MirageGlobe Project

2014-10-13T10:38:00.001+01:00

Our blog posting has grown out of blogger and is now at mirageglobe.com . Many thanks for visiting this site. Blogger mirageglobe will be kept for information purposes.

Please visit www.mirageglobe.com for newer updated posts. Thank you.

SVG graphics with Javascript

2013-12-18T17:55:00.003+00:00

You can generate svg graphics with D3js for graphs. When you are happy with the result, you can use svg crowbar to download the image (and convert it to eps / png etc....). I had a simple html file running raphaeljs with graphaeljs and it generated a svg response.

After having the html file, simply run "python3 -m http.server 8000" to see the result on localhost. (see below).

Once you are happy. Then run crowbar in your browser (http://nytimes.github.io/svg-crowbar/ ). It will then download the svg automatically, which you can convert to other image formats vs inkscape or apache batik. (brew install batik; then run batik)

Hope this helps someone.

FTP Clients for Mac OS X

2013-11-03T18:40:00.000+00:00

All good (s)FTP and ssh visual clients for Mac OS X. Just a reminder for myself :D

http://macfusionapp.org/
https://filezilla-project.org/
http://cyberduck.ch/

or in terminal > shell > new remote connection

Ref:
http://apple.stackexchange.com/questions/52096/is-there-a-gui-ssh-client-for-os-x

Windows Apps on Mac

2013-09-08T01:37:00.001+01:00

Just a note for those who are keen to run window applications on mac or package window apps as a mac app. Many of you have heard of wine (windows emulator) which runs windows applications on linux-based distros or macs.

An interesting application called winebottler allows you to run windows (xxxxx.exe) applications on your mac or package them.

http://winebottler.kronenberg.org/

Programming Languages Compared

2013-08-16T07:26:00.000+01:00

A starter for most is deciding on which programming languages one would like to pick up and learn. A few questions raise to mind:

Which language is easy to pick up; low learning curve
How widely used is the language by industry
How many projects use these languages
How long has the language been around

Born: 1972

Learning curve: Medium

Notable Projects: Linux OS
Mostly used: Embedded Devices, Library extensions for many other languages

C++

Born: 1983

Learning curve: High

Notable Projects: Blender, Apache
Mostly used: Desktop applications, Mobile applications, Financial service platforms

Erlang

Born: 1986

Learning curve: Medium

Notable Projects: ??

Mostly used: Telephony messaging

Java

Born: 1995

Learning curve: Medium

Notable Projects: Hadoop

Mostly used: Desktop applications, Mobile applications

Javascript

Born: 1995

Learning curve: Low

Notable Projects: ??
Mostly used: Web applications

Perl

Born: 1987

Learning curve: Medium

Notable Projects: ??

Mostly used: ??

Php
Born: 1995

Learning curve: Low

Notable Projects: Facebook

Mostly used: Web applications

Python

Born: 1991

Learning curve: Low

Notable Projects: OpenStack
Mostly used: Web applications, Desktop applications, Scientific data crunching

Ruby

Born: 1995

Learning curve: Low

Notable Projects: Twitter

Mostly used: Web applications

Scala

Born: 2003

Learning curve: Medium

Notable Projects: ??
Mostly used: ??

Tcl

Born: 1988

Learning curve: Low

Notable Projects: ??
Mostly used: ??

Developers now will need to know controlled versioning. Controlled versioning is a system that supports multiple developers working in the same set of codes. Common version systems used are.

SVN (born 2000)
GIT (born 2005)
Mercurial (born 2005)

Several good ways to get started.

Online tutorials like nettuts or w3c
Check on developer forums
Participate in developer opensource contributions like Apache Software Foundation, Github, Bitbucket, Google code

Final note is, that like all languages, the syntax is different but the core philosophy is the same. If you can weave comfortably in one language, using the next is mostly about picking up the structure, libraries available for use, and syntax. Good luck!

Hackers Natural Cure

2013-05-10T15:12:00.002+01:00

Bit of a light touch, pun of the day. If nothing else works, sudo crem it. Available at local boots, pharmaceutical , amazon and your IT computer store.

Publishing on Ebooks

2013-04-14T12:35:00.001+01:00

Thought I will keep some directions on desktop publishing. The most widely used format is PDF (portable document format). It is versatile and maintains a structured and consistent layout to the contents.

The other format that is now open is EPUB format. Currently its on 3.0 and now a widely supported open format for digital books using xml. If you are a publisher, so is interested in self publishing, the following opensource tools might help:

https://code.google.com/p/sigil/
http://www.sourcefabric.org/en/booktype/
http://calibre-ebook.com/

Editors such as SublimeText2

2013-03-17T21:00:00.002+00:00

Sublimetext2 is a very nice editor. In MacOSX you can launch it via terminal. First symlink sublime text application. Use the following command to do so.

> ln -s /Applications/Sublime\ Text\ 2.app/Contents/SharedSupport/bin/subl /usr/local/bin/sublime

Now you can launch sublime text2 via...

> sublime [filename]

In terminal if you want to create a file. Something which mac doesnt have "create file". Do the following

> touch [filename]

This should help.

Good Coding Practise

2013-02-17T14:51:00.005+00:00

For all Php Frameworks and .NET Coding

File names - Pascal Case ( applicationStuff.aspx )

Folders - Pascal Case ( /generalFiles/....)

