something to bikeshed. ;-) I'd say let's start with as few features as possible and concentrate on speed (including JIT) and extensibility. With extensibility I mean a) the language itself, i.e. it should be easy to extend the language without having to redesign the core and break everything, and b) easy integration of D and C. JS will become obsolete when other languages can be compiled to webasm, which is the right way to go. I wouldn't put too much effort into JS anymore.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 01:36:48 UTC, Yuxuan Shui wrote: On Tuesday, 14 June 2016 at 00:55:52 UTC, Walter Bright wrote: On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: My *real* dream is for D (or some suitable subset thereof) to replace Javascript in browsers. But that's a very distant, if at all even possible, dream. :-D It's a great dream! Don't we have a pretty efficient JIT javascript implementation in D already? https://github.com/higgsjs/Higgs higgs doesn't work if you want to zerocopy share data between D and the scripting language. so D here is more or less just the implementation language.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 09:30:02 UTC, Chris wrote: Rather than pursuing JS endeavors, I was thinking of creating a light weight scripting language on top of D, something better than JS and not tied to the Web. Something people could use for day to day scripting tasks (like Python or Lua). Maybe we can build on the DMDScript engine, I don't know. you can start with "language report" (i.e. short semi-formal description of language). then people will have at least something to bikeshed. ;-)
Re: I'd love to see DScript one day ...
On Monday, 13 June 2016 at 23:47:40 UTC, Walter Bright wrote: Remember you also have to write the documentation for your own language, and write your own test suite. With Javascript you can rely on existing documentation and test suites, which are enormous time savers. oh, noes. while tests *may* be time savers, ecmascript documentation is surely isn't. es is so full of quirks... sane scripting language can be documented in two pages of text. this may not allow to reimplement it 1:1, but will allow to use it without any troubles. did that IRL. and tests "grows organically", especially with dogfooding. it took me less than a month to make language usable. within next month or two i may be able to write it's own SSA codegen backend. more time here, as i forgot nearly everything i knew about SSA, and have to (re)learn some things. (not that i'll do that, it's just an estimation)
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 08:41:31 UTC, Thomas Mader wrote: On Tuesday, 14 June 2016 at 03:31:10 UTC, H. S. Teoh wrote: It's unlikely to happen in the near future, though, if at all. But one can dream. :-) I think the big plan behind Web Assembly is to allow foreign languages an entry into the browser world. IIRC it's planned that the entire DOM will be accessible via Web Assembly at some point too. But thats quite far in the future. But with that I don't see any reason why D couldn't run directly in the browser and do the same stuff JavaScript does. Well except some technical details which are not clear to me now. :-D Rather than pursuing JS endeavors, I was thinking of creating a light weight scripting language on top of D, something better than JS and not tied to the Web. Something people could use for day to day scripting tasks (like Python or Lua). Maybe we can build on the DMDScript engine, I don't know.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 03:31:10 UTC, H. S. Teoh wrote: It's unlikely to happen in the near future, though, if at all. But one can dream. :-) I think the big plan behind Web Assembly is to allow foreign languages an entry into the browser world. IIRC it's planned that the entire DOM will be accessible via Web Assembly at some point too. But thats quite far in the future. But with that I don't see any reason why D couldn't run directly in the browser and do the same stuff JavaScript does. Well except some technical details which are not clear to me now. :-D
