Rhyous

I have been running the PB-BSD 8.2 and switching between work and home networks.

I have figured out that I have to run these commands when I want to switch over to use wireless:

When switching back to Ethernet, I have use these two commands:

I thought I was done listing the commands that must be run, until I realized that the above commands didn’t change the /etc/resolv.conf in my Ports Jail.

Question: What network settings need updating in the Ports Jail.
Answer: The Ports Jail uses the systems IP Address and Default Route, so only the /etc/resolv.conf needs updated and it appears to be a copy of the system’s /etc/resolv.conf

So I just have to update /etc/resolv.conf and that is all.

Solution

Well, one solution is to run another command that just copies the system’s /etc/resolv.conf to the Port Jail.

But a better solution is to get dhclient to run this command for us. This can be done with the dhclient-exit-hooks.

Step 1 – Create the dhclient-exit-hooks file and make it executable

Step 2 – Enter the function into the /etc/dhclient-exit-hooks to run copy command when dhclient exists

#!/bin/sh
cp /etc/resolv.conf /usr/jails/portjail/etc/resolv.conf

Now when you run dhclient, you will also get a new /etc/resolv.conf in your Ports Jail.

So, when you install using ports, there is a lot of software you may not really want that gets installed. Are you familiar with the term build dependency.  This is software, usually a library or scripting language, that is required for a software to compile from source, but is not needed at run time.  When you always install from ports, your system will eventually have most or all the common build dependencies installed.  Really the only dependencies you need are run-time dependencies.  These dependencies make up the software that must be installed for the desired port to actually function.

Here is what I do:

1. Install ports both in the jail and on my system pretty much at the same time with the same command, just in the different consoles.
   # portsnap fetch update
2. Compile the port in the jail.  This will build and install everything in the jail, including the build dependencies and run time dependencies and of course the port itself.
   # cd /usr/ports/<group>/<port>
   # make install
3. Sync the system’s ports tree with the jail’s ports tree using rsync.
   # rsync -av /usr/jails/portjail/usr/ports/ /usr/ports
4. Install the port in the regular system.  This only installs the run time dependencies and the port itself.
   # cd /usr/ports/<group>/<port>
   # make reinstall

This seems to work, and I don’t get all the build dependencies installed. Plus the rsync command is pretty quick.

I wonder if this is how the PC-BSD team uses it too?

Mono is basically the .NET Framework on FreeBSD or other open source platforms. This allows development in C# on FreeBSD.  C# is an extremely popular language that is not slowing down.  It’s popularity stems from that fact that this language and its features allows for rapid development that is much faster than many other languages.

The version of Mono available in the ports tree is not the latest version available. Just like FreeBSD has a release version and a development version, Mono has a release version and a development version.  The development version is so much newer that it is hard not to recommend it over the release version.

Step 1 – Install the latest ports

This is already documented here:

How to install ports on FreeBSD?

Step 2 – Install portshaker and portshaker-config

The team at BSD# have a tool called portshaker that adds mono ports to the ports tree.  Install it as follows.

Note: Notice I didn’t just install portshaker, I installed portshaker-config which has portshaker as a dependency, so you get both installed with one command.

Step 3 – Configure portshaker

The example portshaker.conf.example is configured correctly for default configurations, so all we need to do is copy it.

Step 4 – Run portshaker

Yes, it is that easy.  Simply run portshaker.

Note: You may be prompted to merge a few files. I diffed and chose either install or continue each time.

Note: Running portshaker uses subversion to download so if you need to use an HTTP proxy, you have to configure subversion to use an HTTP proxy as it doesn’t use the FreeBSD HTTP_PROXY environment variable.

Your ports tree is now updated by portshaker.

Step 5 – Install mono

The mono port should now be updated to the latest version.

Mono is now installed on your system.

There is an example of building a hello world app here:

C# (Mono) on FreeBSD

Firefox 3.6 no longer works with the Java versions in PBIs. Instead the OpenJDK6 must be installed. However, as of writing this document there is not a PBI for OpenJDK6, so we must install it from the ports jail.

