TI OMAP 4460 + Android file system + Android Linux kernel 3.0
I have set CONFIG_LOGO, CONFIG_VT and CONFIG_FRAMEBUFFER_CONSOLE in kernel config, logo image type is CLUT224.
Then I saw strange things, almost 10 seconds later after kernel booting, a broken image appeared, looked like the head of the penguin, only several pixels of lines on the top of sreen. Then I reboot again and agian, sometimes I saw the 'ANDROID' characters, just below the penguin's head, but sometimes not.
see the linkenter link description here
My colleague reminded me I should disable the Android booting logo('ANDROID' characters, to be exactly), because it needs to open a tty console. That might have effect on something?
Besides, why the penguin logo showed up so late? I added some print in drivers/video/fbmem.c
fb_do_show_logo(), this function was called at 3 seconds after kernel booting. Why I saw the logo at 10 seconds?
Related
RStudio is great IDE for R development. I wonder if there is any way for nice support HiDPI resolution?
I currently have 13 inch display and 3200x1800 resolution, it is even hard to read RStudio options to adjust more appropriate setting.
This is of course not an RStudio issue but general issue related to high resolution display which requires HiDPI support to make it reasonably usable.
Posting it here as question because RStudio still doesn't allow to fill questions/issues on github.
RStudio is a Qt application. I've solved this by setting a scaling factor in Qt with this environment variable (note it must be an integer):
export QT_DEVICE_PIXEL_RATIO=2
The Arch wiki has some excellent advice on getting HiDPI working.
UPDATE: This solution is only for WINDOWS!
There is a simple solution for you, it comes from the option "Compatibility" of the execute file.
Close all current RStudio windows.
Right click on the shortcut of RStudio (or the original exe file) and choose Properties
In the RStudio Properties pop-up windows, choose the tab Compatibility
Tick on option Override hide DPI scaling... and then choose System from the drop-down list.
Apply > OK.
(Re)open Rstudio to see the change
Note that, on the HiDPI screen, resolution of modified RStudio is not so good but it really solved the bad scale problem.
You can also apply this technique to other apps without supporting HiDPI.
I use the following bash script to determine the current screen resolution and substitute the scaling factor by 1 (normal resolution) or 1.75 (high resolution) in the RStudio desktop file:
#!/bin/bash
# Determine resolution
width=$(xdpyinfo | grep 'dimensions:' | grep -o -E '[0-9]+' | head -1 | sed -e 's/^0\+//')
echo "Screen width is $width pixels"
if [ $width -gt 3000 ]; then
echo "High resolution detected -> setting zoom level to 1.75"
sed -i 's/view.zoomLevel.*/view.zoomLevel=1.75/' ~/.config/RStudio/desktop.ini
else
echo "Normal resolution detected -> setting zoom level to 1.0"
sed -i 's/view.zoomLevel.*/view.zoomLevel=1/' ~/.config/RStudio/desktop.ini
fi
QT_QPA_PLATFORMTHEME=gtk2 /usr/lib/rstudio/bin/rstudio %F
I am on Ubuntu 17.10. If I want to use the GTK theme, I have to set the QT_QPA_PLATFORMTHEME environment variable.
Closed. This question is not reproducible or was caused by typos. It is not currently accepting answers.
This question was caused by a typo or a problem that can no longer be reproduced. While similar questions may be on-topic here, this one was resolved in a way less likely to help future readers.
Closed 4 years ago.
Improve this question
I want to comment out a block of multiple lines in Jupyter Notebook, but can't find out how to do that in this current version.
It used to be in one of the drop down menus but is no longer there.
How do you comment out multi-line blocks of code at once?
This is not a duplicate because the solution given in the following link doesn't seem to work anymore:
How can I block comment code in the IPython notebook?
Ctrl + / does nothing.
Ctrl + / works for me in Chrome browser in MS Windows. On a Mac, use Cmd + / (thanks Anton K).
Please note, if / did not work out of the box, try pressing the / key on the Numpad. Credit: #DreamFlasher in comments to this question.
I have not yet managed to find the best way possible. Since I am using a keyboard with Finnish layout, some of the answers do not work for me (e.g. user5036413's answer).
However, in the meantime, I have come up with a solution that at least helps me not to comment each and every line one by one. I am using Chrome browser in MS Windows and I have not checked other possibilities though.
The solution:
It uses the fact that you can have multiple line cursors in an Ipython Notebook.