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 03:31:10 UTC, H. S. Teoh wrote: On Tue, Jun 14, 2016 at 01:36:48AM +, Yuxuan Shui via Digitalmars-d wrote: On Tuesday, 14 June 2016 at 00:55:52 UTC, Walter Bright wrote: > On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: > > My *real* dream is for D (or some suitable subset thereof) > > to replace Javascript in browsers. But that's a very > > distant, if at all even possible, dream. :-D > > It's a great dream! Don't we have a pretty efficient JIT javascript implementation in D already? https://github.com/higgsjs/Higgs I know that. What I meant was that instead of writing javascript we'd write D instead, in our webpages. Or at least, some subset of D that's safe for the web. I.e., instead of: ... you'd write ... It's unlikely to happen in the near future, though, if at all. But one can dream. :-) T I love D but I don't see the point of putting it directly in a browser. It's main forces (compile-time stuff and speed) would have no impact at all and the static typing would just make it harder for no good reason. The web is a very dynamic place so I'd rather have a very dynamic language. Maybe not javascript but really no D. Let's keep D where it's strong and accept that it can't be strong at everything and that other languages are better suited for the task.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 03:31:10 UTC, H. S. Teoh wrote: On Tue, Jun 14, 2016 at 01:36:48AM +, Yuxuan Shui via Digitalmars-d wrote: On Tuesday, 14 June 2016 at 00:55:52 UTC, Walter Bright wrote: > On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: > > My *real* dream is for D (or some suitable subset thereof) > > to replace Javascript in browsers. But that's a very > > distant, if at all even possible, dream. :-D > > It's a great dream! Don't we have a pretty efficient JIT javascript implementation in D already? https://github.com/higgsjs/Higgs I know that. What I meant was that instead of writing javascript we'd write D instead, in our webpages. Or at least, some subset of D that's safe for the web. I.e., instead of: ... you'd write ... It's unlikely to happen in the near future, though, if at all. But one can dream. :-) T I know its a dream but really whats wrong with JS in the browser? Now with no need for JQuery and ES6 its not really that bad. Better to use it for a scripting lang and other things. My opinion only.. If were dreamin Id rather see the next Kafka or Elasticsearch big idea written in D...
Re: I'd love to see DScript one day ...
On Tue, Jun 14, 2016 at 01:36:48AM +, Yuxuan Shui via Digitalmars-d wrote: > On Tuesday, 14 June 2016 at 00:55:52 UTC, Walter Bright wrote: > > On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: > > > My *real* dream is for D (or some suitable subset thereof) to > > > replace Javascript in browsers. But that's a very distant, if at > > > all even possible, dream. :-D > > > > It's a great dream! > > Don't we have a pretty efficient JIT javascript implementation in D > already? > > https://github.com/higgsjs/Higgs I know that. What I meant was that instead of writing javascript we'd write D instead, in our webpages. Or at least, some subset of D that's safe for the web. I.e., instead of: ... you'd write ... It's unlikely to happen in the near future, though, if at all. But one can dream. :-) T -- Computers shouldn't beep through the keyhole.
Re: I'd love to see DScript one day ...
On Tuesday, 14 June 2016 at 00:55:52 UTC, Walter Bright wrote: On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: My *real* dream is for D (or some suitable subset thereof) to replace Javascript in browsers. But that's a very distant, if at all even possible, dream. :-D It's a great dream! Don't we have a pretty efficient JIT javascript implementation in D already? https://github.com/higgsjs/Higgs
Re: I'd love to see DScript one day ...
On 6/13/2016 5:13 PM, H. S. Teoh via Digitalmars-d wrote: My *real* dream is for D (or some suitable subset thereof) to replace Javascript in browsers. But that's a very distant, if at all even possible, dream. :-D It's a great dream!
Re: I'd love to see DScript one day ...
On Mon, Jun 13, 2016 at 04:47:40PM -0700, Walter Bright via Digitalmars-d wrote: > On 6/13/2016 2:13 PM, ketmar wrote: > > it heavily depends of what exactly one want and how. by the time > > i'll fully understand DMDScript code, i'd complete my own > > implementation of dynamic scripting language with JIT. > > Remember you also have to write the documentation for your own > language, and write your own test suite. With Javascript you can rely > on existing documentation and test suites, which are enormous time > savers. My *real* dream is for D (or some suitable subset thereof) to replace Javascript in browsers. But that's a very distant, if at all even possible, dream. :-D T -- I am not young enough to know everything. -- Oscar Wilde
Re: I'd love to see DScript one day ...
On 6/13/2016 2:13 PM, ketmar wrote: it heavily depends of what exactly one want and how. by the time i'll fully understand DMDScript code, i'd complete my own implementation of dynamic scripting language with JIT. Remember you also have to write the documentation for your own language, and write your own test suite. With Javascript you can rely on existing documentation and test suites, which are enormous time savers.
Re: I'd love to see DScript one day ...
