2006-10-29

SMS: "ç" = pay twice!

Yesterday i did send an SMS to a German mobile. I was entitled to 160 characters! This should cost 0.102 €. I was confident in the length of the message, until I did send it: the message was cut into 3 messages and did cost me 0.306 €. I tried again to discover what was wrong, and this time a shorter message was cut in two pieces, costing me double the expected price. See the screenshots …

What happened?
Look at the screenshot of the SMS input window below, 38 characters are left.And when I type “ç” (a character used in Portuguese, French and some other languages) this does cost me a second SMS with only 11 characters left!

The same night the girl I did send the SMS to showed me the result on her mobile: they had arrived in reverse order! Hmmmmm … difficult to read!
I tried some other funny characters, and this is the result
ã – ç – û – ë – õ = dangerous characters
é – à – ù – ö – è = seem to be OK
Some characters are very expensive!
Skype 2.6 is definitely still “beta”!

6 comments:

Johan said...

That is a good one to report to the Skype development team!

Jean Mercier said...

Johan,
Ik weet dat ze in Estonia mijn blog lezen. Dus ze zullen wel op de hoogte zijn!
Op de koop toe heb ik een betatester geïnformeerd.
:-)

nirmalya said...

Thanks for this post about having to pay twice for that sybmbol. If I may add, I found sending SMS using Skype to be expensive for some of the countries I normally send SMS to. (mainly New Zealand, Malaysia & Singapore). I did a comparison with a lesser known (but reliable) site called smszilla.com. You can have a look at it here if you want.

Jean Mercier said...

Hello Nirmalya,

i removed the "double" comment and the apologies.

Thanks for the information.

mroach said...

Actually, this makes sense and is not a Skype bug. GSM SMS has its own character set called GSM 03.38. It contains most western European characters, but strangely, not a lower-case c-cedilla. See this chart: http://en.wikipedia.org/wiki/Short_message_service#GSM

So what happens is when you use a lower-case c-cedilla, the whole message has to be encoded using Unicode. Unfortunately UTF-8 isn't supported, so all of your characters get encoded with double-byte characters, thus reducing the number of characters you can type by half.

Possibly put a more clear way, when you use a character not in the GSM 03.38 set, every character in your message now takes up two bytes of space instead of one so you can only type half as much per message.

Jean Mercier said...

Hello mroach,

thanks for the explanation (in the meantime i knew it, but ok, through your comment, my blog is updated!).

But i would not say that "it makes sense"! ;-)