NameSpaces - Pascal Case : CompanyData

Assemblies - Pascal Case + unique to NameSpace : Assembliy.MGDev.dll

Functions - Pascal Case ( GetUserName() )

Function Parse Vars - Camel Case ( GetUserName(int userID) )

Variables -

Html Objects - Lower Camel Case + ux . User eXperience ( uxFirstName )

Classes -

Public vars -

Private vars - UnderScore Lower Camel Case ( _userName )

Other possible methods

lowerCamelCase

HigherCamelCase

Under_Score spaces

Samples

namespace MG.NewProject.BLL{

public class Staff : User{

    private string _myItem = "";

    public sting EveryItem = "";

    public Staff(string UserID){

    }

}
}

LMDE Linux Mint Debian Edition

2012-12-25T21:46:00.001+00:00

My first views on LMDE (Linux Mint Debian Edition) 201204. Linux mint X has always been built on ubuntu series. LMDE is linux mints "vanilla" version built on debian. (sly use of vanilla). This essentially takes out ubuntu as the middle(ware) and removes a substantial amount of bloat.

There are 2 versions of LMDE: XFCE and Mint's own Gnome...called Cinnamon.

XFCE

Virtual Guest Editions (virtualbox) doesnt work out of the box. You will need to install it manually.
Might be the installation but XFCE oddly uses the swap as its installation directory? (might be my bad as this is impossible)

Cinnamon

VirtualBox Guest Editions works. No glitches as far as I know

When you first install. You will need to select /dev/sda2 as Ext4 and "/" which defines it as root partition. /dev/sda1 is a default linux-swap. LMDE does this quite nicely automatically if you allow it to format your disk and setup partitions. All well for my installation. The installation seems very fluid and I didnt find any problems as the step by step was quite intuitive.

Comments welcome on any installation hiccups.

Day 1 Update

I reverted back to XFCE Mint 13 LTS (Long term support). As I was using mint for a media centre, its just too much effort to keep setting the pulse audio settings (which does not save!). I am still sticking with the Ubuntu middleware version for now. LMDE is quite decent but theres a few gritty issues. It shows great potential to be a top desktop replacement.

Heroku Webapp

2012-11-16T14:37:00.001+00:00

Heroku Webapp with Sinatra, Rack, Haml running on Thin.

A typical heroku webapp will consist of the following files:

base.rb - your basic ruby application file
config.ru - your default (similar to passenger)
Gemfile - default bundler file. Note that this has to be upper case 'G'emfile
Gemfile.lock - a generated bundler file. Generated by bundler install
views/index.haml - a default view (MVC) which is using haml language
public - a folder which you can house your javascript, ico, images, css files
Procfile - Procfile defines the process to run the application on remote PAAS

Hadoop Single node 32bit on 12.10 Mint XFCE

2012-11-10T18:04:00.002+00:00

An installation using Linux Mint 13 based on Ubuntu 12.10. Install virtualbox and download Linux Mint XFCE 13. Install Linux mint 13 XFCE on virtual box.

Once installed open terminal and run.

> sudo apt-get update
> sudo apt-get upgrade
> sudo apt-get install build-essential

Get some coffee beans and start brewing :D. Back to terminal.

> su
# adduser hadoop
# passwd hadoop

Into the new users account

> su hadoop
> cd ~
> wget ftp://ftp.mirrorservice.org/sites/ftp.apache.org/hadoop/common/hadoop-1.0.4/hadoop-1.0.4.tar.gz

Note: check from http://www.apache.org/dyn/closer.cgi/hadoop/common/ and see which newer version you want. Here the installation is based on stable 1.0.x.
Check if your java is installed.

> java -version

Unzip the file
> tar xvfz hadoop-1.0.4.tar.gz

This will show the location of your java installation directory
> which java
> grep JAVA ~/hadoop-1.0.4/conf/hadoop-env.sh
> nano ~/hadoop-1.0.4/conf/hadoop-env.sh

and change JAVA_HOME to below.

# The only required environment variable is JAVA_HOME. All others are
# set JAVA_HOME in this file, so that it is correctly defined on
# export JAVA_HOME=/usr/lib/j2sdk1.5-sun
export JAVA_HOME=/usr/
# note that this should be /usr/bin/java , but it seems that /usr/ works on linux mint??

> cd ~/hadoop-1.0.4
> mkdir input

> cp conf/*.xml input

To test..

> bin/hadoop jar hadoop-examples-*.jar grep input output 'dfs[a-z.]+'
> ls -l output

and this will give you some output files.!

Reference: http://www.thegeekstuff.com/2012/02/hadoop-standalone-installation/

Tracing User Behaviour FRUBO

2012-11-04T21:02:00.001+00:00

Frubo is "File Relationship and User Behaviour Ontology". We research on producing high relation models using passive tracing methods.

Background

FRUBO (File Relationship & User Behavior Ontology) also known as previously: AFRO (Attention File Relationship Organiser) System is a user behaviour based document sorting application. By capturing the users document usage patterns, we can build relationships between documents that is essential for document seeking, and grouping. The study exploits usage patterns and enhances a new layer of information built on to current searching tools on Windows / Mac OSX.

Project Status

Please email me at mirageglobe@gmail.com if you are currently testing. Otherwise I will not know.
Currently we are looking for volunteers for data collection. Please contact me at mirageglobe@gmail.com , if you are interested. Many thanks
Improving clustering methods by core seeking

What is this Application Really doing when its running?

It collects your time, date and application name such as "microsoft-word myletter"

The logfile will be like as follows:
===============================
<event><time>1288283690.27</time><app>Unknown</app><winname>Inbox - Microsoft Outlook</winname><handler>3801580</handler><mousepos>(1620, 3)</mousepos></event>
===============================

How do I know if I am safe?

The application does not access the internet or send any details anywhere. All captured data stays in your machine in the file "attention.log". You will need to email this file to the co-ordinator manually. At anytime, you may inspect this file to see its contents.
About Application Details

Dev Libraries: Python 2.7 & SQlite 3.7.4, + Several Python Libraries
Supported OS: WinXP, Win Vista, Win 7

Known Bugs:

Esc Button doesnt work.

Project Team

Supervisor: Prof. Frank Wang (UKent), Dr. Sining Wu (UKent)
Researcher: Mian-Guan Lim, Jimmy
Evaluators: Shawn Yang(UKent), Dr. Jiang Nan (UBournemouth)

How to Use Tracker?

