Sublime Forum

Build 3010

#5

Minor important bug: custom color schemes are not sorted in alphabetical order in menu.

0 Likes

#6

I think this is a bug unless I misunderstand the API:

When calling

view.show(placeToGo)

There is no refresh, as soon as you move the cursor or scroll the view (even a pixel) everything refreshes. The only workaround I’ve found is to set (scroll past the end) to true and manually scroll the view a little. Below is an example of the problem.

[code]import sublime_plugin

class NextInterestingThing(sublime_plugin.WindowCommand):
rex = ‘(^\ \t](class|def)\ \t]+\w+\ \t](.*$|(?!\w\W])|\Z|\n\n\w)’

def run(self):
    view = self.window.active_view()
    # get current line
    curPoint = view.line(view.sel()[0].a).b
    found = view.find(self.rex, curPoint)
    placeToGo = view.line(found.b)
    if placeToGo:
        view.sel().clear()
        view.sel().add(placeToGo)
        view.show(placeToGo)
        for delta in [1, -1]:
            x, y = view.viewport_position()
            view.set_viewport_position((x, y + delta))[/code]
0 Likes

#7

Hey John,

Any chance you can fix this one here:

Thanks.

– Felipe.

0 Likes

#8

Significant indeed :smile:

Are the python console and other panels rendering via the same mechanisms? Maybe it’s something theme specific (I’m on Phoenix Dark) but scrolling in the python console and output panels is much slower compared to regular views.

0 Likes

#9

[quote=“gregor.hoch”]In ST2 there used to be this option for high resolution or retina displays. Is it correct that ST3 sets this option automatically depending on the resolution of your display?

Any chance that you can fix “Look Up”?[/quote]

It’s still controlled by the same setting, gpu_window_buffer, but the default value of auto now applies to more displays.

I suspect I have no control over the “Look Up” functionality, unfortunately.

0 Likes

#10

Performance seems to have improved generally, and the indexer doesnt seem to be spawning so many “rogue” processes.

Now, if only I could get rid of that damn menubar in Linux (please!)

Keep up the good work :smile:

0 Likes

#11

I’m still having issues wih the new indexing with c# code.

ST2: puu.sh/1XcRN/d7a76020f3
ST3: puu.sh/1XcUv/e8c0a5681a

In ST3, it seems to only index the using statements as functions, and doesn’t get any further.

Here is a sample file: gist.github.com/4704955
Trying to lookup @ToString fails.

0 Likes

#12

Seeing some black pixel gaps on the UI… OSX 10.8.2 Non-Retina.

cl.ly/image/0F2r2A1q1h33

Scrollbar gap is sub-pixel and happens on both top and bottom.

0 Likes

Dev Build 3026
Dev Build 3026
#13

Is it possible to “find functions referencing this symbol” as with ctags/gnu global/etc.?

Regards,

-b.

0 Likes

#14

@badeip: Should be possible to implement as a plugin if you really need it.

0 Likes

#15

I’m having trouble with the build system in this one. I have a simple build system file:

{
“cmd”: “make -j7”],
“working_dir”: “/Users/foo/src/myproject/build”
}

Regardless of various permutations (such as /usr/bin/make instead of make, etc.)

[Errno 2] No such file or directory: ‘make -j7’
[cmd: [‘make -j7’]]
[dir: /Users/foo/src/myproject/build]
[path: /usr/bin:/bin:/usr/sbin:/sbin]
[Finished]

The same command works from the command.

0 Likes

#16

Good.
Lack of cross-referencing is a show stopper for me.

0 Likes

#17

@sandover: Different parameters are assigned to different list entries for “cmd”. Thus, you do

"cmd": "make", "-j7"],

because subprocess.Popen formats the parameters automatically depending on which OS you are on.

0 Likes

#18

I’m experiencing frequent lockups when opening and/or closing files. plugin_host often goes wild with 100% cpu usage and keeps allocating memory (last I checked 10 GB).

0 Likes

#19

People have been requesting this since ST2 beta. Good luck.

sublimetext.userecho.com/topic/9 … -on-linux/

0 Likes

#20

Found another XML-related bug: Sublime Text crashes trying to load an empty snippet file: pastebin.com/Kkmv0n53

0 Likes

#21

when I press ctrl+B to run the python file , The console error as below:

Unable to find target command: exec

0 Likes

#22

[quote=“toplinuxsir”]when I press ctrl+B to run the python file , The console error as below:

Unable to find target command: exec[/quote]

Same here on Fedora 17 (64 Bit).

ST2 works fine however.

edit: Same python command but ST3 uses shell_cmd and ST2 uses just cmd in the Python.sublime-build files.

Haven’t tested with other build systems.

0 Likes

#23

I have noted a problem in ST3 in python mode (which I actually just tested in ST2 and there is the same problem):

In the following test code:

class Testing:
	def __init__(self):
		self._name = 'Test'

	def getname(self):
		return self._name

	def setname(self,value):
		self._name = value

	def delname(self):
		del self._name

	name = property(getname,setname,delname)

	# This is just a comment

When I fold (which I am only allowed to do on the class level and not on the method level (unless I manually fold the code) it looks like:

class Testing: ...t

i.e. keeping the last “t” of the comment outside the fold.

I know it is a minor bug, but since I just noticed that it has been there for 2 versions of the editor, I think it should be fixed at some point (and maybe add method folding in Python as well?)

Regards,

Nikos

0 Likes

#24

[quote=“corelon”]I have noted a problem in ST3 in python mode (which I actually just tested in ST2 and there is the same problem):

In the following test code:

class Testing:
	def __init__(self):
		self._name = 'Test'

	def getname(self):
		return self._name

	def setname(self,value):
		self._name = value

	def delname(self):
		del self._name

	name = property(getname,setname,delname)

	# This is just a comment

When I fold (which I am only allowed to do on the class level and not on the method level (unless I manually fold the code) it looks like:

class Testing: ...t

i.e. keeping the last “t” of the comment outside the fold.

I know it is a minor bug, but since I just noticed that it has been there for 2 versions of the editor, I think it should be fixed at some point (and maybe add method folding in Python as well?)

Regards,

Nikos[/quote]

As far as i can see that’s not a bug. From indentation the comment is belonging to the class. I see the same behaviour with Scite and Komodo 7 + 8

0 Likes