1. Install ports on your system as follows: How to install ports on FreeBSD?
2. Open the command prompt.
3. su to root.
   $ su
   Password:
   #
4. As root, get the latest ports.# portsnap fetch extract
5. Install openjdk6.Note: There is an openjdk7 but it didn’t appear to have a web option yet, so hold off until it has one, use openjdk6.

   # #	cd /usr/ports/java/openjdk6 make WITH_WEB=yes BATCH=yes install

6. Run this link command to link the library for the java plugin to your plugins directory.

   #	ln -s /usr/local/openjdk6/jre/lib/IcedTeaPlugin.so ~/.mozilla/plugins/IcedTeaPlugin.so

Testing the Java plugin

You may want to verify that java is working in your Firefox install. You want to know two things:

1. Is the plugin detected by Firefox?
2. Does a java application work?
   1. 1. Open Firefox. If Firefox is open already, close it, and re-open it.
      2. In the URL enter: about:pluginsYou should now see the IcedTea-Web Plugin (using IcedTea-Web 1.0.1).
      3. In Firefox, go to the following URL:
         http://www.java.com/en/download/testjava.jsp 

   Here are steps to make those two verifications. 

   Note; This is pretty much exactly like the handbook states.

   http://www.freebsd.org/doc/en_US.ISO8859-1/books/handbook/desktop-browsers.html

If I open Ports Jail, then a regular konsole, the regular konsole gets the Ports Jail icon.  I can’t really tell the konsole sessions apart.

The same things happens if I do it the other way around.  If I open a regular konsole, then Ports Jail, the Ports Jail gets the konsole icon. Again, I can’t really tell the konsole sessions apart.

This causes me some confusion.  Which konsole am I running?  I have to take a moment to verify or close them both.

Solution: Opening the Ports Jail with a different konsole profile

One solution is to have the Ports Jail konsole session use its own konsole profile that uses a custom konsole profile.

Here is how I configured that:

1. I created a LavenderOnBlack color scheme, which is a copy of the GreenOnBlack with Lavender instead of Green. Normally I use the GreenOnBlack color scheme.~/kde4/share/apps/konsole/LavenderOnBlack.colorscheme

   [Background]
   Color=0,0,0
   Transparency=false
   
   [BackgroundIntense]
   Color=0,0,0
   Transparency=false
   
   [Color0]
   Color=0,0,0
   Transparency=false
   
   [Color0Intense]
   Color=104,104,104
   Transparency=false
   
   [Color1]
   Color=250,75,75
   Transparency=false
   
   [Color1Intense]
   Color=255,84,84
   Transparency=false
   
   [Color2]
   Color=24,178,24
   Transparency=false
   
   [Color2Intense]
   Color=84,255,84
   Transparency=false
   
   [Color3]
   Color=178,104,24
   Transparency=false
   
   [Color3Intense]
   Color=255,255,84
   Transparency=false
   
   [Color4]
   Color=92,167,251
   Transparency=false
   
   [Color4Intense]
   Color=84,84,255
   Transparency=false
   
   [Color5]
   Color=225,30,225
   Transparency=false
   
   [Color5Intense]
   Color=255,84,255
   Transparency=false
   
   [Color6]
   Color=24,178,178
   Transparency=false
   
   [Color6Intense]
   Color=84,255,255
   Transparency=false
   
   [Color7]
   Color=178,178,178
   Transparency=false
   
   [Color7Intense]
   Color=255,255,255
   Transparency=false
   
   [Foreground]
   Color=192,0,192
   Transparency=false
   
   [ForegroundIntense]
   Bold=true
   Color=210,0,210
   Transparency=false
   
   [General]
   Description=LavenderOnBlack
   Opacity=0.92
   

2. Next I created a copy of the shell.profile named PortsJail.profile.
   ~/kde4/share/apps/konsole/LavenderOnBlack.colorscheme

   [Appearance]
   ColorScheme=LavenderOnBlack
   
   [Cursor Options]
   CustomCursorColor=192,0,0
   UseCustomCursorColor=true
   
   [General]
   MenuIndex=6
   Name=PortsJail
   Parent=FALLBACK/
   RemoteTabTitleFormat=%h : %u
   TabBarMode=1
   TabBarPosition=1
   
   [Scrolling]
   HistoryMode=2
   
   [Terminal Features]
   BlinkingCursorEnabled=true
   

3. I then edited the Ports Jail shortcut on the desktop to pass it the approprate TerminalOptions to use the PortsJail.profile. Here is the one line I changed.

