Audio encoding changed, truncated audio files

[RS]

>From: Vangelis forthnet

>Sent: Wednesday, September 14, 2016 23:53

>> Interestingly, all the HAF modes have been removed.

>NO, they have not! You are using GiP 2.95, the "haf"
>radiomodes were implemented in 2.96+ :

>https://github.com/get-iplayer/get_iplayer/wiki/release296#3-combined-dash-haf-and-hlsaac-modes-now-default-for-radio-programmes-dash-preferred

>ergo:
[therefore for those who don't speak Latin]
>perl get_iplayer-296w.pl --type=radio --pid=p045j35r -i | FindStr "modes:"

>modes:          original: 
>dashhigh1,dashhigh2,dashstd1,dashstd2,dashmed1,dashmed
2,dashlow1,dashlow2,flashaacstd1,flashaacstd2,flashaaclow1,flashaaclow2,hafhigh1
>,hafhigh2,hafstd1,hafstd2,hafmed1,hafmed2,haflow1,haflow2,hlsaacstd1,hlsaaclow1

>whereas:

>perl get_iplayer-295.pl --type=radio --pid=p045j35r -i | FindStr "modes:"

>modes:          original: 
>dashhigh1,dashhigh2,dashstd1,dashstd2,dashmed1,dashmed
>2,dashlow1,dashlow2,flashstd1,flashstd2,flashlow1,flashlow2,hlsstd1,hlsstd2,hlss
>td3,hlslow1

Thank you for clarifying this.  I had thought v2.96 was mainly about 
changing defaults.  I thought it had caused a lot of unnecessary confusion, 
and I don't like the idea of removing Flash support in v2.97, and that is 
the reason I have stayed with v2.95.

When I read your analysis and looked again at the v2.96 release notes I 
concluded that HAF support was new in v2.96 and that was something I had 
missed and that was the reason I had been unable to see the HAF modes 
with --info.

Re-reading the release notes yet again I realised that I had still got it 
wrong.  What has happened (I think) is that for radio HLS has been renamed 
HAF, but for television it is still HLS.  That does not explain why I could 
not see what v2.96 calls hafhigh as hlshigh in v2.95, or what hlsaacstd and 
hlsaaclow mean in v2.96, except they are said to be old HLS audio modes 
required for clips.

Maybe I am just more stupid than all the other readers of the list and 
everyone else understands it perfectly.  Even so, I have to ask for whose 
benefit these changes are being made.

I also did not understand the argument that removing the aactomp3 option 
made maintenance easier.  It was not many lines of code.  The --command 
and --preset alternatives are more transparent, but why should users be put 
to the inconvenience of changing?

>The "timeadded" part actually informs you
>of when that specific audio programme first
>populated your local "radio.cache" file;
>it has nothing to do with the actual time
>that audio file was uploaded/modified
>by the BBC on their CDNs!
>E.g., if I "--info" on my machine:

>perl get_iplayer-296w.pl --type=radio --pid=p045j35r -i | FindStr timeadded

>I get:

>timeadded: 10 days 20 hours ago (2016-09-04T02:18:08+00:00)

>(this specific radio show was
>firstbcast:     original: 2016-09-03T06:00:00+01:00)

Yes, I have misunderstood timeadded.  If I had thought it through properly, 
if it had been edited the version would be editorial rather than original. 
I am not sure why I would want to know when an entry was added to the cache. 
However mixing up metadata generated by the BBC with records of functions 
perfomed by get_iplayer in a notation used by the BBC is a recipe for 
confusion.

There is a saying, "Don't shoot the messenger ..."  As always I am grateful 
to you for explaining it all to me.  Please don't think for a moment that I 
am criticising you.