Monthly Archives: April 2008 Issue #86811: Arabic numbers instead of decimal numbers

Issue #86811 is still of my top priorities this week, as it causes a major problem for Hebrew speaking users of As I can’t fix the code myself, I’m focusing on explaining that this is a major problem and it should be fixed in the 2.4.1 release.

Before explaining more about the bug, let me clear a bit some terminologies: Arabic numbers are the numbers you know from English and most languages (including Hebrew). Hindu numbers are the one in use by the Arabic languages (Arabic, Persian, Urdu, etc). Notice that this is confusing as the numbers in the Arabic languages aren’t Arabic numbers but Hindu numbers.

Issue #86811 is about how does handles numbers during import of Microsoft Word documents. The current problem is that during the import, the numbers change from Arabic to Hindu. This makes it impossible for Hebrew speaking users to read Microsoft Word documents which include numbers. See screenshots attached to issue #87625 (dup of #86811).

As in Israel there is still a majority for Microsoft Office over, this makes users isolated from getting documents from other people. This is the real reason for changing the bug priority to a higher one.

Issue #86811 currently have 80 votes (and counting)… Some Linux users are even trying to convince their distributions to exclude version 2.4.0 from their coming release (See Ubuntu bug #210204 and Mandriva bug #38874). I personally think that this is wrong, as 2.4.0 should be included in the upcoming releases, but we should work hard to get 2.4.1 into those distributions when the fix for #86811 is ready.


Hello planet (:



Filed under

Where did the numbers go ? (major bug in 2.4.0)

About two months after the Israeli government stopped the support for localization, we hit a major bug (actaully – a regression) with version 2.4.0 effecting Hebrew users.

In version 2.4.0 numbers written in Microsoft Word documents are shown as Hindu numbers instead of Arabic numbers. There’s a big confusion in the terms, as the number used in Arabic are the Hindu numbers and the number used in Latin languages are the Arabic numbers.

Anyway, this means that, the facto, 2.4.0 can’t show properly Microsoft Word documents which were written in Hebrew. I can’t start to describe how much interoperability problems that causes in Israel.

Which such behavior I can’t recommend to people, and organizations already moved to previous versions of can’t upgrade to 2.4.0 and must stay with 2.3.1. This is also a problem as most Linux distributions (Debian, Ubuntu, Fedora) intend to include 2.4.0 for their stable release.

With all the RTL related bugs, was still usable. But this bug absolutely changes things. I hope it will be fixed for 2.4.1 and won’t be delayed for version 3.0.

I wish I could say this is an April’s fools post, but that bug is a reality… Please vote for this bug.


Filed under