Font of Knowledge

There are many levels of stupidity. Drinking too much on a work night is just plain stupid. Drinking until 3:00a.m., then coming home, turning on all the lights, having another beer while playing computer games with the volume turned up to 11, waking up the guy who drank too much and has to go to work in the morning to pay the bills is really incredibly stupid. But there's a level of stupidity so mind-bogglingly high, so gobsmackingly, migraine-causing, teeth-gnashingly extreme that it makes Steve Martin's character in Dirty Rotten Scoundrels look like the Oracle of Delhi in comparison. I shall let Harish explain it for me in a ticket copied verbatim with only identifying bits munged.

Harish works for $BigBank. $BigBank decided that it would probably be a good idea if the Japanese customers could type in their native language. $BigBank got all Unicodified with the help of our guys on the scene. Huzzah! But then they left $BigBank to their own devices, said devices being sniveling, drooling incompetents who probably type with a pointer stick attached to their safety caps.

We're rolling out $YourBigApp to multiple countries (including Asian countries) and hence, there is a need to enable capability to enter and retrieve multi-lingual characters. To do so, we've made the required changes in all templates by mentioning 'Arial Unicode MS', wherever 'Arial' was specified as the font. It works and we're at UAT stage now.

However, we've observed a strange behavior wherein a few testers reported that they are not able to view the special language characters and instead viewed junk characters (Screen shot 1 attached).

On further investigation, we found that there is a font (Arial Unicode MS), which should be installed on tester's machines. It was installed on their machines, but we copied this font from other tester's machine, where language characters were appearing correctly. On doing so, testers, who reported the problem, were able to view the expected characters (Screen shot 2 attached).

However, to determine the root cause, now when we replace fonts to old fonts, the problem does not persist. We don't know that what has fixed the issue? Is it that when we copied the fonts first time, issue was fixed permanently? Now those testers see junk characters only when we 'Remove' the font altogether from their respective machines. What are your thoughts and what should we take into account while rolling out the application?

Please note that on a Windows 2000 machine, we're navigating to Start->Settings->Control Panel->Fonts to perform the font copy-paste operations.

I was too stunned to smash my head into my desk to relieve the pain. All I could think of was this bash quote. Why is this guy alive? How the fuck does this man remember to take the fork out of his mouth before chewing? How does he remember to breathe in?

If you missed it for all the words in there, let me highlight the important bit:
Is it that when we copied the fonts first time, issue was fixed permanently? Now those testers see junk characters only when we 'Remove' the font altogether from their respective machines.

He installed the font and everything worked. Then he deleted the font and it stopped working. And this surprises him, the "system administrator".

Perhaps this will stop the J'accuse comments about my intolerance. Enough's enough! Give me my fucking Root Cause: 17-Fuckwit already!