Sensible commands like window auto resize, mouse move resize, etc
Note: Binary files aren't shown on the web site. To see all files, please download the extension zipfile.
| Version | Status |
|---|---|
| 39 | Rejected |
| 38 | Rejected |
| 37 | Active |
| 36 | Rejected |
| 35 | Rejected |
| 34 | Active |
| 33 | Active |
| 32 | Active |
| 31 | Active |
| 30 | Active |
| 29 | Active |
| 28 | Active |
| 27 | Active |
| 26 | Active |
| 25 | Active |
| 24 | Active |
| 23 | Active |
| 22 | Rejected |
| 21 | Rejected |
| 20 | Rejected |
| 19 | Rejected |
| 18 | Rejected |
| 17 | Rejected |
| 16 | Rejected |
| 15 | Rejected |
| 14 | Rejected |
| 13 | Rejected |
| 12 | Active |
| 11 | Rejected |
| 10 | Rejected |
| 9 | Rejected |
| 8 | Rejected |
| 7 | Rejected |
| 6 | Rejected |
| 5 | Rejected |
| 4 | Rejected |
| 3 | Rejected |
| 2 | Rejected |
| 1 | Rejected |
1. Selective disable isn't allowed: - line 169 `cmds/win_mouseresize.js` - line 83, 205 `cmds/win_optsize.js` 2. Destroy function should never return (line 54 `prefs/context.js`). 3. Why are you checking something that already you exists (line 46 `shell/keybindingmanager.js`). Please don't send AI code before reading it.
Hi, let me respond to the above: > 1. Selective disable isn't allowed: - line 169 `cmds/win_mouseresize.js` - line 83, 205 `cmds/win_optsize.js` Ans: Neither of these above are selective disable! this.win is nullified on the next line - and that's the only instance that nullies it - so destroy must run to nullify it. It's just that the "end" method can be called with many different actions and not just the destroy, and that is the reason the check is there. If it's already destryoed, then there is nothing to destroy. It's not selective disable. > 2. Destroy function should never return (line 54 `prefs/context.js`). Again, this is also, window is nullified right after. Here again, destory for the window is called in response to several signals and not just gnome, see line 10-12. So, if the window has already been destroyed and nullified, there is nothing to destroy. If you'd really like to, I can extract the destroy into another fn, or just put it in a if { } block, that all of that just makes the code less maintainable? > 3. Why are you checking something that already you exists (line 46 `shell/keybindingmanager.js`). This particular bit was infact manually written. This again is a similar pattern to the above where I prefer to code defensively here. > Why are you checking something that already you exists. This is not accurate. It's only set in enable(). And an extension can be disabled, before the key manager's enable method is called, and then we end up with noisy logs, or worse, the method throwing and results in insufficient cleanup for the reset of the extension, etc.
Another reason I prefer to keep them is also to keep the linters happy. Or I'll have to keep selectively adding exceptions to linters since _these are nullables_ - I do infact avoid redundant checks on non nullables which gnome passes through wherever I can. Are you really that opinioned to prefer that I remove it while fighting the linters?
As mentioned before, destroy function should never return. You may argue that the code currently works fine but since we are reviewing the diff in later updates we can miss that return. For the other selective things, same rule applies and also the entire destroy/disable shouldn't rely with a return. But`this._unsubscribe?.()` is okay. No need to fix then.