Sunday, July 22, 2007

Tedious Defect

Monday (16 July 2007)

This is a hard war. I am working on a defect which is having calculation error in report. At first I suspect there is a double entry error. This could be a new challenge on me as I never work this before. I am spending the whole day to look for the real problem in the source code, resource file and the work flow. Unfortunately all the effort are wasted in this day because nothing can be found.

Tuesday (17 July 2007)

There was no error in double entry because the calculation was a real time process during the "report making" process.

Today I dive into the code to see how double entry was created, how the journal entry is made and how the recap function work. Actually the error is not here. If this is an error, there must be an error in double entry. Since double entry is working fine. There must something wrong in other place. I stop my analysis work on double entry and request help from my senior. After the discussion, I suspect there must be some work need to be done in resource file.

Wednesday (18 July 2007)

I have found something in the resource file. This is very tricky. Every ID defines in resource file represent fields in the data file. This field may have some attribute apply on it to indicate how the field is 'work'. A few changes were made and it works finally. But I need to test other part of the report to ensure my work is complete.

Thursday (19 July 2007)

I am doing the wrong thing. All my works are gone. My effort was wasted. I am not yet clearly understand the real problem. I am not yet clearly understood the work flow. I though I have complete my work. But the fixed I have done will be reactivated in someday. I have no clue how this thing should be fix.

Friday (20 July 2007)

YES!! I have completed the fix. The report is showing the correct value in both invoice and purchase. I couldn't sleep well on yesterday night just because of thinking a 'real' solution for this problem. Now it is working fine here.

Anyhow I found an error during the crash test. I am almost complete my work. At this moment I need to calm down to check every test detail. It should be done in one day effort.

New Explore: MCNLive

On last weekend, I found a mini OS called MCNLive. MCNLive is base on the Mandriva 2007. It just cost me around 440MB and without the need to installed it into hard disk. Everything are load from the CD and execute from RAM when it is needed. Thus I did a test on it. I burn it into CD, and boot up from the CD with my PC which have Windows XP installed.

When it first boot, I see a series of boot option. The first boot option is default option. (I like this option.) Choose on it, wait for a few second and it is ready to be use. One thing I notice that the CD ROM was locked. When U load an application, I saw the CD ROM is spinning. I'll never eject the CD until I shutdown the OS.

I have try on other boot option like "some thing called boot from NTFS hard disk", I don't really remember the full name. With this option, I can eject the CD after the OS is bootup. This is the most interesting part. The other is called "boot from USB". I get stuck the middle during the boot process. I don really know what happen. May be I don have a USB ready during that moment.

At last, I can have an option to remaster the OS. This is really interesting. When I found something is not install in the OS. I can install it and remaster the CD. Something like customization. This is so great. But the OS size might increase depends on the application installed. I never try on this. Will make it one day.