New shiny toy
-
Haha, what? How the fuck does that follow?
Oh well, I knew you wouldn't like that statement.
Microsoft didn't even release C# until the IDE was ready to go.
That's a marketing decision. Rust made a different one that makes sense as well, whether you like it or not: Publish the compiler and build tool first, so that some people can start using the language, then build the rest of the tools around them to make the language available to everyone.
-
Oh well, I knew you wouldn't like that statement.
It's not that I don't like it, it's that I don't even get how the one thing relates to the other. It's like saying, "Seattle has great Thai restaurants, therefore the Express buses always run 14 minutes late." It's a complete non-sequitur.
That's a marketing decision.
No it's not; it's a "releasing a quality product" decision. Something open source developers don't give any shits about.
Rust made a different one that makes sense as well, whether you like it or not: Publish the compiler and build tool first, so that some people can start using the language, then build the rest of the tools around them to make the language available to everyone.
RELEASE EARLY RELEASE OFTEN GUYSSSS!!! Make sure your users use all the most broken shit we have! Break their code all the fucking time! THIS IS OPEN SOURCE PHILOSPHY BROKEBN SHIT ALL THE YITME QUALITY IS FOR SUCCKSKS!
Yes, you're right, I don't like that fucking retarded idea/philosophy. And I'm not sure I'd want to use a programming language from people stupid enough to actually follow it. Even if it had decent tooling, which it never will because OPEN SORUCE GUYZZ!!
-
I don't even get how the one thing relates to the other.
Someone who takes pride in his work and makes sure to polish X will likely also put effort into Y, especially if he states it's important for him.
No it's not; it's a "releasing a quality product" decision.
When you start calling something 1.0 is a pure marketing decision. They decided to call it 1.0 when the language was stable. They know their product is not finished. But they know parts of their product are already usable and people want to use them, so they released those parts.
RELEASE EARLY RELEASE OFTEN GUYSSSS!!! [… rant about OSS quality …]
Bullshit. That argument would be valid if they were releasing half-baked IDE support and claimed it was usable. Which they clearly didn't.
-
Microsoft didn't even release C# until the IDE was ready to go.
I don't care how many websites they have, their priorities are ass-backwards and upside-down in my opinion. Until this great IDE actually exists, it's all just promises and strange asdf logic.
"These languages suck because their open source developers don't understand the importance of IDEs"
"Well these developers have actually written a page on how important IDEs are and they say it will be their main focus"
"Whatever fuck you"
-
When you start calling something 1.0 is a pure marketing decision.
Calling something 1.0 means it is stable, feature-complete, and ready for the intended users/general public to use.They know their product is not finished.
Version 1.0 was finished, and that's what they released. They then moved on to version 2.0.
-
Calling something 1.0 means it is stable, feature-complete, and ready for the intended users/general public to use.
Did they release an IDE plugin? No, they didn't. What they actually released was feature-complete, is promised to be stable and people wanted to use it.
And no, that's not
What you bundle into release 1.0 IS a marketing decision. Even if you originally planned features X, Y, and Z, it's a perfectly valid decision to release only X and Y in the 1.0, and ship Z later, if there are customers who are willing to buy the product with only X and Y.
And don't tell me something about "priorities backwards". Every language ever has been developed in the following way:
- Write initial spec
- Stabilize compiler and basic library
- Complete standard library and write tools (like IDE plugins).
-
Calling something 1.0 means it is stable, feature-complete, and ready for the intended users/general public to use.that's what it's supposed to mean, yes.
that's rarely what happens in commercial products in my experience.
-
Calling something 1.0 means it is stable, feature-complete, and ready for the intended users/general public to use.
IME version 1 means at best that it's "sort of good enough".
-
IME version 1 means at best that it's "sort of good enough".
In the context of an API/tool, it's supposed to mean "good enough that we can promise not to break the API, because there are no major design flaws left". Sometimes, as in this case, also "you can start using it now even if you don't hate yourself".
-
Someone who takes pride in his work and makes sure to polish X will likely also put effort into Y, especially if he states it's important for him.
A person who thinks the CLI program is polished may not even have the skills to be able to determine whether the GUI program is polished or not. There's almost zero relationship between the two things. Hell, look at the delusional people in the various Git threads we have around here who assure me that SourceTree is a good GUI.
So while your statement is correct more-or-less, we don't know if he releases the software when it's polished or what he "feels" it's polished. If the latter, the success of the CLI program has absolutely no bearing on the success of the standalone IDE, should it ever exist which it probably will not.
When you start calling something 1.0 is a pure marketing decision. They decided to call it 1.0 when the language was stable. They know their product is not finished.
Right; it's just a dumb one. "Hey, Rust is 1.0! It's ready to go!" "Ok, where do I download the IDE?" "We don't have one." "But you said it's 1.0." "It is, but it's not done yet. We have a whole webpage about how we don't consider it done until the IDE is done, but we made the language 1.0." "So you're liars, then?" "Yup! Guess so!"
Bullshit. That argument would be valid if they were releasing half-baked IDE support and claimed it was usable. Which they clearly didn't.
