SUSHI - Software Update Service (Hyperlinked, Interactive)

[wardmundy]

We like to test stuff on our own hardware prior to sticking it into the released code. This helps ensure a better SIT fit *Seamless Integration of Technology*

I recall using a similar line in the Dating Game back in the Old Days. Usually only got one of these...

Hopefully, we'll do better with the hardware manufacturers.

 

[mkhurrum]

I like the idea and I will contribute towards it. Some customers have their servers closed and do not allow automatic updates and stuff - in such cases can we allow manual updates using this subscription?

Regards,
MK

 

[wardmundy]

All SUSHI updates will be manual just like update-scripts, update-fixes, and update-source are today.

 

[thunderheart]

Release Ugrade Path

Ward,

You mentioned in another thread recently that you guys might work out an upgrade path from say 1.1 to 1.2 or maybe even 1.1 to 1.3 without the need to reinstall from iso. Will this be something that SUSHI will provide?

Dallas

 

[darmock]

Hi Dallas

At the moment there is no way to convert from PIAF 1.0 or 1.1. That being said there will be the ability to convert your 1.2 box into a 1.3 box. This feature will be available sometime after 1.3 is officially released.

Perhaps if there was a bounty offered of sufficient funds we could offer something like this but right now our resources are concentrated on PIAF 1.3 and also the official release of Asterisk 1.6 non beta.

Tom

 

[thunderheart]

Kewl

Well, I thought that might be the case vis-a-vis 1.0/1.1. I guess what I'm really asking is going forward, will there be an upgrade path between major releases that doesn't involve a re-install that wipes the machine. Excluding 1.0/1.1 of course.

The reason I ask is that on one of my boxes (my home machine) I use samba to serve up my wife's 1 gozillion tropical fish pictures and my music library. Hate having to restore them whenever I upgrade.

Dallas

 

[slickrock22]

Couple suggestions. Have optional tiered pricing.

Required
$30/min

Optional
5-15 Handsets (+$50)
15-30 Handsets (+100)

and so on. That way you have to pay at least $30/bucks but then those users that have a large setup can be provided a suggestion for what to pay. They of course only need to pay the $30 but setting a schedule for larger implementations allows fair people to pay fairly for the value they receive.

Also consider a PLUS solution that adds FreePBX contract support as an options. ($30 + (0-5 Handsets) + $350) annually and then FreePBX gets 100% of the proceeds and the users get support. Just my .02

Ryan

 

[darmock]

Hi Dallas

Yes that is exactly what we will be doing. It is one of the reasons that I rewrote a lot of stuff in 1.2 so the update path could be followed. Of course there are several major exceptions at the moment.

1. There is will be no upgrade path from beta to full release. Thus once Asterisk 1.6 comes out of beta there will be no official upgrade path to this. You will have to reinstall 1.6 final release to get full functionality. This is why it is called beta. We will continue this policy when Asterisk 1.8 comes out of pre-alpha. Of course any policy we post will rankle a vocal minority so they are encouraged to come up with a way of migrating a beta system to a full release.

2. At the moment I cannot foresee an easy way to move from asterisk 1.4 to asterisk 1.6 (final) While the Ast14 > FPBX backup > sneakernet > FBPX Restore > Ast 1.6 may work it is too early to tell. From what I have seen at the core level of 1.6 any upgrade from 1.4 may prove too problematic to be done in a cost effective manner. I have a hunch that a significant bounty might have to be offered and even then it might go unclaimed. Of course you never know Digium might offer an upgrade wizard!

 

[jroper]

Hi Ryan

I'd be nervous about counting handsets, this would suggest that the system would have to "phone home" to tell us how many handsets you have got.

We would would like to build up a level of trust with our users, with no phone home scripts, no root exploits, and not giving the name and extension number of everyone who cares to look at your internet connected PBX.


Joe

 

[andygee]

I believe he was meaning the handsets would be for guidance only, not a hard coded thing. I tend to agree as most installations of 50+ handsets can more than afford $30 and this would give them the opportunity to pay more without forcing the hand.

Hi Ryan

I'd be nervous about counting handsets, this would suggest that the system would have to "phone home" to tell us how many handsets you have got.

We would would like to build up a level of trust with our users, with no phone home scripts, no root exploits, and not giving the name and extension number of everyone who cares to look at your internet connected PBX.


Joe

 

[carlosmp]

Getting sample hardware to test with PBX in a Flash is an ongoing problem, and this funding will certainly help. I would hasten to add that Sangoma has been the hands-down, most generous contributor to the PBX in a Flash project, and we're working hard to include setup scripts just as you have suggested.

Any chance of "fixing" the breaking of the sangoma cards during an update-source? I've resorted to removing the sangoma card drivers before performing, since this has fixed it 9/10 times while performing remote updates. It would be nice to detect and rebuild the sangoma drivers when updating the kernel...

 

[jroper]

Hi Andy Gee

On re-reading, I did jump the gun a bit. However, the opportunity does and will remain to make a voluntary donation if you think it is apropriate.

Joe

 

[DOA62]

Count me IN!!!!!!!!!!!!!!!!!!!!!!
You all have done a Great JOB!!!!

 

[dghundt]

Will pbxiaf 1.3 have update-scripts, update-fixes at first? Then we go to the $30/year model?

 

[wardmundy]

Sure. When we switch to the paid model, it will encompass all of the versions. Until then, everything will remain as it has been.

We're also going to kick things off with a *BONUS* -- with some caveats, we actually have inbound faxing working.

 

[TheShniz]

We're also going to kick things off with a *BONUS* -- with some caveats, we actually have inbound faxing working.

w00t, that's some great news!

 

[wardmundy]

And darn if the outbound faxing isn't working, too... at least so long as you don't send and receive from the same system.

The only wrinkle seems to be VoIP providers. They're all over the map. In fact, the only one that is 100% reliable for faxing (at least for us) is good old TelaSIP.

 

[titodj]

Just one question...
As I said before I will be glad to pay, because I think the system is great.
But in my case I'm not sure that will be able to upgrade to PIAF 1.3 because the system that I use just have 256MB of RAM.
And it works great for me...
3 DIDS 3 Ext 2 Trunks. Nothing fancy, just the phones in our house.
And I'm quite sure there are more users like me.

So will it be a way to keep our boxes up to date, or will it be necessary to upgrade ??

 

[wardmundy]

You can keep it just the way it is, and use update-scripts and update-fixes once a week to keep your system current. The RSS Feed will tell you if there is a security issue with CentOS that needs to be addressed.

 

[undrhil]

Hi there.

$30 a year just doesn't seem right. I mean, I know you guys want to make the money but isn't $30 a year kind of short-changing yourselves? I would gladly go with a monthly fee of $10 or $15 dollars. Sure, the more servers you have, the more expensive it gets.

So, maybe an incentive? If you license more servers, you pay less per server per month? This would allow the PBXiaF project to have a monthly inflow of funds and allow the community to save a little green.

Also, if you stick with $30 a year for the pricing, I think there are quite a few people who will "pay" more just because.