   TerminalOptions=--profile /usr/home/jared/.kde4/share/apps/konsole/PortsJail.profile
   

Now whenever I open the ports Jail, I can easily tell it is the Ports Jail and not the regular konsole because it is using a LavenderOnBlack profile.

I have historically done mock-ups in two ways:

1. Just code it up
2. Paint.NET

Neither are really good solutions.  But as a developer for an enterprise product, the above tools just are not sufficient.

So I have recently introduced to SketchFlow. This is a tool for quickly mocking up your UI.  It has some features that are quite nice and I am impressed.  I can create a UI very quickly and maybe even faster than I could with Paint.NET on the first try.  After using SketchFlow for a while, it will be a lot faster.

There are three features that make SketchFlow an amazing tool for UI prototyping.

1. The ability to map and design at the same time
2. The ability to package it for a demo
3. The ability to export the prototype to word.

These features make it so that for the same or less work that we have to do anyway to provide sample screens, we also get 1) a Map that we otherwise would have had to spend extra time creating in Visio or elsewhere, 2) A packaged demo that we can send to customers for feed back early on before development so we develop it right the first time, and 3) Stub documentation with all the screens and topical guides exported to word in seconds.

All three of these features are demonstrated in this Video.

Note: For my FreeBSD readers...this might not be for you though it would be interesting to see if Moonlight can load this video.

I don’t know what the FreeBSD plan is. Unfortunately, I feel that if there is one, it is not a very well-rounded on. Maybe it only focuses on development and might not have a big focus on also important tasks like advertising and marketing, documentation, and acquiring new committers.  All of which is important because I am not the only that thinks that FreeBSD needs fresh Blood!

Here is a simple step by step plan to make the FreeBSD community bigger. They aren’t all my ideas, many have been mentioned by other people in the forums or elsewhere.

Advertising and Marketing Department

1. Get one. Get an advertising team, however you have to do it.
   1. Maybe the FreeBSD foundation hire a part time advertising professional (or maybe iXSystems could let one of theirs donate some time) who gets as much help as possible from college students majoring in Advertising and Marketing as unpaid internships.
2. Give them goals to 1) get more committers, 2) get more users, 3) Get more Enterprise and Small business exposure and usage.
3. Get a list of companies that use FreeBSD and don’t ask them for donations, ask them to buy feature enhancements. They are more likely to pay more for an enhancement than they are willing to donate.
4. Get a list of all BSD user’s group email. Start using them and encouraging them to hold meetings, improve the user’s groups, etc…
5. Get a list of all college’s and their Computer Science staff members and make it standard practice that CS degrees can get internship credit for working on FreeBSD. Same for English Technical writing and FreeBSD Documentation, same for advertising and marketing and FreeBSD advocacy and marketing.

