For The Record 1.3.3 Released (3 Viewers)

[dvdfreak]

Video stream itself is the same to every single bit, but unfortunately there are differences between

- TsReader has few bugs (mantis)
- live555 library is used (its old and propably contains some bugs)
- winsock 1 is used, so at least hugh bandwidth HD channels can have stuttering / dropped frames
- Seeking and zapping is slower with the RTPS (for many reasons, mainly because its build for network use, extra work & buffering...)

Interesting. I assume most of these things are on the list of things to look at?

So for the single seat users the direct TS buffer access will definately provide a better user experience. Maybe it should be used in 1.3.1?

Hmmmmm.... Perhaps, but then perhaps only by timeshifting the files on a UNC share, like recordings are played back from a UNC share. We never diferentiate between single seat or multiseat systems in For The Record and I would like to keep it that way, it's architecture is built for maximum connectivity.

 

[tourettes]

Video stream itself is the same to every single bit, but unfortunately there are differences between

- TsReader has few bugs (mantis)
- live555 library is used (its old and propably contains some bugs)
- winsock 1 is used, so at least hugh bandwidth HD channels can have stuttering / dropped frames
- Seeking and zapping is slower with the RTPS (for many reasons, mainly because its build for network use, extra work & buffering...)

Interesting. I assume most of these things are on the list of things to look at?

Others than TsReader issues are pretty big changes, so at least 1.0 wont have any of those solved. Also the zapping / seeking speed will newer be as good as on the direct file access.

So for the single seat users the direct TS buffer access will definately provide a better user experience. Maybe it should be used in 1.3.1?

Hmmmmm.... Perhaps, but then perhaps only by timeshifting the files on a UNC share, like recordings are played back from a UNC share. We never diferentiate between single seat or multiseat systems in For The Record and I would like to keep it that way, it's architecture is built for maximum connectivity.

If there is only one client using the TB buffer files directly then there shouldnt be any issues, but multiple clients using same files could cause locked files etc.

 

[dvdfreak]

Others than TsReader issues are pretty big changes, so at least 1.0 wont have any of those solved. Also the zapping / seeking speed will newer be as good as on the direct file access.

Too bad, but I can understand the problem of lack of time completely

So for the single seat users the direct TS buffer access will definately provide a better user experience. Maybe it should be used in 1.3.1?

Hmmmmm.... Perhaps, but then perhaps only by timeshifting the files on a UNC share, like recordings are played back from a UNC share. We never diferentiate between single seat or multiseat systems in For The Record and I would like to keep it that way, it's architecture is built for maximum connectivity.

If there is only one client using the TB buffer files directly then there shouldnt be any issues, but multiple clients using same files could cause locked files etc.

OK, good point, and that really blows. So in that case I would choose to leave things as they are right now.

 

[tourettes]

OK, good point, and that really blows. So in that case I would choose to leave things as they are right now.

You should be aware that the zapping time is pretty critical issue when it comes to live tv. Users get annoyed really quick if the channel changes take 1.5x long as without 4TR.

 

[dvdfreak]

You should be aware that the zapping time is pretty critical issue when it comes to live tv. Users get annoyed really quick if the channel changes take 1.5x long as without 4TR.

The future is multiseat, no? This optimization would only for people using exactly one machine for everything, others will not benefit. Zapping should ALWAYS be as fast as possible, single seat or mulitseat

 

[tourettes]

The future is multiseat, no? This optimization would only for people using exactly one machine for everything, others will not benefit.

Depends on who you ask. I have only one HTPC inmy house, so I definately want to locate my tuners into that PC

Zapping should ALWAYS be as fast as possible, single seat or mulitseat

For the technical differences the zapping cannot be as fast on the multiseat environment as it will be on the single seat. of cource if you have some way to solve / improve the performance please fell free to provide some patch

 

[dvdfreak]

For the technical differences the zapping cannot be as fast on the multiseat environment as it will be on the single seat. of cource if you have some way to solve / improve the performance please fell free to provide some patch

Everything in life is a matter of give and take. We need to outweigh the effort and added complexity against the beneft.

Currently For The Record is perfect for multi-seat use, since the whole architecture is multi-seat from the ground up. You can even install the Web Access site or the IM-Bot on a different machine than your Core Services, and of course as many clients or even recorder/tuners as you wish on any machine on your LAN. Clients can even connect from anywhere in the world over HTTPS (except for watching/streaming of course, but you can edit schedules to your heart's content ).

So we prefer to keep the flexibility like that.

So far For The Record users are very enthousiastic about its stability, and this is certainly in part due to the fact that the architecture is solid and patch-free. So adding specific things for single seat situations may add complexity we prefer to avoid. Just saying that all this needs to be looked at carefully. Stability comes first, features come next. And features on their own need to be outweighed as well of course, because we love killer features as much as the next guy

 

[jonaskp]

Does this mean that the quality when usin multiseat could actually be a little bit worse than with the original TV plugin? I am not talking about freezes and stuff, but just about picture quality.

 

[tourettes]

So far For The Record users are very enthousiastic about its stability, and this is certainly in part due to the fact that the architecture is solid and patch-free. So adding specific things for single seat situations may add complexity we prefer to avoid. Just saying that all this needs to be looked at carefully. Stability comes first, features come next. And features on their own need to be outweighed as well of course, because we love killer features as much as the next guy

When it comes to stability users are much often reporting stability issues with the multiseat (RTPS) than with the single seat installs. So 4TR forces currently single seat users to have less stable systems than without it

 

[tourettes]

Does this mean that the quality when usin multiseat could actually be a little bit worse than with the original TV plugin? I am not talking about freezes and stuff, but just about picture quality.

Image quality itself cannot be worser. User experience can and will.