- From: Nigel Megitt <nigel.megitt@bbc.co.uk>
- Date: Thu, 14 Dec 2017 17:31:38 +0000
- To: TTWG <public-tt@w3.org>
- Message-ID: <D65865CA.50B77%nigel.megitt@bbc.co.uk>
Thanks all for attending today's TTWG meeting. Minutes can be found in HTML format at https://www.w3.org/2017/12/14-tt-minutes.html In text format: [1]W3C [1] http://www.w3.org/ Timed Text Working Group Teleconference 14 Dec 2017 Attendees Present Thierry, Nigel, Andreas, Pierre, Philippe, Cyril, Glenn Regrets Chair Nigel Scribe nigel Contents * [2]Topics 1. [3]This meeting 2. [4]F2F 3. [5]TTML2 Timeline 4. [6]Charter 2018 5. [7]TTML1 agenda issues and pull requests 6. [8]Deprecate offset-time and fraction component in smpte time base #283 7. [9]Clock-time form time expressions allow 60 seconds for media timebase. #210 8. [10]Clarify intended rendering of background color when applied to a span. #209 9. [11]tts:overflow does not apply to the region area #239 10. [12]IMSC issues 11. [13]Clarify that tts:fillLineGap does not hide characters #283 12. [14]TTML2 processor behaviour vs TTML1 13. [15]TTML2 WR labels 14. [16]Agenda Order 15. [17]Meeting close * [18]Summary of Action Items * [19]Summary of Resolutions __________________________________________________________ <scribe> scribe: nigel This meeting Nigel: For today, there were a couple of agenda requests additional to the agenda I sent out: ... Charter and TTML2 publication timeline. ... Also Andreas requested that we move coordination and organisation topics to the end of the meeting. Thierry: I started to draft the Charter, but it needs review, and what needs to be there is ... to focus on the different timelines of the different specifications. We need a timeline ... for each document now. Nigel: I want to quickly confirm the F2F details, too. ... So the order is: ... 1. F2F confirmation ... 2. TTML2 timeline ... 3. Charter ... 4. TTML2 as Edited Rec or go via CR ... 5. TTML1 agenda issues and pull requests ... Then I'm not sure if there are IMSC or TTML2 issues to discuss. Andreas: If time allows, we should cover the pull request about fillLineGap clarification. <Zakim> plh, you wanted to clarify 4 Philippe: We have a process for revising Recommendation, called Edited Recommendation, ... and there are two choices, both of which generate an Edited Recommendation, and you ... don't need to obsolete the previous one because it automatically outdates the previous one. ... If you go to Second Edition you'll get a popup saying not to look at it but to go to Third Edition instead. Glenn: You'll still be able to find the 1st and 2nd Editions? Philippe: That is correct. Glenn: And the generic link will go to Third Edition. Philippe: Correct. Nigel: Any other agenda topics? Glenn: A minor TTML2 item - I'd like to go ahead and convert that over to use the new ... build process with Travis, now that it works for TTML1. Philippe: Port it all to TTML2 spec already? Glenn: Yes. Philippe: OK, if you and Pierre agree that the TTML1 process is working I'm happy to replicate it. ... Open an issue in the TTML2 repo and assign it to me and I'll open a pull request against it. Glenn: OK. ... Nigel we might need your help to tweak the build.xml file Nigel: Okay. F2F Nigel: We have 9-10 Jan 2018 in Building 3 at Apple confirmed. ... Thanks to Apple for volunteering to host. ... Any requests for start time? I imagine a 9am start would probably be okay. Glenn: I could attend an Editor's meeting on the Monday. Cyril: Netflix could host that - let's work out the details offline. Nigel: I expect my flight to arrive late afternoon on the 8th. TTML2 Timeline Andreas: I wanted to discuss what the proposed date means for us - we have set dates ... 4 times before and every time we missed the date. The last time was in January when we ... expected a CR in August or earlier. For the Charter it may be important to think about the ... possible consequence of not reaching that deadline. Otherwise we just keep rolling on each year. ... I would like to start a discussion about that. ... TTML2 is needed for other publications so there are some dependencies. ... One very harsh (I don't recommend it) option is to stop an activity if it does not meet its milestones. Thierry: I'm not sure how we can enforce that deadline or timeline but now in the new Charter ... template we need to specify an expected completion date for each specification, so we ... will need one for TTML2. It will I guess have a stronger meaning because it is in the Charter. Nigel: Philippe, what do the Charter timelines mean? Philippe: They set expectations to the working group, and potentially give permission to the ... Director to say sorry you did not make it but we are not going to extend the WG. ... There are multiple considerations - while you guys are doing the work you might be ... stopping other work for happening, so the strategy guys might pull resources, or you ... may be told to go back to