Errors on Build

A place for developers to discuss development of OSA
Message
Author
User avatar
Mat
Posts: 501
Joined: Thu Aug 09, 2012 2:35 pm
Location: UK

Re: Errors on Build

#11 Post by Mat » Mon Jun 24, 2013 8:54 am

If I was to take a wild guess I would go with:

http://xicore-ip.com/

mstamler
Posts: 20
Joined: Sat Jun 08, 2013 9:44 am

Re: Errors on Build

#12 Post by mstamler » Tue Jun 25, 2013 4:04 am

Hi,

As I said, I am just starting to get acquainted with your code base and would like to compile the entire solution. I will work in individual project as well.

Can't you just update the GIT with your latest link and mark it with a TAG/Branch?

What do you think of my suggestion to establish working branches/tags and then continue development on the master. This is how it is usually done with version control.

My "family of products" are still in the planning stages. I really mean a "system" that will consist of various plug-and-play components, hopefully based on your code.

BTW: Your code is based on Windows .NET. However, most low-cost Home Automation uses Linux/Android. What is your comment on this?

we don't have a site yet but I have a consulting company that is the mother company: www.xicore.net

BR
Michael

Automate
Posts: 1691
Joined: Sat Dec 11, 2010 1:44 pm
Location: US

Re: Errors on Build

#13 Post by Automate » Tue Jun 25, 2013 8:06 am

mstamler wrote:Your code is based on Windows .NET. However, most low-cost Home Automation uses Linux/Android. What is your comment on this?
No one is doing it yet but the code could be modified to run under Mono which will work on Linux.
http://www.mono-project.com/Main_Page

User avatar
Mat
Posts: 501
Joined: Thu Aug 09, 2012 2:35 pm
Location: UK

Re: Errors on Build

#14 Post by Mat » Tue Jun 25, 2013 12:39 pm

Im not sure how familiar with GIT you are but everyone works in there own branch, nobody works on the main branch. When a user is ready to submit there work they create a pull request to the primary branch which is either declined or accepted by some one with the appropriate permission on the main branch. That doesn't obviously help the situation where a mistake makes it into the primary branch... I think it was just an unlucky co-incidence in that there were a number of check-ins going on between Brian and myself in doing work in prep for release. Tags would be a potential option but might result in the overall development cycle becoming less agile. I will try and get the missing file checked in to help resolve your issue.

The second question is a somewhat larger one. From the knowledge I have of OSA I don't think the core system would be overly difficult to get running on Linux, I think the real problem would be with the plugins, where they are using 3rd party products to interact with hardware. To coin the frase "Internet of Things" there is a lot of technologies and options out there all trying to compete for the number one slot in home automation, all pushing very different approaches some of which end up being very costly when you add the bolts and whistles. What OSA is really doing is sitting between those components and providing a mechanism by which they can interact with each other so that you are not forced to used the entire range from 1 manufacturer. Being that OSA has no real restrictions in what you can do your not bound by what the manufacturer says you can do.

Another thing that has been seen recently is that there are a number of systems that initially start as free and they are cloud based and once the user base reaches a certain level they start charging a fee look at COSM as an example.

In terms of Linux there is the obvious option of RaspberryPi, low power consumption, low profile so in many regards a ideal solution. However I work in a large bunch of engineers and a common consensus is often that they are under powered, thats not to say there not good as they are but its worth bearing in mind what they were intended for which is kids in schools. I am aware of people building HVAC systems using a combo of Arduino and Pi and the results from there testing was that the Arduino are rock solid yet the PI was somewhat temperamental and would often go down which is not something you would want from a high availability system. I experienced a number of similar issues myself and was only with a number of tweaking that I have managed to achieve a stable system.

mstamler
Posts: 20
Joined: Sat Jun 08, 2013 9:44 am

Re: Errors on Build

#15 Post by mstamler » Tue Jun 25, 2013 9:22 pm

Hi,

I wrote a reply and posted it but it never registered. Here it is again:

We are developing a system for home automation. I am excited to see this opensource code base that already has a lot of momentum and I am evaluating its efficacy for our needs. I like the support for devices but am concerned at the .NET heavy technology especially in light that embedded product are heavy Linux/Android based these days. What do you have to say about this?

Yes, we are ready to contribute to your efforts with our contributions!!

