Development Machine Build & Setup
I am building yet another development system all tricked out. Rather than write my own install history and steps (which I must have done a hundred times), I thought doing a thread here might bring together some positive collaboration.
Specifically the dev environment I will be building targets these: Debian install, Cross compiler/toolchain (BBB & am3*x), yocto, git, C++ DSP libs, nightly regressions (Jenkins), JS/mysql web, neo4j, LabJack, Node.js, Python, etc. The new parts I will adding this time are R Studio, Eclipse, and Anaconda. (I am a command-line guy, but some of our developers insist on IDEs) I am not an expert, nor am I a newbie, but I do hope that folks here feel that this is an appropriate use of LinuxQuestions.org, and that someone gets some minor amount of help somewhere. Oh; I have a pretty busy schedule so please forgive in advance any tardy and/or terse responses. The Plan: Each post will be a separate part the build, (ie OS install, yocto, etc.) Where possible I will point to existing posts and not copy the content into here. Long live Linux! :hattip: |
Debian Installation
Normally I build VMs, but this time I have need of a field laptop for testing in remote regions of the world. I have a Dell E6410 with Windows7Pro setup in a 128GB partition to start with. It has served me well, but the Ubuntu 12.04 I have on it has been through so much that it's time for a clean start. Now that I have Yocto dialed in on Debian8, I've decided to rebuild the Linux on this Dell.
The system I set up uses a separate partition for the /home directory, but that is irrelevant here. The arrangement I follow on linux is to completely separate my code, data, files, whatever from the OS, by keeping everything tightly confined within the /home directory. Also, all development (except installation of code outside the /home directory) is done with a standard user account. This greatly reduces the chance that some errant program that I write or download will damage the operating system. Specifics: Dell E6410, i7 620M, NVIDIA Quadro NVS 3100M, 8GB, 1TB SSD A long hard lesson about these Dells is that the stock supplied power supply can overload, shut down, and cause cold shutdowns of the kernel that will orphan inodes (bad) and eventually destroy the OS. And you can be misled about it, because once the laptop shuts down, the power supply resets, and you have no clue what happened. Get a hold of a Dell 130 watt adapter if you can. You will need: - An optical disc CD-R/DVD-R with debian-8.7.1-amd64-netinst.iso imaged onto it (https://www.debian.org/CD/netinst/) - A USB 2.0 stick with missing network firmware files, prepared as described here: [SOLVED] How to provide non-free firmware files to the Debian Jessie Installer. Before starting the install, insert this memory stick into some computer and make a note of its name. This will help you identify it in the blkid output later. - A network cable connected to the internet (Optional, but likely better). If you plan to use wireless, you'll need a network to connect to and a password. - An hour or two. Steps to check out the laptop:
Now the install has started. Steps in Beginning of Install:
Now a time of downloading and installing occurs ... Finish installation:
This concludes basic Debian installation. Next we will proceed to install drivers and verify full functionality of all parts of the computer. |
Moderator response
Moved: This thread is more suitable in <Debian> and has been moved accordingly to help your thread/question get the exposure it deserves.
|
Installing NVidia Drivers
Installing the Nvidia driver brings the full functionality of the graphics card to the table.
There are many ways, This page: NvidiaGraphicsDrivers is a good resource. Perhaps one of the easiest ways is this:
Now you'll be rewarded with the nvidia splash screen during boot! Do take the time to look at the nvidia dialog box. Click lower left corner to get Desktop menu and click Preferences > NVIDIA X Server Settings. Best for now not to change anything and simply click quit. |
Enabling sudo
For many reasons sudo is preferred over simply using the root account. Here we show how to add your user account to sudo, while still requiring a password to use sudo. It is possible to configure sudo to allow passwordless use, but this is not recommended unless you have certainty that no one else has access to your machine, either physically or via network.
Ubuntu by default enables sudo for user accounts and disables direct root access, but Debian does not. This is a good page about sudo: How To Edit the Sudoers File on Ubuntu and CentOS. There are many other ways to configure sudo, this method is preferred (by me) for a developer account, as so many different usages occur, and it is just simple. To enable sudo for your user (developer) account: Open a terminal (click System Tools -> LXTerminal). Become root, type: su - and enter password. Type visudo at the root prompt. You will be taken inside the nano text editor (default). Arrow down to the bottom of the file and enter a new line at the end: Code:
myAccountName ALL=(ALL) ALL Type Ctrl-o and press Enter to write the file (/etc/sudoers.tmp), and Ctrl-x to exit the nano text editor. (visudo handles moving sudoers.tmp to sudoers) Now test it. Type exit and press Enter. Type cat /etc/sudoers and press Enter. You will see: Code:
cat: /etc/sudoers: Permission denied Now the contents of /etc/sudoers will be dumped. Henceforth the need to use the root account is essentially never. Lest ye forget, a clever safeguard is to turn both the desktop and the shell command prompt to be red for the root account. Root account login disable is another option, as is shown here: Type: sudo passwd -l root Re-enable by: sudo passwd root |
Add "Empty Trash" to right click on LXDE
You can follow the steps here:
PCManFM empty trash menu But: Beware cutting and pasting the instructions will insert a hyphen instead of a dash, and in the step where you use mkdir, you will in fact create a directory called "–parents" which is not what you want. Corrected instructions placed here: Code:
sudo apt-get install trash-cli |
Set up Clock with NTP to synchronize with internet time servers.
