I have been using the following ir-translate line for ages

B9460D.L=menu,menu,right,OK.L,left,OK

This selects a specific playlist, and selects insert from the Insert menu to put it into the current playlist after the current track.

This has always worked fine.

Having upgraded from Hijack 283 to 325 and v2.00 final, I find that while it still inserts the playlist, it leaves the "Insert Menu" up on the screen.

At first I thought it had missed some of the commands and was waiting for a further selection, but any further button press it acted upon as if the menu was not on screen. I.E if I press the left or right button to move within the Insert menu it simply goes to the previous or next track as it would with no menu on screen. If I hit the top button (to clear the menu) it pauses play.

I have tried adding a ,null at the end of the line and this makes no difference.

I don't think it is a V2.00 final problem as I also noticed this problem when I upgraded to Hijack 320 for a short while still using v2.00-beta13 (320 caused problems with the Sony Stalk ir-translates so I went back to 283 and all was well again).

I notice in the Hijack version list that button_pacing was removed at some point between 283 and 320, I wonder if this has something to do with it.

I've had a closer look at what is actually happening.

I had previously thought that the playlist had been already inserted and the menu just left on screen, but this is not the case.

After the macro is run the "Insert menu" (Enqueue, Append, Replace, Insert) is left on screen and the playlist has NOT been inserted. As soon as any further key is pressed, the playlist is then inserted but the further key press is acted upon as if the menu was not active. (If I press the "right" button it inserts the playlist but then immediately skips to the next track)

It does seem like a timing error in that the menu cannot keep up with the stream of button codes. I would just go back to the earlier version of Hijack, but I really like the "start up volume" feature of the later Hijacks.

Try padding it with NULL.L between "button" presses..

Mark,

I tried that - infact I had NULL and then NULL.L between every button press at different times - still didn't work I'm afraid.

Oh yeah.. I suppose I probably removed the pacing completely or something.

The player has a huge button buffer, so there's no way we're overrunning it. I wonder why it's refusing the buttons?

-ml

Is the fact that it stops after the first .L button code a clue to the problem ?

The player has a huge button buffer, so there's no way we're overrunning it. I wonder why it's refusing the buttons?

I've occasionally seen similar behavior with Hijack ghosting long presses to the player.

For example...

Sometimes (rarely) when it's supposed to be doing the long-press to toggle visuals mode (when I have Restore DC/Car Visuals activated), it will do this:

- It seems to not restore the visuals, as if the long press on the bottom button hadn't happened at all. The screen doesn't change to indicate any keys were ghosted at all.

- The next key I press on the player, even if it's just a knob-turn, the main menu will open, (i.e., it got a short press on bottom button or perhaps just the Key Released event for the bottom button), then that will be followed by whatever key or knob-turn I'd done.

This is very rare and intermittent, I can't get it to happen every time. But it sounds a lot like what GeoffTT was describing.

Mark,

I've been having a play around with this problem and it's seems that it is the .L command that is causing the problems.

I've programmed a much longer test string (11 commands)without any .Ls and it works fine (without any null or null.L strings inbetween) but my original string still "hangs" at the OK.L even with a null.L between ever command. This is the only .L that I use in all of my ir_translates and it is the only one that doesn't work. I have mapped the same string to a Sony Stalk button and this hangs at the same point.

I have tried 330 and that acts in the same way. If I go back to 283 that works fine, but I really like the boot volume command and as I've got my tuner working properly as last, I need the RDS improvements of 330.

You know, I just discovered an issue with an IRtranslate .L string, and it's an interesting one. I wonder if it's related:

My config looks like this:

[ir_translate]
Left.L=Left.L
Left=PrevTrack.V,Left.N,Left.U
Right.L=Right.L
Right=NextTrack.V,Right.N,Right.U
Top.L=Top.L
Top=Pause.V,Top.N,Top.U
Bottom.L=Bottom.L
Bottom=Bottom.V,Bottom.N,Bottom.U

I'm doing this so that the "volume slider dialog-screen thingy" clears and then executes the specified command, instead of just clearing, when you press a button (we discussed this in another thread when mark added the .V feature).