On Monday, 13 June 2016 at 20:19:37 UTC, Walter Bright wrote: Regardless, you're way ahead with DMDScript than starting from scratch. it heavily depends of what exactly one want and how. by the time i'll fully understand DMDScript code, i'd complete my own implementation of dynamic scripting language with JIT.
Re: I'd love to see DScript one day ...
On 6/13/2016 2:21 AM, Chris wrote: Seems like it's a dead race. DMDScript has been in the attic for too long. Trying to play catch up with Google, an Internet company, seems hardly worth the effort. For making your own browser, probably right. For adding scripting support to an IDE, anything with plugins, etc., it's quite good for the job. Regardless, you're way ahead with DMDScript than starting from scratch.
Re: I'd love to see DScript one day ...
On 11-Jun-2016 13:07, yawniek wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript unfortunately this is unmaintained, has no docs and isn't working on linux/os x. having an ecma script implementation that is able to access D data would be very usefull for performant scripting. The only problem I'm aware of is that it's that 64bit are not supported yet. -- Dmitry Olshansky
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 21:32:18 UTC, Ola Fosheim Grøstad wrote: On Sunday, 12 June 2016 at 12:42:23 UTC, Adam D. Ruppe wrote: The standard has moved on, the bar on performance has raised, and dmdscript hasn't even kept up with changes in D. Not too worried about the performance, but EcmaScript 6 has increased the feature set so much that the spec is over twice as large as for EcmaScript 5... And Chrome 52 is apparently shipping with both ES6 and some ES7, so it will be hard to keep up. Seems like it's a dead race. DMDScript has been in the attic for too long. Trying to play catch up with Google, an Internet company, seems hardly worth the effort.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 12:42:23 UTC, Adam D. Ruppe wrote: The standard has moved on, the bar on performance has raised, and dmdscript hasn't even kept up with changes in D. Not too worried about the performance, but EcmaScript 6 has increased the feature set so much that the spec is over twice as large as for EcmaScript 5... And Chrome 52 is apparently shipping with both ES6 and some ES7, so it will be hard to keep up.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 12:05:49 UTC, Ola Fosheim Grøstad wrote: Conformance would be more important so that you can use libraries and compilers that compile to EcmaScript. That's kinda the problem with dmdscript: it is an excellent and conformant implementation of Javascript that performed better than its competitors... many years ago. The standard has moved on, the bar on performance has raised, and dmdscript hasn't even kept up with changes in D. It's still a very good implementation of ECMAScript 3, and with a bit of D stuff, interfacing with it is reasonably easy, but tbh it just isn't something I'd use anymore... my script.d is easier to build and interface for quick toys, and one of the newer JS engines is better at Javascript itself.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 12:30:46 UTC, ketmar wrote: js is fubared from the start. there is no need to follow the broken path. Whatever, the core of JavaScript is close to Self: https://en.wikipedia.org/wiki/Self_(programming_language) JavaScript was just messy in the details. ES6 is fixing a lot of that.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 12:25:00 UTC, Ola Fosheim Grøstad wrote: Well, ES6 is actually reasonably ok. js is fubared from the start. there is no need to follow the broken path.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 12:13:45 UTC, Chris wrote: On Sunday, 12 June 2016 at 11:58:08 UTC, ketmar wrote: On Sunday, 12 June 2016 at 11:23:56 UTC, Chris wrote: I haven't had a chance to look at the source code in detail yet. How hard would it be to integrate JIT and D (and C) interop? not hard. there is extension example (extending engine with D). also, the engine compiles scripts to IR code, which can be JITed as is, or used as a base for some other code representation. the worst thing (for me) is that it is javascript. Yeah, same here! It'd be best to write a proper scripting language based on DMDScript. Well, ES6 is actually reasonably ok. It has both generators and WeakSet (set of objects that is pruned if the object is collected and many many other improvements: http://es6-features.org/
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 11:58:08 UTC, ketmar wrote: On Sunday, 12 June 2016 at 11:23:56 UTC, Chris wrote: I haven't had a chance to look at the source code in detail yet. How hard would it be to integrate JIT and D (and C) interop? not hard. there is extension example (extending engine with D). also, the engine compiles scripts to IR code, which can be JITed as is, or used as a base for some other code representation. the worst thing (for me) is that it is javascript. Yeah, same here! It'd be best to write a proper scripting language based on DMDScript.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 11:23:56 UTC, Chris wrote: I haven't had a chance to look at the source code in detail yet. How hard would it be to integrate JIT and D (and C) interop? Theoretically something like the Derelict-D libs allow for interop with Python and Lua too. I think we could create something really nice and useful here. I don't think JIT is needed to be useful. Conformance would be more important so that you can use libraries and compilers that compile to EcmaScript. But conformance to EcmasScript 6 is a big big big big job. http://www.ecma-international.org/ecma-262/6.0/
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 11:23:56 UTC, Chris wrote: I haven't had a chance to look at the source code in detail yet. How hard would it be to integrate JIT and D (and C) interop? not hard. there is extension example (extending engine with D). also, the engine compiles scripts to IR code, which can be JITed as is, or used as a base for some other code representation. the worst thing (for me) is that it is javascript.
