BED WETTING SIMULATOR updates

[LainIsLain]

BedWetter v0.3.2 and v0.3.34(ProfileCapacityTest) - updated 7/3/21 - https://github.com/DiaperHugger/BedWetter

Bed Wetting Simulator update log. This post is for further discussion about this project including but not limited to ideas and suggestions about the project and it's development. All are welcome to contribute and offer opinions about user experience and features.

I'm currently coding for Arduino platform with the features laid out by @johnzld and @PaSS . I'm limited on what I can do but don't mind learning as I go.

If anyone has some art skills and wants to create a simple logo please chime in!!! https://www.adisc.org/forum/threads/bed-wetting-simulator-art-contest.151880/





This version satisfies the initial function goals. Here is a flowchart of the current operation. Please see the github link for variable description and operation notes. Further refinement will continue. Hardware recommendations and construction details will be refined with user feedback. The next major addition will be a profiles random feature that will give a more tailored random experience. Integration of sensors for feedback will also be included soon. Stay tuned for updates!





What works:

Total Sleep Time / hours​

Initial Delay / minutes​

Wet Intervals​

Interval Durations / seconds​

OutPut Pin select​

Duty Cycle​

Duty Pulse / seconds​

Prime​

LCD backlight control​

Reset Settings​

Random​

Wet Chance​

Profiled wettings​

Secondary relay pin control​

Optional temp and flow sensor​

​

--Update 5/23/21--

​

BedWetter v0.3.1​

​

Integration of secondary relay control. Added options for temp and flow sensors. The secondary relay pin and be controlled by the temp sensor but should only be used a back to a temperature controlled device for heating your water. Secondary relay pin can be activated with the LEFT button on the home screen. Added code to auto detect a compatible temp sensor. Read the README and MENU DESCRIPTION on the github.​

​

--Update 5/18/21--

BedWetter v0.2.6 (PROFILE TEST)​

​

This is the first attempt at the profile system. For a description and discussion read this post.​

BED WETTING SIMULATOR updates

Yes, first Test-runs are very good!!! (testing Normal Mode) ...will update soon... Normal Mode is 100% GOOD and working!!! Awesome thank you. (BedWetter 0.2.54)

www.adisc.org

--Update 5/2/21--

To better understand the features and overall goals I'm going to write a plan of attack for the BedWetter device. There are a few different levels of technical complexity as well as cost associated with the different possible design paths. I'll try to illustrate what I think the possible paths are. I see two possible groups of designs here. The DIY approach including simple and complex designs and the refined prototype/built that may be beyond the typical DIY.

It's cool that others have interest in bringing this thing to life. At this time I don't mind giving some of my free time to develop this project. How far it goes will be determined by the skill set of the people contributing. @Cottontail has even offered his skills around circuit design and possible prototyping. If others have offered to contribute and I missed a post please poke me again. I'm not totally sure how to coordinate this. It feels funny to call myself a project leader but if no one from the original post (@johnzld @PaSS) has any objections I don't mind seeing this through to it's conclusion as long as I have the time to give. I also have a fair bit of skill I can pull from. I can prototype the electronics side and have a small shop with tools. I also have a small MILL/CNC and 3d Printer for fabricating parts. Although the 3d Printer has a bit of dust on it. I can also design and make small prototype PCB's. For accessibility to others though the short term goals should probably involve premade modules or very simple circuits on a breadboard. I will definitely need help with coding. Thanks @PaSS!

Lets start with the overall goal of the device in a simple sentence. Later I'll give my opinion on which path to take and why. These opinions are based on my skill set and personal bias.

-The device should be able to control times and regulate water flow (with and with out heat) with the ultimate goal of causing the user to wake up with a wet diaper for certain or by chance.

