ejeboo

Secure Windows Logon

noreply@blogger.com (Said) — Tue, 22 Sep 2015 10:53:00 +0000

Google Search Term Easter Egg

noreply@blogger.com (Said) — Fri, 03 Apr 2015 11:19:00 +0000

While trying to refresh my memory of what the word askew meant, I turned to Google.

Apparently, Google's SERP will be rendered in an askew way when you search for the word "askew".
That's a pretty neat way to drive home the meaning of a word :)

(It is purely a coincidence that I discovered this easter egg today which for Christian faithfuls is Good Friday.)

svn: The filename, directory name or volume label syntax is incorrect.

noreply@blogger.com (Said) — Thu, 22 Jan 2015 12:29:00 +0000

I recently came up against the following error while checking out code using Subversion (svn client v1.6.17) on a Windows 8.1 machine:

svn up
A    modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img
...
A    modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img\preloader-40x40.gif
A    modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img\sprite.png:Zone.Identifier A    modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img\avatar.png
svn: In directory 'modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img'
svn: Can't move 'modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img\svn-A5B0B865' to 'modules\distribution\repository\conf\ues\jaggeryapps\social\themes\default\img\sprite.png:Zone.Identifier': The filename, directory name or volume label syntax is incorrect.

The error message "Can't move 'path\to\file\img\svn-A5B0B865' to 'full\path\to\file\img\sprite.png:Zone.Identifier': The filename, directory name or volume label syntax is incorrect." wasn't particularly informative.

Attempts to do an "svn cleanup" failed. Based on the suggestions in the answers I found, I decided to delete the "lock" and "log" files in the ".svn" folder of the path mentioned in the error, as well as the contents of the ".svn/prop-base" and ".svn/text-base" folders.

The error remained. The error message sounded like svn was not able to move a temporary file into its final location on Windows. I navigated to the path of the file in the repo I was trying to check out from. It turns out that the file in question has a name that uses characters that would make it an invalid file name on Windows. The offending file was named "sprite.png:Zone.Identifier" -- it contains the colon ":" which is disallowed in Windows file names.

My workaround was to check out the code on a Linux machine and copy it over to Windows.

$15 Million Global Learning XPRIZE

noreply@blogger.com (Said) — Mon, 27 Oct 2014 11:57:00 +0000

The Global Learning XPRIZE which duration-wise, is a very very long challenge (it will be awarded in 2019) launched on September 22, 2014 and it:

"challenges teams from around the world to develop open source and scalable software that will enable children in developing countries to teach themselves basic reading, writing and arithmetic, empowering them to take control of their own learning and ultimately their future. Our goal is an empowered generation that will positively impact their communities, countries and the world."

Similar to the Ansari X Prize, there is a grand prize of $10 million to the winning team. Along the way, there will be a prize of $1 million each to 5 promising teams that are selected as finalists.

The X Prize was inspired by the Orteig Prize which was instituted in 1919. Image comes from Strive Masiyiwa's photo stream.

YUM: Cannot allocate memory

noreply@blogger.com (Said) — Sat, 11 Jan 2014 07:02:00 +0000

While trying to execute yum -y install nfs-utils nfs-utils-lib for the installation of NFS on a test EC2 micro instance running CentOS, I got the following error:

