|
Geometry : |
Subject: Re: [geometry] Crash in rtree after growing shared memory, VS2012 x64, Boost 1.55
From: Tim Finer (tim.finer_at_[hidden])
Date: 2014-11-26 12:59:17
Hi Adam,
On 11/26/14, 3:33 AM, Adam Wulkiewicz wrote:
> Hi Tim,
>
> Tim Finer wrote:
>> On 11/25/14, 4:25 PM, Adam Wulkiewicz wrote:
>>> If you're asking about an info about the change of an internal
>>> structure it isn't mentioned anywhere. There is only an info about
>>> fixing a bug with Interprocess.
>>>
>>> I dissagree that this should be mentioned in the docs, it's an
>>> internal change. Furthermore AFAIU Interprocess doesn't guarantee
>>> that the representation of data will be the same in various versions
>>> of Boost nor it wasn't designed to support versioning so we
>>> shouldn't rely on this. Btw, Serialization supports versioning.
>> Sure, I understand that there are no guarantees between versions. I
>> might've misunderstood your warning though. I thought you were
>> advising not to use 1.55 IPC and bg rtree in general (without
>> updating) because of an internal polymorphic node? If so, then I
>> don't see why the documentation with the sample shouldn't be edited
>> to say something like "don't use this until 1.56" or something.
>
> I was advising not to use 1.54 or 1.55 with Interprocess. It's
> mentioned in the release notes for 1.56
> (http://www.boost.org/doc/libs/1_57_0/libs/geometry/doc/html/geometry/release_notes.html#geometry.release_notes.boost_1_56),
> in Bugfixes section: "rtree crashed in some cases when used with
> Interprocess allocator".
>
> Documentation is built and redistributed with Boost, a copy is placed
> on a webpage. It's not possible to add an info in docs of already
> released versions. To do it we probably would be forced to release
> e.g. version 1.xx.1 and nobody would agree to do that for adding a
> message in the docs, and after a release period. Not to mention that
> there are already 2 newer versions of Boost released. So this info
> could be added in 1.58 but wouldn't be visible in the documentation
> for older versions. It probably isn't a good solution since users of
> 1.58 would just ignore it, and users of 1.55 wouldn't see it.
Ah, OK. Thanks for the clarification. I appreciate all your patience.
Best,
Tim
Geometry list run by mateusz at loskot.net