The value to those who want a device like this will be the careful planning and understanding of the "function" of said device. The goal of "waking up wet" is a simplification of the end result. How this is achieved for best results will be understood best by those who understand the motivation to "wake up wet". This highlights the appropriate steps to make the end result as pleasurable as possible rather than simply just being "wet". For instance I don't fully understand the purpose of a "chance" that one could wake up wet. This being one of the features asked for by @johnzld. The want or need to wake up wet is shared by many on this forum. In my mind if the motivation to wake up wet drove me to use a device like this I would just as well wake up wet when ever I choose to use it and not let it be chance. But this is important to the design. I don't connect with this feature but obviously others do. This input will be valuable to the finished product as we do the heavy lifting for those who just want a working device. This is not to say that some features could be removed or deprecated in favor of a more conforming solution to the user experience. (god this sounds like corporate focus group gibberish)

Group One: DIY Simple

The simplest device has been mentioned on previous posts. At the most basic approach a simple drip feed from an elevated container of water directed by a tube into the diaper achieves the outcome of a wet diaper. The regulation of the liquid is achieved by way of adjusting the tube diameter or restricting the flow with a passive device. This flow would be determined within the use case of the specific system in question. Various containers such as hot water bottles, enema bags, or fish tank and tubing are a simple and relatively reliable method to achieve this. Warm water is achieved with pre-heated water in the case of the hot water bottles or enema bag. And fish tank heater for the fishy tank.

Group One: DIY Intermediate

This project is currently involved in the intermediate stage IMO. This includes but is not limited to, an electronic device that controls both time and regulation of the water. By relying on an electronic device the control over time and quantity of water introduced is greatly higher than the previous method. Features such as timed delay, frequency of wettings, and random chance are all possible. The water flow can be controlled by solenoid or pump. The current platform we're using is the Arduino UNO and the interface is the Arduino compatible LCD shield with buttons. This is a relatively simple and easy to obtain device with a great deal of support. This device may also include a method for regulating monitoring the heating process. This will come with some design challenges. There are devices on the market that can heat water and regulate it's temperature. These are mostly AC driven devices such as fish tank heaters or for food prep/cooking. These would be an addition to the control device rather than a specific built one so the control of these devices will be undoubtedly separate other than some basic on/off control using relay's.

Group One: DIY Advanced

The progression of the device into a more advanced version would necessarily be crafted by design choices beyond the basic functionality. I believe the better user experience is for the device to be compact and easy to setup. This would mean an integrated pump and heater as well as a simple and available water storage source. Things like a custom built on demand heating system with control and monitoring of water temps, flow meter and temp sensors for safety, and possibly a small touch screen UI. These features may most likely require a beefier processor like a small ARM based chip to interface with sensors and the screen. This level may also involve more advanced circuits such as power delivery/switching or custom PCB's. The device foot print would dictate the level of optimization needed depending on esthetics. I would love to see something like a handheld device that includes everything but the water source. This may be difficult depending on the physical size of the heat exchanger.

Group Two: Advanced prototype/product

The end game for a device like this would be a custom built product. This could include plans for building or ordering custom parts such as enclosures, custom PCB's, custom made heat exchanger and so on. Also including a refined UI and possible integration of Bluetooth or other external control schemes. This level of commitment would require a lot of time from those with the appropriate skills.


Considerations for cost.

The overall cost should be considered. The replication of an electronically controlled device should be easy enough. The Arduino ecosystem is vast and accessible. If the heavy lifting is done such as detailed diagrams, coding, and experimentation I believe this will be the best short term goal for those who want to build something that has controls tailored to the experience. A design of a device that can function within more than one level of complexity might be preferable. Such as a control unit that can be used on the most basic of designs but also have the capability of advanced features such as temp control feedback and flow monitoring will be easily integrated if the user wants to add the appropriate hardware. The ancillary hardware at different cost levels may be worth exploring. Things like the hanging bag of pre-heated water with a simple flow control on the low end. An integrated device that attaches to the top of a small fish tank with heater in the middle. A self contained device with pump, heater, and touch screen control on the high end.

That being said I'll offer my opinion here about how I would move forward.

The device I'm most interested in short term is an Arduino + LCD shield controlled pump. This is very familiar ground as I've done several projects similar to this. The coding needed to for the LCD interface is underway and I'm making progress. I will integrate the usage of a timed solenoid with duty cycle like @johnzld asked. However the solenoid as a flow regulator will probably be a dead end in the long run. The solenoid can be great to allow liquid to flow but that's about it. A pump would allow a much greater control over flow rates and quantities and would not cost much more than the solenoid. Coding for a control scheme of a pump should be introduced soon. The code for relay activation could also be used for the pump assuming it's also relay controlled. However a more precise method of PWM would be ideal. This would require the addition of a motor control circuit/module that's DIY or off the shelf parts. The addition of a controllable heater is also something I'm interested in. It will take some time but if something can be made that's small and safe to use I think it will be worth it for possible use with an all-in-one device later.