incubation and be sent back to a different kind of group than ... a Working Group. The risk is for you guys if you don't meet your Charter. Thierry: That doesn't resolve exactly what Andreas pointed out - there are different dependencies ... based on TTML2 - saying no team contact resources does not help issue TTML2 as a Rec. Philippe: Are you saying all the dependencies are in W3C and it will stop other publications? Andreas: There is the dependency to IMSC 1.1 and at the moment some members have ... the requirement only to publish IMSC 1.1 as a subset of TTML2. Philippe: Then you would not be able to publish IMSC 1.1 or TTML2. Andreas: Exactly. What I'm looking for is how you could better achieve milestones. If there ... are no consequences to missing a milestone we can drag on for years and years. Philippe: I agree. Being blunt, no other WG in W3C is dependent on TTML2. It's a lot easier ... for the Directors to close down the WG than, e.g. CSS WG, on which lots of things depend. ... I think I've been pretty clear to Nigel in the past as well, you are going against other ... priorities of the Consortium and TTML is not high in those priorities. Thierry: If video is a first class part of the web then subtitles and captions have to be ... addressed. The web depends on that working. Philippe: No, the web does not depend on subtitles. This is like a lot of other activities ... like MusicML that we don't have any more. I'm being honest here. ... Luckily for you guys I'm not making this decision. Nigel: We had Wendy with us at the F2F and she was happy with the TTWG's work from ... a strategic point of view. What we're discussing here is the consequence of not completing ... our work within the Charter stated deadline. Philippe: Where are you with TTML2? Nigel: We're closed to new feature requests Glenn: We're processing Wide Review comments prior to publishing a CR. Philippe: At some point you may have to cut things out of the spec if you can't get to agreement. Glenn: I agree, I already have a list of some potential at risk features and even some to pull ... from the CR. Philippe: It's always painful doing that, but sometimes you have to. Glenn: The only features to postpone are those where we cannot resolve the spec text in ... a reasonable time to make the CR. I don't have any issues with taking things out that ... fall into that bucket. ... We have at least one proposal for deferring those features. Andreas: Although the process or something that has to do with keeping the WG alive is ... an option to work with, I was hoping to find an agreement between the WG members to ... decide what happens if TTML2 passes a certain fix date when it should be moved to CR. ... The CR date cannot be in the Charter, and anyway should be shortly after the next Charter. ... It should be up to the group members to decide how to meet our milestones. Cyril: I agree the timeline has been postponed, and that is not satisfactory. I don't know ... how to enforce it. For example in the SVG WG, somewhat similar to TTWG, I hope we don't ... get to that state. ... I'm fine with the procedure to remove things from TTML2 if the publication deadline is at risk. ... Or creating modules, for example a Japanese module. Nigel: We discussed the TTML2 timeline already at [20]https://www.w3.org/2017/11/09-tt-minutes.html#item41 ... and we're on target for what we discussed. ... In that discussion we agreed to try to get to CR by Feb 1 for a Rec by July 1. ... Realistically this means we must have agreed any deferrals by the end of the call on 25th Jan. [20] https://www.w3.org/2017/11/09-tt-minutes.html#item41 Pierre: I want to challenge one of the repeatedly made statements, which is inaccurate and ... self-defeating, which is about TTWG not delivering anything. IMSC 1 was published last ... year and has since become the format referenced worldwide. This was followed by IMSC 1.0.1 ... published as a CR about a year later, responding to specific industry requests. So TTWG ... has been active and has successful work output referenced by many other organisations. ... The status of TTML2 should not cloud the success of TTWG elsewhere. ... I would hate for someone to make a decision about TTWG just about TTML2 ignoring the ... group's other successes. ... Secondly, for TTML2 progress there's no secret - we just need to close bugs. For instance ... on TTML1 Third Edition since TPAC we've closed 14 issues and have 13 pull requests ... open and I fully expect all issues on TTML1 to be closed by end Jan, either resolved or deferred. ... I'm completely confident that by October 2018 we'll have IMSC 1.1 published. Andreas: I would agree with everything Pierre said. The existence of the WG and the question ... how we can enforce milestones are separate issues that should not be connected together. ... Going back to the concrete timeline where we say we have a realistic expectation of ... TTML2 reaching