Recent changes on the VM Network

When making a post, please ensure it complies with this site's Main Rules at all times.
  • Did you quote me, Richard?! ?

    Here we go, found your post on Cable Forum:

    (the orange writing on this site are always links and in this case if you click on your name below, it takes you to your original post)

    Sky have confirmed that Sky Cinema Disney (ch. 407) is renaming to Five Star Movies on New Year's Eve, before shutting permanently on 6th January 2021. My prediction had come true. ?

    Thanks Aguero for the info and for Richard for reposting it here.:thumbup:

  • Blimey. Those new encoders they got are obviously more efficient.

    Hmmm! the weird thing being that, now Sky Witness HD has moved transport stream, the STB doesn't have a track record of recordings and is reserving even more space than usual.

    Full disk problems... :D

  • Yes, I don't think that anyone is suggesting that the V6 boxes currently in circulation could be utilised, they don't have a return path and DOCSIS isn't one way traffic. It would need a modified box with a return path.

    If you look at the configs using the "DL_Sam_TiVoBL_16MB_CM" stream, there are a few pointers that suggest a different model (in other words, not the V6 model currently in mainstream use). It is level 155, which is a unique configuration not used for any of the other NIT Linkage streams, so something "different". It is only added to the Knowsley DMC (Hayes_RHE) network, an internal test network, which suggests that it could be a test config for a new box. The stream service name includes the "TiVoBL" element, which is the same element used in the streams thought to support the V6. As with DL_Sam_TiVoBL_10011 & DL_Cisco_TiVoBL_10066, the configs started at version 1 and haven't changed, suggesting that this isn't being used for incremental firmware versions.

    I guess this, together with the "CM" in the name has fuelled the speculation that this could be a test config for a modified V6 box (V6-CM maybe?) with a return path. As far as I know, nobody has taken a look at the SID 10068 stream to see if there are any clues as to what the stream supports, nor are there any plans to do so.

    As far as the 8 tuners are concerned, what's point with a 1TB harddrive? It would just encourage customers to record more of the dross and then complain that the box was under-resourced as the harddrive filled too quickly. Maybe enable them on a modified box with a larger harddrive or enable them on a modified box with a return path and bond a couple of DOCSIS channels together for more download capacity...

    With hindsight, could this have been for the TV360? Also, SP said at the time that it was unlikely that the 2 unused tuners on the V6 would be utilised with the TiVo software, but "who knows after that". I wonder if the TV360 (SP said the other day that this is what the new box is now referred to as they no longer use the H word ?) will eventually have 8 tuners?

  • With hindsight, could this have been for the TV360? Also, SP said at the time that it was unlikely that the 2 unused tuners on the V6 would be utilised with the TiVo software, but "who knows after that". I wonder if the TV360 (SP said the other day that this is what the new box is now referred to as they no longer use the H word ?) will eventually have 8 tuners?

    You may have seen my reply elsewhere, but it's silly saying they won't use the Horizon word, TiVo was still mentioned when talking about V6 as it's the software and in use in other markets. If they renamed Horizon nto TV360 in all markets then fair enough, but as far as I'm concerned it's like buying a Dell Inspiron with Windows installed and calling it Dell OS.

    If I remember the context of JJ20X's post, the possibility suggested was that it might be evidence of what VM could do to replace Original VM Tivo boxes for customers who don't take broadband. The point being V6s don't have modems and Original VM Tivos do. TV360 blxes still need the customer to take VM broadband, so it's not the currently released TV360 service/product. It could be a future one. Or something else entirely, since it wasn't clear what was being seen.

  • With hindsight, could this have been for the TV360? Also, SP said at the time that it was unlikely that the 2 unused tuners on the V6 would be utilised with the TiVo software, but "who knows after that". I wonder if the TV360 (SP said the other day that this is what the new box is now referred to as they no longer use the H word ?) will eventually have 8 tuners?

    Probably not, no new linkage segments have appeared for the new hardware, so it would seem that it only updates via the internet. Which the V6 did too but, seemingly, with an initial linkage config. possibly for initial setup. I only see the config, not the data, so it's also possible that the data streams have been adapted for the new boxes. NIT linkage was always a bit beyond the scope of the analysis.

    Link SIDs were renamed some time ago but, as far as I remember, the "DL_Sam_TiVoBL_16MB_CM" config was eventually removed.

