I have a script named my_script.m, which has been mistakenly replaced by another file with the same name. However, prior to overwriting the script, I have run it in MATLAB console. Hence, I have the following history.
>> my_script
>>
Is there any way to recover that history file?
The reason why I think it is still possible is that I do have the run history of that script in my current console. If only I had selected all the script and run it! That way I would have every command in the console history. But now, it is simply one line as above.
Check if you had diary mode on, then everything is there:
get(0,'Diary')
You can open the file in windows explorer, right mouse button, properties. Then check if there are any other versions of the file in the Previous Versions tab ;)
Sometimes matlab writes backups, the file would be named my_script.asv. Check if it exists, it's located in the same directory.
Related
Which is the final outcome of the VScode [issue] #14298 (https://github.com/microsoft/vscode/issues/14298) ?
Is it: "No, we will NOT issue a warning, if an external app modifies a VScode opened file, like ALL other editors do ? (except Notepad)"
Up to now, I cannot find such a setting...
I have never seen a prompt when an open file is changed elsewhere. What I have seen is a warning when I try to save a file that has been changed elsewhere. I have seen this: preventing dirty writes
VS Code will show you an error message when you try to save a file
that cannot be saved because it has changed on disk. VS Code blocks
saving the file to prevent overwriting changes that have been made
outside of the editor.
In order to resolve the save conflict, click the Compare action in the
error message to open a diff editor that will show you the contents of
the file on disk (to the left) compared to the contents in VS Code (on
the right):
Until those issues have been resolved I believe that is as close as you are going to get.
Is there a way to manually enter perforce commands in the p4v GUI?
I'm relatively new to Perforce and find the GUI useful for most of what I need to do, but I can see situations where a one line command would be much easier than navigating through the GUI.
Example:
Opening a file for edit that is nested deep in the source tree; this takes many clicks when done in the GUI but would be a relatively short command with environment variables set for my most commonly used directories.
p4 edit -c NNNNNN $DIRECTORY/file
It would be great if the command entered was also added to the Log window. I really like being able to scroll through the Log to see everything I've done in a session.
The closest that exists right now is "File"->"Open Command Window Here" when you have a file selected in the file browsing pane. It will open a command line in that file's directory with all of the relevant environment set up.
Unfortunately, this won't give you unified Log, and you'll have to juggle a second window.
Recently I have been having an issue with desktop save mode where it will not actually save my desktop. In the echo bar it says "Error while saving the desktop..." After typing no it says "Opening output file: no such file or directory, then gives the location to the path of the file". After saving a .emacs.desktop file then restarting emacs I noticed that it is saving the buffer locations in that file but is not loading that file. Thanks. Also I am not sure what has caused this to happen as it was working a couple weeks backs and nothing has changed that should make a difference.
The only thing i have in my .emacs for the desktop mode is
(desktop-save-mode 1)
Looking at the code for desktop.el here, it looks like the error is bubbling up from desktop-kill, which runs when you exit Emacs. The first thing I'd try is to check that the directory where it tries to save the desktop is sane.
Looking at the code in desktop-kill, it only tries to do anything if the variable desktop-dirname is non-nil. But that only gets set when you run M-x desktop-save for the first time: are you sure that it's set to something sensible? To check its value quickly, you can type M-: desktop-dirname RET and it should appear as a string in the message area.
If the directory is something sensible (the directory exists and you can write to it...), then I'm not sure. You'll probably have to give more information to get a solution, and it's not really clear that it's an ideal question for StackOverflow.
i should hazard that you get this error by creating a shortcut in the windows start manual via clicking addpm.exe in the ...\emacs-version\bin\ folder.
you can further modify the shortcut. go to its property->shortcut tab, you will find that the Target has value like ...\emacs-version\bin\runemacs.exe, while Start in is void. try to fill Start in with the corresponding folder ...\emacs-version\bin (actually most directories would be fine, just don't leave it blank), then everything is fine. still, the machanism behind this remains unclear to me.
or you could always creat your own shortcut manually, only make sure that the target is runemacs.exe, not any other exe file.
It used to be that when I made changes to my .py file in Eclipse, then the changes would be automatically saved when I then ran the program in debug mode. I liked this since it meant that what it was debugging was the same as what I was looking at. Now it no longer does that, which means that's it's actually running an old version, unless I manually save the file first. The first time I ran the program, Eclipse asked me whether I always wanted any changes to be saved when the program was run, and I answered yes. I don't know why it doesn't do it anymore or how to get it back.
This seems to have happened after I started using Mercurial TortoiseHg with BitBucket. As part of that, I did move some files around, but everything is back and named the same as before.
Of course, I can just do change, save, run, but I'd rather not have to remember the extra save step.
Edit: Actually it's a bigger problem. It also shifts my breakpoints. In regular edit mode, I 1) set a break point and Save. 2) Add a new line above the breakpoint. Everything looks fine. 3) Save. Now the first breakpoint shifts position.
From your description, it seems that you're always opening the file as an external file and not a file within your workspace (the fact that breakpoints don't get updated is the major clue here).
You can check if this is what happening from the title that's shown in Eclipse (if it's a full filesystem path and not relative path considering your workspace location, this is what's probably happening).
How are you opening the file you're running? Are you opening it from the PyDev package explorer or dragging from the filesystem? (also, it'd be interesting knowing which eclipse/pydev/tortoise versions are you using)
Look at your Preferences -> Run/Debug -> Launching, you probably have "Never" set for Save Required Dirty Editors before launching.
I am using Matlab on a remote server from my MacBook. Partly because I find the Matlab GUI clunky under X11, I use it in the terminal window (by running matlab -nodesktop). This has been working fine, until I now needed to copy some stuff from the command history.
It seems like history.m is only being saved on exit, even though when I check the settings in the GUI, it says that it is being saved for each command. It also seems like exit statements are included in history.m when Matlab is run from terminal, but not in the GUI, so it seems like the settings in the GUI don't apply there. I cannot find anything in the Matlab help pages on how to change the settings for the terminal window.
I need to access the command history for my session while it is still open. Does anybody have an idea about how I can do this -- short of starting to use the GUI?
You might try the diary function. It logs both commands and their (text based) results to a file. On my system (Mac OSX with MATLAB R2011b), the output does not show up in the diary file immediately, but it does update when you turn off the diary. Which you can do repeatedly if you need without exiting matlab.