Documentation Department

1. Make documentation contributions easy.
   1. Make the FreeBSD documentation pages wiki editable. Anyone can edit the page, like Wikipedia. Yes, it would be fine if a page could have an person or team who has to approve the change, because we have to let the localization teams know that a change was submitted.
   2. Make the man pages available as a wiki, and the man page updates are included in the next release.
2. Make these training videos:
   1. Getting started with FreeBSD – Install and Usage.
   2. Getting started with FreeBSD – Documentation
   3. Getting Started with FreeBSD – User land development and debugging
   4. Getting Started with FreeBSD – Kernel Development and debugging
   5. Getting Started with FreeBSD – Contributing to the KDE on FreeBSD project.
   6. Getting Started with FreeBSD – Contributing to the FreeBSD GNOME project.
   7. Getting Started with FreeBSD – Remote debugging
   8. …continue as needed
3. Get a community site up. Not just a forum site, but a community site.
   1. Write a link between IRC and the community site so a persons IRC posts are also on the community.
   2. Write a link between the mailing lists and the community so emails are on the community.
   3. Provide FreeBSD User’s groups free sites and mailing lists but encourage them to get $10 a month in contributions each year to sustain it.
4. Support the Advertising and Marketing team.
5. Implement Kanban as a process for managing your work.

Development Department

1. Join the desktop world in a big way. FreeBSD is not just a server. It must have a solid desktop distribution. PC-BSD exists (though the name is not marketing friendly). Get a marketing friendly OS Name, and a marketing friendly logo, and create a desktop distro that has a chance at attracting new users. Let the Advertising an Marketing team determine the name and brand of the new desktop. (I don’t care if the distro is PC-BSD as long as it is re-branded.)
2. Build a graphical newbie proof installer that can install the server or desktop. (I know, many of us have come to love the simplicity of sysinstall, but I still remember my first year with FreeBSD when I hated it.)
3. Include Mono in the install of all desktop versions of FreeBSD. Lets face it, C# developers are everywhere and if the user land were C#, it would be much easier to get dev work done.
4. Support the Advertising and Marketing teams by sponsoring internships, maybe at your Alma mater.
5. Create a ready-to-dev on Virtual Machine.
6. Support the documentation team by making docs “Getting Started” easier.
7. Make the FreeBSD build easy to replicate in a graphical IDE. The new generation has few individuals who want to learn vi and gdb or otherwise work from the command line. They want to build and debug elsewhere.
   1. Get the FreeBSD projects building with a Graphical IDE. Lets leave the ability to build with make but add an a project file. I don’t care what graphical IDE is chosen, Eclipse, KDevelop, Code::Blocks, MonoDevelop, etc…Just pick one and make it happen.
   2. Have a master project file at the root of the build that will do everything, build the entire FreeBSD source, in debug or release.
8. Update the ready-to-dev on Virtual Machine to be include the graphical IDE and be ready to build and debug.
9. Maybe there is a reason to make a few different types of ready-to-dev virtual machines.
10. Implement Kanban as a process for managing your work, I recommend a web site version for open source communities. Each committer can have a view of their work just for them.

Well, that is the plan that has formed in the back of my head the past ten years that I have been involved with FreeBSD. Unfortunately, most of the above is not happening. PC-BSD is helping with #1 and #2 of the development plan, which is great!

Click to read more ideas for improving FreeBSD.

Hey all,

I was going to write an article on this, but it was so easy, there is no need for one.

Windows 7 doesn’t know how to deal with PC-BSD, but PC-BSD knows how to deal with Windows 7.

So all I had to do was install Windows 7 first (I chose a 50 GB partition) and then I installed PC-BSD on the remaining space. I did click to install the PC-BSD boot manager as it is not selected by default, and now I can dual boot Windows 7 and PC-BSD.  Which is a must for a guy who’s job is writing C# on windows.

There is a lot of work to do in this world, and there are plenty of open source projects. However, there are still plenty of projects that need help. There are still plenty of projects that haven’t even been started yet.

I just read this post and I am quite in agreement with it: FreeBSD needs fresh Blood!

Here are a list of projects ideas or projects that need contributors that could really help FreeBSD go to the next level.

