This thread looks to be a little on the old side and therefore may no longer be relevant. Please see if there is a newer thread on the subject and ensure you're using the most recent build of any software if your question regards a particular product.
This thread has been locked and is no longer accepting new posts, if you have a question regarding this topic please email us at support@mindscape.co.nz
|
I have some fairly large SASS files, which everytime I save takes forever to compile. Visual Studio save icon won't go away. Takes up to 5 minutes. Is there something I can check to fix this? |
|
|
Hi, Could you let us know what version of the extension you're using? We had issues like this a few months ago and resolved them with an async build process (you can check the existing version and if there are updates in the Visual Studio Extension dialog). A quick note on the extension manager: When you go to the updates tab it takes about 20 seconds to actually check if there are updates. Don't be fooled by their initial state of 'No updates found'! :-) If you still have issues, could you attach your SASS files? If they are sensitive please send them to support@mindscape.co.nz. Thank you, John-Daniel |
|
|
I think it relates to this problem |
|
|
Yes, we think your diagnosis is correct -- it does seem to be related to the same issue. We don't have a workaround in your case unfortunately but we are investigating to see if there is anything we can do to fix it. |
|
|
Hey guys, this is becoming a bigger issue now that I've started to use the SASS version of Twitter Bootstrap. On a Core i5 2500k, it takes about 15-20s to compile my styles. The C# rewrite of IronRuby is supposed to be faster: http://ironjs.wordpress.com/2012/04/19/why-not-f/ I don't have Pro, so I can't test and see if the LESS Bootstrap is better performing. Is this more of a SASS-compiler issue (i.e. is it also slow in Ruby?)? |
|
|
That link refers to IronJS, not to IronRuby. We've investigated tip versions of IronRuby and they don't address the regex performance issue. We are evaluating a performance enhancement for the Sass compiler at the moment and will update this thread shortly. The Less compiler runs under the V8 JavaScript engine and its performance characteristics are entirely different from those of Sass on IronRuby. Our experience is that compiling Twitter Bootstrap using Less takes a couple of seconds. |
|
|
Subkamran, which of the Sass Twitter Bootstrap ports are you using? I'd like to evaluate performance with and without our candidate fix, but there seem to be about half a dozen Sass ports...! |
|
|
It is sort of amazing how much time I waste waiting for CSS to compile now. I'll just leave this here. https://github.com/hcatlin/libsass "Sass was originally created by the co-creator of this library, Hampton Catlin (@hcatlin)." It only supports the modern syntax (not the braceless one) but I bet it can transform 1000 lines of text in under 10 seconds. Edit: here's the wrapper that calls the library and a 1.8MB test file "huge.scss" that sets up an entertaining "time spent compiling, before-and-after" blog post. |
|
|
I'm using this one: https://github.com/jlong/sass-twitter-bootstrap Also nice post about the C/C++ compiler... maybe that would be better than IronRuby; I hate the braceless syntax so no loss for me. |
|
|
I believe I heard on of .NET podcasts (Herding Code?)that you guys have a Sass parser written in F# for intellisense. Would you be able to extend this solution so you can offer e2e Sass compiler that is written in managed code? I can imagine it will be much faster than Ruby regexes. |
|
|
Hi PawelPabich, Sorry for not replying to you on twitter about this -- your tweet triggered a fair bit of discussion and I totally missed coming back on there with the outcome. We're evaluating different options as yes, Ruby is slow on Windows (to be fair it's pretty slow everywhere, Windows particularly so. The usual Ruby story of 'cache all the things!' doesn't really work when you're talking about compiling a file). What we're reluctant to do is write our own port, or to use a port that looks relatively unstable. The reasons are:
I hope this transparency in our decision making helps and we really appreciate you all making suggestions! We'll get it quicker, we're just deciding what path to take. I hope that helps, John-Daniel |
|
|
Thanks for the explanation. This might be more of a reason to switch to LESS, since it's so much faster. The compilation time is killing my productivity and since I'm working on my own app, it wouldn't be much of a pain to switch. |
|
|
Thanks for that. I fully understand your decision. I mentioned fully managed version because I thought that you guys are already half way there with your parser. I'm only interested in Sass :) so I would not mind if this was the only one you implement. |
|
|
I didn't realize how similar LESS and SCSS are. I renamed main.scss to main.less, changed $ to @ and everything worked. On a 2000 line file, compiles take a tenth of a second instead of 9 seconds (on a SSD). I was only using imports, colors and nesting, so YMMV. (But if you're willing to wait 10 seconds just to see if a div moved one px to the left and now lines up, we're probably not going to be friends anyway.) |
|
|
Hey folks, There is now a nightly build which uses a different way of invoking the Sass compiler. For cblaze's files it dropped the compile time on my machine from several minutes to under a second. We're not pushing this out to the VS gallery yet because it's a significant change and we want to check that we haven't overlooked any compatibility issues e.g. behaviour with non-ANSI text, but we'd encourage Sass users to give it a go and report whether they run into any problems. For info about downloading and installing the nightly build, please see the FAQ sticky thread. Please do try it and let us know how it goes! |
|
|
Hi, Thanks guys you have managed to cut down one of our largest files from ~16sec -> ~8secs. Please keep up the good work. Will the command line tool be updated to this version? Regards, Brett. |
|
|
Yes, the updated command line tools should be in the store now. |
|