Total                                                                                                              1.6 MB/s | 644 kB     00:00
Running rpm_check_debug
Running Transaction Test
Transaction Test Succeeded
Running Transaction
Warning: RPMDB altered outside of yum.
Installing : libgssglue-0.1-11.el6.i686                                                                                                      1/7
Error in POSTIN scriptlet in rpm package libgssglue-0.1-11.el6.i686
error: Couldn't fork %post(libgssglue-0.1-11.el6.i686): Cannot allocate memory
Installing : libtirpc-0.2.1-6.el6_4.i686                                                                                                     2/7
Error in POSTIN scriptlet in rpm package libtirpc-0.2.1-6.el6_4.i686
error: Couldn't fork %post(libtirpc-0.2.1-6.el6_4.i686): Cannot allocate memory
Error in PREIN scriptlet in rpm package rpcbind-0.2.0-11.el6.i686
error: Couldn't fork %pre(rpcbind-0.2.0-11.el6.i686): Cannot allocate memory
error:   install: %pre scriptlet failed (2), skipping rpcbind-0.2.0-11.el6
Installing : libevent-1.4.13-4.el6.i686                                                                                                      4/7
Error in POSTIN scriptlet in rpm package libevent-1.4.13-4.el6.i686
error: Couldn't fork %post(libevent-1.4.13-4.el6.i686): Cannot allocate memory
Installing : keyutils-1.4-4.el6.i686                                                                                                         5/7
Installing : nfs-utils-lib-1.1.5-6.el6.i686                                                                                                  6/7
Error in POSTIN scriptlet in rpm package nfs-utils-lib-1.1.5-6.el6.i686
error: Couldn't fork %post(nfs-utils-lib-1.1.5-6.el6.i686): Cannot allocate memory
Error in PREIN scriptlet in rpm package 1:nfs-utils-1.2.3-39.el6.i686
error: Couldn't fork %pre(nfs-utils-1:1.2.3-39.el6.i686): Cannot allocate memory
error:   install: %pre scriptlet failed (2), skipping nfs-utils-1:1.2.3-39.el6
Verifying : keyutils-1.4-4.el6.i686                                                                                                         1/7
Verifying : rpcbind-0.2.0-11.el6.i686                                                                                                       2/7
Verifying : libgssglue-0.1-11.el6.i686                                                                                                      3/7
Verifying : 1:nfs-utils-1.2.3-39.el6.i686                                                                                                   4/7
Verifying : nfs-utils-lib-1.1.5-6.el6.i686                                                                                                  5/7
Verifying : libevent-1.4.13-4.el6.i686                                                                                                      6/7
Verifying : libtirpc-0.2.1-6.el6_4.i686                                                                                                     7/7

Dependency Installed:
keyutils.i686 0:1.4-4.el6

Failed:
libevent.i686 0:1.4.13-4.el6           libgssglue.i686 0:0.1-11.el6       libtirpc.i686 0:0.2.1-6.el6_4       nfs-utils.i686 1:1.2.3-39.el6
nfs-utils-lib.i686 0:1.1.5-6.el6       rpcbind.i686 0:0.2.0-11.el6

Complete!

The error was exactly that: the VM did not have enough memory due to the already running Apache, MySQL and Tomcat services I had running on VM with only 613MB of RAM.

Killing Tomcat freed up enough memory to allow the command succeed.

RIP: iPhone 4S

noreply@blogger.com (Said) — Mon, 09 Sep 2013 16:19:00 +0000

It's been exactly 3 weeks now since my iPhone 4S over-heated - it became unbearably hot that I couldn't even hold it in my hand. It has not shown any signs of life ever since.

Even the 2 different technicians that have taken a look have not been able to pin-point what the cause is ...

You will be sorely missed, iPhone :(.

SSH Brute Force Attempts

noreply@blogger.com (Said) — Wed, 22 May 2013 12:37:00 +0000

I found an interesting entry in the output of the netstat utility while trying to see the actively running services on an Amazon VM running CentOS.

netstat -pant
Proto Recv-Q Send-Q Local Address               Foreign Address             State       PID/Program name
tcp        0      0 0.0.0.0:27017               0.0.0.0:*                   LISTEN      996/mongod
tcp        0      0 0.0.0.0:80                  0.0.0.0:*                   LISTEN      1017/nginx
tcp        0      0 0.0.0.0:28017               0.0.0.0:*                   LISTEN      996/mongod
tcp        0      0 0.0.0.0:22                  0.0.0.0:*                   LISTEN      858/sshd
tcp        0      0 127.0.0.1:631               0.0.0.0:*                   LISTEN      803/cupsd
tcp        0      0 127.0.0.1:25                0.0.0.0:*                   LISTEN      952/master
tcp        0     48 10.204.214.243:22           173.203.XXX.XXX:43344        ESTABLISHED 21748/sshd
tcp        0    784 10.204.214.243:22           218.109.6.241:56570         ESTABLISHED 25742/sshd

The IP in the last line clearly isn't one I own. I did a quick IP2Location lookup and it is a Chinese IP address.
I did a Google search to learn more about what this IP is up to from the system logs and came across the 3 commands below.

grep -ir ssh /var/log/*
grep -ir breakin /var/log/*
grep -ir security /var/log/*

Running the first command on 2 different servers (CentOS and Ubuntu respectively) revealed loads of automated attempts at breaking into the servers I tried the command on.

