1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 <2018> 2019 2020 2021 2022 2023 2024 | Index | 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 <2018> 2019 2020 2021 2022 2023 2024 |
<== Date ==> | <== Thread ==> |
---|
Subject: | Re: Why is post_event deprecated? |
From: | Dirk Zimoch <[email protected]> |
To: | Till Straumann <[email protected]>, <[email protected]> |
Date: | Thu, 8 Feb 2018 17:57:24 +0100 |
Hi Till,I don't know if there are any plans to skip support for post_event(). Probably not soon because nobody knows how much code there is using it.
But how to tell people to switch to the newer function if not marking the old one deprecated? If the compiler does not warn, many people would probably not even notice the new feature.
The compatibility implementation of post_event() is a bit of a hack and as seen in the thread "Scanning on event number fails with EPICS 3.16" it is not without issues.
But when committing the fix, I will probably remove the deprecated marker. (Tomorrow, now it's 6pm.)
Dirk On 08.02.2018 16:45, Till Straumann wrote:
On 02/08/2018 12:45 AM, Dirk Zimoch wrote:Hi Tim Isn't that what "deprecated" means: supported but discouraged?I thought it also indicates that support will go away (which would make maintaining applications that need to be built against newer as well as older versions of epics a pain). A side-effect are many compiler warnings that could obscure more serious issues.Dirk On 07.02.2018 00:43, Till Straumann wrote:I'm using some (old) code with 3.15.5 and learned that post_event() had been deprecated in favour of postEvent(). While I have no real problems with the new flavour which in some cases might be really nice - I do question the decision to deprecate post_event() (rather than, let's say discourage its use for new applications but continue to support it). 1.) There are probably many existing applications out there using post_event(). People have better things to do than migrating to the new version. 2.) The backwards-compatibility solution currently in place is cheap. 3.) People might have to use code with both, 3.15+ as well as older versions. It is a pain to have to maintain #ifdefs for this. My two cents... Best - Till