Subject: Re: I18n floating point conversions
From: Mike Nordell (tamlin@algonet.se)
Date: Mon Nov 20 2000 - 19:13:42 CST
> On Mon, 20 Nov 2000, Mike Nordell wrote:
[i18n floating point string representation]
Martin Sevior wrote:
> You can easily tell if the internationalization of Abi is
> sufficiently advanced to tell the difference between "." and "," based
> locale. Just look at the default tabs entry in the Tabs dialog.
Misery. :-(
It's not up to speed, though that code would temporary fix my problem for
0.7.12, but I'm not the one to give the call. However, my call is "Hold!",
with a very visible exclamation mark behind itī. If we're working our a**es
off to get a UCS-2, ASCII, Kanji and so forth relesee going, and then it
turns out we can't even handle the C (since we're forbidden to use the C++
locale) library in UI, wouldn't that make some users look at us like "what
*are* these persons doing?"?
So, how do we handle this. I'm in favor of a 0.7.12 release RSN, but on the
other hand I know that a release somehow gets developers going on new stuff
and sort of "forgets" about the earlier problems.
I don't see this as a 0.7.12 showstopper, but it would be a stopper for .13
or .14 since we know about it.
Words anyone?
Should we release a 0.7.12 with a known (hardcoded fix!) anomaly regarding
floating point values in this very limited area, and then *require* that
anyone ivolved in floating point display or parsing have a look at their
current usage and if appropriate use what we decide upon, or should we
simply hold 0.7.12?
> This issue never occured to me before so I haven't checked the code.
This might (I'm not saying it does) display that similar problems might be
present elsewhere.
/Mike
This archive was generated by hypermail 2b25 : Mon Nov 20 2000 - 19:13:33 CST