CR by Feb 1. Can we then agree that in 25th Jan meeting if that would ... not happen then we can discuss other options for example modularising TTML2. We have ... had those deadlines for years before and knew about the strategies but did not meet them. ... Can we agree to discuss other options if we see we won't make it on 25th Jan. Nigel: We are not passive here - we are doing the work, so if we need it to be done in a ... particular time we need to provide the input. If we restructure the spec on 25th Jan ... that will need significant editorial input to get the CR version ready for 1st Feb. Pierre: When it comes to input, Movielabs is committed to providing resource to meet the ... IMSC 1.1 publication timeline of October 2018 as it has been doing for the past couple of years. Philippe: For TTML2, you guys can make the decision to say at the end of the F2F any non-closed ... issue on TTML2 will not be part of TTML2, and that will let you move to CR by end of Jan. ... If you really want the feature in TTML2, put the effort in so that there is a resolution for ... the issues by the F2F. That's a simple formula. If at the end you don't like it and want to ... formally object to moving to CR then you know why. The Process can't help you on that. Nigel: Summarising, there is no change to the timeline for TTML2 that we agreed in the TPAC F2F. Charter 2018 Thierry: We don't need to talk about this today, but I'd like you to review the document. [21]Draft TTWG Charter [21] https://github.com/w3c/charter-timed-text/blob/gh-pages/Draft-2018-TTWG-Charter.html Thierry: Please review for a discussion later. Nigel: I want to add this to the F2F agenda max 30 minutes at the end of the meeting. Thierry: As a minimum we need to agree which specs are in the Charter. ... Let's defer to a further discussion. Pierre: Let's discuss on the reflector please. Nigel: Please raise issues on that github repo. TTML1 agenda issues and pull requests Pierre: Thank you everyone for helping close issues - it's been productive so far. Nigel: +1 Deprecate offset-time and fraction component in smpte time base #283 github: [22]https://github.com/w3c/ttml1/pull/283 [22] https://github.com/w3c/ttml1/pull/283 Pierre: There's universal consensus that the things being corrected here were errors in TTML1 ... - two syntaxes that were permitted erroneously. Glenn: We are not talking about removing a syntactic feature per se but clarifying that ... certain semantics are not defined and usage is not intended. ... The only issue in my mind is how to resolve that. Some people want to deprecate it and ... I don't so I'm fine with adding text that makes those recommendations against usage ... and points out that its undefined. Pierre: There's consensus that this syntax has no meaning and probably should not have ... been allowed in the first place. There are three paths to correcting: ... 1. Informative text "It is not recommended..." ... 2. Deprecation "You should not use it and it will probably be forbidden in the future" ... 3. Clearly prohibit it in Third Edition reflecting that there's no semantic. ... We have to pick one of those paths. Philippe: I assume no implementations try to do something with that syntax? Glenn: Not quite - the syntax is valid but in a different semantic context. We're not talking ... about it being wholly undefined. Philippe: Do we have implementations that follow that semantic? Glenn: yes two Nigel: I don't think so. Pierre: Let's not muddle the issues. Specifically when timebase="smpte" it is not forbidden ... to use fractions. Can there be any document or implementation out there that uses fractions ... of seconds in a SMPTE timecode timestamp? Philippe: If there are no implementations and usage then it makes no sense to have ... it permitted in normative text. I don't understand the ambiguity. Andreas: Thanks for those three options Pierre. From my side, the first two options would ... be fine. I find the third option problematic because it is another addition and I'm not sure ... about the consequences on validation. Philippe: I don't understand how you guys could recommend something with no implementations ... and when there are never going to be any implementations. Pierre: So you prefer option 3? Philippe: Yes, option 3. ... I could live with option 2. Nigel: I think I asked for option 3 and I think it's my preference. Glenn: Option 3 means going through the CR process, so it's a trigger on a CR, whereas ... without it we would not have to do a CR. That's a big bump in my opinion. Up to now ... we have attempted to ensure that anything that looks like a substantive change in TTML1 ... is not written in a way that it is a substantive change. We have long avoided deprecation ... in TTML1 and tried to be very careful about not removing something. In this case I don't ... think it's a matter of removing it. It's not that it's recommended, it's just something the ... spec is silent about. Implementers have ignored it because it makes no sense. ... There is no implementation of fraction in SMPTE mode as far as I'm aware. However the ... syntax and the feature itself cannot be removed because it is relevant in media timebase. ... I'll argue this is an opportunity to clarify that it is not well defined and that's consistent ... with how we handled this kind of thing in Second Edition. Making it stronger doesn't ... have any purpose here. Philippe: You are going to make the note informative, but we don't want anyone to use it, ... so if you say "authors should not..." I'd be willing to live with that but that's not what you ... proposed. Glenn: If we add a should or should not then that's a substantive change. Philippe: Yes, and I don't think we should hold ourself to a lower standard because of ... W3C Process. Are we here to serve the community or meet an arbitrary publication timeline? Glenn: Every time we've had this kind of discussion we decided to use an informative note. ... It isn't winning anything to go through CR and do more work. Philippe: It's not substantive. Pierre: I echo Philippe, going through the extra work of publishing a CR is really worth the effort. ... The gymnastics in the past have lead to a hard to read and maintain specification. In this ... particular case it behoves the group to be clear with implementers. The effort needed to ... go through CR will not be meaningful compared to correcting the spec. Glenn: The group can agree to go through a CR process. I don't think it's worth it. Andreas: I partly agree with Glenn here, I'm not sure if it is worth making this a substantive ... change and discussing if it should go through the CR process or not. TTML1 is very ... focused on processor conformance and a bit loose on document conformance. It doesn't ... really say what syntax should be rejected, just what processors need to support. ... It is also difficult because we don't say anywhere in the TTML1 document where there ... are dependencies between syntax and attribute values. For example we don't say that ... if you use pixels you have other constraints. ... So options 1 and 2 can be done by saying "if you do this no processor will support it or ... need to support it" and that's sufficient for me. Philippe: Clarifying my position, in terms of how I see the priorities, I say publish often. ... At the end of the day we have a Rec on the website. If we do nothing it is no worse than ... what we have today. If we need more time we can take it without blocking Third Ed so ... I could live with doing nothing. Cyril: Netflix's priority is TTML2. Anything that would potentially delay TTML2 including ... more work on TTML1 is not the best option for Netflix. However I understand that TTML1 ... is also used, but in terms of these options I would select the one that gives us least work ... so we can focus on TTML2. Pierre: This particular issue is not the one that will trigger CR - there are other patches ... being made that already are likely to trigger CR, so this is unlikely to be the lynchpin. ... So it's not a good decision making criterion for this issue. ... Considering the different preferences (Nigel can you live with option 2?) Nigel: Yes, deprecation would be okay Pierre: Glenn could you live with option 2? Glenn: I would object to labelling it as a deprecation. Pierre: If the text says authors should not use it would you still object? Glenn: It would be useful if someone could point out the changes that would trigger CR. Pierre: Even lineHeight - definition of substantive in W3C Process includes resolving ambiguities so ... it is easy to trigger it. Glenn: Triggering CR is a departure from previous. As Andreas points out we generally ... don't indicate what authors should do. It's okay to inform the author that something has ... no meaning or is undefined and let them make their own conclusions. I'm not willing to ... go with the deprecation. ... If we have to go through CR I might be willing to accept something that looks like a ... substantive change expressed as what the processor does. Pierre: How would you phrase that? Glenn: If a fraction is present the processor should treat it as if it were zero. Pierre: What about "consider it an error"? That's used elsewhere. Glenn: That would be fine if we are going the CR route. Philippe: Can we put that question on the side and come back to that on the basis that we are going to CR, and revise it if we are not. Nigel: Assume we are going to CR right now and consider the solution. Glenn: If you change "deprecated" to "not defined" then I could accept it under the assumption that we're going through CR. Pierre: I'm happy to change that in 8003 and 8005. Glenn: And in the note change "deprecation" to "not defined status". Pierre: Can folk live with that? Nigel: So it's not a prohibition. Pierre: And it can not make existing implementations