XEnd TAP (similar to XStart)

[jumbo]

Is it possible to write a TAP (a la XStart) that will terminate (or disable) another TAP at a specific (elapsed) time?

[asrael]

[quote=""jumbo""]Is it possible to write a TAP (a la XStart) that will terminate (or disable) another TAP at a specific (elapsed) time?[/quote]

Yes, if the TAP is cooperative (TSR commander compatible).

Regards,

asrael

PS: For what reason do you want this?

[jumbo]

[quote=""asrael""]Yes, if the TAP is cooperative (TSR commander compatible).

Regards,

asrael

PS: For what reason do you want this?[/quote]

Excellent, who is the best person to ask for such a TAP? Firebird comes to mind

The reason I'm interested is that I get my TV EPG info from the web (not OTA). However, I do not get my Radio info from the same source, i.e. it is OTA. Consequently my Radio data is incomplete using EPG Navigator (a problem that I'm sure you are aware of). Improbox/JAG/EPGNavigator have support for Description Extender TAP which addresses this problem but, in my case, I can't use it as it'll mess up my TV EPG (which has "proper" extended info). If I can start with Description Extender and after a specified time stop/disable it, then I might be able to achieve my goal.

[asrael]

[quote=""jumbo""]Excellent, who is the best person to ask for such a TAP? Firebird comes to mind

The reason I'm interested is that I get my TV EPG info from the web (not OTA). However, I do not get my Radio info from the same source, i.e. it is OTA. Consequently my Radio data is incomplete using EPG Navigator (a problem that I'm sure you are aware of). Improbox/JAG/EPGNavigator have support for Description Extender TAP which addresses this problem but, in my case, I can't use it as it'll mess up my TV EPG (which has "proper" extended info). So if I can start-off with Description Extender and then after a specified time "stop" it then I might have the best of both worlds.[/quote]

Is Description Extender TSR commander compatible? Can you stop it in TAP commander?

Regards,

asrael

[jumbo]

[quote=""asrael""]Is Description Extender TSR commander compatible? Can you stop it in TAP commander?

Regards,

asrael[/quote]They are both written by simonc, so I'd like to say "probably" (not having tried it for sure). There's no "menu" to exit the TAP.

PS Description Extender source is in berlios.

[Sigittarius-E]

[quote=""asrael""]Yes, if the TAP is cooperative (TSR commander compatible)[/quote]

or the TAP has a TAPCOM interface for disabling.

[quote=""asrael""]Can you stop it in TAP commander?
[/quote]

is there any TAP out there which cannot be stopped by TAP commander? (AFAIK: TSR commander disables EVENTS to the disabled TAP)


If you look into the FireBirdLib manual you'll find a possibility to terminate or to disable TAPs.

HDD_TAP_Terminate (with the comment "untested") and
HDD_TAP_Disable

I think it shouldn't be a big problem to create a TAP (not configureable ) which runs TSR, checks the current time and disables a TAP at a specific time.

[Elle4u]

[quote=""Sigittarius-E""]
HDD_TAP_Terminate (with the comment "untested") and
HDD_TAP_Disable[/quote]

But terminate is not nice.

If a hook is set, it is not unset for example.
And memory is not given free.

The TAP_Commander command is the better way (if supported)

[Sigittarius-E]

Recently I saw that jumbo created the same thread in the uk forum. I think we should answer in the uk forum to reduce redundancy.

[jumbo]

Thanks everyone for your comments/suggestions. It's given me some food for thought.

[quote=""Sigittarius-E""]Recently I saw that jumbo created the same thread in the uk forum. I think we should answer in the uk forum to reduce redundancy. [/quote] UK forum is not a replacement of the German forum; it complements it well - so no danger of redundancy there.

[asrael]

[quote=""Sigittarius-E""]is there any TAP out there which cannot be stopped by TAP commander? (AFAIK: TSR commander disables EVENTS to the disabled TAP)
[/quote]

As far as I know any TAP can be disabled by TAP commander, but not each TAP can be stopped (e. g. (if I remember correctly) PiP, 3PG).

[quote=""Elle4u""]And memory is not given free.
[/quote]

From http://toppy.xwiki.com/xwiki/bin/view/Main/TAPMemFree:

The firmware is known to keep a list of which memory block is assigned to which TAP, presumably so that it can free all memory allocated to a TAP when that TAP unloads regardless of whether the TAP cleaned up properly. Of course this is no excuse to allow your TAP to leak memory...

Does this mean that, after killing a TAP with HDD_TAP_Terminate from the FireBirdLib, the Firmware will free the memory by itself or does this work correctly only if the TAP calls TAP_Exit()?

Has somebody tested it yet?

Regards,

asrael

[Sigittarius-E]

[quote=""asrael""]
Does this mean that, after killing a TAP with HDD_TAP_Terminate from the FireBirdLib, the Firmware will free the memory by itself[/quote]

if this is is correct - would it be possible to write a "mem watchdog" TAP similar to MemWatcher which checks the memory of each TAP and terminates a TAP if this TAP needs too much memory (maybe because of a memory leak) and would lead into a freeze of the PVR?

[jumbo]

[quote=""Sigittarius-E""]if this is is correct - would it be possible to write a "mem watchdog" TAP similar to MemWatcher which checks the memory of each TAP and terminates a TAP if this TAP needs too much memory (maybe because of a memory leak) and would lead into a freeze of the PVR?[/quote]
What an excellent idea! Firebird to the rescue - again

[R2-D2]

[quote=""Sigittarius-E""]if this is is correct - would it be possible to write a "mem watchdog" TAP similar to MemWatcher which checks the memory of each TAP and terminates a TAP if this TAP needs too much memory (maybe because of a memory leak) and would lead into a freeze of the PVR?[/quote]Yes, the TAP memory and file lists can be examined (see here) but as mentioned at the start of that thread the mechanism is not robust so should not be relied on to work properly (the information in the lists may not be complete). In particular, if a TAP is misbehaving enough to be leaking memory then killing it and hoping the system clears up is probably wishful thinking. But, of course, using these lists would be a good way of spotting which TAP is to blame and hopefully the author would be able to fix it. And this approach would be superior to that used in FireBird's MemWatcher v2 (especially if the firmware was patched to work properly...) because the memory would be attributed to the correct TAP even in the presence of hooks like DescriptionExtender.

[asrael]

[quote="jumbo"]Is it possible to write a TAP (a la XStart) that will terminate (or disable) another TAP at a specific (elapsed) time?[/quote]

I just stumbled over this old thread. If you are still interested: with current version of BootMenu you can terminate TSR compatible TAPs after a specific time.

Regards,

asrael

[Sigittarius-E]

I just stumbled over this old thread.

I guess you searched for XStartTAP because of this thread

[asrael]

I guess you searched for XStartTAP because of this thread

Yes, Sir Sherlock! :

Regards,

asrael