Regarding GIT:
I will compile separate projects but the solution should also build. I am in the phase of learning your system and code and need a complete system.

You really should create a stable version, some version, that has ALL of the components and that builds cleanly and then tag with some branch label. This is how it is done in all revision control software. Then you can continue to develop with or without errors on the master.

So, can I get the missing file/files/distribution so that I can build?

Thanks
Michael

mstamler
Posts: 20
Joined: Sat Jun 08, 2013 9:44 am

Re: Errors on Build

#16 Post by mstamler » Tue Jun 25, 2013 9:27 pm

Hi,

I just got your response so my latest post is superfluous.

You got our site right but we use www.xicore.net instead. Soon this will change to www.xicore.eu as we are also Paris based.

You know, for a commercial product it has to be cost-effective, so a ARM/Raspberry/Linux/Android approach makes the most sense here. I have been working for many years with ARM with very heavy technologies, such as video processing, and I have never run out of horsepower with the ARM.

BR
Michael

User avatar
Mat
Posts: 501
Joined: Thu Aug 09, 2012 2:35 pm
Location: UK

Re: Errors on Build

#17 Post by Mat » Wed Jun 26, 2013 5:55 am

I would for the most part agree as nobody wants to spend more money than they have to. Something to bare in mind however if you asked a less technical person what Linux was they probably couldn't tell you, where as for the most part people have used windows and are familiar as to how it works. And I think that is one of the strengths of OSA it allows less technical people to Automate there home.

One of the key points you make is about "commercial product" which OSA is not, it is open source and is at no cost to the user.

The most recent video from Intel on automatedhome re-iterates my point in that they smaller hardware components do not have the processing power to achieve some of their goals and as such it gets offloaded to a central system to be processed.

fiveHellions
Posts: 86
Joined: Wed Feb 06, 2013 12:12 pm

Re: Errors on Build

#18 Post by fiveHellions » Wed Jun 26, 2013 6:43 am

I agree with Mat that if people had to set up a linux box or rasberry pi or something similar to be able to run this it would make getting started alot more difficult for alot of people.

The only real reason I would see to make it run on ARM/Raspberry/Linux/Android is if the goal is to have small appliance that you can sell to people that they can just plug in and not have to worry about installing something. Having an appliance would make it much more consumer friendly.

The average consumer doesn't want to run products like this on their computer. They want an appliance that they can just plug in and use. Just look at windows media center, it is arguably one of the better dvrs out there, but it was a total flop for Microsoft because your average person just wanted to plug a box into their tv. They didn't want to set up software, and add hardware tuners and find a way to connect a computer to a tv.

That being said I don't think OSA's goal (as least not currently) is to be a consumer product. I think that it is built mainly around the tinkerer that likes to have total control of what their system can do. The people who want to do more than buy a fancy light buld just so they can push a button on their smartphone to turn it on and off (cool but not very useful, especially considering the cost).

I think they aim to make it more user friendly as it evolves, but I see that as a ways off. I still see the focus on making the system powerful more that user friendly. And I don't really see it straying from a windows/.net product anytime soon. Most of the current intended audience will have no problem running thier central server on a windows box.

So I guess what I'm trying to say is, Welcome. Feel free to come in and learn, contribute and share ideas. But don't expect the entire project to change direction just to fit your commercial needs. If a commercially available appliance running on ARM/Raspberry/Linux/Android is what your after then feel free to learn parts of OSA and take concepts from OSA and use that to develop your own product written for those systems.

Vaughn
Site Admin
Posts: 1432
Joined: Thu May 13, 2010 2:17 pm

Re: Errors on Build

#19 Post by Vaughn » Wed Jun 26, 2013 3:42 pm

Well, with new Windows 8 Haswell tablets and laptops, we could put out something like an appliance, for a few hundred bucks and still stay on the Windows platform. With the miniscule ability of Rasberry Pi's, etc, it would really be a completely different product filled by another project out there.

I like the Windows .NET platform, which is why I work on these projects, else I may have just ran Mr. House or something. Now, at the same time, I am all for the integration of these different platforms. I would love to see a java plugin running on a Linux box. Maybe even the DB hosted on a Linux box, as long as I do not lose my ability to run Windows plugins against the same system.

Vaughn

Post Reply