The technical write up including diagrams and setup for a working device will need to be placed on the github page. Right now it only functions as the most basic controller with hard coded times. https://github.com/DiaperHugger/BedWetter

I will continue to work on the code for now in my free time and post it to github.

Any design consideration and feature tweaks are welcome! I'll check in and consolidate them as I have time.

 

[PaSS]

Following

 

[LainIsLain]

First attempt at an inline heater was a bit of a failure. I kinda knew this going in but now I know I need much more dwell time for the water to gather heat. I tried using two of the little heater blocks by drilling a 1/4" hole along side the block and inserting some brass tubing. Even with the blocks at 200F the water is much too cold. I can get some warm water by blocking all but a micro jet of water at the end but it's still barely room temp. The heating elements are probably going to work but a different heat exchanger needs to be found/made. I'd like to keep it small and off-the-self. Ideally I'd like it to be the same size as the pump but it will probably need to be a small reservoir/exchanger with massive surface area. For testing purposes I might just make something.

I got a small 12v liquid pump from Amazon. https://www.amazon.com/dp/B01LXCB9GC?psc=1&ref=ppx_yo2_dt_b_product_details
This is both nearly silent at low voltage and has surprising pressure! Like full spray when blocking the end of the holes pressure!

I'll shelve it for now and work on some code.

 

[LainIsLain]

@johnzld
Question about how you imagined the "probability" part?

You want a percent chance that you will wake up wet. What does the percent chance effect? Does it determine a total amount of liquid?

So lets say that 50% is chosen. Does that mean 50% chance of a max amount of liquid? 50% chance that any wet event will be the full duration? What if the random number doesn't land within that 50%? Is there any wet events if the chance doesn't happen?

This is how I imagine it. Lets use two parameters. 4 random wettings, and chance of fully wet with a target of 1 liter. What I want to know is what happens if the 50% doesn't land?

100% chance of fully wet w/random times: Full liter

|--#---#-#------#---|10sec
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|0hr-------------8hr|0sec

50% chance of fully wet but doesn't happen w/random times: 1/2 liter

|-------------------|10sec
|--#----------------|
|--#------#---------|
|--#------#---------|
|--#------#--------#|
|--#------#-----#--#|
|0hr-------------8hr|0sec

25% chance of fully wet but does happen w/random times: Full liter

|---#--#--#--------#|10sec
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|0hr-------------8hr|0sec

The number of wet events to reach a full liter would be determined by the flow rate. 4 - 10 sec wet events would need to drain 250 ml into the diaper. So lets say a standard can of soda in 10 secs. Trial and error of course but that's a lot of liquid to 10 secs.

- In order to avoid flooding/leaks/waking up, let the solenoid actuate open/closed to slow down the flow during that 10 seconds activation

So will this be how you will regulate the flow? Are you using any other kind of flow regulation? The on/off times of a gravity feed system will need to be determined with a working system. However if the flow it too much and the solenoid needs to be actuated many times it might be loud enough to wake you up especially if it's close to you. Both the relay and solenoid will make noise and if the actuations are rapid it might make a real racket. Ultimately the solenoid would only be used to allow flow and a restriction of some kind would serve as a flow regulator. If you truly want to solenoid to act as the regulator I would need to know what kind of parameter you want to adjust. Something like a duty cycle and pulse time over the 10 sec duration.

10 second full amount:

50% duty 1 sec pulse: Solenoid on for 0.5/sec and 2 actuations each/sec
25% duty 2 sec pulse: Solenoid on for 0.5/sec and 2 actuations every 2/sec

 

[johnzld]

@johnzld

Yes, found this post and also following.

Up to the 2nd goal LainisLain? ;-)

2nd goal/expansion of the program can then be:
- In order to avoid flooding/leaks/waking up, let the solenoid actuate open/closed (to slow down the flow) during that 10 seconds activation

 

[LainIsLain]

Sorry was updating while you replied! Had questions! Read above.

 

[ade]

Loving the effort (and the code ).

