an interesting discussion over here
Decent and next level of excellence

with people giving their wish-list for things they'd like changed on the Decent.

I thought I'd let the discussion run its course, and then give some of my thoughts. As a vendor, I am not allowed to post in the community forum, so I'm writing my thoughts here, where I am.

GDK wrote: 1. Lose the ceramic drip tray and make it from nice sturdy plastic. This would provide benefits:
a. easier operation/workflow:
- tray will not break easily while handling it
- it will be lighter
- it will slide in and out more smoothly (no need for tape)
b. less shape variation which should eliminate that imperfect fit and wiggle with the tray grid on top
c. there may be cost benefit (?)
2. Have a plastic water reservoir. Ceramic may be less susceptible to grime etc. but again the benefits of plastic outweigh that, adding a cover may become possible too.

We will absolutely never use plastic for a water tank or drip tray. I don't want to ever think about water leaching.

That's why we went with ceramic baking dishware, made by Loveramics.

Also, plastic ages terribly, and that's doubly so for clear plastic. It tends to get brittle or cloudy with just a few year's age, and that's not great for a water tank.

Ceramic, if you don't drop and break it, lasts forever and ages beautifully over the years.

Also, ceramic is resolutely high temperature dishwasher safe. I'm not sure if other espresso machines's plastic water tanks are.

However, we would like to:
1) lower the weight of our water vessels
2) make them stronger (ie, withstand dropping on the floor)
3) make the shape more consistent (shrinkage variation is a thing with ceramic)

To that end, we hope to be able to move to ceramic-coated, stamped aluminum.





These have the same food safety merits of solid ceramic, but:
1) they're lighter (often sold for camping)
2) much stronger
3) consistent shape, due to the inside stamped aluminum shape

The reason we haven't already moved to this, is that we haven't found a manufacturer who will tell us what's in their coating. They consider this to be a trade secret, and, to be blunt, we don't trust that.

So, if we manage to find a company who will be transparent with us about what they're doing, we'll move to this in the future. But, I'm not sure we'll succeed, in which case we'll stay with our current approach.

ShotClock wrote: 6 - Easier way of adjusting headspace. I use a spacer/dispersion block from Sheldon which is great. I believe that there is a new block incoming from decent which will address this soon. Would be great if this could be adjusted shot-to-shot, but I'm not sure if this is feasible.

In June, we'll be releasing new group head inner parts, that reduce the headspace again some more, about to whatever Sheldon's lowest setting is. We're now following the E61 headspace standard.

ShotClock wrote: 7 - A method of plumbing the machine that completely bypasses the reservoir. Currently, it looks like both refill and plumbing kits just refill the reservoir automatically. To avoid biofilm etc., it would be great to bypass the reservoir completely.

That's not going to happen, because want the water reservoir to be cleanable.

I realize that people like directly hooking up a boiler-based machine to water, but have you thought about the downside? You still have a water reservoir, but now it's the boiler, and you can't get in there to clean it.

For me, it's a major plus that the water reservoir in the Decent can be taken out and fully cleaned, unlike a boiler.

ShotClock wrote: 8 - Separating the update of profiles from the rest of the software would be a great addition. Currently, if you want to test the "80s espresso" low temperature profile (or any other experimental profile that decent is developing), you either need to move to a less stable build of the decent app, or download the profile via a third party app such as visualizer. Reducing the friction of trying new profiles would be a nice step in my opinion.

Two things on this:
1) firstly, new profiles deserve a beta cycle, just like the app, so I don't want to release an experimental profile into general release.
2) secondly, there will be a better way to share profiles coming in an app upgrade.

Quester wrote:I've also never been able to get my two DE1s to behave similarly enough that the same grinder, bean, and recipe produces the same result. Lot's of work with calibration has helped, but they still behave differently. This would be unfortunate in a cafe setting. But John and others have done this a lot and not had any issues.

Using the same model and calibrating flow and pressure on them, to your location, will get you very, very close to having identical machines.

As the discussion mentioned, your two machines are different models, separated by a few years, and yeah, they do have differences.

Randy G. wrote: My only major thought in terms of such improvement would be to have the plumbed drip tray made so that it 'plugged into' the machine or had a simple quick-disconnect of some sort so that the hose did not interfere with the removal of drip tray so that the reservoir could be more easily accessed for removal and cleaning.

In our planned for 2024 "Bengle" model we'll likely move to a stainless steel drip tray, as it won't be visible like the ceramic is. We'll then be able to have a more robust tubing solution. The ceramic drip tray can't.

GDK wrote:There is no such thing as too much steaming power on the DE1XXL. AFAIK, it will have a feature (FW upgrade in progress) where you can preprogram the power to match the typical quantities that you use - you have whatever you want, which is great and I do not believe any other machine has it!

We're working on a "wattage divider" feature in the firmware, which can downgrade the higher power steam heaters so they are more gentle.

You can currently adjust temperature and flow rate of your steam (down to 0.4ml/s) which makes it fairly gentle, but on the DE1XXL, steaming a small 50ml of milk well is still difficult.

pizzaman383 wrote:How about positive displacement gear pumps instead of solenoid pumps so that they directly control flow instead of modeling flow.

We've been 2 years in R&D designing our own pump, which you can think of a "small auger". Internally it's called "RayPump" (guess who invented it).

