Dev Build 2083 is out now, with some important fixes to issues introduced in 2081 - as far as Iâm aware, this should sort out the last of them.
Dev Build 2083
Build 2083
Release Date: 15 July 2011
Build 2082
Release Date: 14 July 2011
Build 2081
Release Date: 13 July 2011
Will this pattern continue into the weeked? Soon weâll expect a new version twice a day!
Just tried it and itâs a little bit erratic:
If you drag slowly the mouse cursor out of the text control, first the scrolling is very fast, then suddenly it almost stop and then it become to accelerate again.
Not a big deal.
Every others issues Iâve noticed on previous builds look resolved (mouse cursor shape, selection lost with page down/up, âŚ)
Thanks Jon for your quick fixes.
Still canât make mouse_wheel_switches_tabs work. Added to User global settings, with no avail.
Also, when setting show_tab_close_buttons to false, thereâs a disturbing and inconsistent shift in title letters of tabs when hovering them. Hereâs what I mean: screenr.com/HBUs
Can we have a setting to disable close buttons on tabs forever? I keep on hitting them no matter what. Could be nice if âhover to show close buttonâ was configurable too.
Assuming youâre on Linux and trying to disable it, thereâs no reason it shouldnât work if placed in the user global settings. You can type this in the console to verify the applied value:
sublime.load_settings('Global.sublime-settings').get('mouse_wheel_switches_tabs')
sublime.load_settings('Global.sublime-settings').get('mouse_wheel_switches_tabs')
Windows 7 64bit:
sublime.load_settings(âGlobal.sublime-settingsâ).get(âmouse_wheel_switches_tabsâ)
True
Scrolling over tabs still does not change tabs though. Am I missing something? Hereâs my user global prefs for reference:
// Place user-specific overrides in this file, to ensure theyâre preserved
// when upgrading
{
âremember_open_filesâ: true,
âfolder_exclude_patternsâ: â.svnâ, â.gitâ, â.hgâ, âCVSâ],
âfile_exclude_patternsâ: â.pyc", ".pyoâ, â.exe", ".dllâ, â.obj", ".oâ, â.a", ".libâ, â.so", ".dylibâ, â.ncb", ".sdfâ, â.suo", ".pdbâ, â.idb", â.DS_Storeâ, ".pdfâ, â.class", ".pngâ, â.jpg", ".gifâ],
ânew_window_settingsâ:
{
âside_bar_visibleâ: false,
âstatus_bar_visibleâ: false,
âshow_tabsâ: true,
âshow_minimapâ: false
},
// Set to 0 to disable smooth scrolling. Set to a value between 0 and 1 to
// scroll slower, or set to larger than 1 to scroll faster
âscroll_speedâ: 1.0,
âshow_tab_close_buttonsâ: false,
âmouse_wheel_switches_tabsâ: true
}
Itâll only work if you rotate the mouse wheel while the mouse cursor is over the tabs, and if the mouse wheel is detected as such. On Windows this means it must send a WM_MOUSEWHEEL message, and not use fractional scroll amounts.
This functionality isnât new, the only recent change was turning it off by default on platforms other than Linux. It only exists to keep in line with typical GTK behaviour, and the only reason thereâs even a setting for it is so that it can be turned off on Linux too - itâs not a nice bit of functionality IMO.
If youâre on Windows, a better option is to leave the setting off, and switch tabs instead by holding down the right mouse button and using the mouse wheel - it works without having to have the mouse over the tabs, wonât cause inadvertent switching, and works with fractional scroll amounts.
A small drawing problem with tabs:
If you start ST2 with tabs hidden and turn them on after, thereâs a kind of transparent line in the bottom of them.
startup, version: 2083 windows x64 channel: dev
Iâve finally just started moving from ST1 to ST2 and felt I should mention a couple issues:
Noticed some behavior with the minimap that doesnât seem right. When clicking somewhere outside the box that indicates the visible area, the place you end up scrolled to isnât where you clicked.
The larger the file/ the farther youâre moving the box the more drastic the error seems to be. It doesnât happen when the map is small enough for the whole thing be shown.
See attached GIF.
(And, while Iâm on the subject of the minimap Iâll mention again that it would be nice to be able to scroll past the end with the map like you can with the scrollwheel or scrollbar)
Keep up the good work.
- toothrot
Thatâs working as designed - when the minimap doesnât show the entire file, something has to give. Thereâs a setting to change the trade-off in S1, but not in S2.
Ah, okay. However, I donât see the usefulness of the current behavior, itâs not really different than the scrollbar at that pointâŚ
I would be glad to see the old behavior as an option in ST2, if not, I guess Iâll have to get used to dragging the view rectangle to where Iâm looking on the map.
Canât please everyone I suppose⌠thanks for the prompt explanation.
Addendum: I searched the forum before posting that, however I failed to look on userecho, where I see now that itâs mentioned:
(http://sublimetext.userecho.com/topic/43437-click-on-minimap-does-not-highlight-the-clicked-part-on-large-files/
http://sublimetext.userecho.com/topic/26724-shift-click-minimap-move-to-shown-location/)⌠whoops.
Actually, something wrong with it. Now I am using build 2083 with two columns layout. In left column files automatically saved sucessfully, but not in right column. Files in second column lose unsaved circle symbol but physically not saved.
Addition to my previous post. This problem occurs sometimes but not all times. It seems when file changed only in second column all saved ok, but when files changed in both columns not all files saved. Strange, I canât find system in it.
Oh! Last file in second column (rightmost tab) not show âeditedâ circle symbol when file edited and this file not saved when focus lost. Problem definetly with rightmost file (tab) in second column when in two columns layout.
Iâm not able to replicate this - files are consistently marked dirty for me when using a two column layout, including the rightmost one. Iâd personally be surprised to find that the layout or position of a file influences if it gets marked dirty or not when editing.
Hm⌠After restart all works again, but some time latter this strange behaviour is repeated. I not understand how and when it happens again. Itâs like mysticism. But this problem definitely exists
Just to be clear, youâre finding that after some time, you can type in some files, and they donât get marked as dirty?
Next time you see it, could you check and see if itâs just a case of the dirty indicator being clipped in the tab label, and the dirty indicator is perhaps still showing in the window title?
Yes. And when it happens in this file I have problems with undo: it undos not last changes but some past of them. From 15 july I was see this many times, but have not time to post here. One guess I have - it is somehow connected to ZenCoding, becouse sometimes ZenCoding wrap command make so strange things it hard to discribe and after that I canât undo this corrupted result. It just stuck in this broken code condition and I must revert file from last save.
Ok, next time I try to closely watch all what will happens.