True; your hypothetical situation would be worse.
"Well these developers have actually written a page on how important IDEs are and they say it will be their main focus"
Anybody can put up a website about anything. It's just words. Words don't matter; only actions do.
Did they release an IDE plugin? No, they didn't. What they actually released was feature-complete, is promised to be stable and people wanted to use it.
But you said they have a whole website where they say having an IDE is a huge important priority to them. But not important enough to bother finishing before they call the product 1.0? Again: this "marketing strategy" (in my eyes) makes me want to call them liars. Not applaud their efforts.
(The reality is I'm guessing there is no marketing strategy at play here, and the website is just feel-good gibberish about an IDE nobody's actually working on, and they're doing the typical open source release early release often bullshit without even bothering to engage their brain cells and think about it. And asdf is covering for them.)
And don't tell me something about "priorities backwards". Every language ever has been developed in the following way:
Write initial spec
Stabilize compiler and basic library
Complete standard library and write tools (like IDE plugins).- That's not even slightly true. VBA wasn't written that way. Neither was HyperTalk or AppleScript. Or Microsoft Basic on Mac. Hell, JavaScript. Or probably a million other examples if I took the time to think of them.
- That IS the backwards way.
Look, it's simple. Answer this question:
- What is the user experience of using Rust?
The answer, as-is, is: "we have absolutely NO FUCKING IDEA". Some people are maybe using Vim to write their Rust. Maybe some people are using SublimeText. Maybe some people are using Notepad++. Some people might be building the project by creating a SublimeText plugin, others might be using PowerShell, others might be using Bash.
You don't even KNOW what the user experience is. Needless-to-say, you also don't control the user experience.
If they gave even the SLIGHTEST care for usability, at the bare minimum they'd want to have some form of control over the user experience. But they've already told the world: "nope! We don't care! No shits given here!"
But now you're in this forum saying, "but guyz, but guyz but guyz, it's coming soon! Coming soon! Coming soon!" Well that's great, but we don't need to see this hypothetical IDE to know that the creators of Rust have put absolutely ZERO thought into what the user experience of their language was. Because they've released it to the public.
This is the Git thing all over again. These open source-y people build about a third of a product, then slap a 1.0 on it and pretend it's done forever. No, it's not done. It's barely even started. You haven't even done a working lunch and discussed what you want the user experience to be like, as is obvious to everybody who's been forced to actually use the broken thing.
So anyway, it's not a big deal, whatever. I wasn't even angry about it until asdf started making promises that are almost guaranteed to be broken. (Someone set a calendar item to come back to this thread in a year and show me this great Rust IDE.) Sadly, in the current climate, I just expect software to be released when only a third done.
-
I get your point, even if I don't fully agree, but:
This is the Git thing all over again.
The major difference is that the core Git developers don't even care about UX. They don't even want to ship a usable GUI, they don't care enough to provide a good library and even the command-line interface is confusing and inconsistent (
-avs-A,~vs^, etc.).Rust, OTOH, seems to have some developers who actually care about tooling. So far, I've never needed to execute a weird series of arcane commands for basic tasks. It was always "type
cargo xy, which will do the obvious thing and you're done".
-
The major difference is that the core Git developers don't even care about UX.
Since the creators of Rust have release 1.0 of their product without thinking about UX, I don't see the difference.
Like I said above, it's not even like they've released their idea of what the UX should be and it's bad. They've released something that shows they don't even care what the UX is.
Say what you want about, say, Borland Builder, but at least when Borland released Builder, they were saying to the world: "hey world, here's the best way we can think of to write GUI apps in C++. The whole burrito, back to front, here in one package."
So far, I've never needed to execute a weird series of arcane commands for basic tasks. It was always "type cargo xy,
Yeah; but you're also delusional.
cargo xyIS an arcane command.
-
-
Like I said above, it's not even like they've released their idea of what the UX should be and it's bad. They've released something that shows they don't even care what the UX is.
Gotta agree on this.
Let's put aside the vanishingly small number of languages that are so highly specialized that they literally are better than anything else for the job. COBOL, let's say. Okay, all that's aside.
What you're left with is this huge group of Programming Languages. And they all do exactly the same thing. They take some sort of syntax, and turn it into something compiled for the CPU to do. There may be some squabbly differences, but 99.9% of a programmer's interaction will be in the IDE, writing that syntax.
So if you don't have an IDE, you don't have a language. Maybe Rust does something slightly better than c#. But C# has an IDE that reduces the amount of dumbshit work I have to do by orders of magnitude. Therefore, any language I can use Visual Studio for is better than Rust.
How you program in that language is just as important-- if not more so-- than the language itself. If I can create an entire full-fledged application in Language A in the same time it takes me to even figure out how to set up an environment to write in Language B, then fuck Language B until it's ready.
-
So we'll have to stick with C# for quite some time then?
-
@Lorne_Kates said:
But C# has an IDE that reduces the amount of dumbshit work I have to do by orders of magnitude.