At the end of 2023, we'll start making coffee with it. I don't think it'll make its way into a public machine until 2024, though, because it's certain to have many issues, as it's a new invention (well, it exists in larger form for the oil industry, but nobody has downscaled the design before).

luca wrote:Software - There is still ample room to improve the software.

-UI could do with a bit of a think through and a tidy up. It has a lot of functionality, but it's not very polished. I'd like bigger "power off" and "settings" buttons, for example, since I tap them a lot. In the insight skin, I'd love a one tap button to toggle between my two most frequently used profiles, since I usually start the day with an espresso and then the next coffee is a filter coffee. There are lots of things that probably would come up if I really thought about it.

-Visualiser is really a thing that Decent should build and maintain. It strikes me as something that has some features that are maybe OK, but not really a lot that renders it all that useful. Decent really needs a web enabled way to access your data and go through it, and an integrated way to take notes. There is just so, so much clunk factor in taking notes. Everything is so hard to do that I use a whiteboard to keep tabs on my grind setting, not the software.

-The software should really now be developing "shot improvement" features. By this I mean something like a stats package that shows you your rolling average of what the shot output is like. So it could compare AUC for your shot volumes against your rolling average and give you your standard deviation, to give you some actual stats about how good you are at making consistent shots. Decent had the market lead on the software platform for years, and squandered it a bit, so that there are now third party platforms and gadgets that people are retrofitting onto commercial machines to get stats and improve shot-to-shot consistency in their cafes. Decent keeps on saying that they want to be in cafes, and this was the clear killer feature that they could have built and had as a point of difference. It sort of builds on top of having an OK interface for inputting data, which Decent doesn't really have. It's kind of insane that Decent had the absolute best platform around to develop these features, but now if you're running a cafe, you can get these sorts of stats easily by retrofitting some stuff onto a Linea or whatever, but you can't actually even get them on Decent. So, crazily, Decent is now trailing its competitors on the software stuff in some respects.

Greater visualizer integration is in progress, and I'm working with Miha on that. You'll have each shot's visualizer chart viewable inside the de1app, and a click-on-it will move you to Chrome and the right webpage for that shot.

Stats are an interesting problem, because what is good for a cafe is not what a home user needs. Specifically, cafes want total consistency, and don't change much, so (say) a "shot times since the past 30 days" chart is quite useful in a cafe. At home, it's likely useless as we home baristas change beans, grinders, profiles, way too often.

Anyway, no matter, as data visualization is a huge passion of mine (hello Tufte!) and I do plan to spend quite a bit of time on this in the future.

A "one page" version of Insight is coming, which will also borrow (cough) some of the best ideas from MimojaCafe (ie, such as quick profile changing)

Nobody's mentioned it here, but a mobile-phone friendly version of the de1app is also coming.

InfamousTuba wrote:Wouldn't using a gear pump require a redesign of the water mixing? Or would they put 2 gear pumps in the machine and raise the cost of the pumps dramatically. I think eventually they will change the pump system but the redesign would take a long time

Pflunz wrote:Principally two gear pumps would work, but you have to change how you control the pumps (PWM instead of skipping pulses). Water mixing could be the same.
As far as I remember, John said there will be a new version of the decent with different hardware, with another water mixing solution. There will only be one gear pump with two outlets, and an additional motor in the pump will control the water ratio of those outlets. So that you only need one of those pumps.

Yes, the RayPump:
1) is largely silent
2) knows exactly how much water has been moved
3) uses a different water mixing approach

yoshi005 wrote:The MARO is trying to surpass the Decent by its UX:
https://www.kickstarter.com/projects/ma ... so?lang=de
video

I'm not sure how purple and grey bubbles along a line is a better user experience than a chart showing real data, with X and Y labelled axis, but you are free to disagree.




But I am impressed by the UI on the Meticulous machine, and there are lessons for Decent there.

On the Diaspora user forum, there are a number of UI/UX professionals, drawing up ideas, but so far, nothing is a clear win. Everything has a compromise. For example, while the Meticulous UI supports direct-manipulation (very nice!) this comes at the cost of precision. They're quite good designers, so they might have a solution, but as a rule, all design is compromise.

mikelipino wrote: Like Luca above, I quite like the ceramic water tank. Both the water tank and the drip tray aren't as fragile as I initially thought. They feel just like any ceramic baking dish and are likely as durable. They certainly won't bounce if they hit a hardwood floor, but I also don't need to treat them delicately.
On the topic of a cover, I created a DIY cover from a silicone stretch food cover. The 185 mm x 185 mm cover fits nearly exactly, and is food safe, heat safe, and reusable. Just cut some holes for the water inlet and return and stretch over the water tank. Be sure that the water tank and stretch cover are bone dry (wet silicone slips, dry silicone clings) and it stays put.
I would use this cover if I were plumbed in to keep particulates out. Since I'm not currently plumbed in, it's more of a hassle to refill the tank so I've set it aside for if I get the catering kit. And my environment isn't particularly dusty nor spore-y, so a weekly wash of the tank hasn't been a huge deal.

In 2023 we'll be releasing a silicone water tank lid. It's fairly rigid, unlike your stretchy design.

What we're working on currently is making it easy to refill the water tank manually, without needing to take the lid off. Here's a video of a recent test of the latest revision. You can see that we have a "canal" that you can pour into, that then goes into the tank.
This part will be very low priced.