Possible new projects

• FreeBSD Mentoring Project – The primary goal is to get FreeBSD developers developing immediately and maybe have a VirtualBox VM that is ready to download and dev on immediately.
• A Windows driver for FreeBSD’s UFS, maybe using the Installable File Systems (IFS) Kit
• A Windows driver for ZFS, maybe using the Installable File Systems (IFS) Kit.
• A GNOME based FreeBSD Desktop distribution (Note: PC-BSD and/or GhostBSD might fill these need in the future.)
• Contributors to DesktopBSD.
• A FreeBSD phone operating system to compete with Android, Windows Phone 7, etc… I think a clever name for phone OS based on FreeBSD would be Angel.
• A new desktop that isn’t GNOME or KDE but is written using Mono (C#).
• VirtualBox BootCamp for FreeBSD – Ability to boot to a VirtualBox image, similar to BootCamp on a MAC. So you can boot to a different OS, or while running FreeBSD you can load the OS as a virtual machine.

Projects that could use more Contributors

• FreeBSD Advocacy and FreeBSD Marketing – We need actual advertising and marketing people here, not coders or techies.
• FreeBSD ACPI - I would recommend a focus on improving ACPI support for laptops running FreeBSD.  The Sleep and hibernate ability are important.
• Mono on FreeBSD or BSD#
• Ports on FreeBSD – There are a lot of unmaintained ports, such as K-3d and you could adopt a port.
• KDE on FreeBSD – The KDE support is in need of help.  The network manager hasn’t worked in years.
• GNOME on FreeBSD – There is not an installable FreeBSD desktop distribution that focuses on GNOME.  GhostBSD, a live-CD, is based on GNOME.
• GhostBSD – They could use some help being not just a live-CD but an installable distribution.

Business ideas based on FreeBSD

While free contributions from the community is nice, for FreeBSD to really thrive, it needs more full-time developers. There is not substitute for actual paid employees who spend their days working on FreeBSD.  The only way get those type of employees is to get some companies out there making money. For this reason, I would recommend that the FreeBSD Foundation looks at helping start new businesses based on FreeBSD.

• BSD Appliances – Maintain appliances, both hardware and virtual appliances, for all the common server types. This is a hardware and support model, the software is of course free.
  • A complete suite of BSD appliances from Web Servers to Firewalls, to NAS, to backup, to networking, etc…I own the URL and thought about starting this project, but never really had the bandwidth.
  • Also, outsource appliance maintenance so other companies can have their proprietary appliance maintained by this company.  I know my company, LANDesk, has considered outsourcing our Management Gateway appliance. I heard a rumor KACE was too, though that rumor remains unverified.
• BSD Business – A software company that makes a server and applications for common businesses. This is a hardware and support model, the software is of course free.
  • A Business Server (three version Small Business, Medium Business, Enterprise) that has everything a business needs. (Maybe this server is an appliance maintained by BSD Appliances
  • Common businesses means businesses that exist in every city, law offices, dental offices, eye-doctors, chiropractors, restaurants, hotels, mechanics, gas stations, etc… There are some applications that are ubiquitous to their fields. To use an example local to me, think of what Dentrix is to the dental field. I have always thought of an ultra-secure desktop for lawyers running an very secure version of FreeBSD.
• Point of Sale system based on FreeBSD. This is a hardware and support model, the software is of course free. Why let Red Hat and CentOS have this who market.
• Gaming on FreeBSD – A gaming console based on FreeBSD to compete with XBox, Playstation, etc…Gaming has an amazing power to drive improvement.
• A support center where companies can get enterprise FreeBSD support, development, etc…
• Animation on FreeBSD – A film studio that makes animated movies and runs everything on FreeBSD.
• Feature4Hire – A web site that maintains a list of enhancement requests and people or companies can submit how much they are willing to pay for the feature. Then as soon as the dollar amount is worth it to someone in the world, they develop the feature and get paid, with something like 3% going to Feature4Hire. Even if $10 is attributed to a feature, some one in a 3rd world country might find some small dollar amounts worth it.  Ok, this site might be for all open source projects, but it could run on FreeBSD.

There are a lot more ideas out there and anybody can be a part of them.

I recently posted about the GhostBSD 2.0 beta.

Review: GhostBSD 2.0 Beta 2 – A FreeBSD LiveCD

Well, the product is no longer beta as they just released GhostBSD 2.0.

If you didn’t take a look when I posted about the beta, take a moment to take a look now!