  • Wednesday's changes:

    SID 2107 InternalF (phew! that's been around for a while) was renamed Virgin TV Ultra HD - no surprise there, it points to the same streams. Moving from channel 296 to channel 1. Hmm!

    The genre changed from "Lifestyle" to "Entertainment". The HD EPG flag was added and the "hide reversal" flag was removed. This service is only added to the NIT of several internal networks and isn't in the config. of customer facing networks. Also, one of the SDT descriptors, not usually included in the report, was removed from the SDT.

    SID 810 HUM Europe ended the closure notification stage. It had the "hide reversal" flag removed, hiding the service from the EPG as it enters the placeholder stage. Placeholder and End of Service flags had already been added when normal broadcasts ended.

    SID 522 Sky Documentaries had the PIN Access flag added, which is normal for Sky channels.

    VOD transport streams for the Scottish ex-Telewest networks finally reduced from 2 streams to 1. Changing frequency to co-locate with TSID 514 (TSID 512 is the usual ex-Telewest default, so a departure from the norm). That leaves only NGRHEE Knowsley with 2 active frequencies in the config. and VM Hammersmith with a unique (unshared) VOD frequency - both of these are internal networks. Ex NTL Langley networks are not monitored.

    There was another look overnight at changing the analysis database to reduce the database size. One attempt split the SI tables up into segments, not very productive as it exceeded the maximum file size working on current data - before adding brought forward data. A second attempt started to look at a reduced overall parse and rebuilding data by parsing mini sections. It cuts down on file size quite significantly but needs multiple processing runs, which is never popular with the analysts. The database will now be frozen with the amendments made earlier in the week. Further changes may be made when more analysts are available (so not in the immediate future).

  • ITV Hub App is finally available in Scotland on VM. It was previously a Hub-branded folder for VM-hosted service.

    It's probably related to the change of streams you reported earlier JJ20X, not that the app uses them.

    Also STV Player app has changed as it was un-Favourited and it's icon had changed when I went to reinstate it.

    So in theory I should be able to see all the ITV Catch-up content now on my V6.

  • ITV Hub App is finally available in Scotland on VM. It was previously a Hub-branded folder for VM-hosted service.

    It's probably related to the change of streams you reported earlier JJ20X, not that the app uses them.

    Also STV Player app has changed as it was un-Favourited and it's icon had changed when I went to reinstate it.

    So in theory I should be able to see all the ITV Catch-up content now on my V6.

    The changes may have been provisioned at the same time but, other than that, aren't really related.

    Speculation about the Scottish ex-Telewest networks includes a theory that they could be about to link ex-NTL Glasgow into the group, although there is no evidence to suggest that is actually happening.

    Minor changes were made to the analysis database overnight adding extra code to the initial parse routines. That may seem odd as it increases the file size but a secondary module has also been changed to work with raw SI data with an overall reduction in file size. Further minor changes are now expected to be made to other secondary module code routines, eliminating as much use of the parsed data as possible. When everything is working as expected, chunks of parse code will be removed. Quite important as around 200,000 SI records are parsed for both current and previous scan periods on each run. As stated yesterday, no changes will be made if they require additional processing.

  • ITV Hub App is finally available in Scotland on VM. It was previously a Hub-branded folder for VM-hosted service.

    It's probably related to the change of streams you reported earlier JJ20X, not that the app uses them.

    Also STV Player app has changed as it was un-Favourited and it's icon had changed when I went to reinstate it.

    So in theory I should be able to see all the ITV Catch-up content now on my V6.

    The ITV Hub/STV Player app changes must be for the V6/TiVo platforms only just now as there have been no changes to the TV360 app line up - still only STV Player and no ITV Hub.

  • I'm now wondering if VM messed up or STV and ITV made a deal to simplify their lives on VM. I had time to look this morning and STV Player no longer has Channel 3 content. Either the two Channel 3 companies decided it wasn't economical to develop two versions of their apps for VM each or someone at VM accidentally put the down south app configuration on my area overnight.

    Can someone else on VM in Scotland, preferably not Tayside, confirm they have the same change?

  • There were around 20 layout versions of the analysis database released over the last 24 hours, so a lot of progress. A 6% file size saving and several long form tables have been adapted to give analysis reports on the initial processing run, even if the database tables won't fully update until there is a second processing run. The initial parse routines have been largely stripped out, although complicated ones that would otherwise need 2 processing runs have been left in place. Some work will be done next year to eliminate many of these.