non-conformant. Glenn: If this turns out to be the only issue driving CR I reserve the right to come back to it. Pierre: I agree Glenn! Andreas: This is an example of an issue that has taken a long time to cover, so I would ... encourage everyone to take the same approach of compromise in the spirit of meeting our timelines. Pierre: Thank you I would also propose to use the same approach for #210 (leap seconds) Clock-time form time expressions allow 60 seconds for media timebase. #210 github: [23]https://github.com/w3c/ttml1/issues/210 [23] https://github.com/w3c/ttml1/issues/210 Pierre: As per pull request #283 I will propose a similar solution. github, end topic Clarify intended rendering of background color when applied to a span. #209 github: [24]https://github.com/w3c/ttml1/issues/209 [24] https://github.com/w3c/ttml1/issues/209 Pierre: There's quite a discussion on this. Andreas commented a couple of hours ago, ... suggesting the issue can be closed. Glenn: I agree. Pierre: Even though the comment Sep 19 implies that there was an errata, it sounds like ... there was further discussion. Can we point to the London meeting. Glenn: The comment would require paraphrasing XSL-FO or CSS so I would be in favour ... of closing. Nigel: Looking at [25]https://www.w3.org/2017/01/13-tt-minutes.html it seems we agreed ... not to change the previous agreement to add a note. [25] https://www.w3.org/2017/01/13-tt-minutes.html Pierre: Can we agree to change our minds and close this based on Glenn's comment and ... Andreas's request? Nigel: Yes it seems we have consensus to close this. RESOLUTION: Close this issue tts:overflow does not apply to the region area #239 github: [26]https://github.com/w3c/ttml1/issues/239 [26] https://github.com/w3c/ttml1/issues/239 Pierre: Some background - I had filed the issue not because the text was unclear but ... because the result was surprising. However nobody thinks it is surprising, so I'm happy ... to close the issue. Now Andreas has come in and we may need to make a change. Glenn: I'm happy to close this. Both Andreas and I need to take a note that the default ... behaviour in XSL-FO is "auto" whose meaning is implementation dependent, with a ... suggestion to use scroll. I would prefer to close this issue and open a new issue focussed ... on giving a better definition to "auto" if we want to do so. Nigel: The issue may have diverged a little but I haven't seen anything to suggest that ... the default value needs to be anything other than "visible". Glenn: [explanation of parent providing space to allocate space] ... In our case the region's extent overrides, in that it sets an ultimate maximum for the ... content to be laid out. I'm not sure that visible is the right default. Nigel: Then I support creating an issue for this so you can add that comment Glenn. Andreas: I think this can be fixed here. Because all the formatting objects other than region ... have no overflow property specification then we just need to define what auto means ... in TTML1. If we just say that we treat it as "visible" that fixes it and I see no other issue ... with that. I would be happy to propose some text for that. Then Glenn if you find any ... problems with it then we can work through it. Glenn: My concern was tentative so I'm willing to speculate that visible will work, so I agree ... that if we have an issue for resolving auto's meaning in TTML1 that's what I was suggesting. ... Without concrete evidence that visible is wrong I will go with that. Clearly the CSS folk ... think that's the right meaning. If that's somehow broken I would expect more clarification ... from the CSS folk. Andreas: Glenn would you propose the fixing text for that? Glenn: Yes I can propose something very simple. That would be another substantive change ... in clarifying a potential ambiguity. Pierre: Keep it in the same issue please. Glenn: Fine, but we should add the comment in the issue. SUMMARY: @skynavga to propose text clarifying that XSL-FO overflow property default "auto" should be processed as "visible". Glenn: I prefer to create a new issue and pull request if that's okay. Pierre: okay so we're closing this issue #239. Nigel: I've assigned this issue to you Glenn - when you create the new issue please ... reference this and close it. Glenn: OK Pierre: If you create the text proposal in the issue then I'm happy to create the pull request. IMSC issues Clarify that tts:fillLineGap does not hide characters #283 Pierre: This was a WR comment which needed a lot of discussion so we could understand ... the comment. We decided to generate an informative note, and since then it was pointed ... out that the note did not cover all the corner cases. From my perspective the note is now ... trying to say what the spec says, which we should avoid. ... I don't think we even know what the commenter really wanted. My proposal is to go ... back to the commenter and ask for a test