Click on the link below to download (Stable version is recommended)
Save the file on your desktop
Unzip the file with Winzip or Peazip
Go into the folder and click on "afrotrack.exe" to run
You may want to minimize the tracker window.
Thats it. Carry out your general usage of the pc and AFRO will track your behaviour in the background.
After 2 weeks - 4 weeks, go into the folder and check the file "attention.log"
Email this file to our Research Group to mirageglobe@gmail.com (M.Guan Lim Jimmy)
A nice to have will be feedback of the system. You may want to feed back on improvements anytime during the volunteering process. Many thanks again for your contribution/help towards research in UK

Download Location

License: Free Closed Source Testing At the moment

Frubo Tracker Beta (Standalone) v0.3.1 (Current Stable Version)
Frubo Tracker Beta (Standalone) v0.3
Frubo Tracker Beta (Standalone) v0.2

Sponsors/Partners:

This project was made possible by the following groups/institutions:

FRUBO by Jimmy Mian-Guan Lim is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.

Installing Nginx on MacOSX

2012-11-04T16:43:00.001+00:00

Installing NginX

If you are installing Nginx on MacOSX, i recommend using homebrew. Once you have home brew running, install nginx via:

brew install nginx

Once you have it installed, there are a few notes of where you will need to get your nginx running. You nginx config file will be in:

/usr/local/etc/nginx/nginx.conf

if you are install it on Linux either debian or ubuntu. Just do this in terminal:

sudo apt-get install nginx

Setting up Php for NginX

Once you have installed Nginx, you may want to get Php5.4 running. Using homebrew, you can install php-fpm (Php fast process manager) that is a process listening on your (default) port 9000. You can now set nginx to forward any php script requests to this process by adding the following: PHP-FPM Configuration in the nginx.conf

location ~ \.php$ {
#root html;
fastcgi_pass 127.0.0.1:9000;
fastcgi_index index.php;
#fastcgi_param SCRIPT_FILENAME /scripts$fastcgi_script_name;
fastcgi_param SCRIPT_FILENAME $document_root$fastcgi_script_name;
include fastcgi_params;
}

You might find that the config may have this section of code commented out already. Just uncomment and restart your nginx server.

sudo nginx -s reload

Something interesting is to check your hosts file (if you are using mac), it will be the file at /etc/hosts. To edit, type this command:

sudo nano /etc/hosts

Some configuration parameters which you will find useful.

index index.php index.htm index.html; #Can be placed on the Http level if FPM is activated.

root /Users/[your user id]/Sites/; #This will point to your sites folder on mac. Note [user id], use yours! This can be placed in the Server level

Installing Ruby on MacOS X

2012-11-03T10:31:00.001+00:00

Installing RVM for ruby is a better choice than using homebrew on MacOS X. You can install Ruby through terminal by running the following if you have homebrew installed. Homebrew is an application manager for MacOS X. Some have opted for homebrew ruby + rbevn combo:

brew install ruby

Homebrew is good.....I just found the gem executions path is a bit confused with homebrew installed ruby. The alternative is to use ruby version manager (RVM) instead by:

curl -L https://get.rvm.io | bash -s stable --ruby

You may see common errors when you type any commands from a fresh install like "shell-init: error retrieving current directory: getcwd: cannot access parent directories: No such file or directory. Just

cd /

Once installed you can add in a few good gems for Ruby such as:

Sinatra (Web framework for Ruby, which influenced Flask.py web framework for python)
Thin (Web HTTP server for Ruby)
Shotgun (Auto reload web apps for development. Good for Sinatra)
HAML (html templating for ruby apps. similar to jinja)
Passenger (Platform for Node / Ruby / Python Apps)

For those interested in cross compatibility with Java <-> Ruby. try Jruby. You can install using RVM command:

rvm install jruby

Once you have installed ruby via RVM. You can install gems such as listed above by:

gem install [package]

For those interested in running an app platform for nginx with passenger, try:

rvmsudo passenger-install-nginx-module

Turn off annoying compress outlook message

2012-08-20T13:24:00.001+01:00

Ever got this: "To free up disk space, Outlook Express can compact messages"

With Service Pack 2, Outlook Express is set to ask about Compacting messages after every 100th closing of Outlook Express. Before Service Pack 2, the default behavior had Outlook Express set to compact messages in the background which has lead to the corruption of the message store. Compact the messages manually, reset the Registry value which keeps track how many times Outlook Express has been closed.

Look for this key in the Registry: (run regedit by going to start > run > type in "cmd")

HKEY_CURRENT_USER\Identities\{GUID}\Software\Microsoft\Outlook Express\5.0
Double click the Compact Check Count
Select the Base option for Decimal and change the value to 0
This should change both the Hexadecimal and decimal values to zero.
Close the Registry.

Reference: http://forums.techguy.org/business-applications/479919-solved-free-up-disk-space.html

HTML5 Apps on Cordova PhoneGap with MacOSX

2012-08-18T23:12:00.000+01:00

I am not a fan of eclipse but lets try to get over this nightmare. (fan of lightweight editors: nano and editra)

Installing Eclipse on Mac

First download Eclipse Classic from http://www.eclipse.org/downloads/
Unzip the file and drop the folder into your applications folder.
Launch "eclipse.app" from the app folder.

Setting up Eclipse on Mac

Go to Help > Install New Software > Add
Type in Android ADT and paste in this url:

https://dl-ssl.google.com/android/eclipse/

Now to add Android SDK.
Goto and download the Mac SDK from: http://developer.android.com/sdk
Unzip the file and place the folder in your user directory. (if you click your mac hdd, in places, you should see your login. There should be a android-sdks. Open it and place it there.)

