You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@mesos.apache.org by "Adam B (JIRA)" <ji...@apache.org> on 2017/08/03 23:16:00 UTC
[jira] [Assigned] (MESOS-3449) Expand the range of integer
precision in json <-> protobuf conversions to include unsigned integers
[ https://issues.apache.org/jira/browse/MESOS-3449?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Adam B reassigned MESOS-3449:
-----------------------------
Assignee: (was: Joseph Wu)
> Expand the range of integer precision in json <-> protobuf conversions to include unsigned integers
> ---------------------------------------------------------------------------------------------------
>
> Key: MESOS-3449
> URL: https://issues.apache.org/jira/browse/MESOS-3449
> Project: Mesos
> Issue Type: Bug
> Components: stout
> Affects Versions: 0.25.0
> Reporter: Joseph Wu
> Priority: Minor
> Labels: json, mesosphere, protobuf
>
> The previous changes (MESOS-3345) to support integer precision when converting JSON <-> Protobuf did not support precision for unsigned integers between {{INT64_MAX}} and {{UINT64_MAX}}. (There's some loss, but the conversion is still as good/bad as it was with doubles.)
> This problem is due to a limitation in the JSON parsing library we use (PicoJSON), which parses integers as {{int64_t}}.
> Some possible solutions or things to investigate:
> * We can patch PicoJSON to parse some large values as {{uint64_t}}.
> * We can investigate using another parsing library.
> * If we want extra precision beyond 64 or 80 bits per double, one possibility is the [GMP library|https://gmplib.org/]. We'd still need to change the parsing library though.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)