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 10:51:16 UTC, Walter Bright wrote: On 6/12/2016 3:17 AM, Chris wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS > once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript I just tried to compile it on Linux with dmd v2.071.0 in 64bit mode. The compiler emits loads of deprecation warnings concerning module import, Yeah, the import lookup was recently changed. and it seems to be geared towards 32bit: That's right. It should be fixed. Nevertheless, these are minor issues. If you try to create a new script compiler, you're in for a heluva lot more work than fixing some bit rot. Sure, but then we should turn it into a first class citizen and update it with each version of dmd and prevent bit rot. I haven't had a chance to look at the source code in detail yet. How hard would it be to integrate JIT and D (and C) interop? Theoretically something like the Derelict-D libs allow for interop with Python and Lua too. I think we could create something really nice and useful here. By the way, provided we go ahead with this, wouldn't the name DScript be catchier than DMSScript? Although I don't want to bikeshed about the name now.
Re: I'd love to see DScript one day ...
On 6/12/2016 3:17 AM, Chris wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript I just tried to compile it on Linux with dmd v2.071.0 in 64bit mode. The compiler emits loads of deprecation warnings concerning module import, Yeah, the import lookup was recently changed. and it seems to be geared towards 32bit: That's right. It should be fixed. Nevertheless, these are minor issues. If you try to create a new script compiler, you're in for a heluva lot more work than fixing some bit rot.
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote:
On 6/10/2016 3:55 AM, Chris wrote:
> Cool. I'd love to see `DScript` one day - and replace JS once
and for all ...
> well ... just day dreaming ...
Dreams are reality:
https://github.com/DigitalMars/DMDScript
I just tried to compile it on Linux with dmd v2.071.0 in 64bit
mode. The compiler emits loads of deprecation warnings concerning
module import, and it seems to be geared towards 32bit:
[...]
engine/source/dmdscript/dstring.d(953,16): Deprecation: module
std.string is not accessible here, perhaps add 'static import
std.string;'
engine/source/dmdscript/expression.d(216,9): Warning: statement
is not reachable
engine/source/dmdscript/expression.d(306,28): Deprecation: module
std.ascii is not accessible here, perhaps add 'static import
std.ascii;'
engine/source/dmdscript/expression.d(318,9): Error: static assert
(8LU == 4LU) is false
dmd failed with exit code 1.
The error message refers to:
`override void toIR(IRstate *irs, uint ret)
{
static assert((Identifier*).sizeof == uint.sizeof); //
<==
if(ret)
{
uint u = cast(uint)Identifier.build(string);
irs.gen2(loc, IRstring, ret, u);
}
}
`
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 08:32:10 UTC, Dicebot wrote: On 06/11/2016 01:01 AM, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: Cool. I'd love to see `DScript` one day - and replace JS once and for all ... well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript On a related topic - has anyone looked into what needs to be done to support D + WebAssembly combo? WebAssembly is currently close to asm.js. In fact I believe WebAssembly is distilled from the asm.js backend. Single threaded, no garbage collection support.
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... My crazy dream (or cool idea :): to use the DMD interpreter as the script engine!
Re: I'd love to see DScript one day ...
