2002 2003 2004 <2005> 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 | Index | 2002 2003 2004 <2005> 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 |
<== Date ==> | <== Thread ==> |
---|
Subject: | Re: Agenda change |
From: | Ralph Lange <[email protected]> |
To: | "Dalesio, Leo `Bob`" <[email protected]> |
Cc: | EPICS Core Talk <[email protected]> |
Date: | Wed, 21 Sep 2005 14:58:39 +0200 |
Dalesio, Leo `Bob` wrote:
Good. I hope not just some small print on the last slide. More than in a formal disclaimer, I'd be interested in how it's presented during the discussion. I would like to hear: hey, this is our first idea on how this could look like - how do you like it? What is missing? And not: This is it. Please send questions and complaints to the support department. Well - except for: - What's the string interface? - What's the array interface? - What does the surveyor interface look like? - Is the type description mechanism enumerated or functional? - Is there support for unsigned integers or not? Except for those basic issues, yes: it is pretty well set. This exactly is the still open issue that I was trying to get on last week's VC agenda: If Jeff's CA interface is the user interface and it uses DA, DA will also be the user level interface. If Kay's mortal data container is to be the mortal user interface, how can it be used with Jeff's CA calls which are using DA as their interface to data? Will there be a mortal Channel Access interface on top of Jeff's stuff that will use Kay's mortal containers instead of DA? Or will Kay's dead easy containers be implementing DA as well so that they can be used in the original CA calls? I doubt that we have an agreement on even the basic structure of the client lib. We have seen in mails of the last days (!!) that Jeff and Kay have different impressions of where their respective APIs are located. I fear that multiple talks on APIs held by a bunch of people that don't even share a conception of where these APIs will be located might be more confusing than helpful to a broad audience. Ralph |