John Bowen Synth Design

chapolin wrote:I really believe in open source things, but I think that there is not enough users to put the Solaris Os in the open source world.

It might not be an instant silver bullet (it would still take time, and one owner with the right skills, or at least the will to acquire them or hire someone who does), but it would at least put the power to do something squarely into the hands of users, and completely future-proof the Solaris forever.

I believe the small number of owners can be an issue for the pay-for-fixes model rather.
How many months of a software engineer's salary can several dozen (maybe a few hundred) potential software update purchasers realistically support ?
It could work as one-off push for bugfixes that add to the instrument's value though (because of that I'd probably be ready to pay 100-300ish), but it doesn't sound very sustainable in the long run, because people aren't going to pay hundreds every few months are they ? Well, maybe they would, I don't know ...

Also, do you make this donation-based, with the update eventually available to everyone ?
Or are new owners automatically entitled to the fixed OS but others stuck with whatever came with what they bought ? It's morally tricky to support, especially if we're talking bug fixes not features.

John Bowen wrote:Holger is the ‘new programmer brought on board’, and he has produced a number of betas, up to v1.2.9 now.

What's been fixed in the versions between 1.2 and 1.2.9?

I am pretty sure the open source road is not possible as John stated earlier. He does not own the code and therefore cannot decide to make it open source (and assume he knows the owners don't want to), even if he wanted to (I guess he would) so let's focus on other business models.
I agree it should be fair, but on the other hand if you make it 'too fair' you won't get anywhere.
Companies charge for updates, do discount and renew models without consulting their new/old users all the time. It feels unfair to buy machine 1 one week before machine 2 comes out. But that shouldn't stop NI from doing so. New users pay more for a Solaris anyhow (prices have gone up steadily) so I would be fine paying for an update for which future users don't have to pay extra.
I do think all current users should be given a choice. E.g. Solaris 2.0 is €250 and Everyone decide if they want to purchase it or not. Early investors (people willing to pay up front) get a discount. Just like it used to be with the Solaris itself. If this turns out to be successful, repeat it. Not monthly but eg yearly would be fine. And more users means lower prices. And people love updates so I think few users could resist purchasing the update (I know I couldn't)

I agree that it would be fine to charge for updates, especially updates that add features but also smaller charges for bug fixes. I just don’t know what price to put on it. Probably $200 or $250 would be fine but it really depends on what is added. Multitimbral capability is a big one and might cost more. But I’d guess something like oscillator slop is smaller. The only thing I’m not sure should cost is fixes for the detented encoders.

sam wrote:

John Bowen wrote:Holger is the ‘new programmer brought on board’, and he has produced a number of betas, up to v1.2.9 now.

What's been fixed in the versions between 1.2 and 1.2.9?

There was a MIDI loop going on with Program Select, so if the Solaris was connected to a DAW both in and out over MIDI, sending a program change would come back to the Solaris and request a program change, which would cause it to send another program change, which would....etc., etc. The Solaris would slow down completely, and just ‘choke’ in this situation, This was the only bug fix he did to first test the coding system (to see if he could make a new OS. The software tool had been in Julian’s control prior to that). He then started to look at the other MIDI stuff, when we came to the new production that uses detented encoders. There appeared a strange result when a sustain pedal was connected, and Sustain assigned to the pedal. It was a very difficult search for this, ending with him basically rewriting the entire message handling portion of the code, which is the v1.2.9 now. The iterations in between were just trying different ways to address the Sustain bug, without rewriting.

OpenSource is not a silver bullet - it only works when there is enough momentum and general interest. Having just a few hundred users, out of which only a handful are programmers, would make the whole thing moot.

I like the idea of putting some money into this thing but how much? Well, that is in the stars…

I suppose $100 for the bug fix and 300$ and upwards for new features, depending on the features. Seriously, I will pay whatever the price will be. I don't care. I *really* like the Solaris and I think it is well worth some extra dough. It is a keeper and as such I am rather willing to spend on it.

An enthusiast's synth - you hardly come by those these days. Never let go!

koenig wrote:OpenSource is not a silver bullet - it only works when there is enough momentum and general interest. Having just a few hundred users, out of which only a handful are programmers, would make the whole thing moot.

I forgot the most serious argument against OpenSource here: this is a DSP solution, not your general Java hack. It is a totally different beast and to come by good DSP programmers is not an easy task.

I would thus venture to say that there is perhaps one Solaris owner that is capable of even understanding the code, less develop it further. And to add insult to injury; a development environment for a DSP is probably not something you would just download from the internet, is it? I have no idea. Sometimes are highly specialised development environment equally expensive.