On 06/11/2016 01:01 AM, Walter Bright wrote: > On 6/10/2016 3:55 AM, Chris wrote: >> Cool. I'd love to see `DScript` one day - and replace JS once and for > all ... >> well ... just day dreaming ... > > Dreams are reality: > > https://github.com/DigitalMars/DMDScript On a related topic - has anyone looked into what needs to be done to support D + WebAssembly combo?
Re: I'd love to see DScript one day ...
On Sunday, 12 June 2016 at 00:37:50 UTC, Jonathan Marler wrote: On Saturday, 11 June 2016 at 10:34:41 UTC, Chris wrote: On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: [...] No, JS is here to stay, unfortunately. Good news, take a look at asmjs.org In the future you'll be able to compile your D code into javascript assembly, and run it in the browser. I predict LLVM will have a javascript assembly backend, so LDC would get this for free. Pretty cool stuff going on in the web world right now. Emscripten has been able to compile C++ to asm.js for a while - uses modified llvm back end and ported C runtime. I guess web assembly will take off, but somebody would still need to port the runtime for D and integrate with new llvm wasm back end when it is stable. See binaryem in meantime. I am not sure anyone is working on this for D currently, but we might put this up somewhere as a project in case volunteer comes along. Then when that's done still need to write a library to do web things like provide access to DOM from D. Emscripten already looks quite nice, and I was so wanting to be able to use it for serious work - it just doesn't seem to be there just yet. But I agree with your basic point, and it would be great not to have to write js.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 10:34:41 UTC, Chris wrote: On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: [...] No, JS is here to stay, unfortunately. Good news, take a look at asmjs.org In the future you'll be able to compile your D code into javascript assembly, and run it in the browser. I predict LLVM will have a javascript assembly backend, so LDC would get this for free. Pretty cool stuff going on in the web world right now.
Re: I'd love to see DScript one day ...
Having DScript running in the browser and D on the server would be an absolute dream to develop with. It might also turn a few heads our way.
Re: I'd love to see DScript one day ...
On 6/11/2016 3:07 AM, yawniek wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript unfortunately this is unmaintained, It was working last year, when I got it up on Dub. Dmitry ported it to D2 a while ago. There has been a lack of interest in it, and so it doesn't get exercised much. But it is a fully functional implementation of Javascript. For anyone wanting to add a scripting language to, say, an IDE written in D, it would be ideal. It would also be ideal as a starting point if one wanted to invent yet another scripting language. has no docs There is not much to document. Feed it a Javascript source file and it executes it. and isn't working on linux/os x. http://bugzilla.digitalmars.com/issues/buglist.cgi?quicksearch=. It has in the past. Probably something trivial. having an ecma script implementation that is able to access D data would be very usefull for performant scripting. Pull requests are welcome.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 18:36:23 UTC, Ola Fosheim Grøstad wrote: On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: I have given up hope for different browser languages than Javascript. The problem is not finding a language. A lot of people would love to have Lua, Python, or something else and it has not happened yet. Dart was the most serious attempt. They had full-time people at Google working on it and they gave up. They didn't give up as they use it internally for their advertising application. Dart 1.7 was out 2 days ago: http://news.dartlang.org/2016/06/dart-117-more-performance-improvements.html They gave up last year [0], when they abandoned Chrome-integration: “In order to do what’s best for our users and the web, and not just Google Chrome, we will focus our web efforts on compiling Dart to JavaScript,” Dart co-founders Lars Bak and Kasper Lund wrote today. “We have decided not to integrate the Dart VM into Chrome.” [0] http://news.dartlang.org/2015/03/dart-for-entire-web.html
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 17:18:59 UTC, Observer wrote: On Saturday, 11 June 2016 at 09:23:55 UTC, Chris wrote: However, with `DScript` I meant a new scripting language that can draw on the power of D, not necessarily a re-implementation of JS. Adam[1] and ketmar[2] have already worked on D based scripting languages. I wonder, if there is interest in creating a D-based (not `debased`) scripting language. D has features that are perfect for scripting or DS languages. Why not try something new? Having a child language might also help with D development in general, who knows. I'm confused. I thought I read in TDPL (don't have it handy here to check) that you could just slap a shebang line on a D program and you then have a runnable script. Or something very close to that construction. So don't we already have what want? this way (she bang rdmd) is for turning a module into something like a shell script. You cannot use it dynamically in an application e.g with binding to the application variables and to the applications functions (like LUA in game or JS in web browsers).
