RFR: 8065138 - Encodings.isRecognizedEnconding sometimes fails to recognize 'UTF8'

[Martin Buchholz]

Great!

On Fri, Nov 21, 2014 at 3:16 PM, Magnus Ihse Bursie
<magnus.ihse.bursie at oracle.com> wrote:
>
> On 2014-11-21 22:40, Martin Buchholz wrote:
>>
>> A high-level followup ...
>>
>> Running most text-based OS tools, including sed and sort, is risky
>> because the user's encoding may be different from the encoding of
>> source files in the JDK (of course, this is particularly problematic
>> with properties files, which must be ISO-8859-1).  These tools will
>> probably work better when run with LC_ALL=C.  We should consider
>> changing the definition of SED and SORT to something like
>> $(ENV) LC_ALL=C sed
>
> I agree. We should use LC_ALL=C on a more generic level, preferrably by
> prefixing all, or a selected set of tools with that. I opened
> https://bugs.openjdk.java.net/browse/JDK-8065704 to track that.
>
>>
>> As for set -o pipefail, it is sad that every single pipeline in the
>> makefiles is vulnerable to failure to detect broken builds.
>
> Fortunately, we already require bash for building, so given that -o pipefail
> it is not a too new invention in bash, we can enable it quite easily. Erik
> has already opened a bug for that,
> https://bugs.openjdk.java.net/browse/JDK-8065576.
>
> /Magnus