Just charge for the thing. If you want it, you buy it. If you don't want it you either live with the bugs or sell it. Easy choice, at least for me.

Go for it John!

Well I say that the focus should be on bug fixes. The Solaris - as it is right now - is a fantastically deep and complex machine. Surely no-one can seriously suggest that they have got to the point where it holds no more surprises for them. The creative potential is enormous. If the bugs were fixed it would be an utter monster. It wouldn't do everything but does it really NEED to? More features would mean more bugs to fix.

A constant demand for more features will weigh this project down, in my opinion. With a good - and good willed - programmer on the case it may be possible to get the bugs out of the system and then start to think about more features, if you absolutely must! Overwhelm him with requests to "fix this, add that, more more more" and he will be running away to live on a desert island.

This instrument should be learned like the cello or the piano - studied at length, for years. Fix any obvious "breakages" and troubling design flaws then study the thing. Don't constantly strive to change it, please.

No offence intended. £200 sterling for bug fixes only. I don't want new features. It's got plenty.

I would pay a couple hundred for an update. A unison spread/panning feature would be worth quite a bit alone to me.

Reliability bugs should be fixed for free. When I prepaid my early Solaris, I did not expect any bugs. I've been waiting for a full working unit for years, not only a hardware. Sustain pedal Midi issue, bad sample index, freeze on startup, stolen notes, disappearing sound, freezes... all these should be fixed before we talk about something else. They make Solaris unusable live. I want a reliable Solaris with initial specs.
Then talk about paying updates for specs that were not promised in the beginning and remaining bugs correction. I would pay anytime for 4-part multimode, slop / VCO drift, unisson stereo spread and better FX (8 FX busses with serial / parallel routings + reverbs).
For sure John WANTS to do all this since the first fay, I want to believe he CAN.

Nice to see some discussion in this topic for a while. Just want to say that if that is what it takes, I am ready to pay 100 usd for a new OS version that has the bugs fixed. I do not necessarily need new features, but I can consider paying more for a new OS version that would have new features, such as real linear FM. However, I consider a new OS with bugs ironed out the first step that should be taken. The Solaris really deserves a stable OS. It has plenty of featuers built in. Count me in.

John, can you contact all Solaris users in a more direct way to possibly speed up the progress of OS development? I mean, not everybody read the forum often and might be interested in supporting such progress. I believe the OS situation could be fixes faster if all users would recognize that a new version would be possible sooner, if it was a paid upgrade...

F5D wrote:John, can you contact all Solaris users in a more direct way to possibly speed up the progress of OS development? I mean, not everybody read the forum often and might be interested in supporting such progress. I believe the OS situation could be fixes faster if all users would recognize that a new version would be possible sooner, if it was a paid upgrade...

Yes, I realize this...I can try to gather up as many names as possible for a group update, but first I want to get a response from Holger about this idea.

I am willing to pay $100 - $200 for an updated OS that fixes the current bugs. I think that we would all need to agree on the list of bug fixes.

My bug fix requests would be around MIDI and other basic fixes - sustain pedal issues, ribbon sending and receiving midi, joystick receiving midi, receiving channel AT, booting issues, etc.

There are also changes to the NRPN messaging that I would like to see changed - buttons sending NRPN messages, knobs resending the NRPN number vs. running status, etc. This would help with automation in a DAW.

I would also agree to pay again for enhancements - random parameters for voice pitch, filter, panning, etc. to emulate older synths.

I love this synth - the sound continues to inspire me. The programming is deep, rewarding, and I really want to see it reach it's full potential.

John

John,

I have told Holger we need to send messages with the switches/buttons. I also found that there is some unfinished parameter lists (such as Osc Type) which are not handled correctly.

RE: ribbon sending/receiving MIDI. This is a bit more complicated, as the ribbon doesn’t have anything assigned to it for MIDI, and it is controlled by a separate PCB. There is an Atmel microprocessor on that PCB that I believe would need to be reprogrammed for the ribbon to handle MIDI, but I need to confirm that.

RE: receiving channel AT - I have not seen that as a problem. Could you provide me with more details?
RE: Booting issues - there is a v1.2.9 Beta that resolves the ‘freeze on Waldorf screen’ bug. I believe it should handle most any freeze on boot issue. Send me a private email with your serial number so I can confirm.

John Bowen wrote: RE: receiving channel AT - I have not seen that as a problem. Could you provide me with more details?

John - I just retested and I am not able to get the Solaris to recognize Channel AT from another keyboard or when I record it in my DAW. Poly AT from my Midiboard works fine.

John