That's kinda the point of most modern languages as well: Reduce the amount of dumbshit typing and complexity and debugging by making smart language design choices that eliminate bugs and reduce the necessary amount of code.
BTW: I didn't bring up the comparison with C# and I'm not even sure Rust is trying to compete with it. I only compared it to Go, which is definitely worse.
-
<HTML> Hello world. </HTML>
-
IMO you should either go full backend render or full SPA. All these MVC applications that also use ajax to load up dialog views and duplicate code in 3 different languages are the devil.
No more client side validation!!!
-
``` <HTML> Hello world. </HTML> ```
Errors found while checking this document as HTML 4.01 Transitional!
3 Errors, 3 warning(s)- Unable to Determine Parse Mode!
- No DOCTYPE found! Checking with default HTML 4.01 Transitional Document Type.
- No Character encoding declared at document level.
- Line 1, Column 1: no document type declaration; implying "<!DOCTYPE HTML SYSTEM>"
- Line 2, Column 1: character data is not allowed here
- Line 3, Column 7: end tag for "HTML" which is not finished
0/10 see me after class.
-
Kinda why I find HTML to be a frustrating technology.
-
Reduce the amount of dumbshit typing and complexity and debugging by making smart language design choices that eliminate bugs and reduce the necessary amount of code.
Yes, but we were talking about Rust.
-
Google search.
Proper HTML HelloWorld
Result
HTML
not a programming language.
That's not a sufficient reason to exclude it. Anyhow, ColdFusion is based around it, thus one can "print" just by stating the string:
Hello WorldHard to beat that.
-
Introduction to HTML
W3Schools offers free online tutorials, references and exercises in all the major languages of the web. Covering popular subjects like HTML, CSS, JavaScript, Python, SQL, Java, and many, many more.
<!DOCTYPE html> <html> <head> <title>Page Title</title> </head> <body> <h1>My First Heading</h1> <p>My first paragraph.</p> </body> </html>
-
IMO HTML is stupid and should be a binary language you compile to.
But then I've always had some hate for text-based languages in general.
-
text-based languages
...It's a markup language. For marking up text. Not sure how you'd do that without text, exactly.
-
So we'll have to stick with C# for quite some time then?
I hope these open source language designers eventually get a clue and we get a plethora of high-quality super-efficient languages and accompanying tools for each one.
Looking unlikely though.
-
I meant for languages that the computer has to read, not languages that you write in.
You could write in an HTML-like text language, then let the computer compile to a "binary HTML" file (or it could be text-based, but there's no reason to do that). Now you can change one side of the language without touching the other one, and the developer never has to see a doctype.
Having the language you write in be the same language the computer at the other end reads means that language is both an user interface for the page creator and a file format for the computer to interpret, so it has to do two different things at once, and will inevitably have to compromise somewhere in the middle.
With all the javascript and CSS minimizers and generators, you pretty much do this already, so why not go all the way?
-
Compile HTML to binary? OK. But only if you redesign all the browser dev tools.
-
The binary format could have the same concepts as current HTML (divs, spans, styles, attributes etc) so the development tools would require very few changes.
Or, realistically, it could just be a different serialization of HTML5 so it wouldn't require any changes at all to the browser other than a new parser.
Always reuse your old standards when making a new one
(I think we're going a bit off-topic here?)
-
Compile HTML to binary?
HTML -> Server-side rendering engine
Server-side rendering engine -> ImageMagik
You'll just need to emit Javascript to handle clicking the image, so that if someone clicks on a picture of a hyperlink, it will properly redirect via document.location. You can obfuscate that javascript enough that it might as well be binary.
(Bonus points if the click x/y is sent server-side and the URL is returned)
-
You laugh, but how do you think some people are "porting" desktop apps to the web?
Run the app on the server, stream video to the browser.
-
You laugh, but how do you think some people are "porting" desktop apps to the web?
Run the app on the server, stream video to the browser.
=(
-
>HTML
not a programming language.
That's not a sufficient reason to exclude it. Anyhow, ColdFusion is based around it, thus one can "print" just by stating the string:
Hello WorldHard to beat that.
Lol @ quote bug.Isn't there a language where printing Hello World is a one character instruction?
Edit: it's HQ9+
-
I do realize I'm a month late but I feel like this definitely qualifies as a shiny toy and I wanted to post it:
It might seem like yet another Scratch clone to teach programming to children. But:
- When you try it you realize the UI "library" is so weird that only an expert programmer can figure it out (doh!). Try to make a simple calculator app in it and you'll see what I mean. It's generally more suited to making games.
- It has a buttload of advanced features: debugging, cloud functions, libraries, OOP, object decorators, asynchronous functions, unit tests, accessing phone features (camera, calls, acccelerometers, NFC) through Apache Cordova, run code in Arduino, run code in Minecraft, export to Azure web app...
- It runs in the browser, and it even (mostly) works offline. So if you're ever stuck somewhere with just your phone and feel the urgent need to write a "99 bottles of beer" program in a new language, now you can.
-
I suspect that consuming the 99 bottles of beer would be widely considered a better option when stuck somewhere.
-