The problem happens with the Right.L=Right.L and Left.L=Left.L. For some reason, I need to have these in the config or it doesn't work. But anyway, here's what happens when I use them to FF/REW: Sometimes the "key released" command for the button never reaches the player. I begin to rewind from the front panel and it works, but then it never stops rewinding when I let go of the button. I have to press and release the key again to make it stop. And yes, I'm sure it's not hardware, the button itself is not sticking.

Mark,

Is there any further testing I can do on this to pin down the problem ? I would gladly put some effort into this myself rather than just keep bugging you.

I'm ignoring it, in hopes that it will go away.

But if you bump it up once in a while, I may eventually have a look.

Can't right now, my devel system is down for a memory transplant (installing RH9).

Cheers

I'm ignoring it, in hopes that it will go away.

LOL - I can certainly understand that view.

[grovel] I don't know how you find time to do any paid work with all the excellent work you do on the various Empeg projects.[/grovel]

I will bump this every now and again and see if I can find a work round until you get a chance to have a look at it.

Turn on the hijack ir_debug flag (check the FAQ?) and post the resulting trace.

Thx.

OK Mark here goes

ir_translate line

Visual=menu,menu,right,OK.L,left,OK

Debug data

114824: IA1:0020df15,dk=00ffffff,dr=0,lk=0

114824: IA2:0020df15.P,dk=00ffffff,dr=0,lk=0
114824: IA2: tflags=fe, flags=94, match=0
114824: ENQ.I: 0020df12.0
114824: ENQ.I: 8020df12.0
114824: ENQ.I: 0020df12.0
114824: ENQ.I: 8020df12.0
114824: ENQ.I: 00000002.0
114824: ENQ.I: 00000003.0
114824: ENQ.I: 0820df12.0
114824: ENQ.I: 8020df12.133
114824: ENQ.I: 00000004.0
114824: ENQ.I: 00000005.0
114824: ENQ.I: 0820df12.0
114825: ENQ.P: 0020df12.0
114825: ENQ.P: 8020df12.0
114826: ENQ.P: 0020df12.0
114826: ENQ.P: 8020df12.0
114826: ENQ.P: 00000002.0
114826: ENQ.P: 00000003.0
114827: ENQ.P: 0020df12.0
114827: ENQ.P: 8020df12.133
114827: ENQ.P: 00000004.0
114827: ENQ.P: 00000005.0
114828: ENQ.P: 0020df12.0
114828: DEQ.P: 0020df12.0
114829: DEQ.P: 8020df12.0
114829: DEQ.P: 0020df12.0
114829: DEQ.P: 8020df12.0
114830: DEQ.P: 00000002.0
114830: DEQ.P: 00000003.0
114830: DEQ.P: 0020df12.0
114841: IA1:8020df15,dk=0020df15,dr=0,lk=0
114841: IA2:8020df15.R,dk=0020df15,dr=0,lk=0
114841: IA2: tflags=fe, flags=94, match=0
114841: ENQ.I: 8020df12.0
114842: ENQ.P: 8020df12.0

Mmmm... there should more data after that.

Speaking of which.. if that's the only ir_translate line that's bugging you, there's now a MUCH better way to do it! Use the VisualSeek virtual button, or the PrevVisual and NextVisual virtual buttons.

Cheers

By the way, here's a trace of the output when I get the "runaway fast forward" that I described earlier in the thread. This is me pressing the right front-panel button, holding it for a moment to induce the longpress, then releasing it as it the longpress gets sent to the player. Then, it begins fast forwarding and does not stop, so I press it once quickly. This is the debug trace from that action :


