Sorry for the late answer, 2009/1/29 BOUWSMA Barry <freebeer.bouwsma@xxxxxxxxx>: > On Thu, 29 Jan 2009, Christoph Pfister wrote: > >> > I intend to take Christoph's files and massage them to add >> > bits of info, reviewing the info by hand, adding missing info > >> I don't mind adding those further bits. They need to be after the main >> block in the file, so that they don't get overwritten when those files >> are updated e.g. because of a new pdf. > > Hmm, actually, the first thing I was planning to do would be > to sort the entries by, for lack of a better term, provider. > That is, roughly, ARD multiplex when appropriate, ZDFmobil, > and regional Dritte multiplex everywhere, and in selected > regions, the remaining of the seven assigned GE06 allocations > or, in short, private providers. > > This is intended to provide an overview of these allocations > and the particular sites where they can be found, as well as > to handle the potential cases of frequency re-use in widely- > separated areas between two muxes with incompatible > parameters -- how often this will occur, I cannot say, as I > do not yet have a complete overview. > > This also can help the case of Tobi, who would prefer to use > two or three transmitter-site files, in that it would be easy > to see which frequencies would be ``local'' (shades of Royston > Vasey here, ``You'll Never Leave''). > > But I'm not sure how well this would work with a PDF-skimming > application... I see some possible ways to deal with this. The one I already mentioned is to separate the auto-generated and the manual part of a file. You could provide further information about the transmitters with comments in the second part of the file. If you want to relax the constraints, you could allow re-ordering of the transmitters either within the auto-generated section or within the whole file (in this case you need to mark the individual transmitters). Actually I don't care about the arrangement, as long as there are machine-implementable rules for the update. <snip> > And to Christoph, the danger, as I'm sure you know, of > pulling from a PDF or other single source of information, > is that those will not always be infallible, as we've seen > with a centre-frequency ending in `3' and paste-errors > for Hamburg between VHF and UHF parameters, as well as > absent info for the specific cases of the frequency change > in HH; or for BW, the change planned for Aalen plus the > to-be-in-service status of Bad Mergentheim. They've updated their file and the mapping between channel numbers and frequencies are fixed. So I consider it as a good base point (I need only some fractions of a second to generate updated files :), will commit them soon). The additions are up to people who know more about the story ... > I suppose I ought to write the originator of the PDF to > point out errors, but I suspect I'll get at best a reply > from TV Licensing saying they have no record of me in > their database and that as a non-resident, I have no > grounds for criticism of their offering and I should > stick to the programmes for which I pay the license fee. > (How else can I explain the years-running errors in > linking of teletext pages, that are again changed to > errors when those pages are relocated? ZDF, I'm looking > at you for failing to list your extended programme guide > following ttx page 380...) > > > > Oops, off-topic again. > > >> They shouldn't be too excessive, I meant the number of transmitters, not the size of the file. > Oh dear. The question is, how do I skate the thin line > between providing too much information, and failing to > include useful information? I guess, it depends on your > interest. If all you care about are just the frequencies, > and if I were to say to you `ERP' you would say, ``Oh, do > excuse yourself, you glutton'', then the existing files are > fine. But if you need transmitter sites and info to > help you orient whatever antenna you use, or are otherwise > technically inclined, then perhaps I do not provide enough > info... I'll have to rely on the feedback of others for > this, sad to say, being myself a technical geek... > > >> but for example I prefer if you add the Leipzip transponder >> to the de-whatever file instead of creating a new de-Leipzig file, so > > No worries, I'd add additional information to both the > appropriate Bundesland, as well as to de-Leipzig. Wait, > I just updated overnight, and de-Leipzig no longer exists! > Aieee!@ Good thing I made a backup copy of that directory > before I updated hg has a good memory as well :) >> this point shouldn't cause trouble to you. People don't have to scan >> every day, so it doesn't hurt if the scan time is increased by some > > Another thing that I don't include is the precise channel > layout, as this can churn over time -- particularly with the > private local (non-RTL/Pro7Sat1) multiplexes, as participants > drop out, go bankrupt, lose their licenses, or whatever; or > when a national or regional PSB makes small adjustments. > > After all, the results of the scan, if correct, will show > who is present... > > Of course, terms such as `ZDFmobil' may not immediately > ring bells with the general public as including 3sat, or > `arte' typically meaning the ARD multiplex, but these may > be the same people who complain they can't see their quiz > call-in shows on the ARD mux, and they know that next time > they will win for sure, guaranteed, if they can stop staring > at the topless announcer long enough to dial the number which > will cost just as much as the last thousands of times they > tried that week... > > > > Ah yes, looking at Bayern, there is an overlap between > the uncommon private multiplexes. I've read that Pro7Sat1 > is considering ways to increase bitrate to make available > AC3 audio together with higher quality video; this may be > a switch to 64QAM, a reduction of the guard interval, or > lowering the FEC error correction, which would make those > parameters incompatible with the others -- and I'll need > to manually add the transmitting sites to get an overview > of where it can be received. > > > Anyway, many thanks to Christoph for merging that info > into convenient files, and I'll post -- if only for the > benefit of the zero who care -- the fascinating info that > I can find based on that. > > > barry bouwsma > still writes too much, about nothing Christoph -- To unsubscribe from this list: send the line "unsubscribe linux-media" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html