Hi Steve

I made a new board called "Wish List" at the root of the forum. A lot of people use the Wish List on the blog, but the blog gets spammed so much I keep it in approval mode for new users. This will be better as the forum is all automatic.

Thanks Steve

Hi Steve

Wow - 385! You're probably about on par with me. I have 467, but I got a head start. Mine date back to about 6 months before release to December of 09, once I finally got the database all wired up. Thanks for sharing - love to hear stuff like that!

Copy Roast - thanks for suggesting this! I've been wanting to add that for 2 updates now, but I try to keep myself focused on things people are asking for the most. Now it has become official! Mainly though, I've held off on because I didn't want to sacrifice screen space on the roast list for iPhone users by adding a bottom toolbar. I think I might end up doing is putting a "Copy Roast" like you suggest in the roast details screen. I think that's less cumbersome than forcing a choice every time you start a roast. Either way, I will do that soon. That bugs me as well when I roast with my Quest. I usually end up doing a few batches of the same thing back to back. That will come in very handy.

Now - templates. The next update is ready to go - I'm just waiting on a small issue to be fixed with drivers for the new iPhone 5 for the new data-logging option. That update will introduce Curve Templates in profiles and programs for folks who use curves. If curve templates are defined, with one tap all of the curves are set up and ready to go. I find that to be the most cumbersome of data entry. Do you use curves? And if so, would that combined with a "Copy Roast" feature be along the lines of what you're looking for, or are you thinking more of a strict template, in addition to?

Now that work on data-logging is mostly over for the time being, I want to switch my head to productivity-enhancing features like you're suggesting - things that speed workflow. The new roast analyzer coming in this update is one of those features - basically the current full screen graph, but greatly enhanced to allow quick comparisons between roasts. It will be a great thing for folks who use curves.

Please advise about the Curve Templates if you can. It will help me greatly in determining how folks would like to see this implemented.

Thanks
Danny

Hi Pierre

You record variables like that to curves. In the roast, tap the blue curve button to create a new curve. Set it to "Control" curve. The only downside would be that curves function by a numerical range. You can accomplish what you're after by setting the max value to 100, and then entering 0, 33, 66, 100 throughout the roast to represent off, low, medium and high. Or, come to think of it, maybe setting the max value to 3, and just using 0,1,2,3 to represent those values would be easier. Whatever you choose, it will graph the same, based on the max value you define in the curve.

From there on out, just tap the digital readout on the roast console, with that curve selected, to record a time-stamped entry in the roast.

The update I'm wrapping up at the moment introduces "Curve Templates" to profiles and programs. They're a way to define blank curves that will automatically populate a roast so that you don't have to keep defining them manually over and over again in each roast. I got tired of doing that really quickly in my own roasts.

Thanks for the kind words - I'm glad to hear you're liking it! Any more questions just let me know.

Danny

Hi Chipbutty

I'm so sorry not to have replied sooner - I've had my head deep in the code of the next update for the past two weeks. I was hoping to get it to Apple last week (crunch time), but found a bug that needed fixing with probes. This update will contain support for wireless probe readings via Phidgets hardware, and has been a long time in the making. Unfortunately, fixing the bug required a huge overhaul of the way I had structured them. There have been a lot of late nights lately. A week and a half later, though, it's fixed, which is a good thing, just a little setback. Now to start testing again

Roasting curves are designed to hold one of two types of data - control data (like amps, fan, etc.) or reading data that you measure with a thermometer or probe. You define them in a roast, and record values to them throughout the roast. If you happen upon a combination you like, or that cups well, you can copy and paste those into a profile or program. In future roasts, when you tag that profile or program in a roast, those curves will be viewable inside that roast. You can then define new reading curves inside the new roast and enter data to match against any "target" reading curves contained in the profile or program. You'll then have two reading curves and a good way to gauge the progress of the current roast against the target profile or program you're using, plus hints from Roastmaster for any "control" curves they contain, as to when to change the settings on the machine.

