I'm not exactly sure where I stand on this whole issue so I thought I write it down to see if I could come to a conclusion.
Customer Perception
On one hand, I'd rather not have any of my customers reading Ward's article without an understanding of the OSS world, and then either complaining about the shiny new system I've sold them, or indeed not buying it, because they are worried about not being able to upgrade.
I can counter that argument by saying that exactly the same situation will exist in the proprietary market, except that the pace of change is not so quick because they have less developers, and a smaller testbed, so it's not so noticeable.
At least asterisk is standards based, so we are not going to have to throw away all our phones when we upgrade, unlike many proprietary PBX systems
It is likely that there is not a major upgrade path with a proprietary system, yes, there will be fixes and additions, but I would doubt if there is a major upgrade available on the same hardware.
However, having said that, PiaF, FreePBX and Ward in particular has coped with the dial plan issues, and although it has been a right royal pain, it has been done in a very very short space of time in the grand scheme of things, and very successfully
There in lies the beauty and curse of OSS software.
General Development Issues
When you program something, or provide a product, quite often you find that your users do not use the system in the way that you had envisaged, or having released it, you think of a better way of doing stuff, and you have to adjust your path to take advantage of this. Take text messaging as an example, that came about by accident, when it was released, they did not even charge for it.
Thus, you often find that you have programmed yourself into a corner, which you can only get out of by throwing the whole lot up in the air and starting again.
I know that Philippe at FreePBX is suffering from programming decisions made some years ago, which makes some of the features in FreePBX difficult to achieve without convoluted and complicated code.
However his view is to keep working round the problem, and rewriting things very carefully so as not to break it. He cannot throw things in the air and expect to get away with it , as he has not got the programming resource that Digium has.
So on one hand, I want to see the Asterisk base code improve with extra features, and I do understand that you have to break stuff in order to improve the system. However, I do wonder if stuff is being broken for the sake of it.
e.g. when you get a message like this one: -
|
Code:
|
pbx*CLI> sip debug
SIP Debugging enabled
The 'sip debug' command is deprecated and will be removed in a future release. Please use 'sip set debug' instead. |
Both commands work interchangeably now in that version - why can't they be left working in future versions.
Competition issues
This morning, I got an email from Digium entitled what's new at Digium, and top of the page is their SwitchVox appliance, which is a PBX designed for small and medium business.
This smacks of a wholesaler operating in a retail market. We base our systems on Asterisk, and we supply a PBX for small and medium business (and more).
Should I be worried that in the future, when/if Asterisk is the dominant player in the PBX market, which may be in part due to our (community) efforts, that they will do things to the underlying core of Asterisk to make it difficult for FreePBX, and thus make SwitchVox, their own product, more attractive to the purchaser?
Will they stop supporting Digium cards because we've installed FreePBX on Asterisk, because financially, that makes them better off?
If I was a hard-nosed business investor looking for maximum return, who had invested in Digium, I'd try and make it as hard for everyone else as I could, in order to reduce the competition.
Am I being paranoid or cynical if I wonder whether these changes are the symptoms of making life difficult for us, in order to reduce the competition?
I know that they have to make money to survive and grow, but I'm not sure that running with the hare, and hunting with the hounds is right.
Passing the Buck
We have been accused of passing the buck to FreePBX, Digium, and other players in this space when there is a problem. We are not. PiaF just drags a whole lot of good stuff together, and presents it as a package.
However, we are in a good position to see problems, and yes we may be able to fix those problems with an update-fixes, but it is much better to report those issues back to the original development team so they can fix them for everyone, and for future releases. That is how the software gets better. Incidentally, they also report back to us if we have assembled something wrongly, and that's our problem, so we fix it.
That is exactly what Ward has done here, and clearly someone is taking notice, the latest contributor to comment on his article is Leif Madsen, co-author of TFOT
In Conclusion
Ward's article has raised the concerns of the community, and particularly those who want to invest time, money and energy in this, from forum members, hobbyists, VAR's and fortune 500 companies.
To use the car analogy used by Jared when cars were first manufactured, they had all sorts of different ways of controlling them, tillers instead of wheels, different pedal arrangements, different power plants, e.g.steam, and it was not until the late 20's that cars were standardised to be the way they have been for the last 75 years or so, e.g. wheel in each corner, driven by petrol engine, and an ABC pedal arrangement.
Can we take it then that Asterisk development is still in the car equivalent of the 1920's, and we can expect in the future that they will release versions where they don't swap the pedals around. And now they have decided on the formula, will Asterisk just get better and better?
If this is the case, and this is an end to major syntax changes, and from now on (V1.6 forward), they are adding new stuff, not breaking existing stuff, as far as is possible, then I find that quite acceptable.
Joe Roper
Linear Mode