/var/log/secure:May 22 08:05:00 ip-10-204-214-243 sshd[23741]: Invalid user server from 218.109.6.241
/var/log/secure:May 22 08:05:00 ip-10-204-214-243 sshd[23745]: input_userauth_request: invalid user server
/var/log/secure:May 22 08:05:00 ip-10-204-214-243 sshd[23745]: Received disconnect from 218.109.6.241: 11: Bye Bye
/var/log/secure:May 22 08:05:02 ip-10-204-214-243 sshd[23747]: Invalid user share from 218.109.6.241
/var/log/secure:May 22 08:05:02 ip-10-204-214-243 sshd[23751]: input_userauth_request: invalid user share
/var/log/secure:May 22 08:05:02 ip-10-204-214-243 sshd[23751]: Received disconnect from 218.109.6.241: 11: Bye Bye

/var/log/auth.log:May 22 04:43:08 ejeboo sshd[18376]: Failed password for invalid user test1 from 198.23.229.207 port 44166 ssh2
/var/log/auth.log:May 22 04:43:08 ejeboo sshd[18378]: Address 198.23.229.207 maps to host.colocrossing.com, but this does not map back to the address - POSSIBLE BREAK-IN ATTEMPT!
/var/log/auth.log:May 22 04:43:08 ejeboo sshd[18378]: Invalid user test1 from 198.23.229.207
/var/log/auth.log:May 22 04:43:08 ejeboo sshd[18378]: pam_unix(sshd:auth): check pass; user unknown
/var/log/auth.log:May 22 04:43:08 ejeboo sshd[18378]: pam_unix(sshd:auth): authentication failure; logname= uid=0 euid=0 tty=ssh ruser= rhost=198.23.229.207
/var/log/auth.log:May 22 04:43:11 ejeboo sshd[18378]: Failed password for invalid user test1 from 198.23.229.207 port 44599 ssh2
/var/log/auth.log:May 22 04:43:11 ejeboo sshd[18380]: Address 198.23.229.207 maps to host.colocrossing.com, but this does not map back to the address - POSSIBLE BREAK-IN ATTEMPT!

It's a sad world out there :(

Disable The Lock Screen in Windows 8

noreply@blogger.com (Said) — Wed, 13 Mar 2013 10:44:00 +0000

In case you ended up with a new PC running Windows 8 and need to disable the hugely annoying lock screen (might be good for tablets but mostly certainly lame on PCs) when all you want to see is the login prompt then you are in luck.

If you are running Windows 8 Pro edition then you can follow the instructions here: http://blog.laptopmag.com/how-to-eliminate-the-win8-lock-screen

If you are running Windows 8 (aka Core) edition then the instructions linked to above won't work for you since you'll likely not have gpedit.msc in your copy of the OS. Instead you'll need to tinker with the registry using the steps that follow.

1. Start the run dialog by pressing WINKEY + R, then type “regedit” and hit ENTER
2. It is strongly recommended that you backup your registry by using the "File" -> "Export" menu.
3. Navigate to HKEY_LOCAL_MACHINE\SOFTWARE\Policies\Microsoft\Windows\Personalization
4. You may have to create the final “Personalization” key. Just right click on “Windows” and select “New” and “Key”
5. Create a new DWORD (32-bit) with the name NoLockScreen and set the value to 1.

6. Once your changes appear like in the screen shot above, close the registry editor.
7. Confirm that the lock screen is disabled like in the screen shot below. You can do this using the Windows Charms menu (WINKEY + C) , click "Settings", then "Change PC Settings" at the lower right hand corner of the pane.

Credit goes to "Mike" in the comments on the original article.

Adobe, Quit Being Annoying by Asking Me to Upgrade to No End!

noreply@blogger.com (Said) — Thu, 20 Sep 2012 10:59:00 +0000

Obviously, the "Do not remind me about this update" check box doesn't work as I've clicked it for about a year now and it keeps coming back to harass me into upgrading Flash.

Adobe Flash remains the number one app that causes my MacBook fan to spin up whenever I open up certain Flash sites in Firefox. Adobe, fix your act!

Yahoo Mail 404

noreply@blogger.com (Said) — Thu, 30 Aug 2012 12:00:00 +0000

Today, I tried to check my Yahoo! mail account from my Acer Aspire One and this is what I got:

Not sure if this was a temporary change but http://mail.yahoo.com/ gave a benign not found error when I tried. I visited the Yahoo! homepage then accessed my mail from there instead. Reloading the URL worked afterwards.

