Globalization Eye Opener
I took a special class today at work that was specifically about testing for internationalization and "localizabilty", which means you are "faux localizing" your resources and testing before your translation is completed.
I've now been testing for almost 8 years! I tried to count, and I can safely say that this is at least my tenth product cycle. I've tested in most every language (Most Latin Based languages, the Americas, Japanese, Chinese, Korean, ect). I've never tested in Middle Eastern languages or Cyrillic (Russian, Eastern Europe).
Shockingly, having never tested on many features where sorting was of great significance, there is an entire area of testing that is brand new to me. Searching, GUI changes, Truncation, consistency, encodings,..these are all old news. Here is the biggest thing I learned all year: Not everyone sorts alphabetically in ways that make any sense to me. I've never even considered how in the heck to sort Asian characters.
It just makes my week to learn something so key, that I've been missing from my testing knowledge all of this time without ever knowing it! It wasn't just a lightbulb moment for me, it was one of minor shame that I could miss getting such an important and obvious concept. I knew this was important for search, for metadata, and have written great bugs about those items. Right now I can't wait to get in there and use my new information to break some sorting. I'm going to be looking like mad for ways that sorting is possible.
It has been so long that I forgot, many years ago I created a hilariously geeky set of test files which tested input/output of every possible UTF-16 character possible both in the file name and in XML content and allowable tag names each individually named so that it was very quick to identify problems. I think input/output is a wonderful area to automate, and I had to share that I have had good results when automating mundane tasks such as verifying correct encoding and other functionality tests. Automation is not so good at finding bugs, but it is very good at doing some functional testing, so long as the expected result is pretty well defined.
In other thoughts, MS Arial Unicode is the biggest test font I've found for doing unicode testing yet, but is there anything as large and complete in number of characters for the Mac platform specifically? Help me oh fellow lovers of Steve.
I've now been testing for almost 8 years! I tried to count, and I can safely say that this is at least my tenth product cycle. I've tested in most every language (Most Latin Based languages, the Americas, Japanese, Chinese, Korean, ect). I've never tested in Middle Eastern languages or Cyrillic (Russian, Eastern Europe).
Shockingly, having never tested on many features where sorting was of great significance, there is an entire area of testing that is brand new to me. Searching, GUI changes, Truncation, consistency, encodings,..these are all old news. Here is the biggest thing I learned all year: Not everyone sorts alphabetically in ways that make any sense to me. I've never even considered how in the heck to sort Asian characters.
It just makes my week to learn something so key, that I've been missing from my testing knowledge all of this time without ever knowing it! It wasn't just a lightbulb moment for me, it was one of minor shame that I could miss getting such an important and obvious concept. I knew this was important for search, for metadata, and have written great bugs about those items. Right now I can't wait to get in there and use my new information to break some sorting. I'm going to be looking like mad for ways that sorting is possible.
It has been so long that I forgot, many years ago I created a hilariously geeky set of test files which tested input/output of every possible UTF-16 character possible both in the file name and in XML content and allowable tag names each individually named so that it was very quick to identify problems. I think input/output is a wonderful area to automate, and I had to share that I have had good results when automating mundane tasks such as verifying correct encoding and other functionality tests. Automation is not so good at finding bugs, but it is very good at doing some functional testing, so long as the expected result is pretty well defined.
In other thoughts, MS Arial Unicode is the biggest test font I've found for doing unicode testing yet, but is there anything as large and complete in number of characters for the Mac platform specifically? Help me oh fellow lovers of Steve.
Try http://www.code2000.net/">http://www.code2000.net/">http://www.code2000.net/ (has a good set of test pages too!). There's also http://www.alanwood.net/unicode/fonts.html">http://www.alanwood.net/unicode/fonts.html">http://www.alanwood.net/unicode/fonts.html, which has a decent list. You also should be able to use the Arial font on your Mac OS X system, at least from a technical perspective. It natively supports TrueType (not to mention .dfont, Multiple Master, OpenType, TrueType Collection, and Type 1 PostScript).
Good luck!
Reply to this
Okay, let's try this again, since it obviously eats tags:
Try http://www.code2000.net/ (has a good set of test pages too!). There's also http://www.alanwood.net/unicode/fonts.html , which has a decent list. You also should be able to use the Arial font on your Mac OS X system, at least from a technical perspective. It natively supports TrueType (not to mention .dfont, Multiple Master, OpenType, TrueType Collection, and Type 1 PostScript).
Good luck!
Reply to this