    The (rather odd) scan results:

    SID 322 and 326 are additional versions of ITV Be +1 - both listed as EPG 120.

    SID 10628 and 10630 are additional versions of Channel 4 - both listed as EPG 104.

    SID 10269 and 10631are additional versions of Channel 4 Audio Desc - both listed as EPG 854.

    SID 620 is an additional version of BBC ALBA listed as EPG 161.

    None of the above have been added to the NIT (unless added to ex NTL Langley networks which we cannot see).

    The hidden version of BBC Alba (bbcalbahid) has moved from SID 612 to placeholder SID number 650 - still listed as EPG 75, beyond the range of customer devices. This change has "broken" the analysis database as the NIT entries seem to remain on SID 612 but with no SDT entries, which isn't supposed to happen.

    Sky Premiere 1hr (placeholder) has moved from SID 322 to placeholder SID 358. SDT and NIT records updated.

    SID 1806 Love Nature HD has had the "hide reversal" or "hide override", or whatever you want to call it, flag removed. Reactivating the hide flag to remove the channel from the EPG, so no more of the classy redirection slate.

    That's it! It seems like a lot more has happened with most of the analysis modules having been run to test the new database routines, especially for the ones re-written because of changes to the initial parse routines and to extract analysis data from run 1 long form tables.

  • BBC Alba needs moved into the Top 24 positions on the EPG by January, as per Ofcom requirements. It may only have to be in Scotland, so maybe that's the reason for another feed? They already moved it to 117 on Sky recently.

  • BBC Alba needs moved into the Top 24 positions on the EPG by January, as per Ofcom requirements. It may only have to be in Scotland, so maybe that's the reason for another feed? They already moved it to 117 on Sky recently.

    A third version popped up yesterday, so that makes it even harder to explain. Multiple versions of ITV Be+1 and extra versions of Channel 4 (and its AD service) are also a mystery. I thought all of the Channel 4 regional versions were added or, at least, allocated service numbers even though the HD versions carried the London service. I've not had a chance to look into it as the database has been down for development. The new slimmed down version should be available today after any unforeseen problems created by the code changes have been resolved.

  • Saturday's changes:

    SID 612 BBC ALBA was added and assigned to channel 161 - the third version of BBC ALBA.

    The old SID 612 (bbcalbahid) which remained in the NIT after the service name was removed from the SDT, transferred to SID 650, regaining the service name.

    SID 1302 Virgin TV Highlights changed active hours to 18:30:00-18:00:00

    There were no changes for Sunday.

    The analysis database hit beta layout 31, so quite a few changes over the last few days and now a significant file size reduction. Much of the parse code has been removed from the SI table analysis and is applied to smaller sections / extracts. Code has been added to analyse pretty much everything on the first processing run, so some of the tables used for secondary analysis have been removed. The coders stopped short of converting NIT Linkage analysis to a single processing run as they said it was too complicated. I added the code to do that on Sunday by parsing the long form data with the same routines as the short form data, then extracting data for analysis from the long form tables. It also needed new, additional, reference codes to match the records. NIT Service count still needs 2 processing runs because the data isn't included in the SI data. The code physically counts how many networks each service is carried on and marks the last network / final count. That data is produced on the first run and then analysed on the second. There is no way around that.

    A few problems have had to be ironed out following the changes. Deletion of primary SI parse routines accidentally removed time / date stamping, which has now been added back by directly accessing the details from the SI tables rather than using parse code. The removal of primary parse code means that data for full analysis of the SDT tables is no longer available. New reference codes allow the data to be extracted from the SDT long form parse extract / table. 3 codes being allocated for the SDT conditional access descriptors because the SI tables use the same reference for both channel and pack versions and for the null version sometimes used for hidden channels. Three tag codes have been included in the reference numbers to allow the details to be extracted by the correct module. The three tag method wouldn't work for full analysis of the whole SDT table. As a workaround a further reference code has been added to analyse the whole table. This reference code doesn't read the descriptor data to check if the descriptor has pack or channel details, it simply marks the first conditional access descriptor found on each SID with a -1 tag and the second one with a -2 tag. A "blind" extraction and matching of data. Eventually, this may allow auto-matching of full SDT analysis but that isn't going to be a priority. The analysis database is now considered to be "mature" with little further development possible. As a consequence, we may lose the dedicated coder but this is under review. Hopefully, code changes will now slow down and stop over the Christmas period. There is still a lot of experimental and test code in the database, which needs removing or developing further if the analysts want the extra data it provides. Removal of test and duplicate code will give a further reduction in the file size.