If this change is permanent, them it might not be unconnected to Yahoo!'s recent hiring of Marissa Mayer from Google. They might want more users to hit the Yahoo! homepage rather than bypass it to read emails on their free account as that is what I mostly use them for these days. The more eyeballs that hit the homepage the higher their income from ads running on the Yahoo! network.

Working with IBM DB2 Express 10.1

noreply@blogger.com (Said) — Thu, 28 Jun 2012 23:08:00 +0000

IBM's DB2 database is not like any of the other RDBMS products I've used in the past.
It's a monster.

I'll try to relive some of the things I've learnt while trying to find my way around it for the past 2 days now.

I strongly recommend reading the DB2 documentation thoroughly when you run into a snag as it will immensely help you in crafting the right blend of keywords to use when searching for a quick solution. At the very least, you should read this page table space design which is a nice intro on the design philosophy used by DB2. A lot of DB2 documentation will refer to the terms you'll learn from that page so it is important you understand them.

While doing a database import:
db2 -stvf C:\Code\BOT.ddl > C:\Code\BOT.log | more

I got bitten by a silly block commenting mistake described in this question.
I was using "/*" and "*/" which sometimes worked but most of the time didn't. It often failed with:

SQL0104N An unexpected token "*/" was found following "BEGIN-OF-
STATEMENT".
Expected tokens may include: "". SQLSTATE=42601

The fix is to use a pair of "(=" and "=)" block comment delimiters since the SQL block that I intended to comment out was being executed from the Command Line Processor (CLP).

Link to the very well written DB2 documentation regarding CLP options is here.

Installing the 64-bit JDK 6 an Ubuntu 11.04 Slice

noreply@blogger.com (Said) — Thu, 17 May 2012 12:47:00 +0000

After reading the instructions here and here, below are the steps of what I did to get the 64-bit version of JDK 6 installed on a 256MB VPS from Slicehost running Ubuntu 11.04 (aka Natty).

NOTE: You will need to first of all download the x64 binary using a browser that supports cookies in order to correctly accept the Oracle terms of use before the download can commence.

$ wget http://download.oracle.com/otn-pub/java/jdk/6u31-b04/jdk-6u31-linux-x64.bin
$ chmod u+x jdk-6u31-linux-x64.bin
$ ./jdk-6u31-linux-x64.bin [Press 'Enter' when installation completes]
$ sudo mv ./jdk1.6.0_31 /usr/lib/jvm/
$ sudo update-alternatives --install "/usr/bin/java" "java" "/usr/lib/jvm/jdk1.6.0_31/bin/java" 1
$ sudo update-alternatives --install "/usr/bin/javac" "javac" "/usr/lib/jvm/jdk1.6.0_31/bin/javac" 1
$ sudo update-alternatives --install "/usr/bin/javaws" "javaws" "/usr/lib/jvm/jdk1.6.0_31/bin/javaws" 1
$ sudo update-alternatives --config java

TopCoder, Inc. 2001-2011

noreply@blogger.com (Said) — Mon, 28 Feb 2011 22:59:00 +0000

TopCoder, Inc is 10 years this year!

I'm willing to bet this year's TCO come September 25 - 28, 2011 will be really awesome! They are ditching Las Vegas, Nevada as the traditional venue for TCO finals this year in favor of Fort Lauderdale, Florida.

Another significant milestone is the number of SRMs: TopCoder's 500th SRM will come up on Saturday March 19, 2011.

Informix Tips

noreply@blogger.com (Said) — Thu, 27 Jan 2011 12:33:00 +0000

Informix command-line tips:

export CLIENT_LOCALE=EN_US.UTF8
export DB_LOCALE=EN_US.UTF8

: ensures any client can correctly connect to the server (see: UNIX setup guide)

oninit

: starts the server

onmode -ky

: kills the server

onstat -

: display server online status

oncheck -pr

: displays detailed status including dataspace info

onstat -g env all

: displays database server start-up environment settings

Executing:

cd /opt/IBM/informix
mkdir dbspaces
cd dbspaces
touch chunk1
chmod 660 chunk1
onspaces -c -d datadbs -p /opt/IBM/informix/dbspaces/chunk1 -o 0 -s 204800

will create a data space named chunk1 with size 204800 but will give a warning message like below which you can ignore.