Running HTML5 cordova on Android

http://docs.phonegap.com/en/2.0.0/guide_getting-started_android_index.md.html#Getting%20Started%20with%20Android

DBAccess using SQLite on Android

http://docs.phonegap.com/en/2.0.0/cordova_storage_storage.md.html#Storage

Choosing and using Latex

2012-08-15T09:34:00.002+01:00

Choosing and Using Latex

Personally i feel that latex is a bit bloated and too confusing for first time users. If more effort can be put into front end. It will definitely be the choice of writing. http://openwetware.org/wiki/Word_vs._LaTeX

For LaTEX, I recommend the use of texmaker http://www.xm1math.net/texmaker/ . It just makes everything look more manageable just like word.

Adding URLs in Bibtex with Latex

There has been quite some discussions about how to add urls/websites in your bib file. some use natbib which supports @online or @webpage.

If you prefer not using natbib the other options, you can do this:

@misc{jones07anatomy,
    Author = {M. Tim Jones},
    Title = {IBM Anatomy of the Linux file system},
    Howpublished = {\url{http://www.ibm.com/developerworks/linux/library/l-linux-filesystem}},
    Month = {October},
    Year = {2007},
    Note = {Accessed: 12/03/2010}}

Do note that in howpublished which is your url. You may notice I have used \url.

For some apparent reason, if you url has strange symbols or is like a protected word for example www.asd.com/DS_129385. PdfLatex throws up an error when it tries to automatically insert $ before these characters and } at the end. If you use \url, it encapsulates this and makes everything in order.

Reviewing Javascript World

2012-08-08T21:30:00.000+01:00

As of writing, realised that I have been out of the javascript world for sometime. HTML5 and canvas has gone leaps and bounds.

Important technologies that we should be concerned on is as follows when WEBGL becomes the de-facto standard.

WebGL Defacto standard: http://en.wikipedia.org/wiki/WebGL
Web Data using html5 Css: http://d3js.org/
Etherpad which is an online editor: http://typewith.me/p/test
and 3D generation: http://en.wikipedia.org/wiki/Three.js
Processing 3D and 2D: http://processingjs.org/

and interestingly: http://www.chromeexperiments.com/detail/web-lab/?f=webgl

and quite impressively threenode.js: http://idflood.github.com/ThreeNodes.js/public/index.html

Draw up an Ubuntu Server Environment

2012-07-29T14:23:00.000+01:00

Ubuntu Server is becoming a de-facto linux platform due to its frequent updates and wide adoption. This makes it one of the most simpliest to manage in Linux distributions. Other popular distributions include Debian (which ubuntu is built upon), Fedora, Suse. With cloud computing, you can launch an instance very easily. An instance is just like an empty hardware box, with the only difference of it being virtual. When you start off with a newly installed machine, lets see how we can set it up. We will focus on Ubuntu server edition here.

Choosing a version

For ubuntu, there are a few versions. It runs in twice a year release cycle, thus 10.04 (is around april) and 10.10 (is around october). So why are versions important? The release cycle keeps the basic packages updated. If you are running a server and is not looking to update your ubuntu installation every 6 months, use a LTS (Long Term Support). The LTS versions have a gap of about 2 years with 4 years support. 10.04 (Lucid) and 12.04 (Precise) is LTS versions as of writing.

Getting the basics up

Now we have a command prompt blinking in our face. What now? Type in the following.

> sudo apt-get update
> sudo apt-get upgrade

this will run loads of installs and type "y" for yes if you have to. After installing, reboot.

> sudo reboot

Now you need to setup your date time. there are other methods but this method adjusts and prevents time drift. You can make this more complicated via NTP daemon. Read this https://help.ubuntu.com/12.04/serverguide/NTP.html

> nano /etc/cron.daily/ntpdate

Enter this line if its not there, without the quotes! "ntpdate ntp.ubuntu.com". Save and exit using ctrl-x.

> sudo chmod 755 /etc/cron.daily/ntpdate

setting local of system, which adds compilers to your server. Notice we use Sudo alot? Its meaning to run the command as the super user (Super User DO).

> sudo apt-get install build-essential

Change the default shell to bash

Nano is a great text editor that is simple and easy to use. You will need to frequently use ctrl+x to save and quit once you finish your changes. Here we use nano to edit the password file of your server. After entering the command below, find your username and change from bin/sh to bin/bash

> sudo nano /etc/passwd

This command is to load the default shell as bash. Bash shell has more commands and is easier to use.

Installing Security Packages - ClamAV and NMAP

ClamAV is a general antivirus which is constantly maintained by public and a group of dedicated team. Kudos to them! To install and start a basic scan, enter in terminal and run,

> sudo apt-get install clamav
> sudo freshclam -r -i

There are a few methods to make clamAV a default daemon mode so you can invoke this via other applications. Mailservers can find this very useful. Another use is if you have upload folders.

NMAP is a port scanner to see if you have open ports

> sudo apt-get install nmap
> sudo nmap [IPAddress]

Installing UFW (Uncomplicated FireWall)

In command line, key in the following.

> sudo apt-get install ufw
> sudo ufw enable
> sudo ufw status
> sudo ufw logging on [this enables logging]
> sudo ufw allow ssh
> sudo ufw allow 80/tcp
> sudo ufw allow ftp
> sudo ufw allow smtp

To enable FTP to successfully go through your firewall, you might need to open certain ports. The example below opens from 20000 to 20999.

> sudo ufw allow ftps
> sudo ufw allow proto tcp from any to any port 20000:20999
> sudo ufw delete allow ftps [this is to delete a rule]
> sudo ufw status [this is to view the firewall]

Now reboot your machine. you can do this easily via command line below.

