51–60 of 81 posts
I don't see how you can really use this without already being very familiar with Javascript. While you're writing your code you'd need to be constantly aware of how the code will be translated to js (so you can use the correct attributes and define datatypes correctly), and you'd also need to be aware of avoiding. NET library classes that can't be used by the translated js. It seems easier to just learn to manage lar…
Looks impressive! I'm curious how you're doing the parsing. I see that you reference both NRefactory (which reads C#) and Cecil (which reads IL) -- can you talk about how are each of these used and how you combine them? Also, did you consider using Roslyn (Microsoft's official C# parser)? Obviously it's still beta, but once it's complete would it remove the need for Cecil and NRefactory completely?
Earlier quoted context omitted.
No, one should limit the level of abstraction to the fastest one that is reliable enough to get the job done. There's nothing intrinsic about high abstraction layers that say they must suck for debugging what you need debugged. It just depends on what you need to do with it. If it gets your job done faster and it's reliable enough for your job, then why not? That's why higher abstractions are created in the first pla…
Yes no limits to abstractions always works perfectly.. Check out this stack trace: http://www.ibm.com/developerworks/java/library/os-ecspy1/sta... Now imagine that across three different languages.
I don't see how you can really use this without already being very familiar with Javascript. While you're writing your code you'd need to be constantly aware of how the code will be translated to js (so you can use the correct attributes and define datatypes correctly), and you'd also need to be aware of avoiding. NET library classes that can't be used by the translated js. It seems easier to just learn to manage lar…
Another thing static typing gives me that I really wish I had in JavaScript, is (almost) worry free refactoring.
In Java, I can extract methods, constants, delegates, change method signatures, rename members, fix typos in code, and never to worry I copy pasted something wrong.
I have yet to find such a way to do it safely in JavaScript / CoffeScript or any other dynamically typed language. (feel free to point if I'm missing something, I said I'm not that a great of a developer)
So maybe I'm just not good enough as a developer, but I like IDEs to protect me, perhaps I'm not smart enough to use VIM / Emacs / TextMate, and I can't memorize APIs. Static typing + functional features make me the most productive. So C# / Scala / Java 8 javascript compilers are on the top of my wish list.
Looks impressive! I'm curious how you're doing the parsing. I see that you reference both NRefactory (which reads C#) and Cecil (which reads IL) -- can you talk about how are each of these used and how you combine them? Also, did you consider using Roslyn (Microsoft's official C# parser)? Obviously it's still beta, but once it's complete would it remove the need for Cecil and NRefactory completely?
Why would I want to use the proprietary Roslyn when I have a working solution with the open-source NRefactory and Cecil?
Earlier quoted context omitted.
I don't see the problem. - If I had written Javascript, I would have to debug the Javascript that is running in the browser. - If a tool writes Javascript for me, I have to debug Javascript running in the browser. A better wish would be: I hope that your tool outputs nicely formatted and easy to debug Javascript.
But it will not write JavaScript the way you would write it - and (avoiding arguments about personality in code) that JavaScript will be way harder to follow and debug.
The first argument is that "the style is different". I use a js generator language (not this one, I don't want to shift attention), and it does "nice" things that would be a pain for me to do manually... like setting up namespaces for code, spacing everything perfectly. Even if the output is not to your taste, at least you know that the style will be 100% consistent in style and implementation. I never have problems debugging.
The other argument is that "the generated code is complex/ugly". Unfortunately, "fast" code is often ugly. Static compilers can do some optimizations by default (e.g. optimized for loops), and you only need to worry about the "pretty" code at the higher level of abstraction. I doubt you will ever have to debug one of these optimizations, any decent compiler should get them right. You merely have to understand why they were used, which often makes you a better programmer to boot.
I don't see how you can really use this without already being very familiar with Javascript. While you're writing your code you'd need to be constantly aware of how the code will be translated to js (so you can use the correct attributes and define datatypes correctly), and you'd also need to be aware of avoiding. NET library classes that can't be used by the translated js. It seems easier to just learn to manage lar…
The reason I'm looking for either C# / Scala JS compiler is that I'm a static typing fan, I am an old school developer who likes IDEs and like discovering API by pressing ctrl+space, instead of mentally knowing it or constantly checking the API docs, I know it's against the trend, and not considered cool, but this gets me working software out the door. I'm much more productive and have much less bugs while writing ol…
Regarding refactoring, it's hardly worry free for me. I always have to double check what VS is going to do when I use its refactoring tools, because sometimes it makes a significant mistake. Now, that's often caused by the way I wrote some code and I have to fix it up as part of my refactoring, but I would have discovered that while refactoring manually too, and if I trusted VS I wouldn't have discovered it at all. So VS helps a bit with the typing, but it's not much faster than I'd be without it. A lot of that is probably because I learned to refactor using vim, grep, sed, and other Unix command-line tools, so I have no problem doing large-scale refactoring changes touching thousands of lines over a 100k+ line codebase. Extracting a method is childs play in comparison, and I don't mind the extra few seconds it takes to do it manually exactly the way I want it.
There is definitely a need for better javascript editing tools. VS, with the right plugins, wouldn't be bad, but I'm not sure Intellisense can really cut it. When I'm putting some code into an MVC template or js file and I want to refer to a jQuery plugin I know is loaded by my layout template, there is no way for Intellisense to know about that plugin or its API, so it can't help me. In contrast if I load my app in Firefox and use Firebug's console and debugger, I can type in some code and it's able to show me the API in real-time because it's reflecting the live javascript objects as I type. That's the kind of intellisense I want.
I don't see how you can really use this without already being very familiar with Javascript. While you're writing your code you'd need to be constantly aware of how the code will be translated to js (so you can use the correct attributes and define datatypes correctly), and you'd also need to be aware of avoiding. NET library classes that can't be used by the translated js. It seems easier to just learn to manage lar…
I think what you're saying is a given...
I've had to work with a bunch of C, C++, Java, Perl, and C# developers who were good at backend stuff and just started to learn HTML, CSS, and Javascript, and they're mostly pretty awful at it until they give in and dedicate themselves to the frontend web developer platform and environment. Prior to that they'd grasp at tools like this one that seem to offer hope that they won't have to learn that nasty javascript.
How does this compare to http://sharpkit.net (free for small and open source projects) Feature wise? SharpKit seems more complete (language support and has source map debugging support on Chrome). Also support for all popular JS libraries.
When it comes to the language support, I think Saltarelle is slightly ahead, as it supports goto and yield (I don't count SharpKit's implementation of yield which translate 'yield return "x"' to '$yield.push("x")' and then returns the $yield array).