Press the Alt button and keep holding it. The cursor should change its shape into a big plus sign. The next step is, using your mouse, to point to the beginning of the first line you want to comment and while holding the Alt button pull down your mouse until the last line you want to comment. Finally, you can release the Alt button and then use the # character to comment. Voila! You have now commented multiple lines.
Try using the / from the numeric keyboard.
Ctrl + / in Chrome wasn't working for me, but when I used the /(division symbol) from the numeric it worked.
Quick Addition to Top Answer: CTRL + / is nice because it toggles back and forth between adding and removing # at beginning of all selected lines. Didn't see that exact nuance mentioned so just wanted to add it here. (This worked in Firefox Developer Edition 54.0b12 on Windows 7).
On a Finnish keyboard use Ctrl + ' to comment on multiple lines and use the same keys to de-comment.
Ubuntu 14.04 Google Chrome
TL;DR:
Using MacBook Pro with Spanish - ISO Keyboard.
Solution: Ctrl + -
Full story
This is an old post but reading it got me thinking about possible shortcuts.
My keyboard is a Latin Apple MacBook Pro, which is called Spanish - ISO. I tried the changing keyboard distribution to U.S. solution... this works but with this solution I have to switch keyboards every time I want to comment which... sucks.
So I tried ctrl + - and it works. The - is where the / is located in an english keyboard but doing Cmd + - only changes the Chrome's zoom so I tried Ctrl which isn't as used as Cmd in macOS.
My takeaway with this would be: if I have more shortcut problems I might try the original shortcut but using the key where the U.S. keyboard would have it.
Select the lines you want to comment out. Then press:
Ctrl + #
I tried this on Mac OSX with Chrome 42.0.2311.90 (64-bit) and this works by using CMD + /
The version of the notebook server is 3.1.0-cbccb68 and is running on:
Python 2.7.9 |Anaconda 2.1.0 (x86_64)| (default, Dec 15 2014, 10:37:34)
[GCC 4.2.1 (Apple Inc. build 5577)]
Could it be a browser related problem? Did you try Firefox or IE?
Use triple single quotes ''' at the beginning and end. It will be ignored as a doc string within the function.
'''
This is how you would
write multiple lines of code
in Jupyter notebooks.
'''
I can't figure out how to print that in multiple lines but you can add a line anywhere in between those quotes and your code will be fine.
Fn + Cmd + / in Safari browser on MacOS
On MacOS 10.11 with Firefox and a German keyboard layout it is Ctrl + ?
Select the lines on windows jupyter notebook and then hit Ctrl+#.
I add the same situation and went in a couple of stackoverfow, github and tutorials showing complex solutions. Nothing simple though! Some with "Hold the alt key and move the mouse while the cursor shows a cross" which is not for laptop users (at least for me), some others with configuration files...
I found it after a good sleep night. My environment is laptop, ubuntu and Jupyter/Ipython 5.1.0 :
Just select/highlight one line, a block or something, and then "Ctrl"+"/" and it's magic :)
After searching for a while I have found a solution to comment on an AZERTY mac. The shortcut is Ctrl +/= key
I am using chrome, Linux Mint; and for commenting and dis-commenting bundle of lines:
Ctrl + /
For a Dutch keyboard layout (on Debian 9 in Chromium 57) it is Ctrl + °
Another thing to add, in the version I'm using, the code has to be initialized in order to be to comment it out using CTRL and / . If you haven't ran the code and the code isn't colorized it wont work.
If you have a Mac and not a English keyboard:
Cmd-/ is still easy to produce.
Follow the below steps:
Just go into the Mac's System Settings, Keyboard, tab "Input Sources" or whatever it might be called in English
Add the one for English (shows up as ABC, strange way to spell English).
Whenever you want a Cmd-/, you have to change to the ABC keyboard (in your menu row at the top of your screen,if you have ticked it to be shown there in the System Settings - Keyboard tab).
Cmd and the key to the left of the right "shift key" gives you Cmd-/.
P.S: Don't forget to switch back to your normal keyboard.
I have a Friendly-ARM embedded board with a W35 3.5" Touchscreen. You can see the board via the following link: http://www.friendlyarm.net/sites/products/mini2440-35s.jpg. Also I write QT program for that using Qt Creator. I have to write a GUI using QT with Width x Height = 240 x 320. I mean width = 240 and height = 320. According to what I found at various online documents and pages, the dimensions of W35 are 320 x 240, it means width = 320 and height = 320. So when I run the program, there are large margins at left and right and some part of GUI is cut at top and button. How do I exchange the board width and height?