> sudo reboot

Using CHKROOTKIT

> apt-get install chkrootkit
> chkrootkit

Moving around Ubuntu Server 12.04

The following are places which are commonly accessed by admins

/etc/init.d/... - directory where all your application controls are stored
/etc/... - directory where your applications are stored
/var/www/... - directory where your web server html/php files are stored
/var/log/... - directory where your log files are stored
/home/.... - directory where home directory of your users are held
/etc/passwd - file which stores the password
/etc/shadow - file which I cant remember what this is for
/etc/group - file which shows how groups and users are defined for this system

Some nice Commands

if php5-cgi is installed, you can find the version of Php: php-cgi -v
copy command: sudo cp /mysource /mytarget
changes the permissions of the folder or file: chmod -R 777 myfolder
secure access into another machine: ssh asd@123.45.67.89
view memory usage: free -m
shows which group user mydomain is in: grep mydomain /etc/group
usermod -a -G adm demo
Top / ps -ef
vmstat 1 [si so is the swap]
crontab -e [to edit]
crontab -l [to list]
To find out if a file exists and create one if there isnt one: touch filename

References

Ref: https://help.ubuntu.com/community/InstallingSecurityTools
SECURE KEY GENERATION: ADMIN USER + PUBLIC KEY
http://cloudservers.rackspacecloud.com/index.php/Ubuntu_-_Setup
http://www.cyberciti.biz/faq/howto-linux-add-user-to-group/
http://www.kalzumeus.com/2010/12/12/staging-servers-source-control-deploy-workflows-and-other-stuff-nobody-teaches-you/

Draw up a Lightweight Mail Server

2012-07-29T14:21:00.000+01:00

Mail servers or services are not like before. In the old days, any administrator will like to install and run their own mail servers. Likewise today, there are still quite a handful doing this. In today's land, there are 2 options to do this: either commercial methods like using google mail servers or installing dedicated server software. The most common commercial mail software is microsoft exchange servers and the alternative is sendmail.

Opting for the opensource method would be most ideal for most administrators. This will include installing www.dovecot.org or www.microsoft.com exchange server. My view is the setup and maintenance being horrendous and unmanageable if you are an SME. I personally prefer some open lightweight solutions. I will still advise mail servers to be installed on a separate server by itself. It has a high chance of screwing up your configs.

Google Public Apps for Business

So what is advisable? Before you start, make sure you have a domain name, something like myfirstwebsite.com. You can purchase one from large domain registras such as www.godaddy.com, www.register.com or even google themselves sell domains.

When you have a domain, use Google Apps to host 50 free email accounts. Effectively its a personal mail server that hosts (xxxxxx@yourdomain.com) 7gb free, upgradable to 25gb. Mine is a small development group so 50 is enough. If you are running a medium to large company considering to pay google per account to sort out the problems may be an option or using dovecot setups. Lets focus on the SME approach. Go to Google Apps Standard Edition: http://www.google.com/apps/intl/en/group/index.html

Follow googles instructions of setting up the environment. There will be a point you will have to configure your mx records. Google requires to validate domain so you will have to place html files into your ftp or change cname. Just do the html method.

When you are adding MX records, place in your purchased domain and googlemailservers as mx record. TTL 300, priority as defined.

Notes

In setting up the valid domains for google apps for business, you might find it easier to use the html method. If you upload the file on to your server to be hosted and its not displaying, most of the time is your permissions problems. While on your own server, in your command line do > sudo ls -l and see if the group of the file is www-data and your user is part of www-data group.

Draw up a Web Database

2012-07-29T14:20:00.002+01:00

Classic SQLDB
Mysql DB

when running or building an application, you will need to have some sort of storage. Be it games, software, there will need be data to be stored. The common large scale database is mysql. Personally I do not prefer it, but hey there are so much applications using it, its not a bad idea just to get an idea of how to use it.

In Linux terminal, typing the following command will install mysql.

> sudo apt-get install mysql-server

follow instructions and enter new password for mysql admin. A good front to access your MySQL installation is phpmyadmin. However before you do that, you will need to install a web server and php support.

PhpMyAdmin for MySQL

To administer MySQL painlessly, its ideal to install a webfront for MySQL. The most widely used is phpmyadmin. To install make sure you have a http server such as apache, nginex, lighttpd or cherokee. Again in Linux terminal run the following command.

> sudo apt-get install phpmyadmin

Embedded SQLDB
SQLite a rising favourite for Apps and Websites

Now, databases need not be large DBs. There are cloud DBs and NoSQL DB. SQLite I recommend here is based on traditional SQL based commands. The only difference is that its an embedded DB. It doesnt need a server like MySQL.

So what is SQLite? Many languages have it already built in. To get an idea, you can install Mozilla Firefox and download the manager SQLiteManager. All you have to do after you have installed it, is to create a local file. YES, just a local file and that is your database. You can access it via python, java and php. Have a look at their website: http://www.sqlite.org/

NoSQL DB
(Not only SQL) WebDbs on CouchDB

Apache CouchDB is based on JSON which is a document format data. Meaning this is human readable and easily passed around. CouchDB boasts non-locking or non-blocking architecture which allows concurrency in DB. http://couchdb.apache.org/

Php CMS and Frameworks

2012-07-29T14:19:00.001+01:00

What is a php framework? A php framework is a set of prebuilt libraries that allows any php developer to use and develop websites with. The main difference between CMS and Frameworks is the flexibility. Php frameworks just extends on the current php language so as to provide a quicker way for developing websites or applications. Thus these frameworks are more commonly known as Rapid Application Development. Most frameworks support normal php code as well as extended php functions or classes; and quite a number of them boasts a learning curve of 1 to 2 days (considering if you have experience with php beforehand).

