Re: Getting server crash on Windows when using ICU collation

Hi,

On Sat, Jun 17, 2017 at 6:38 PM, Peter Eisentraut
<peter(dot)eisentraut(at)2ndquadrant(dot)com> wrote:
> On 6/16/17 23:46, Amit Kapila wrote:
>> I have just posted one way
>> to determine if icu library has support for ucol_strcollUTF8, see if
>> that sounds like a way forward to you.
>
> I'm not in a position to test such patches, so someone else will have to
> take that on.

Well, I have tested it from my end. I've basically tried to test it by
running multi-byte tests as Amit suggested and by verifing the things
manually through debugger. And, i had mistakenly attached wrong patch
in my earlier email. Here, i attach the correct patch. Sorry about
that.

>
> It might not be worth bothering. ICU 50 is already about 5 years old.
> If you're packaging for Windows, I suspect you have the option of
> bundling a version of your choice. The support for older versions is
> mainly to support building on "stable" Linux distributions, and even
> there the window of usefulness is closing. (CentOS 7 has 50, CentOS 6
> has 4.2 which is too old for other reasons, Debian stable has 52 (and it
> will become oldstable after today)).

Yes, it's hard to find any user's having ICU version < 50 installed on
their system. But, having said that, it's always good to have such
detective checks basically considering that we already have such
configure check for Linux platform. Thanks.

--
With Regards,
Ashutosh Sharma
EnterpriseDB:http://www.enterprisedb.com