The closest page I found on Friendly-ARM site is: http://www.friendlyarm.net/forum/topic/2881.
At this page someone mentioned, there is a file s3c2410.c at drivers/video directory, or there is file mach-mini2440.c at arch/arm/mach-s3c2440 directory and we should tweak some C #define, but I don't have neither one on the board kernel. What should I do?
1) Reinstalling the Kernel?
2) Writing program for 320 x 240 instead of 240 x 320
3) Changing the touchscreen to similar ones like X35 or T35
FYI, when the board starts up, there is Qtopia with right dimensions.
TIA,
-- Saeed Amrollahi Boyouki
First of all cross-compile your Qt with -qt-gfx-transformed option.
Method 1:
You can rotate your Qt Application using
./myApp -qws -display ":Transformed:Rot90:0"
Method 2:
You can set Display Dimension using
export QWS_DISPLAY=Transformed:Rot90:0
and start your application using
./myApp -qws
What you need to look for are instructions to rotate the display from "landscape" mode to "portrait" mode. I am unsure if there is a hardware option on the ARM processor included in the FriendlyArm, but that gives you a place to start searching. I'd also look in the Qtopia forums for a similar switch, ie http://qt-project.org/forums/viewthread/8875 or similar.
I noticed this question has been asked many times in the past and surfing the web I found many pages about it. However, it seems like the proposed solutions rarely work and, in my case, the problem does not refer to a program that I wrote. So I'll give it another try here.
I recently installed Linux Mint 14 on my laptop. After the OS was spick and span, I started to install the software I need, and among these netgen (a Mesh Generator software). I tried both ways: download+unpack+compile+install and synaptic. Either way, this is the output I get when I try to execute the program
X Error of failed request: BadAlloc (insufficient resources for operation)
Major opcode of failed request: 154 (GLX)
Minor opcode of failed request: 3 (X_GLXCreateContext)
Serial number of failed request: 490
Current serial number in output stream: 491
As I said, I surfed the web, and apparently, this is thought to be linked to some problem in the X server configuration. And here start the mess. Someone says I should modify /etc/X11/Xorg.conf, adding the lines
Option "Videoram" "65536"
Option "Cachelines" "1980"
Under the section "Device." Unfortunately, I have no such file, as apparently in recent distros, the X configuration file has been moved to /usr/share/X11/xorg.conf.d/* and it's now split into different files. The one about the monitor and graphics should be called 10-monitor.conf...which I don't have. I tried to create one, following the instruction at this link, and then add those lines, but nothing happened. To be fair, I'm not 100% sure I generated the file correctly since I am not sure how to detect the driver for my graphics card.
I don't know how much and which information people would need to have an idea of how to fix this problem. Here's what I see might be useful.
The output of 'lspci | grep VGA' is
01:05.0 VGA compatible controller: Advanced Micro Devices [AMD] nee
ATI RS880M [Mobility Radeon HD 4200 Series]
My current /usr/share/X11/xorg.conf.d/10-monitor.conf is the following
Section "Monitor"
Identifier "Monitor0"
Modeline "1920x1080_60.00" 172.80 1920 2040 2248 2576 1080 1081 1084 1118 -HSync +Vsync
EndSection
Section "Device"
Identifier "LVSD"
Driver "fglrx" #Choose the driver used for this monitor
EndSection
Section "Screen"
Identifier "Screen0"
Device "LVDS"
Monitor "Monitor0"
DefaultDepth 24
SubSection "Display"
Depth 24
Modes "1920x1080_60.00" "1366x768"
EndSubSection
EndSection
Under the section "Device." Unfortunately, I have no such file
Try creating your own xorg.conf file, placing it in this location will override your X settings after restarting X or simply by restarting the computer.