Zend Php(Free License) - ZendPhpFramework

Qcodo(Fully MIT) - Qcodo

Symfony(MIT) - Symfony

CakePHP(Fully MIT) - CakePHP

Wiki Software
Content Management System
Setting up Wordpress

Dokuwiki(Fully MIT) - Monobook for Dokuwiki, Dokuwiki

Mediawiki(Fully MIT) - Mediawiki

Joomla - Joomla

Typo3 - Typo3

Mambo

[Another big topic. please read other installation guide]

[ftp into your main domain]

[download the latest version of wordpress]

[upzip and upload to public folder]

[the following copies the wordpress setup from mydomain.com to mydomain2.com]

> cd /home/mydomain2.com/public

> cp -R /home/mydomain.com/wp wp

> chown -R mydomain2.com wp

[remember to regenerate API keys.]

[remember to create new db and update config file]

[recommended plugins: zero conf mail, configure smtp, google anilicator]

[make sure ownership is correct.]

[this might come in handy: > chown -R mydomain.com:www-data /home/mydomain.com/public]

[Adding lighttpd supported permalinks]

[Ref: http://www.cyberciti.biz/tips/lighttpd-and-wordpress-setup-clean-seo-friendly-urls.html]

Setting up Joomla

Go to Joomla website (http://www.joomla.org/)
Download the latest 1.0x version stable full package from the site.
Unzip joomla onto your local desktop and rename the folder to joomla
Setup a new connection with WinSCP to your webserver and make sure you are in the htdocs folder
Drag the unzipped folder of joomla into the folder.
Now try to access the joomla site with your browser. Eg (www.myjoomlasite.com/joomla)
You should get a new installation screen right now with all the requirements.
Going back to your winscp connection, create a folder called tmp in joomla folder on the server.
Now right click on the tmp folder and set permissons to 777 by checking all the checkboxes
Now create and upload a file called configuration.php. make sure its empty
Rename htaccess.txt to .htaccess
Edit the htaccess file and add the following lines to the end of the page

php_value register_globals 0
php_value session.save_path '/tmp/'

open globals.php and change

define( 'RG_EMULATION', 1 ); to
define( 'RG_EMULATION', 0 );

now alter all the folders and set the permissions to 777.. make sure all the requirements on the page is green :D
now set up and follow the sql server instructions
Finally when you have finished, remember to delete the installtion folder and all its contents. and volia!.. you have installed Joomla CMS

Php on IIS 6.0

I was trying to enable an extension (xml-rpc) on iis6 with php, but i didnt want to restart iis. The steps is to:

create a php file in your root folder or subfolder with a php code "phpinfo();"

navigate to that page and search for the extension you want.

locate the php.ini file on your server (check the phpinfo to get hints on where this is)

now enable the extension by removing the semi colon.

go to iis6 manager on your server

right click on default application pool

and select recycle.

check your phpinfo file again and the extension should be loaded

Drawing up a Web Service

2012-07-29T14:13:00.001+01:00

So you have a machine setup running which is a server. Your server should have an external IP address. This makes your machine address seekable. An IP v4 would look like 123.45.67.89. Nowadays IPv6 is both applicable. Well, we are not very interested about this here.

What we are looking at is DNS. Domain Name Server. There are a few out there such as GoDaddy, Register.com etc. Most are good. Avoid small unknown name servers. The prices are competitive so going to larger companies is the way forward.

Setting up DNS using your Domain Host (such as GoDaddy)

In your domain, just set the IP address to point to your new server IP (given to you by your provider, Rackspace or Amazon. You can do this by going on to the machine and typing

> ifconfig

Setting up DNS using RackSpace's servers

Some DNS guides at rackspace are rather complicated and hard to manage. I would leave it to using the domain providers or your hosting provider's DNS system. At rackspacecloud; and i presume most DNS settings have similar setting options, heres how its set:

you need

1x A record [which is IP],
1x CName record [which is www.mydomain.com],
1x MX record [which is mydomain.com; this allows asd@mydomain.com], and thats it.

if you need more sub domains you can add more CNames e.g. abc.mydomain.com

A record

Name = mydomain.com;
Content = 123.45.67.89; [content is your ip address]
TTL = 300;

CName record

Name = www.mydomain.com;
Content = mydomain.com;
TTL = 300;

MX record:

Name = mydomain.com;
Content = mydomain.com;
TTL 300;
Priority = 10 [if you use google apps, they will give you a huge list of 6 mx records, I know. its a pain. Enter them all by changing the Content, such as Content = aspx.googlemail.com]

Take a note of TTL which is time to live in seconds. 300 seconds would be 5 minutes. If DNS is changed it will take 5 minutes to propagate throughout the world. Normally people put 84600 which is 24 hours.

Install a HTTP server

You might notice why i call it a http server and not web server. Web server works but to make the term clearer so you remember what it is for, we stick to HTTP server. Apache / Cherokee / Lighttpd / thttpd and many more are just HTTP servers. All they do is sit as a service at port 80 (default) and listen to incoming requests.

So what about Php and LAMP stack etc? Those are good terms and technologies but i see them confusing lots of people. Php is just an interpreter. It takes in php scripts and processes it. Occassionally its used to spurt out HTML code. Aha! thats why you need a HTTP server to parse the HTML code to the world!

So with this in mind. You dont really need php to do server side scripting. Python, Ruby, even C is possible.

Installing Cherokee with Php5
Cherokee (is the alternative of Apache). Its non blocking style of http service make is outperform Apache in response. (http://www.cherokee-project.com/). To install it, open up terminal and type in the following,

> apt-get install cherokee

To call up an interface, run the following

> sudo cherokee-admin -b

The rest should be quite self explanatory.

Installing Lighttpd with Php5

apt-get install lighttpd php5-cgi

/etc/lighttpd/lightly-enable-mod fastcgi [1]

/etc/init.d/lighttpd restart [4]

Now we can proceed to create a skeleton for your domains. This is so you can reuse it when you are adding new domains. Ignore this if you have just 1 domain.

cd /etc/skel

mkdir -p {backup,logs,private,public/uploads} [5]

now every time you create a new domain user, this will create a new structure. If you go to your domain from any browser http://123.45.67.89 you should see a lighttpd page.

cd /etc/skel/public/uploads

nano index.html

add in the following:

<h1>Site under maintenance, please come back later</h1> [press crtl-x to exit nano]

Install php-myadmin

> apt-get install phpmyadmin

select lighttpd using spacebar then hit enter. when it asks to configure database for phpmyadmin, select yes. it will then ask for your mysql password, and ask you to set phpmyadmin password .

