Tuesday, May 22, 2007

No Narrow Escape

There are a few benefits to not being single, physical comforts ranking rather high among them. There are also the downsides, and her shitty immune system is one of them. At this time tomorrow I expect to be laying there feeling like death and unable to breathe, much like she was this morning when I left for the office. This is not the sort of time off work I look forward to, and almost as bad as feeling sick as a dog is knowing that within the next 12 hours that's how I'm going to feel.

I wouldn't have come in today but for a couple conf calls which are simply easier to do when I have three work computers in front of me and I don't have to access virtual servers through a VPN. Connecting through the VPN at home on my personal boxen would kill some of those torrents as well as my FTP server. $MegaCorp blocks everything.

Everything except fuckwits.

we are currently use version 3.4.g, we have installed new server where are 9 languages, we have an issue with JPN. JPN server comes up , but all the screen Names are showing up as ??? marks however the data in the data bse shows up in Japaneese language

we checked all the components and check files and it shows all the latest pacthes has been applied as below

please adivce on this issue

A glance at the screenshots tells me all I need to know. Where characters should be are boxes. Incompatible, non-compliant font. They never modified the templates to use "Arial Unicode MS". I sent him easy-to-follow instructions. It's nothing but search and replace in a few text files.

We see your answer and making changes to all template file but see screenshot we are still have problem. We need solution because this is wrong with not seeing the texts.
That's unpossible. There's only one thing that causes boxes to be displayed instead of characters: the wrong font. Either they don't have a font which can display the character (also unpossible since they're using a native Japanese OS) or they specified the wrong font. That's it. There are no other possibilities. If you see boxes instead of characters, you have one of these two simple font problems. Clearly search for 'Arial' and replace ALL with 'Arial Unicode MS' wasn't complicated enough for them and therefore couldn't possibly be the solution, and I'm betting they got clever.

All I need is the font directory listing and templates. Nothing else unless they did something seriously fucknutted like changing the font substitution list in the registry. Still, they've already shown themselves to be thoroughly incapable of following simple instructions so it's time to put them to work and add eggs and milk. Along with what I actually needed I got server logs, screenshots, a registry export (just in case), a database extract... all in all about 600MB, of which I actually needed less than 14KB.

And it was right there.

{ font-family: ‚l‚r ‚oƒSƒVƒbƒN, Arial Narrow,Helvetica,Sans-Serif; font-size: 8pt; }

Mangled thanks to their editing it in fucking Notepad and saving it as an ANSI text file rather than Unicode. I recognised the mangled bit quickly: a Japanese font. Still, I had to unmangle it before I could properly chew them out.

Step 1: Open the file with a hex editor:
0000008bh: 2D 66 61 6D 69 6C 79 3A 20 82 6C 82 72 20 82 6F ; -family: ‚l‚r ‚o
0000009bh: 83 53 83 56 83 62 83 4E 2C ; ƒSƒVƒbƒN,

Step 2: Remove familiar and unnecessary crap which served to delimit the mangled garbage:
82 6C 82 72 20 82 6F = ‚l‚r ‚o
83 53 83 56 83 62 83 4E = ƒSƒVƒbƒN

Step 3: String the useful hex together:
82 6C 82 72 20 82 6F 83 53 83 56 83 62 83 4E

Step 4: Guess which encoding they're using
For Japanese, it's usually going to be Shift-JIS or Unicode. If it was the latter you'd see high bytes in the $30s, $40s, $50. These are all in the $80s which screams "Microsoft codepage 932: Shift-JIS"

Step 5: String bytes pairs together and hope your endianness is correct:
826C 8272 20 826F 8353 8356 8362 834E

Hmm... there's a space in there. I'm thinking this will be "MS Mincho" or "MS Gothic", typed by some halfwit in half-width.

Step 5: Character look-up:
At least Microsoft is helpful with I18N issues. They have a neat chart for codepage 932. First byte takes me to plane 82, position 6C: an "M", but not a standard ASCII "M". That would've made sense. No, this is a Japanese "Fullwidth Latin capital letter M" which has little in common with a normal 0x4D ASCII M. An "S" comes six positions later and I can count well enough in hex that I don't need to look up the second byte.

Then there's that "20", a space. Which fucktard out there decided computer font names should have spaces?

Working through it all I ended up with "MS Pゴシック", the "Gothic" font I expected. And right behind it, "Arial Narrow". Why? Because, well... maybe because Japanese people have some weird halfwidth font fetish.

I thought about the time I'd spent on this ticket so far and how much more time I was likely to have to waste in explaining to them what they needed to do, namely, use the right fucking font, stop adding wrong ones, and save text files in the correct code page.

Dear Sparky-san,

Please zip up and send me ALL of the HTML and CSS files from $Our$Big$App for analysis.


"For analysis" prevents questions.

Two days later they sent what I'd asked for. I fixed the dozen or so files which specified fonts inside 10 minutes. And as I finish writing this I'm confronted with the fuckwitted subject of tomorrow's entry if I manage to get out of bed. ファクヰト everywhere. ルウトコズ:十七
x-posted from HuSi, where there's a pointless poll.

Labels: , , , , ,


Post a Comment

Links to this post:

Create a Link

<< Home

In compliance with $MegaCorp's general policies as well as my desire to
continue living under a roof and not the sky or a bus shelter, I add this:

The views expressed on this blog are my own and
do not necessarily reflect the views of $MegaCorp, even if every
single one of my cow-orkers who has discovered this blog agrees with me
and would also like to see the implementation of Root Cause: 17-Fuckwit.