Lms 7.8?

[Mnyb]

Enhancement request 142 is a problematic one for a reason I think.

As the server software is committed to a solution where it never ever write to the files directly and this is very dear to many users. This paradigm wins in this situation.

This will undoubtedly always lead to a situation where you loose all ratings, Erland has gone to some lengths in his plugins with an externally available trackstat backup . But this is for technical reasons not safe either as there are circumstances where the file can't be 100% identified . One way it works is by the extremely geeky music brains id and some other means .

There are great ideas that could be implemented if the server did not treat the dB as a temp a mere reflection of the tags .
But that would need a whole other level of programming where the dB always could be recovered and translated between schema changes and not needed to be " cache cleared" aka tossed away for every minor scanner issue and it would need to hash every file during scan to
create unique identities but eventually for safety reason it would need to write an Id to the file .

Then the dB could be augmented by all sorts of info and associated functionality not only ratings that would only be the 1% of what would be possible .
A greatly enhanced music exploring application that many wants.

but even this is a bit over the top for most users that are not committed music nerds you must keep the file collection in order and tagged.
There is a lot of pull in the other direction where people just want to play a jumble of misshaped files by folder and file name and those users already treats music as a disposable not a cherished collection.

This is eventually also outdone by all that relies on a service to even have files is a minority solution.

I would settle for if they actually fixed bugs in whatever functionality they decided the server should have.
As it is now I'm irritated by the fact they do build functions that are buggy and never fix them instead they develop next function that is also buggy ad naseum . And it detoriates, old functionality brakes and is never fixed they are not on top of the maintenance aspect of this .

And in many cases this does not lead to high vote counts a prime sample is that album artist issue .
Other examples include where less tech savvy user usually think they have done something wrong or never realise how it really should work .

Another example it has becoming popular by for example Bandcamp and some other vendors to use separate track art in addition to album art . This is great now every song has a different artwork and the album as a whole one. this actually worked before 7.6 but is broken now the presence of track art even brakes the functionality of album art and this issue has all of 3 votes or similar.

So please please fix total lack of quality controll . Do a rudimentary QC before every release have a functional spec and test every function against that .

[Mnyb]

Here is the artwork issue http://bugs.slimdevices.com/show_bug.cgi?id=17634 FYI

[jmschnur]

Market survey makes sense plus bug fixes.


J

[Mnyb]

As the dev resources are as limited as they are can't we just scrap the video and picture support and upnp/dlna thing ?
To make this actually work properly you must commit even more resources than you sunk into the whole project already ? This seems extremely costly to me .

[reniera]

Enhancement request 142 is a problematic one for a reason I think.

As the server software is committed to a solution where it never ever write to the files directly and this is very dear to many users. This paradigm wins in this situation.

This will undoubtedly always lead to a situation where you loose all ratings, Erland has gone to some lengths in his plugins with an externally available trackstat backup . But this is for technical reasons not safe either as there are circumstances where the file can't be 100% identified . One way it works is by the extremely geeky music brains id and some other means .

There are great ideas that could be implemented if the server did not treat the dB as a temp a mere reflection of the tags .
But that would need a whole other level of programming where the dB always could be recovered and translated between schema changes and not needed to be " cache cleared" aka tossed away for every minor scanner issue and it would need to hash every file during scan to
create unique identities but eventually for safety reason it would need to write an Id to the file .

Then the dB could be augmented by all sorts of info and associated functionality not only ratings that would only be the 1% of what would be possible .

I fully agree with your last statement.

But 142 is not problematic at all. Solution is to store info in hidden files in the media directories ( in addition of database). See how Picasa works with photos. No need to reinvent the wheel.

[Mnyb]

I fully agree with your last statement.

But 142 is not problematic at all. Solution is to store info in hidden files in the media directories ( in addition of database). See how Picasa works with photos. No need to reinvent the wheel.

Workable yes but.

Some aluminium foil hat people prefer the directories read only .

A one album one folder approach ? Not all people do this .

And finally does picasa account for filename changes and file changes aka tagging done by other software ?

Remember that the music file can change name and size and timestamp due to external tagging operations and also location how would an external file track that ?

[mherger]

> I guess what I would like to see is that Logitech continue to develop
> features they've already added and not just abandon them as soon as
> they have enough to put a sticker on the box.
>
> Regarding anything related to browsing or ratings/smart playlists, I've
> just given up, I would be extremely surprised if Logitech would put any
> focus in this area, because it's something that would make people
> recommend the product but it's not something that's easy to advertize,
> so it's not in line with what their strategy seems to be.

Remember my initial response? Nothing earth shakingly new I can promise
for 7.8. But improved UX. Some of what you wish for might actually be part
of the plan. Be prepared for the unexpected.

Oh, did we ever mention we're hiring?

http://hire.jobvite.com/CompanyJobs/...ion&j=oRydWfwi
http://hire.jobvite.com/CompanyJobs/...ion&j=oVydWfwm

--

Michael

[erland]

Oh, did we ever mention we're hiring?

http://hire.jobvite.com/CompanyJobs/...ion&j=oRydWfwi
http://hire.jobvite.com/CompanyJobs/...ion&j=oVydWfwm

Excellent news!

[MrSinatra]

hopefully one of the jobs they'll do is implement ONE top ten voted bug per official release.

but thats impossible or so i'm told.

[Schindler]

@Michael, did you move from Bern to Fremont, CA, United States?