Before you start with .htaccess

If you don't have a .htaccess file in your directory just create a simple textfile with the name .htaccess, remember to save the .htaccess file with a "." dot. In Windows Notepad save as ".htaccess" (with quotes in the save dialog).

Lines starting with # indicates a comment and will note be parsed.

Prevent downloading of your SQLite database

If you place your SQLite database in a public www directory it could be downloaded by a visitor. See the SQLite quick tips for creating a SQLite database in a safer way. Add or remove file extension if needed. Place the following code in your .htaccess file.

Custom error pages, "404 page not found"...

If you want to set your own error page for "404 page not found" and for other HTTP status codes you can do that easily in a .htaccess file.

ErrorDocument 404 /yourOwn404Page.html

Just replace the "404" with the the code you need and the file that should be shown.

Disable directory browsing

If you have an image directory that shows all your files when accessing it, you can disable this by placing an .htaccess file with the following code (notice the minus sign)

# - (minus) disable browsing. + (plus) enable browsing
Options -Indexes

If you place a plus sign in front of Indexes it will list your files again. A simpler variant is just to place an index.html file in that directory, it will do the same thing. You decide :-)

Change default index page

Want to have another index page than the regular index.html, index.php page? Change the myStartPage.html in the following code and then add all other index pages you need. If myStartPage.html is not found it will try index.html and then index.php (and so on).

# reading left to right
DirectoryIndex myStartPage.html index.html index.php

Reference

Shamelessly quoted from: http://www.litewebsite.com/?k=tips

[1]some blogs say that we should add fastcgi-server={… config. The above was done without problems. perhaps bug fixed.]
so where are my http files? its at /var/www/
some blogs report memory leaks of lightly like a seive. Bahh. add in weekly cron server restart
[4] this will stop 403 errors when you try to access phpmyadmin remotely
[5] if you make a mistake you can use > rm -r xxxx to remove the xxxx directory.

Designing your webserver and access

The file you need to alter here is /etc/group

make sure theres a www-data group in the list. A record like "www-data:x:33:mywebsite.com,mysite2.com" means that mywebsite.com is a user of www-data

www-data is the group that controls your web sites. The idea is to add your first user account which is the administrator. "madmin" is my first user. I added this user to www-data to help ease traversing and controlling the websites by command line.

I have then made individual users by their website name to help ftping.

1) Have 1 admin user who is not the root

2) Have users for each domain to access FTP when needed.

Adding a new domain website

> useradd -d /home/mydomain.com -m mydomain.com

[note that adduser is different. it requires more details]

[rm -r to remove the directory from /home/mydomain]

[userdel to delete the user if wrongly created]

> passwd mydomain.com

> usermod -a -G www-data mydomain.com

[this adds your user to the www-data group]

> chgrp -R www-data /home/mydomain.com/public

> chmod -R 2750 /home/mydomain.com/public

> chmod -R 2770 /home/mydomain.com/public/uploads

[if you copy stuff from another directory using root remember to use following]

> chown -R mydomain:www-data /home/mydomain.com/public

> chgrp -R www-data /home/mydomain.com/public

> chmod -R 2750 /home/mydomain.com/public

[otherwise you may get ftp errors]

[lightly configuation and logs]

> nano /etc/lighttpd/lighttpd.conf

[find a nice place: scroll all the way down till: #### handle Debian Policy Manual, Section 11.5. urls]

[add in the following]

$HTTP["host"] =~ "(^|.)mydomain.com$" {

server.document-root = "/home/mydomain.com/public"

server.errorlog = "/var/log/lighttpd/mydomain.com/error.log"

accesslog.filename = "/var/log/lighttpd/mydomain.com/access.log"

server.error-handler-404 = "/e404.php"

}

> mkdir /var/log/lighttpd/mydomain.com

> chown -R www-data:www-data /var/log/lighttpd

[now lighttpd own this logs directory]

[you have to run this everytime you add a new domain]

> /etc/init.d/lighttpd restart

[this should restart the server nicely. if you have installed vsftp you should be able to put in a test.html file in the public folder and test it.]

[at this time you can configure your dns for mydomain.com to point to this ip address. This would be the A record]

FAQ:

Q: Why asd.html file is not found but sdf.html file is found when i type in 123.45.67.89 when both files are in the directory?

A: File permission settings might be wrong. check that www-data is the group and its chmod -R 2750 to the public folder

Install ftp server

> apt-get install vsftpd
> nano /etc/vsftpd.conf

[change the following: you have to remove # from the front of the line]

* anonymous_enable=NO
* local_enable=YES
* chroot_local_user=YES

[press crtl-x to exit and save]

> /etc/init.d/vsftpd restart [1][2]

NOTES:

[1] just by running vsftpd restart will give you a OOPs 500 error. You need to put "service" before vsftpd
[2] or you can run > service vsftpd restart
seems like vsftpd or other ftp software hooks onto user accounts. so for web servers, each domain should have a ftp account
remember to set chroot
by default openssh should be installed on ubuntu

FTPS for VSFTPD (Very secure ftp daemon)

Note that ftps is different from sftp.