Verifying physical disk space, please wait ...
Space successfully added.

** WARNING ** A level 0 archive of Root DBSpace will need to be done.

UPDATE: Here's a useful IBM article on Informix usage: http://www.ibm.com/developerworks/data/zones/informix/library/techarticle/0305parker/0305parker.html

Girl and Vulture

noreply@blogger.com (Said) — Thu, 30 Dec 2010 17:38:00 +0000

Still disturbing as when I first saw it in Time magazine several years ago...
Background. Original image was sourced from here.

Ways to Create Logical Backups for an MS SQL Server 2005 Database

noreply@blogger.com (Said) — Wed, 01 Sep 2010 17:15:00 +0000

There are three ways, that I know of, that allow you to create logical backups of an SQL Server database.

(1) Detach and copy the data and log files (.mdf) from the data directory;
(2) Create a backup file (.bak) from the Tasks -> Backup menu when you right-click on the database in SQL Server Management Studio (Express) and;
(3) Create an SQL backup file in DDL-format (.sql) using the Microsoft SQL Server Database Publishing Wizard;

Truncate or Delete All Tables in an Oracle Schema

noreply@blogger.com (Said) — Tue, 31 Aug 2010 12:40:00 +0000

It took a lot of poring over forums, faqs and of course a visit to the awesome programmer wiki to finally get a single block of PL/SQL code that did just this!

Most of the code comes from this answer and its basic algorithm is already described in this OTN forum post.

CREATE OR REPLACE PROCEDURE sp_truncate AS 
BEGIN
-- Disable all constraints
FOR c IN
(SELECT c.owner, c.table_name, c.constraint_name
FROM user_constraints c, user_tables t
WHERE c.table_name = t.table_name
AND c.status = 'ENABLED'
ORDER BY c.constraint_type DESC)
LOOP
DBMS_UTILITY.EXEC_DDL_STATEMENT('ALTER TABLE ' || c.owner || '.' || c.table_name || ' disable constraint ' || c.constraint_name);
DBMS_OUTPUT.PUT_LINE('Disabled constraints for table ' || c.table_name);
END LOOP;

-- Truncate data in all tables
FOR i IN (SELECT table_name FROM user_tables)
LOOP
EXECUTE IMMEDIATE 'TRUNCATE TABLE ' || i.table_name;
DBMS_OUTPUT.PUT_LINE('Truncated table ' || i.table_name); 
END LOOP;

-- Enable all constraints
FOR c IN
(SELECT c.owner, c.table_name, c.constraint_name
FROM user_constraints c, user_tables t
WHERE c.table_name = t.table_name
AND c.status = 'DISABLED'
ORDER BY c.constraint_type)
LOOP
DBMS_UTILITY.EXEC_DDL_STATEMENT('ALTER TABLE ' || c.owner || '.' || c.table_name || ' enable constraint ' || c.constraint_name);
DBMS_OUTPUT.PUT_LINE('Enabled constraints for table ' || c.table_name);
END LOOP;

COMMIT;
END sp_truncate;
/

In case you want to be able to issue a rollback then you should change the TRUNCATE TABLE (a DDL statement) to a DELETE FROM (a DML statement) so that Oracle can recover the data from its redo logs.

Executing this stored procedure is as simple as:

exec sp_truncate;

Run PL/SQL Fragments Directly in SqlPlus*

noreply@blogger.com (Said) — Tue, 31 Aug 2010 09:59:00 +0000

BEGIN
FOR i IN (SELECT table_name FROM user_tables)
LOOP
EXECUTE IMMEDIATE 'TRUNCATE TABLE ' || i.table_name;
END LOOP;
COMMIT;
END;

Ok so you want to run the fragment above that you got from a forum post somewhere and only have access to the command line SqlPlus* utility to do so.
You add more semi-colons, press the enter key repeatedly and hope it magically parses the snippet and determines that you must want to execute the contents of the BEGIN ... END block right?

Unfortunately it doesn't work that way :). You need to tell it to parse the contents of the buffer by adding the forward slash / to the end. Voila!

