SSlmixed.com Forum
XK-System (community project) => Software Development => Topic started by: Artur D'Assumpção on June 08, 2014, 09:00:40 PM
-
Hi guys,
Good news, we will soon have a fully working hardware early prototype of the new computer! \o/\o/
We'll then start to develop the software including the main desktop application. We plan to support both OS X and Windows, but since our time and resources are limited we need to give priority and base our development in one of the operating systems and port it to the other on later stage. The question I am asking you guys is which operating system should have priority, OS X or Windows?
I am personally a Windows user, but I know that perhaps I am not the norm here. I would like to hear what's your opinion on this so we can get most of you guys covered on this early stage.
Thanks!
Cheers,
Artur
-
Mac-User here...
-
running it on windoze would scare me.
most people would like windoze because they are used to it and its familiar.
running it in a unix/linux environment might be marginally better, that means Linux or OSX.
You are developing this as a non dedicated machine? If it was dedicated, then Linux might be better because it could be a fixed release machine and then years from now we could still load and run it (maybe). Will we be able to run it on windoze 11? won't have support for the old OS as MS wants to kill off their old OSs. ie run it on XP, ha. OSXI, no idea.
Linux on a mac mini appeals to me.
some random thoughts on a beautiful sunday afternoon
Cheers
Alan
-
Mac user here (hackintosh too!)
M.
-
Mac also.
Agree with Alan, a Mac Mini, is the perfect host for this kind of thing. Cheap (relatively) and silent running until massively taxed.
-
Hi guys,
I think there might be some confusion.
This is just for the Desktop client (desktop environment). The actual computer is running on Linux on a separated machine that's inside the I/O computer rack. This is running with a realtime kernel so it's very very reactive and as near zero-latency has it's possible in a Linux realtime dedicated system.
All the rest will connect to the computer through TCP/IP, this includes the desktop client (GUI stuff). We are also planing other interconnections through the network such as displays (as many as you want) and remote controllers.
Regarding the desktop client (main application) you can install it wherever you want, each configuration with its advantages/disadvantages.
As I see it, one of the main advantages of installing the desktop client exactly on the same machine where you run your Protools (or other DAW) installation is the following:
- One less computer to have (one less keyboard, one less mouse, one less monitor)
- Everything's integrated in one place
- Time Code slaving/Transport integration can be achieved without use of any cables:
- This means you'll be able to slave the SSLMixed Computer to TimeCode support without using any SMPTE or MTC/MIDI cable in the scenarios where your DAW is Time Code master (most of them I imagine). This is possible because we'll develop a virtual midi MTC IN where you can send your TC reference directly from DAW to the client inside the same system without having to get the signal out through a cable and back in.
- This is exactly the same approach for the Transport integration support.
- Of course if you still use a different system/appliance other than your DAW to provide TC reference on your studio you'll still be able to use SMPTE or MTC through a cable to sync the system. This will be possible through a special purpose I/O card we'll design and that you'll slot in the computer. In this scenario the desktop client application will get TC reference through TCP/IP.
Regarding the OS X/Windows debate I think we need to be cautious with a decision in here. Not all userbase is OS X, although I agree the great majority of it is.
You have the example of Apogee that ditched the support of Windows and I hated that because I really can't use their gear and hate to be forced to use a specific system, specially when I don't use it. For this reason I would prefer to support both systems, even if that means a bigger workload in the future. This is still under consideration at this stage of course...
Also regarding the application code maintenance for the future, that is something that will be always necessary. An application written for OS X 10.8 with certain framework dependencies will NOT work on OS X 10.3 or 10.10. Look at what happens with Protools, everytime a new major version of OS X is released they have to do something in the code to make it 100% stable and supported. So there's no such thing as having an app that will work forever without any maintenance, unless you're doing only POSIX compliant C applications and with shell command line look and feel... I bet you guys would hate that and kill me! ;)
I have to be honest I am currently having a big big dilema. I am not a OS X user and I was willing the start developing the systems for OS X, using a virtual machine as my developing base. But the Yosemite (OS X 10.10) really is coming as a big show stopper to me due to SWIFT, the future new developing language by Apple.
Apple is killing Objectve-C (which SUCKS!!!) and slowly giving place to SWIFT. It is foreseeable that in a near future Apple kills the Objective-C support. Should we invest lots of time and effort developing on a language that has its future condemned?
I was thinking that we should take the plunge and wait for Yosemite (I can apply for the beta to get it earlier) and look into the SWIFT thing. We could start the development directly on this language with the upside that we would be investing in the future and downside that users pre-Yosemite can't use it.
This is my dilema and what at the same time pushes me to start the development on a Windows system, while waiting to see what's the big deal with the Yosemite+SWIFT.
Let me know your thoughts.