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 18:36:23 UTC, Ola Fosheim Grøstad wrote: On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: I have given up hope for different browser languages than Javascript. The problem is not finding a language. A lot of people would love to have Lua, Python, or something else and it has not happened yet. Dart was the most serious attempt. They had full-time people at Google working on it and they gave up. They didn't give up as they use it internally for their advertising application. Dart 1.7 was out 2 days ago: I meant Dart 1.17... Duh. Interestingly there was an interview related to Dart and Adwords UI http://news.dartlang.org/2016/03/the-new-adwords-ui-uses-dart-we-asked.html I found these quotes interesting: «One thing we found out was that developers preferred even stronger type checking than what Dart was providing. So, we are very excited about the work on Dart Strong Mode. We are also looking forward to cross-browser, fast edit refresh with the upcoming Dart Dev Compiler.» and «With Angular2, the application performance and code size are very similar between the JS and Dart versions. So any team considering Angular2 should consider Dart as well.» Dunno. Still not sure about the future of Dart, but stricter static typing is a good selling point.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: I have given up hope for different browser languages than Javascript. The problem is not finding a language. A lot of people would love to have Lua, Python, or something else and it has not happened yet. Dart was the most serious attempt. They had full-time people at Google working on it and they gave up. They didn't give up as they use it internally for their advertising application. Dart 1.7 was out 2 days ago: http://news.dartlang.org/2016/06/dart-117-more-performance-improvements.html I think most people outside Google has given up though. I've picked up Typescript and I suspect many others have as well. Typescript is close enough to Javascript to make debugging reasonable.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 09:23:55 UTC, Chris wrote: However, with `DScript` I meant a new scripting language that can draw on the power of D, not necessarily a re-implementation of JS. Adam[1] and ketmar[2] have already worked on D based scripting languages. I wonder, if there is interest in creating a D-based (not `debased`) scripting language. D has features that are perfect for scripting or DS languages. Why not try something new? Having a child language might also help with D development in general, who knows. I'm confused. I thought I read in TDPL (don't have it handy here to check) that you could just slap a shebang line on a D program and you then have a runnable script. Or something very close to that construction. So don't we already have what want?
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 13:41:32 UTC, ketmar wrote: i'd also say: "JIT it from the start!" targeting LibJIT is very easy, and you will have nice machine code for x86, x86_64 and ARM, plus IR interpreter for any other "unsupported" arch. Yes, I was thinking along the same lines. JIT from the start.
Re: I'd love to see DScript one day ...
p.p.s. i'm slowly working on my own SSA-based codegen library too. no timelines are set, tho.
Re: I'd love to see DScript one day ...
i'd also say: "JIT it from the start!" targeting LibJIT is very easy, and you will have nice machine code for x86, x86_64 and ARM, plus IR interpreter for any other "unsupported" arch.
Re: I'd love to see DScript one day ...
p.s. i did a D wrapper for LibJIT, it is in DACS repo. i also talked with LibJIT maintainer, and that wrapper may be included in next LibJIT release.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 10:14:25 UTC, qznc wrote: Do you want a scripting language in general or a browser language? I have given up hope for different browser languages than Javascript. The problem is not finding a language. A lot of people would love to have Lua, Python, or something else and it has not happened yet. Dart was the most serious attempt. They had full-time people at Google working on it and they gave up. The problem is that Javascript is tightly coupled to the DOM data structures. Disentangling that und inserting a layer in between there is the hard part. Oh, and you have to do this for all the major browser engines. No, JS is here to stay, unfortunately. If you just want to have a general purpose scripting language (e.g. for a game) you can use Lua, Javascript, TCL, and others already. How would you improve upon them? Yep, something like that. We could look into what a scripting language based on D could offer that others can't - based on D's features. If you have an easy to use scripting language, people in science and data processing could use it for fast development. If speed is crucial, modules could be written in D or even C[1]. Just because we already have Python and Lua doesn't mean we shouldn't try to create another one. [1] Interfacing to C in Lua and Python is a bit annoying. `DScript` could offer both access to D and C. In fact, `DScript` could also interact with Lua and Python (cf. PyD and LuaD).