In days past, I tinkered about with various mechanical types of 'wetting', but was hampered with delaying and unpredictability issues (not to mention the availability of materials). With such simple set-ups, I was at the point of resolving it down to a hot-water bottle laid in bed with me (for to be 'accidentally' squashed as I turned over in my sleep), a suspended hot-water bottle with a very slow constant drip and, the hardest to buy for, a hot-water bottle with a water soluble material acting as a timer; and all suitably plumbed, of course.
A lot of trial and error, with many accidents not of my design nor want

You inspired me to have a quick gander at what's available on the market, now, and 'holy diapers, batman!' there's some good stuff about: soluble bags (for fishing, but could be put to you-know-what use), soluble fabric, etc.
And these:

https://www.amazon.co.uk/Automatic-Irrigation-Watering-System-30-Day/dp/B07KYBL8CP/ref=mp_s_a_1_4?dchild=1&keywords=micro+drip+timer&qid=1619953771&sr=8-4

 

[johnzld]

So will this be how you will regulate the flow? Are you using any other kind of flow regulation? The on/off times of a gravity feed system will need to be determined with a working system. However if the flow it too much and the solenoid needs to be actuated many times it might be loud enough to wake you up especially if it's close to you. Both the relay and solenoid will make noise and if the actuations are rapid it might make a real racket. Ultimately the solenoid would only be used to allow flow and a restriction of some kind would serve as a flow regulator. If you truly want to solenoid to act as the regulator I would need to know what kind of parameter you want to adjust. Something like a duty cycle and pulse time over the 10 sec duration.

10 second full amount:

50% duty 1 sec pulse: Solenoid on for 0.5/sec and 2 actuations each/sec
25% duty 2 sec pulse: Solenoid on for 0.5/sec and 2 actuations every 2/sec

Yes indeed, the idea is to regulate the flow with this part of the program.
Solenoid 10sec full open will let through to much fluid at once i think. And a change it will be a to big/strong stream that would wake you up.
I know both the relay and solenoid will make noise and if the actuations are rapid it might make a real racket.....but i take that change in this fase of the project. Improvements can be added later on. And yes eventionally i think it would be something as "Ultimately the solenoid would only be used to allow flow and a restriction of some kind would serve as a flow regulator"
I think the 25% is very good for a starter and would try that one for now:
-> 25% duty 2 sec pulse: Solenoid on for 0.5/sec and 2 actuations every 2/sec

 

[johnzld]

@johnzld
Question about how you imagined the "probability" part?

You want a percent chance that you will wake up wet. What does the percent chance effect? Does it determine a total amount of liquid?

So lets say that 50% is chosen. Does that mean 50% chance of a max amount of liquid? 50% chance that any wet event will be the full duration? What if the random number doesn't land within that 50%? Is there any wet events if the chance doesn't happen?

This is how I imagine it. Lets use two parameters. 4 random wettings, and chance of fully wet with a target of 1 liter. What I want to know is what happens if the 50% doesn't land?

100% chance of fully wet w/random times: Full liter

|--#---#-#------#---|10sec
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|--#---#-#------#---|
|0hr-------------8hr|0sec

50% chance of fully wet but doesn't happen w/random times: 1/2 liter

|-------------------|10sec
|--#----------------|
|--#------#---------|
|--#------#---------|
|--#------#--------#|
|--#------#-----#--#|
|0hr-------------8hr|0sec

25% chance of fully wet but does happen w/random times: Full liter

|---#--#--#--------#|10sec
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|---#--#--#--------#|
|0hr-------------8hr|0sec

The number of wet events to reach a full liter would be determined by the flow rate. 4 - 10 sec wet events would need to drain 250 ml into the diaper. So lets say a standard can of soda in 10 secs. Trial and error of course but that's a lot of liquid to 10 secs.




So will this be how you will regulate the flow? Are you using any other kind of flow regulation? The on/off times of a gravity feed system will need to be determined with a working system. However if the flow it too much and the solenoid needs to be actuated many times it might be loud enough to wake you up especially if it's close to you. Both the relay and solenoid will make noise and if the actuations are rapid it might make a real racket. Ultimately the solenoid would only be used to allow flow and a restriction of some kind would serve as a flow regulator. If you truly want to solenoid to act as the regulator I would need to know what kind of parameter you want to adjust. Something like a duty cycle and pulse time over the 10 sec duration.

