186 replies to this topic

[sambir]

Bought a unicable II lnb and would like to compile my own image for a Xtrend ET10000, Mutant HD2400 and VU Zero since this will not be included in the master GIT.

 

Where do I start?

[littlesat]

I just saw the jess branch is removed... 

[sambir]

What does that mean? So I will be unable to compile an image? Never tried it tho...

[littlesat]

You need to clone our git... Find Huevos patch and commit it in the clone and then try to build your image...

[sambir]

Well I will look at it this evening. I cannot find Huevos commit. To be honest I never built something from Github ever...

 

Only downloaded the images from the openpli site. 

[sambir]

this looks like Huevos's commit:

https://github.com/H...7a4a2599ffd4d08

[Huevos]

@Sambir, I wouldn't bother. The commit was in sync with PLi at the time but it was never accepted and now after this amount of time you are going to have major problems with merge conflicts.

 

I would just use an image that has this technology already built in and maintained, and team members who have the hardware and are familiar with it (OpenViX, OpenATV, OpenMIPS).

[littlesat]

Inderdaad an alternative, but the code really sucks... It hurts that they don't care.... But we can't change it...
Sorry for the hard words als I still understand it is a lot of work...
It also reports slower zapping

Edited by littlesat, 11 April 2016 - 20:19.

[sambir]

I guess I will load openatv on the ET10000 and HD2400 for JESS and let the dm800 and VU-ZERO work on legacy (unicable 1) frequencies.

[Huevos]

OpenViX covers those 3 boxes and is a very user friendly image.

[littlesat]

You can say that as you're one who tried to get jess into openpli...

[littlesat]

And here there exists a less complicated jess patch https://github.com/a...d8efdbc6ee721bd

as far I can see this one is really stripped to the real core...

Edited by littlesat, 11 April 2016 - 21:36.

[Huevos]

You can say that as you're one who tried to get jess into openpli...

Littlesat, I say it with all due respect for OpenPLi.

[littlesat]

I understand... no big words.. See he patch from adhoik... As far I can see this does not include the timing trick extra's and others... So it will also quickly switch... Also for the normal legacy stuff...

[Huevos]

And here there exists a less complicated jess patch https://github.com/a...d8efdbc6ee721bd

as far I can see this one is really stripped to the real core...

Look like the initial patch.

[littlesat]

P.s. It would be nice if someone who has a jess config could check (and debug) adhoiks patch...

[littlesat]

At least as far I can see adhoiks JESS patch suggestion it does not effect the current 'system'....

[arn354]

The one from VU is only the first part of it. Not everything will work - i do not need to check, as i've done that a year ago.

Additional changes were not done just to do something and to make things more complicated - they were needed.

If you think you can improve that, you should do it based on HUEVOS JESS branch as base and add the last one or two additional fixes first.

Edited by arn354, 12 April 2016 - 14:59.

[littlesat]

As far I can see VU's patch is very clean... and there are not timing changed. So as UNICABLE works, that patch should work too..

 

All/Most further additions seems to be timing related and that give me the impression it does not work on specific hardware...

 

I understand the work that was done trying to fixed things... But I'm afraid it went into a non structured direction.... It starts to look like spaghetti code... (But I'm afraid to say that because some can see this as an arrogant argument and I really understand the big work)....

 

In our team we do not own a JESS system and we want to test it first before we commit it... And the current patch is such big it does not get anymore a clear overview what is included here...

 

At least the timing fixes (when really required) and the real JESS patch should be splitted to get good overview... And when asking this we ask again someone who owns and can check JESS needs to put tons of hours into this project.

 

But as far I can see the commits the 'basic' patch should work for the VU basic image.... As the basic VU image did furthermore nothing with it...

 

It would be nice when we could get a good clear patch... I suggest you are more aware about it...

Edited by littlesat, 12 April 2016 - 15:24.

[littlesat]

P.S. Can you at least explain why e.g. we need guard_offset.. TAKEOVER_COUNTDOWN, TAKEOVER_MASTER, TAKEOVER_SLAVE, TAKEOVER_RELEASE...

 

Here I get a bit lost... and as far I can see this is one of the extra's added later...