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 09:23:55 UTC, Chris wrote: However, with `DScript` I meant a new scripting language that can draw on the power of D, not necessarily a re-implementation of JS. Adam[1] and ketmar[2] have already worked on D based scripting languages. I wonder, if there is interest in creating a D-based (not `debased`) scripting language. D has features that are perfect for scripting or DS languages. Why not try something new? Having a child language might also help with D development in general, who knows. The source for the old MiniD language[1] is still available at DSource. Jarret moved on and converted it to C++ as Croc[2], but it might be an interesting project to get the old MiniD code compiling again as a starting point. [1] http://www.dsource.org/projects/minid/browser/trunk [2] https://github.com/JarrettBillingsley/Croc
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 09:23:55 UTC, Chris wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript Thanks for opening a new thread about this topic. Does DMDScript work with the latest version of D2? I once wanted to use DMDScript as a server side JS with vibe.d, but there were issues concerning versions of D etc. so I dropped the idea. However, with `DScript` I meant a new scripting language that can draw on the power of D, not necessarily a re-implementation of JS. Adam[1] and ketmar[2] have already worked on D based scripting languages. I wonder, if there is interest in creating a D-based (not `debased`) scripting language. D has features that are perfect for scripting or DS languages. Why not try something new? Having a child language might also help with D development in general, who knows. Do you want a scripting language in general or a browser language? I have given up hope for different browser languages than Javascript. The problem is not finding a language. A lot of people would love to have Lua, Python, or something else and it has not happened yet. Dart was the most serious attempt. They had full-time people at Google working on it and they gave up. The problem is that Javascript is tightly coupled to the DOM data structures. Disentangling that und inserting a layer in between there is the hard part. Oh, and you have to do this for all the major browser engines. If you just want to have a general purpose scripting language (e.g. for a game) you can use Lua, Javascript, TCL, and others already. How would you improve upon them?
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript unfortunately this is unmaintained, has no docs and isn't working on linux/os x. having an ecma script implementation that is able to access D data would be very usefull for performant scripting.
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript Thanks for opening a new thread about this topic. Does DMDScript work with the latest version of D2? I once wanted to use DMDScript as a server side JS with vibe.d, but there were issues concerning versions of D etc. so I dropped the idea. However, with `DScript` I meant a new scripting language that can draw on the power of D, not necessarily a re-implementation of JS. Adam[1] and ketmar[2] have already worked on D based scripting languages. I wonder, if there is interest in creating a D-based (not `debased`) scripting language. D has features that are perfect for scripting or DS languages. Why not try something new? Having a child language might also help with D development in general, who knows. [1] Adam https://github.com/adamdruppe/arsd (jsvar.d, script.d) [2] ketmar http://repo.or.cz/dacs.git https://www.gnu.org/software/libjit/ http://repo.or.cz/gaemu.git
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript You have 2 readme files, so on github it looks like you have only one very short one (README.md) and that makes a poor impression.
Re: I'd love to see DScript one day ...
On Saturday, 11 June 2016 at 04:13:26 UTC, Jack Stouffer wrote: On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: Cool. I'd love to see `DScript` one day - and replace JS once and for all ... well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript I have a feeling this will end up like your compiled Java story. Just my pessimism speaking. Depends on whether it is conforming to edition 6. Modern frameworks use edition 6 or 5, and edition 6 may become ISO/IEC 16262:2016. For Go a similar project seems to be the most popular interpreter project: https://github.com/robertkrimen/otto Benchmarks, conformance comparisons?
Re: I'd love to see DScript one day ...
On Friday, 10 June 2016 at 22:01:53 UTC, Walter Bright wrote: On 6/10/2016 3:55 AM, Chris wrote: Cool. I'd love to see `DScript` one day - and replace JS once and for all ... well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript I have a feeling this will end up like your compiled Java story. Just my pessimism speaking.
I'd love to see DScript one day ...
On 6/10/2016 3:55 AM, Chris wrote: > Cool. I'd love to see `DScript` one day - and replace JS once and for all ... > well ... just day dreaming ... Dreams are reality: https://github.com/DigitalMars/DMDScript