BEGIN
FOR i IN (SELECT table_name FROM user_tables)
LOOP
EXECUTE IMMEDIATE 'TRUNCATE TABLE ' || i.table_name;
END LOOP;
COMMIT;
END;
/

Please see this post for truncating/deleting all tables in a schema using a stored procedure in PL/SQL.

Changing the Author Name in Xcode

noreply@blogger.com (Said) — Fri, 02 Jul 2010 16:49:00 +0000

So you want to change the author name from the default (the full name of the logged in Mac user) from:

// Created by Firstname Lastname on 6/21/10.

to something more creative:

// Created by SpeedingLunatic on 6/21/10.

Here's how. Run this from a Terminal:

defaults write com.apple.Xcode PBXCustomTemplateMacroDefinitions '{ "FULLUSERNAME" = "SpeedingLunatic";}'

Charsets, PageEncodings and a JSP page.

noreply@blogger.com (Said) — Tue, 27 Apr 2010 22:08:00 +0000

I recently had an issue where the prototype HTML page for a JSP displayed OK on my browser but the converted JSP rendered some non-ASCII characters with weird symbols. To top that up, the JQuery animation refused to work.

The fix was simple, I had to re-save the .jsp file by providing a pageEncoding attribute at the top in my (Eclipse) IDE:

<%@ page language="java" contentType="text/html; charset=UTF-8" pageEncoding="UTF-8" %>

so that the page characters will not be saved with the default JSP encoding of ISO-8859-1.

The charset (contentType="text/html; charset=UTF-8") is the encoding with which the rendered JSP (final output) will be served to a browser while the page encoding (pageEncoding="UTF-8") is the encoding which the JSP compiler will use to understand the contents of the JSP page (saved as a text file) so that it can compile it correctly to a .class file.

View Status of Network Services

noreply@blogger.com (Said) — Thu, 22 Apr 2010 19:12:00 +0000

netstat is a simple command line utility for viewing network status available on Windows, Linux and Mac OSX.
To view network status of services using TCP sockets:

netstat -abn -p tcp

for UDP sockets:

netstat -abn -p udp

Tested on both Windows and Mac OSX.

netstat -pant

for TCP Sockets. Tested on a RedHat Linux VM

Debugging Password-less SSH Authentication

noreply@blogger.com (Said) — Sat, 03 Apr 2010 22:07:00 +0000

Whenever I need to frequently log into a Amazon EC2 server running Linux, the first thing I usually do is configure password-less login more properly known as public key authentication over SSH.

The steps described here are clear and to the point to help you get started.

The problem I had this time round was due to the folder/file permissions defined for the Linux user (apps) under which I was trying to login with. After some googling, I discovered that you could track SSH authentication attempts by 'tailing' this file: [tail -f] /var/log/secure which on some systems is: [tail -f] /var/log/security and then I noticed the error message --

Apr 3 17:12:36 domU-XX-XX-XX-XX-XX-XX sshd[2626]: Authentication refused: bad ownership or modes for directory /home/apps/.ssh

Rather than 'waste' time solving the permissions problem (after a few attempts I didn't make much progress) I settled with passwordless login as root! I know it's not exactly a good practise but hey the instance is not exactly running mission-critical apps anyway nor does it contain sensitive financial information ....

Neat Data Visualization Libraries for Flex

noreply@blogger.com (Said) — Sun, 28 Feb 2010 22:14:00 +0000

Here are some really nice data viz libraries for Adobe Flex that sport line, bar, waterfall, cluster, bubble charts and a lot more without requiring that you have a Flex Builder 3 Professional license:

(1) Axiis

(2) BirdEye

There are open source projects so you should be able to get the code for Axiis here and for BirdEye here (SVN checkout).

Multiple undo keystrokes and bam all your day's coding is gone!!!

noreply@blogger.com (Said) — Mon, 08 Feb 2010 21:48:00 +0000

Working on a fresh project all day that hasn't been setup in a Subversion repository and while testing out stuff in Flex Builder (which is essentially Eclipse under the hood) I made some Ctrl+Z key strokes to undo some commented lines and I noticed the build started to fail whenever I saved the reverted changes. (Flex Builder 3 is set to build on each save by default).

I ignored the initial build errors thinking heck, Eclipse had gone gaga from prolonged use and it started to freeze after brief shots of inactivity so the error would definitely go away once I perform a Clean then a Build. How wrong I was indeed.