10 second full amount:

50% duty 1 sec pulse: Solenoid on for 0.5/sec and 2 actuations each/sec
25% duty 2 sec pulse: Solenoid on for 0.5/sec and 2 actuations every 2/sec

Hi, great to have such active conversation here and discuss options.
Initially the Probability Percentage was the change the Solinoid was activated once during the night(1 pee) or not (dry night)
But let's take your question, and that is even far better thought!

- So lets say that 50% is chosen. -> 50% chance that any wet event will be triggered during night.

- This is how I imagine it. Lets use two parameters. 4 random wettings, and chance of fully wet with a target of 1 liter.
Yes, 100% ok and nice the amount of 4 random wettings
- What I want to know is what happens if the 50% doesn't land?
That i do not understand...sorry...i really want to understand your question and respond with the correct answers you need.

- 4 random wettings -> That 4 is a fixed value in the program then, correct?
- What I want to know is what happens if the 50% doesn't land? -> can you try to explain in other words so i can understand better, thanks

 

[johnzld]

Decision Tree Diagram

1) Give in Probability Percentage (Dry or Wet night)
-> eg. 80%
-> 80% change that wetting will occur during night -> if true see 2)
-> 20% change that the person will stay dry (not pee at all) during the night -> if true end program
With this 80/20 So most probably the person will wet but also a small change wil be staying dry

2) Go to the program you are creating ....so above would be step1 and above all others in my opinion.
-> within your program then offcoarse all the features you already have and we are talking about how to address them
-> within your program we can then indeed see how many wetting, eg. 4 random onces, and the amount fluid/duration of each pee....

 

[LainIsLain]

Ok! So wetting of ANY amount will be determined by the percentage? I was assuming that if the device was being used that some amount of wetting would happen and the percentage was a chance of a FULL wetting. So unless you choose 100% there will be a chance that no wetting will occur. Is that right?

It's late and I'm heading to bed. I did some work on a basic menu navigation using the LCD. Still no data input method or save and recall but working on it.

You inspired me to have a quick gander at what's available on the market, now, and 'holy diapers, batman!' there's some good stuff about: soluble bags (for fishing, but could be put to you-know-what use), soluble fabric, etc.

Now that is an interesting product! I wonder how precise the timing is. Would be funny if there were different timed bags and you could pick one at random to preserve the random chance of a night time wetting. LOL

 

[LainIsLain]

Thanks @ade I know about the drip systems. These micro drip things are basically what we're building minus the specific timing features and heat of course. Still pretty cool though. This ones battery powered! https://www.amazon.com/Auburet-Auto...words=micro+drip+timer&qid=1619962704&sr=8-12

 

[johnzld]

Ok! So wetting of ANY amount will be determined by the percentage? I was assuming that if the device was being used that some amount of wetting would happen and the percentage was a chance of a FULL wetting. So unless you choose 100% there will be a chance that no wetting will occur. Is that right?

It's late and I'm heading to bed. I did some work on a basic menu navigation using the LCD. Still no data input method or save and recall but working on it.

View attachment 54356


Now that is an interesting product! I wonder how precise the timing is. Would be funny if there were different timed bags and you could pick one at random to preserve the random chance of a night time wetting. LOL

" So unless you choose 100% there will be a chance that no wetting will occur. Is that right?"
Yes, that was my idea/wish...
....but " my will is not a law" I think you are having great ideas and things i did not even think about.
So please do not rule out your wishes/ideas!

Great seeing the LCD and choices.
This really is going to be a success.
...and will end in some wet nights..HAHAHA
....some with just a little dampness, some fully dry, but also changes that some nights will be fully wet.

 

[Abdlchriscrinkle]

This is such a cool thread, I can't wait to see the mastered end product. You know damn well there are people that will be building their own too lol

Only problem for me would be how much I toss and turn in my sleep, I could see tangling issues with it, as my bed usually looks like a tornado went thru in the morning.

I won't need to make one though, as I need no help in making my diaps wet by morning...

 

[johnzld]

This is such a cool thread, I can't wait to see the mastered end product. You know damn well there are people that will be building their own too lol

Indeed, it took me 2 tries to revive this thread but it really breathes now ....and soon it will spray liquid

"Hi, i really liked the approach of "logancody16" with the Arduino project and use of a Solenoid.
unfortunately that user is not an active member anymore, i would liked to build also such project.
I would like to revive this idea and experiment some with it."