Control curves are typically weight-dependent. Professional roasters will probably opt to roast simply against reading curves, unless they don't vary the weight they typically roast. Reading curves are NOT weight-dependent - i.e. what temp readings occur with 400g shouldn't differ from what you want to occur with 50kg.

The update has a number of enhancements for curve-related workflow, not the least of which is automatic probe readings. Once that goes live, I will focus on some videos. I've put the update in priority over the videos because any videos produced now won't make a lot of sense once the update goes live. A number of things have been improved upon.

This is just a brief intro. I'd be happy to answer any specific questions you have. Please let me know exactly what you're trying to achieve, and I'll shed some light. Again I'm sorry for the delay, and I promise I'll be able to get back to you quickly.

Cheers
Danny

I've been working on adding data-logging support since right after the last update went live. I just got the preliminary approval from Apple today that this scenario is allowed, so I'm pretty excited. It gets hairy in iOS with 3rd party devices. As it stands now, it's limited to ethernet hardware, so I can bypass a lot of Apple's legal red tape. I'd say I'm about 70% done with the code for this update. I'll write a full write up on the blog soon.

The way it will work is that you will be able to install software probes in Roastmaster, and bind them to curves. I've beefed up profiles and programs a bit with curve templates that will create empty, pre-configured curves into roasts that use that profile or program. This will greatly speed up workflows for everyone that uses reading curves, but makes working with probes extremely easy.

So, for instance on my Quest, I choose my roaster, then my profile, then all 4 curves are there and ready to start: 2 control curves for heat and fan, and 2 empty reading curves, already reporting my roaster bean temp and drum temp in the readout below the gauge. When I start my roast, these two curves start sampling data every so many seconds (configurable) and track the roast curves themselves without any data entry.

So... in a nutshell, I now roast from my deck (or office or couch) only running back to the roaster as the temp nears 195 for first crack or 220 for second crack! It's a blast.

I still have some work to do, and initial support will be limited to Phidgets open-source hardware only. This has its pros and cons. The upside is the hardware performs beautifully, and is lightening fast - giving real-time readouts. It's also very economical for users wanting multiple probes - especially roasting shops. (It would be possible to build a probe array of 24 probes for around $1,000 USD, plus the prices of the probe leads).

I'll write a full write-up on the blog soon to explain everything to folks once I get a little closer. I've had to cool my heels on development for a little while until I got the preliminary approval that this was allowed by their rules. Now that I've finally got that, I'm very excited get these live!

Danny

Sorry you've had problems.

That error occurs because Roastmaster has lost the "pointer" to the database file. It should not ever happen, so it's a bit baffling. One user reported it happening to him once, but I've not heard of it since. I saw it once on my test unit a couple of years ago, but that instance ended up being a Core Data bug that was cleared with a restart.

If I know why it's occurring, I can tell you how to avoid it. Once the "pointer" is lost, that error will happen every time Roastmaster tries to save data, but nothing will be saved.

Can you tell me what you were doing right before the error happened? The only place where I can imagine something could go wrong would be in the creation, deletion, renaming or switching of a database. What had you been editing before the error popped up?

If it's easier to talk through it, give me a ring on the number in yesterday's email. I'll be up til 11 EST.

Danny

Hi Michel - thanks for the kinds word.

I wouldn't be against adding a preference setting to make the timers/readouts report reverse time since start IF you're using Behmor curves, a few folks have asked about this. I, personally, like to think of my Behmor roasts in terms of forward time, but I'm sure not everyone does.

The crack times would still need to be entered as real time, if you enter them manually. The easiest way, of source, is by using the crack buttons - they make that point moot anyway. Using those would be still be automatic. It could still report Behmor time, for cracks, though.

One reason I've steered clear of that is that the Behmor's timer (at least on my machine) seems to vary with the line voltage. I've noticed it can be off by as much as 15-20 seconds by the end of the roast. I always thought it would be more disconcerting to work with the timing discrepancies if you were expecting Roastmaster and the Behmor to read the exact same time.

I'll add it to my backlog list. I keep tallies of the number of folks who ask for what, and try to go in order.

Thanks for the feedback!!