By the time, I got over my drunkenness three source files in newly created sub-folders had vanished from my file system. My God. No. This can't be happening to me... all of todays's work gone foreever????

I figured OK maybe Eclipse was playing pranks by somehow hiding the files so I hit Terminal to do an:

ls -lpA

to view all the hidden files in the parent directory containing the missing files/sub-folders. No dice.

Next thing I'm doing

apropos undelete

and surprisingly enough I get an entry for the undelete C library function on BSD systems. How convenient. I have to dust off my C skills to write an undelete utility to avoid loosing my mind from lost work? No way.

Enter: Google
I fire up my EVDO connection and hit the search box "eclipse deletes source files" trying to point fingers squarely at Eclipse for my misery. The first result, a mailing list post from way back, precisely in 2001, where the author wrote that it was dirt easy to lose changes while working in Eclipse with someone chiming in that it is a known issue that is planned to be addressed in version 2.0.

How refreshing. This post is from 2001?? and this is like 2010 and Google somehow thinks this will solve my issue? Rather than jump off to the next result or crafting another search query, I decided to continue reading the thread a bit more as to why this ancient post should be ranked higher than the rest, I might actually learn something. Interestingly, I hit this little nugget of information regarding Eclipse:

A couple of workarounds until the issues are resolved...

As a side note, the same thing happened to me today. I was working on a
class and deleted it by mistake. What I did was load an old version from the
repository and then choose the class in the Navigator and selected "Replace
with edition from Local History". That seemed to work fine.

Then I quickly recollected how I used to ignore that menu item "Restore From Local History..." when you right-click items in the project navigator.

I highlight the parent folder, right-click to 'Restore From Local History...' and bam my files re-appear. They all contained all my most recent edit. Hurray!!!

Google's search technology is still unarguably their most valuable piece of IP on top of which everything else powered by Google resonates which is why they won't risk it being stolen if their continued stay in China still poses such risks.

I can't say for sure if I accidentally hit some keystrokes other than Ctrl+Z while working on this project causing Eclipse to think I wanted those newly created source files deleted or that multiple undo key strokes in quick succession can cause source files that were creating within a coding session (i.e. Eclipse hasn't been restarted) to vanish. Either way, I'm too scared to try to find out until I've backed up all my changes safely.

Enough pontificating, it's time to get back to work.

Installing Oracle 10g (10.2.0.1.0)

noreply@blogger.com (Said) — Fri, 05 Feb 2010 18:51:00 +0000

While installing Oracle 10g on a Windows 2003 (32-bit) server instance on Amazon EC2 I got a warning from the installer after it ran some diagnostics:

Checking Network Configuration requirements ...

Check complete. The overall result of this check is: Failed <<<< Problem: The install has detected that the primary IP address of the system is DHCP-assigned. Recommendation: Oracle supports installations on systems with DHCP-assigned IP addresses; However, before you can do this, you must configure the Microsoft LoopBack Adapter to be the primary network adapter on the system. See the Installation Guide for more details on installing the software on systems configured with DHCP.

If I had chosen to ignore the warning the installation would have still succeeded anyways but here's how to fix the warning message.

Oracle Database Preinstallation Requirements: 2.6.5.3 Installing a Loopback Adapter on Windows 2003, Windows Server 2003 R2, or Windows XP offers 21 steps to help resolve the warning on Windows 2003, Windows Server 2003 R2, or Windows XP. It also contains directions for other Windows variants (Vista etc).

I decided to shave off a few steps by going the command line route since I might have to perform a number of installations.

Step 1
Download the DevCon package, a self-extracting zip archive containing an utility provided by Microsoft to allow you perform tasks that you would otherwise do from the Device Manager (and a little extra). You can read more about the tool DevCon command-line utility.

Step 2
After extracting DevCon to a location on your machine, open a command prompt window and change to the directory containing the actual devcon.exe file.

Step 3
Steps 1 to 12 from 2.6.5.3 Installing a Loopback Adapter on Windows 2003, Windows Server 2003 R2, or Windows XP can thus be condensed to these two simple steps using DevCon:

devcon -r install %WINDIR%\Inf\Netloop.inf *MSLOOP

devcon restart =net @'ROOT\NET\0000

In case you loopback adapter device address isn't "ROOT\NET\0000" you can check what value it is via:

devcon find *MSLOOP

Step 4
You can now go ahead to complete the rest of steps 13 to 21.