> mkdir /etc/vsftpd
> openssl req -x509 -nodes -days 365 -newkey rsa:1024 -keyout /etc/vsftpd/vsftpd.pem -out /etc/vsftpd/vsftpd.pem
> chmod 600 /etc/vsftpd/vsftpd.pem
> nano /etc/vsftpd.conf

[Add in or make sure the following is uncommented.]

ssl_enable=YES
allow_anon_ssl=NO
force_local_data_ssl=YES
force_local_logins_ssl=YES
require_ssl_reuse=NO
# need require ssl reuse otherwise you get errors when you are editing files using ftps
# error "vsftpd: SSL connection failed session reuse required"
ssl_tlsv1=YES
ssl_sslv2=YES
ssl_sslv3=YES
rsa_cert_file=/etc/vsftpd/vsftpd.pem
listen_port=21
#you can use any port, filezilla defaults to port 21. So does wordpress
pasv_min_port=20000
pasv_max_port=20999
# you will need this pasv, otherwise when you set up firewall, you get ftp connection problems. as a sample use 20000 - 20999

> /etc/init.d/vsftpd restart

now you can ftp into your system using port 990 with ftps

NOTES:

Ref: http://ubuntuforums.org/showthread.php?t=518293

Draw up a Virtual Cloud Server

2012-07-29T14:12:00.003+01:00

Getting up your personal server pages is aimed to help novice users or technology hobbyists setup a personal server on the cloud environment. It is also aimed to help web developers setup a simple easy to manage web server commandline way style! Prepare the get your hands dirty.

If you have tried VPS, Shared Hosting and is fed up with preinstalled rubbish, we are on the same boat. This guide is to help you venturing into cloud virtual servers www.rackspacecloud.com and Amazon EC2. Alternatives are slicehost (which are mentioned alot and i have heard good stories). Rackspacecloud / Amazon EC2 has a low pricing structure at 0.02 cents per hour. With either, I recommend the ubuntu server LTS series. Its fast, does the job and supported by a huge fan base.

Why is this good? A simple answer is that bugs get fixed very quickly and new improvements keep happening. So you see all the big names on the internet, but what do all of them mean?

RackspaceCloud Hosting [256mb ram / 10gb HDD] / Amazon Micro Instance
Ubuntu Linux Server LTS Edition 12.04
LightTPD 1.5 (webserver used by youtube meebo wikimedia)
MySQL 5.x (www.mysql.com) / SQLite 3.6.x.x (www.sqlite.org)
Php 5.2.x (www.php.net)
vsftpd 2.2.x (vsftpd.beasts.org)
Python 2.6.x (www.python.org)
Wordpress 2.8.4 / 2.9.x / 3.x (www.wordpress.org)

The setups are done simply and mostly default. So it would work if your version is not too far behind or forward in time. Before we start, I assume you have account with RackspaceCloud / Amazon EC2. (other providers or personal machine may work; but I will be biased towards my installation methods)

Your options of choosing cloud instance (Virtual Machine)

As you venture into the cloud, you have a few options of all the yadada: "SAAS / PAAS / AAAS / DAAS....Private, Public..."

Here we summarise it in 3 categories:

Private Cloud (For in house use / Accounts / Sensitive Data)
Public Cloud (For general use / Website / Webservices)
Hybrid Cloud (Web management of inhouse stocks)

For a general web application or site service, lets focus on a public use of cloud, which is currently dominated by RackspaceCloud and Amazon.

We will not be stating too much on setting up your own Virtual Instance on Private Cloud however you will be able to get more information from googling OPENSTACK or DEVSTACK or OPENNEBULA

Virtual Instance on RackSpace

RackSpaceCloud ( www.rackspacecloud.com ) is a major player of cloud enabled web service. It has partnered with NASA to provide one of the largest open source cloud technology OpenStack.

At Rackspacecloud website, select order a new account and the product "Cloud Server". Follow the instructions to get an account, username and password. Once you get an account, you will need to follow the given instructions at the site and create an instance of the webserver.

For RackSpace users, in the control panel when you setup a new instance, you get an option to choose your favourite linux distro. I recommend ubuntu server LTS (Long Term Support).

> ssh root@123.45.67.89

Setting up an instance in Amazon

Amazon web services provides a range of web enabled services. Of which the one we should focus on is Amazon EC2. It is essentially a black naked box (nothing installed), allowing us to set it up as we wish. Amazon ( aws.amazon.com/ )

For Amazon EC2 users, you get to choose from a number of distros when you create an EC2 instance. I recommend reading the Amazon EC2 starter guide to setup your cloud https://help.ubuntu.com/community/EC2StartersGuide and getting your EC2 instance from http://cloud-images.ubuntu.com/releases/. As of writing, Amazon AWS(Amazon Web Services) does provide free micro instances for use and default Ubuntu Server LTS instances.

Lets whiz through the setup procedures. Register an account on Amazon AWS. The setup a primary key pair. You will have to download the key pair file, which is your private key to your machine. AWS will prompt you to download a PEM file. Keep this safe!

Next setup an instance and get an elastic IP. Get the basic EC2 Micro instance with Ubuntu Server LTS. Move on to the elastic IP and bind this ip address to your instance. Lets assume your IP is 1.2.3.4. On your local machine, open up terminal and type in the following (note that the IP address should be your elastic IP; ALEX is your username; mykey.pem is your AWS private key which you downloaded before.)

> ssh -i /Users/ALEX/Desktop/mykey.pem ubuntu@1.2.3.4

say yes if fingerprint RSA Key is needed prompt pops up. You should be in. Lets update the server packages while we are here. Type in the following.

> sudo apt-get update
> sudo apt-get upgrade

this will run loads of installs and type "y" for yes if you have to. After installing, reboot.

> sudo reboot

You will be thrown out of your server when you reboot. Wait for reboot and login using ssh. Welcome to your first virtual cloud server! Well done. Read more on Setting up Ubuntu Server.

Common Problems

While you are trying to SSH on to the machine, if you get errors, try this. on YOUR LOCAL MACHINE goto hosts file and delete the hostsrecord; look and delete the line with the same ip and delete it> nano ~/.ssh/known_hosts