  • Saturday's changes:

    ...SID 1302 Virgin TV Highlights changed active hours to 18:30:00-18:00:00

    And with Saturday's Fight Night preview out of the way, SID 1302's active hours reverted to 10:00:00-06:00:00 with 6am to 10 am reverting to the fitness promo slot.

    No other changes for Monday.

    A further sizeable reduction in the file size of the analysis database this morning as I deleted development work for tracking the more obscure descriptor length data. If the analysts wish to pursue that exercise it can be done externally.

  • A further sizeable reduction in the file size of the analysis database this morning as I deleted development work for tracking the more obscure descriptor length data. If the analysts wish to pursue that exercise it can be done externally.

    And I spotted a pattern with the new way of extracting data. NIT Services works in a very strange way now, a hybrid reference code is made up of TSID + SID + NIT and this is then expanded to recreate the fields and to lookup the data rather than carrying all the data fields in the table. Because it's a single field, it's much easier to compare previous and current references rather than compare the whole table for the analysis. So, in future, that's what the analysts will compare and then the analysis routines will automatically build the reports from the auto-matched reference codes. I then did something similar for the full SDT comparison. These aren't auto-match routines in the same way as the individual SDT reports, which match everything by the unique service id. USID is an inappropriate field for matching NIT records as USID doesn't really exist in the NIT. Also, if there is a NIT record without a corresponding SDT (and therefore USID) record, it would break the analysis. Similarly, the full SDT analysis is supposed to be an overview and shouldn't match records by USID.

    Ok, I added a lot of beta code to duplicate existing analysis modules to run auto-comp in parallel with the existing method. The file size is now pretty much as large as it was before the coders got to work on the database. I guess I'm not very popular right now. ^^

    Fortunately, when the new routines get past beta testing, the old code and routines can be deleted. At least this way, the code can be tested without breaking the system. The same can't be said of the live changes the coders have made over the last few days. :P

  • The auto-match routines set up yesterday seem to be working pretty well. The additional Full SDT reference codes extracted the data but weren't very good at matching records. However, as soon as the data is extracted, the information is there to build the correct reference codes which can then be used for matching. So, the module now uses 2 reference numbers, one for data extraction, the other for matching records in the analysis module.


    Changes noted on Tuesday:

    SID 809 Sky Sports Arena (Sky Arena) rebranded as Sky Sports Darts (SkySpDarts) on channel 518

    SID 2303 Sky Sports Arena HD (SkyArenaHD) rebranded as Sky Sports Darts HD (SkyDartsHD) on channel 508

    SID 10264 Channel 4 Audio Desc (Midlands) gained a few extra unprintable characters, with the descriptor length changing from 20 to 23 characters (this usually means that code is added to identify the character set). The service names are tidied up on extraction to remove these extra characters.

    SID 612 BBC ALBA

    SID 620 BBC ALBA

    SID 10268 Channel 4

    SID 10269 Channel 4 Audio Desc

    SID 10270 Channel 4

    SID 10271 Channel 4 Audio Desc

    All of the channels in the above list had the channel sections of the SDT conditional access descriptors removed.

    Also, in the SDT root, run status changed from 1 (not running) to 4 (running).


    SIDs 322 & 326, the additional versions of ITVBe +1, had pack codes added for M/Player and above.


    In the NIT entries for internal network Code Download Test (Portsmouth):


    SID 322 ITVBe +1 replaced SID 314 ITVBe +1

    SID 612 BBC ALBA replaced SID 611 BBC ALBA

    SID 10268 Channel 4 replaced SID 10223 Channel 4

    SID 10269 Channel 4 Audio Desc replaced SID 10263 Channel 4 Audio Desc


    The Channel 4 versions removed are for the South region, which also covers Wales. So the changes currently testing on Code Download Test (Portsmouth) are probably in preparation for mandatory changes to EPG slots in Wales.

    Quite a few odd changes there. At least, with more auto-matched segments, the job is now much easier for the analysts.

    Edited once, last by jj20x (December 15, 2020 at 5:54 PM).

Participate now!

Don’t have an account yet? Register yourself now and be a part of our community!