Cheers,
Artur
-
Good news, you can start developing in SWIFT on OS X 10.9.3 if you install XCODE 6 Beta.
This is good news since I could tackle the SWIFT language right now without having to wait for Yosemite (october I think).
Cheers,
Artur
-
in that case, Mac as most of the clients i have that use SSL and PT are Mac based.
Alan
-
Alan,
Yes, I am now 95% inclined to start the development with Mac, I am already studying and prepping the environment on a virtual machine (I don't have an apple computer)
Cheers,
Artur
-
Hi Artur
Over the years, I have visited loads of studios and practically all of them have Mac computers, even non SSL ones. The last one I saw which had a PC was in 2002 (Enya's studio in Ireland).
Surely Mac is the way to go and, as someone once said to me, a Mac is built as an aircraft but a PC is built as a car with wings! Macs are arguably less liable to crash and have have viruses, easier to learn and are much more affordable these days. They used to be associated with graphic design many years ago but, with the arrival of the old colourful iMac, much later followed by the iPad and iPhone, their use has increased dramatically today.
At SSL, the mix software was complied on a PC but Colin Sanders always had a Mac handy in the background (Classic SE I think?) and preferred that system ideally even though it couldn't handle the SSL requirements at the time (1980 ish).
Like others on here, I would strongly suggest going with the Mac operating system. Hope this helps get the vote up to beyond 97 - 98%? Regards from Jim.
-
I vote for osx.
-Paavo
-
Eheheh I see that I am the only one here using Windows! ;)
Honestly I never had any problems... since Windows 7 it's pretty stable and works flawlessly. :D
-
I think the future of Mac and Win is actually a very open question regarding PT.
Look at the current offering from Apple for computer and project forward. No slots on the top end. No slots in the iMac, no slots in the Mini. Must be native or expansion chassis. And where does your interface attach? Its all stand alone thru the Thunderbolt conn or USB.
Where will this lead, too early to say. I stopped guessing years ago after being wrong so many times. (I started supporting digital audio in 1991).
Heck, we could all get pushed off the major OSs because they all morph into something that is hard core interfaced with "smart devices". Who knows.
thoughts for the day!
Cheers
Alan
-
I'm intrested what happend to web browser interface plan? That would work with pretty much every os.
-Paavo
-
The web browser is still there for configuration purposes, it was never meant for the controller. We need something that's more real time and the HTTP protocol + web languages (html + css + javascript) is not adequated for that.
-
Sorry to dig up this old thread.
Just one question :
why not using libraries/tools that are available on both (all) systems, like g++ ?
I know there are libraries that make an application look 'native' on the OS it's compiled on (I am thinking of wxwidgets at the moment).
It'll also be available on Linux sytems as well and this is free software.
Which, in turn, would mean, that you'd have to respect the GPL (or whatever is used for these libraries).
Just my 2ct
Matt
-
Hi,
We've thought about this issue and came up with a conclusion:
At this moment the Manager component (software that runs out of the I/O racks) will run on Linux, on a very lightweight Linux distribution, such as Lubuntu or Xubuntu. We are still testing both and will eventually decided which one we prefer. Of course there will be no problem in running it on any other Linux distribution if one wishes to, we are just low balling the hardware requirements so you can run the Manager on a small format (cheap) dedicated computer. - remember we are aiming to an affordable solution.
We have discarded OS X and Windows for several reasons, one, because you will want to run on separate computer so you have keyboard focus all the time, second, also to avoid having to maintain 2 different software versions, third reduce costs (licenses). For this reason, Linux was an obvious candidate. The main application lib we'll be using for the Manager (and other apps) is Qt.
The main trigger for this decision, was when we decided we wished the user to be able to interact with the manager both through the console keyboard and the computer keyboard. So, if one wants to use the computer keyboard, one need keyboard focus, so if you have the Manager installed on the same computer as your DAW (which was the original idea) you would then have to switch between applications to have keyboard focus, which really sucked!! Having a separated computer was a must for this.
To keep costs and hardware/OS requirements low, Linux was an obvious choice. It will also have the benefit we could concentrate only in one software version and not two (Windows/OS X).
Cheers,
Artur
-
Hi Artur,
I have absolutely no problem with the software running on Linux. There are very good reasons for that as you pointed out.
In fact, I have been a long term Linux user and always watched the developments regarding i.e. a Linux daw.
Qt is fine - and is also available on many OS.
What I couldn't find out yet is, if this computer (and the software) will be a commercial product in the end or will it be freely available regarding schematics (and layout) of the boards as well as the source code for the software ?
Maybe that's already stated somewhere in this forum but I couldn't find it right away.
Thanks
Matt
-
Hi Matt,
Thanks for you comments!
Since we are going to much off topic, to an important subject I decided to create a new thread where we can continue this discussion: http://forum.sslmixed.com/index.php?topic=843.0
Cheers,
Artur