Only problem for me would be how much I toss and turn in my sleep, I could see tangling issues with it, as my bed usually looks like a tornado went thru in the morning.

I won't need to make one though, as I need no help in making my diaps wet by morning...

We will see to find a sollution, when time comes and starting some trial-runs offcoarse (also mentioned in the original thread we will address options of tubing and how to fixate it, etc.

 

[Cottontail]

I’ll be interested to keep tabs on the hardware requirements. If at some point it becomes desirable to integrate all of the control electronics onto a single compact PCB, build enclosures, etc. I can help. I won’t claim to be the best with 3D stuff, but I get by. I do have both a 3D printer and a laser cutter on my desk, and I lay out / assemble PCBs all the time.

I’d suggest attempting to construct a list of user scenarios for the device, along with an overall system diagram, both as non-technical as possible, and revising as the discussion advances. I believe blog posts can be edited indefinitely. This would help ensure relevant contributions from newcomers, including technical laypersons who might otherwise get exhausted trying to decipher all the talk of code and whatnot. Easier to discuss the merits of a high-level bullet point than of a chunk of Arduino sketch.

 

[PaSS]

Bed Wetting Simulator update log. This post is for further discussion about this project including but not limited to ideas and suggestions about the project and it's development. All are welcome to contribute and offer opinions about user experience and features.

I'm currently coding for Arduino platform with the features laid out by @johnzld. I'm limited on what I can do but don't mind learning as I go.

Code can be found here: https://github.com/DiaperHugger/BedWetter.

View attachment 54310

You could use a switch statement for your event check. It would optimize it and make it run smoother. Its a option. If I wasnt on my phone I would code it for you and send it here but I am.

Another optional change would be making a void for changing solenoid state.

void solenoid_state(Boolean state) 
{
    if(state)
    {
           BedWetter[0].SolenoidState = 1; //on
           Serial.println("on");
    }
    else if(!state)
    {
          BedWetter[0].SolenoidState = 0; //off
          Serial.println("off");
    }
}

 

[LainIsLain]

You could use a switch statement for your event check. It would optimize it and make it run smoother. Its a option. If I wasnt on my phone I would code it for you and send it here but I am.

Another optional change would be making 2 voids. One for solenoid_on and solenoid_off. Basically move the two lines in the event check and move them to their own sperate voids and call them instead. It would look neater but is optional

Ya I've used switch functions in my other code and I will probably do that. The V0.1 was just a proof of concept. The code will get loads more complicated with the interface. One of my shorts comings is thinking through and understanding logical constructs. Sometimes it's the hardest part for me to figure out how to get the results I want. My fall back is ballooning IF/ELSE statements. Like I said I can usually brute force my through.

@Cottontail Yes I agree. A full write up and description is needed. Ya the blog post can be editing indefinitely.

I'm out and about now but will try to consolidate when I get back.

 

[PaSS]

Ya I've used switch functions in my other code and I will probably do that. The V0.1 was just a proof of concept. The code will get loads more complicated with the interface. One of my shorts comings is thinking through and understanding logical constructs. Sometimes it's the hardest part for me to figure out how to get the results I want. My fall back is ballooning IF/ELSE statements. Like I said I can usually brute force my through.

@Cottontail Yes I agree. A full write up and description is needed. Ya the blog post can be editing indefinitely.

I'm out and about now but will try to consolidate when I get back.

If you have any questions code wise you can private message me. I am fairly decent at it. Also i edited my post you replied to after you read it.

Also if you keep the github updated i would be more than willing to help you code it

 

[Cottontail]

Sometimes it's the hardest part for me to figure out how to get the results I want. My fall back is ballooning IF/ELSE statements. Like I said I can usually brute force my through.

@Cottontail Yes I agree. A full write up and description is needed. Ya the blog post can be editing indefinitely.

I suspect most of the logic will fall right out of the write-up. My current boss is totally nontechnical, which at first seemed like a curse. A few years in, though, and it was clearly a blessing in disguise. By the time I’ve conjured a plain-English spec that he can understand, the implementation has become a no-brainer.