The default Debian/LXDE does not synchronize the clock with internet time servers.
These steps change the clock to synch automatically: Code:
sudo apt-get install -y ntp |
Backup dual OS system
Now that the basics are done, it's time to back it up.
Perhaps the best solution for these older laptops is to buy a dock which has USB3.0, eg: The E-Port Replicator with USB 3.0 from Dell™. 2.5 inch notebook disks >= 2TB with eSATA are hard to find. This product offers the best of both worlds, eSata and USB 3.0: MiniPro 2.5-inch eSATA 6Gb/s, USB 3.1 External Hard Drive Enclosure To take it on the road, the USB2.0 on the E6410 is just too slow. What I have handy is the HornetTek Viper SSC 3.5" Enclosure (HT-316SSC). I have inserted a 2TB 7200RPM drive into the enclosure for use in backing up this laptop. To use this drive as eSATA requires an eSATA cable and then connecting it to the one shared eSATA/USB on the laptop. Remember that by default, SATA devices are not hot-swappable, and must be connected with the laptop powered off. The back strategy I will implement is to divide the 2TB drive into two partitions. One will hold the disk image that will be made with Clonezilla live. The other partition will incremental backups of $HOME made by rsync. This post describes the Clonezilla live portion, while a later post (after git is in) will described the rsync portion. You will need:
Steps to set up "Clonezilla live"
Steps to back up disk image: This page: Clonezilla live doc under Save disk image"" describes the process of creating an image backup of the internal hard drive. Steps to mount incremental partition: The goal here is to mount the second partition we created on the eSATA drive at a known location within the file system and then repeatedly back up to it. To accomplish this we will use the system known as UUID, which is essentially a unique number assigned to a drive.
|
Install and Configure minicom on Debian 8 Jessie
An important tool for embedded development is a terminal emulator. There are many, and I use minicom. Here we will set it up and create a profile for talking to a Beaglebone black computer.
You will need:
Steps:
|
Memory Upgrade!
Never never never expose your precious linux installation to new, untested memory modules! That's a hard lesson.
I use memtest86 to check new memory DIMMs, etc. Steps: Download memtest86 for linux here: http://memtest86.com/download.htm. Scroll down to "Image for creating bootable USB Drive" and click to download. List current disk storage: ls /dev/sd*. In my case: Code:
/dev/sda /dev/sda2 /dev/sda4 /dev/sda6 /dev/sdb /dev/sdb2 Noe type: ls /dev/sd* again. Code:
/dev/sda /dev/sda2 /dev/sda4 /dev/sda6 /dev/sdb /dev/sdb2 /dev/sdc1 Change directories: cd ~/Downloads Unpack tarball: tar xvfz memtest86-usb.tar.gz. NOTE the next step will completely overwrite the USB stick! Type: sudo dd if=memtest86-usb.img of=/dev/sdc and enter your password. This will take a few moments. Once completed, unmount it: sudo umount /dev/sdc1. Pull out memory stick. Now power off computer: sudo poweroff Install new memory Insert USB stick with memtest86 on it and power up computer. Press F12 to enter boot select menu. Arrow down to select the "UEFI BOOT:" option and press Enter. |
Code:
# Clarkman _______________________________________________ |
Install tightvncserver
|
Install Acrobat Reader on Debian 8.x
For schematics and PDF documents such as Texas Instruments' series of ARM TRMs, there is no substitute for Acrobat Reader.
It's a 32bit executable, so install this way:
The following is copied from one of TI's manuals: Acrobat Reader Tips Acrobat includes two methods to search for words in a PDF: • The Find toolbar provides a basic set of options to locate a word in the current PDF. • The Search window lista words or partial words that match your text in the current PDF. These guidelines apply to Acrobat Reader 5.x, 6.0, and 7.0. For more information on Acrobat Reader search features, see the Adobe Reader Help. To search for words in a document using the Find dialog box: 1. Open the document. 2. To display the Find toolbar, right-click in the toolbar area and select Find. 3. In the Find box, type the word, words, or partial words for which you want to search. 4. From the Find Options menu, select options as desired. 5. To view each search result, click the Find toolbar, the Find Previous button, or the Find Next button to go backward or forward through the document. To search for words in a document using the Search PDF window: 1. Open the document. 2. Click the Search button on the File toolbar or right-click on your document and select Search. 3. Type the word, words, or part of a word for which you want to search. 4. Click Search. 5. The results appear in page order and, if applicable, show a few words of context. Each result displays an icon to identify the type of occurrence. All other searchable areas display the Search Result icon. 6. To display the page that contains a search result, click an item in the Results list. The occurrence is highlighted. 7. To navigate to the next result, choose Edit > Search Results > Next Result (or Ctrl+G). 8. To navigate to the previous result, choose Edit > Search Results > Previous Result (or Shift+Ctrl+G). Navigate through your previous view To retrace your path within an Adobe PDF document: • For the previous view: Choose View > Go To > Previous View or Alt+Left Arrow. • For the next view: Choose View > Go To > Next View or Alt+Right Arrow. The Next View command is available only if you have chosen Previous View. If you view the PDF document in a browser, use options on the Navigation toolbar to move between views. • Right-click the toolbar area, and then choose Navigation. • Click the Go To Previous View button or the Go To Next View button. NOTE: This navigation tip is useful to return to your previous view after clicking on a register call hyperlink. |
All times are GMT -5. The time now is 07:18 AM. |