475148: IA1:00000002,dk=00ffffff,dr=0,lk=0
475148: IA2:00000002.P,dk=00ffffff,dr=0,lk=0
475148: IA2: tflags=ff, flags=94, match=0
475250: ENQ.I: 00000002.0
475250: ENQ.P: 00000002.0
475254: DEQ.P: 00000002.0
475292: IA1:00000003,dk=00000002,dr=0,lk=0
475292: IA2:00000003.R,dk=00000002,dr=0,lk=0
475292: IA2: tflags=ff, flags=94, match=0
475292: ENQ.I: 00000003.133
475294: ENQ.P: 00000003.133
475528: IA1:00000002,dk=00ffffff,dr=0,lk=0
475528: IA2:00000002.P,dk=00ffffff,dr=0,lk=0
475528: IA2: tflags=ff, flags=94, match=0
475538: IA1:00000003,dk=00000002,dr=0,lk=1
475538: IA2:00000003.R,dk=00000002,dr=0,lk=1
475538: IA2: tflags=ff, flags=94, match=0
475538: IA2: tflags=fe, flags=94, match=0
475538: ENQ.I: 0420df11.0
475538: ENQ.I: 8420df11.0
475538: ENQ.I: 10000002.0
475538: ENQ.I: 10000003.0
475538: ENQ.I: 20000002.0
475538: ENQ.I: 20000003.0
475539: ENQ.P: 00000002.0
475540: ENQ.P: 00000003.0
475542: DEQ.P: 00000003.133
475543: DEQ.P: 00000002.0
475543: DEQ.P: 00000003.0

For reference, my config looks like this:

[ir_translate]
Left.L=Left.L
Left=PrevTrack.V,Left.N,Left.U
Right.L=Right.L
Right=NextTrack.V,Right.N,Right.U
Top.L=Top.L
Top=Pause.V,Top.N,Top.U
Bottom.L=Bottom.L
Bottom=Bottom.V,Bottom.N,Bottom.U

Mark,

The string has got nothing to do with Visuals, I just randomly picked the Visual button on the Handheld remote to map to for the debug test. In the car the string is mapped to one of my Sony Stalk buttons, but as I was using the serial port on my desk I didn't have the tuner or stalk connected.

The string inserts a specific playlist into the current playlist after the current track. I find it really useful as I have the Empeg playing all tracks shuffled randomly most of the time but then have a "current favourites" playlist which I can then insert at the press of button.

I will have another go tomorrow and see if I was missing any data.

Okay, fixed.

Hijack v331 will be out in a few minutes.

The only change is that the ".L" translations in ir_translate have been fixed.

Cheers

For the curious, this bug affected ANY usage of .L to the right of the '=' in ir_translations. Basically, Hijack was forgetting to wake-up the player after the 1.3sec .L delay expired.. so the player never received the button release until the next button press by the user. This could have all kinds of nifty side-effects.

Cheers

I think hijack v331 fixes these problems also. Let me know if it does not.

Cheers

Mark,

Thanks for the fix!!!!


Geoff ,

Thanks for bugging Mark until this got fixed.

Now I can finally use Right.L=Bottom,Bottom.L,Top,Right,Bottom,Bottom,null to jump to the next playlist.

I think hijack v331 fixes these problems also. Let me know if it does not.

Your description of the fix sounds like it definitely would fix it. I'm going to try it now, I have a high degree of confidence it'll take care of the problem. Cool! THANK YOU!

You da man, Mark. You da man.

Yeah seems to fix it. ROCK!

Mark,

Thanks very much, that works great. You are a STAR.

Geoff

Just went from Hijack v324 to v331. I regularly use the NextSrc feature in the Popup0 menu. With a tuner fitted and working correctly, NextSrc went from MP3 to Aux and back again.

Tuner was definitely working since selecting it from the remote worked as per normal. I have no IR translations whatsoever currently. Worked fine in v324 (MP3->Tuner->Aux->MP3).

I do NOT see that behaviour here.

Anyone else?

Did you fail to use NextSrc to find the tuner before or after selecting it manually once? I think I saw the former behavior once recently, but not with 331 (which I haven't loaded yet, but will today)

Now I can finally use Right.L=Bottom,Bottom.L,Top,Right,Bottom,Bottom,null to jump to the next playlist

NICE little macro there... why didn't i think of that!!

Oh wait, I DO see that behaviour: Select MP3 with the remote, unplug the power, reconnect power, hit knob->nextsrc, and watch it cycle MP3 - AUX - MP3 - AUX ...

I'll have a closer looksee.