case that would trigger the concern. ... We should say "We've tried really hard to resolve it but we don't understand it, so please ... provide a test case and then we'll address it" Andreas: I think that's a good option. How long should we wait for feedback? Pierre: I think the Chair should send an email to the commenter directly with say a 2 week ... period. We have documentation that we really tried hard. Philippe: Two weeks starting today - there are holidays so I'd prefer 4 weeks. Pierre: Let's do Jan 5 or Jan 6. Philippe: That's harsh! Nigel: That gives us time to see the response before the F2F. Andreas: Richard Ishida actually filed it so we should check if the commenter intends to reply. Nigel: I'll write that email and CC Richard. Pierre: I'm going to close that PR. Nigel: We can always reopen it if it turns out we were on the right lines! TTML2 processor behaviour vs TTML1 action-510? <trackbot> action-510 -- Glenn Adams to study TTT codebase and report on use of ttp:version to select processor behavior -- due 2017-11-23 -- PENDINGREVIEW <trackbot> [27]http://www.w3.org/AudioVideo/TT/tracker/actions/510 [27] http://www.w3.org/AudioVideo/TT/tracker/actions/510 Glenn: I did that and sent an email about it. close action-510 <trackbot> Closed action-510. Glenn: There are some follow-up questions from Cyril that I haven't got to yet. TTML2 WR labels action-508? <trackbot> action-508 -- Thierry Michel to Check if there are editorial/substantive labels for ttml2 issues and add if not. -- due 2017-10-12 -- OPEN <trackbot> [28]http://www.w3.org/AudioVideo/TT/tracker/actions/508 [28] http://www.w3.org/AudioVideo/TT/tracker/actions/508 Thierry: I don't think I've done that, I will look at it tomorrow. Agenda Order Andreas: I just wanted to pick up some comments from Pierre last meeting about how much ... time we spend on coordination. Especially at the beginning, and sometimes there are ... staff/Chair issues that could be handled offline. In other groups all that is done at the ... end as AOB. For us to make better progress on other issues I would propose to cover ... other issues first and then coordination. If more time is needed for this then I'm sure the ... people needed would stay on the call. Philippe: I agree you guys spend a lot of time talking about process rather than issues. ... However in Nigel's defence last time I was at the meeting you guys wanted an agenda ... before a F2F meeting so he was respectful of that. In my experience putting it at the ... beginning leads to discussion like today but at the end means it turns into a backlog. ... Nigel if you want to talk to staff outside meetings you can do that. Nigel: Yes, I'm aware I can do that. I'm not aware of spending a lot of Chair + staff discussion ... time in meetings but I can take that on board. Pierre: I'd like to echo Andreas and say administrative stuff should be dealt with on the ... reflector or offline. I'd like to jump straight into issue resolution after the agenda bash. <Zakim> plh, you wanted to suggest an other possible approach Philippe: Just throwing ideas out - if you can discuss on the reflector or github then ... fine, another way is to say the first 15 minutes are timeboxed for admin and if the time ... is up then move onto other issues. Pierre: Ideally we'd spend no time on administrative issues. Nigel: I'm not sure I agree with this all but I'm happy to try it but reserve the right to move ... admin issues up the agenda if they've been pushed off the backlog for too long and the ... priority has gone up. Some things are admin but best dealt with in a call. Andreas: I'm happy with that or to put a timeout on the discussion as Philippe suggested. Philippe: CSS WG uses this technique, setting a 10 minute timeout for admin stuff. Nigel: OK seems like a worthwhile approach. Meeting close Nigel: Thanks everyone. Next planned meeting is 4th Jan. If anyone wants a meeting before ... then, I won't object, otherwise have a good break. [adjourns meeting] Summary of Action Items Summary of Resolutions 1. [29]Close this issue [End of minutes] __________________________________________________________ Minutes formatted by David Booth's [30]scribe.perl version 1.152 ([31]CVS log) $Date: 2017/12/14 17:29:01 $ [30] http://dev.w3.org/cvsweb/~checkout~/2002/scribe/scribedoc.htm [31] http://dev.w3.org/cvsweb/2002/scribe/ ---------------------------- http://www.bbc.co.uk This e-mail (and any attachments) is confidential and may contain personal views which are not the views of the BBC unless specifically stated. If you have received it in error, please delete it from your system. Do not use, copy or disclose the information in any way nor act in reliance on it and notify the sender immediately. Please note that the BBC monitors e-mails sent or received. Further communication will signify your consent to this. ---------------------
Received on Thursday, 14 December 2017 17:32:24 UTC