mkdir -p /etc/X11/xorg.conf.d/
cp /etc/X11/xorg.conf.d/xorg.conf /etc/X11/xorg.conf.d/xorg.conf.bk # in case it exists
cp /usr/share/X11/xorg.conf.d/10-monitor.conf /etc/X11/xorg.conf.d/xorg.conf
The content of /etc/X11/xorg.conf.d/xorg.conf would look like (adding your options):
Section "Monitor"
Identifier "Monitor0"
Modeline "1920x1080_60.00" 172.80 1920 2040 2248 2576 1080 1081 1084 1118 -HSync +Vsync
EndSection
Section "Device"
Identifier "LVSD"
Driver "fglrx" #Choose the driver used for this monitor
Option "Videoram" "65536"
Option "Cachelines" "1980"
EndSection
Section "Screen"
Identifier "Screen0"
Device "LVDS"
Monitor "Monitor0"
DefaultDepth 24
SubSection "Display"
Depth 24
Modes "1920x1080_60.00" "1366x768"
EndSubSection
EndSection
This also could be related to the driver you're using, there are other common drivers available like
Driver "fbdev"
Driver "vesa"
Driver "fglrx"
The fbdev driver supports all hardware where a framebuffer driver is available.
The vesa driver supports most VESA-compatible video cards. There are some known exceptions, and those should be listed here.
fglrx is a X.org(7x) driver for ATI (Mobility(TM)) RADEON® and (Mobility(TM)) FireGL(TM) based video cards. The driver provides hardware acceleration for 3D graphics and video playback. It includes support for dual displays, TV Output and as of version 8.21.7 also OpenGL 2.0 (GLSL).
Depending on which driver you choose, certain options/functionality/compatibility would be enabled or not, you could change the driver and test with the options you said would work.
Finally, you have hundreds of options here to play with X11.
When I had made written a lot text to my program I took similiar error. When I reduce the text size the error was disappeared. I think you should reduce too showing things or reorganise how they are looking on screen.
I hope I could help you with this broken English ;)
The log rotation for Plone product installation would be a nice feature. What are the current best practices regarding the log rotation integration into Plone?
I found this article: http://encolpe.wordpress.com/2010/06/17/how-to-get-log-files-rotate-in-zope-with-buildout/ but as there are no documentation on plone.org I'd like to ping the community for the good known best practices not to fill up their hard disks.
ZConfig has support for the standard library RotatingFileHandler and TimedRotatingFileHandler. Taking an example from the ZConfig tests:
<eventlog>
<logfile>
path /path/to/file.log
level debug
when D
interval 3
old-files 11
</logfile>
</eventlog>
This will roll over the logs every three days, keeping 11 old files.
You place these config snippets your buildout using the event-log-custom/access-log-custom parameters in your instance recipe. plone.recipe.zope2instance
Similar to what Laurence said above but keeps size under 10mb and saves only 1 old file.
<eventlog>
level INFO
<logfile>
path /path/to/plone4/var/log/client1.log
max-size 10mb
old-files 1
</logfile>
</eventlog>
plone.recipe.zope2instance can generate this now. For example, you can specify the following options:
event-log-max-size = 10mb
event-log-old-files = 3
Here's what we do, it's simple but works:
In your buildout you add this part:
[logrotate]
recipe = collective.recipe.template
input = ${buildout:directory}/templates/logrotate.conf
output = ${buildout:directory}/etc/logrotate.conf
And in templates/logrotate.conf
rotate 4
weekly
create
compress
delaycompress
missingok
${buildout:directory}/var/log/instance1.log ${buildout:directory}/var/log/instance1-Z2.log {
sharedscripts
postrotate
/bin/kill -USR2 $(cat ${buildout:directory}/var/instance1.pid)
endscript
}
${buildout:directory}/var/log/instance2.log ${buildout:directory}/var/log/instance2-Z2.log {
sharedscripts
postrotate
/bin/kill -USR2 $(cat ${buildout:directory}/var/instance2.pid)
endscript
}
Add whatever other log rotations you need. Then it's about linking /etc/logrotate.conf to the generated file.
Mikko, you should ask me directly ;)
My blog article is still good and this feature is still undocumented in Zope2. It can be used with Plone 4 and Plone 5. The extension iw.rotatelogs was designed for Plone 3 only.
Using logrotate is not fine because it doesn't handle the case of logwriting during the rotation. It can make your instance crash silently and writing logs in RAM until your restart the instance.
I've been using iw.rotatezlogs since at least early Plone 3 very successfully. I see from your link that I should no longer need iw.rotatezlogs.
I don't like to rely on logrotate, since it's not usable on the one Windows server I have to deploy to.
As near as I can tell, the pure-zope solution (which I stil haven't tested) doesn't do logfile compression (at least I can't see it in ZConfig/components/logger/handlers.xml which is where I think it should be defined), so I suspect I'll be staying with iw.rotatezlogs.