Duet Network Sucks

Well I value my time and asked specific questions about support and time. I even mentioned I was staying with Smoothie because I knew it. But I listened to the "experts" here and wasted a lot of time. I am now tuning my machine and we will see if the hardware is any good. The Documentation isn't nearly as good as smoothies.

please go back to smoothie and leave the duet to users who actually know how to use then...

I'm a little reluctant to dive into this snit-fest, but despite cwaa's abrasiveness, he does have a good point - a lot of products (not just open source ones) have really poor documentation. There's the assumption that certain things (network configuration, how & where to make connections, what have you) any halfway competent person should be able to figure out, but that's only if you happen to already be familiar with those things. People are coming to 3D printing from all sorts of backgrounds (artists, old school engineers like me, etc.) who may have no clue about a lot of stuff. There's also the real problem of simply referring users to some generic third party source of information, leaving a ton of gaps in how to use the actual product sitting in front of you. I've whined about documentation before, and the common response is, well, technology is changing so fast with one thing leveraged on the next, that it is impractical to keep things well documented for the uninitiated. I, for one, would favor a slower rate of change if it means more stable, well documented and easy to use products.

The DUET is over-engineered and overpowered for most home applications. There's no reason development couldn't be slowed in order to create more thorough documentation in my opinion.

Thanks for the comments on the documentation. I agree we can improve it and I am doing so (albeit slowly!). It is a slow process and possibly not as simple as some other firmwares due to the increased flexibility of RepRapFirmware. Our forum (forum.duet3d.com) also provides a place for people to seek clarification and advice specific to their setup, we try to respond as promptly as possible to questions there.

Edited 1 time(s). Last edit at 07/01/2018 09:16AM by T3P3.

Quote
LoboCNC
I'm a little reluctant to dive into this snit-fest, but despite cwaa's abrasiveness, he does have a good point - a lot of products (not just open source ones) have really poor documentation. There's the assumption that certain things (network configuration, how & where to make connections, what have you) any halfway competent person should be able to figure out, but that's only if you happen to already be familiar with those things. People are coming to 3D printing from all sorts of backgrounds (artists, old school engineers like me, etc.) who may have no clue about a lot of stuff. There's also the real problem of simply referring users to some generic third party source of information, leaving a ton of gaps in how to use the actual product sitting in front of you. I've whined about documentation before, and the common response is, well, technology is changing so fast with one thing leveraged on the next, that it is impractical to keep things well documented for the uninitiated. I, for one, would favor a slower rate of change if it means more stable, well documented and easy to use products.

I think thats a pretty good analysis of the situation

Edited 1 time(s). Last edit at 07/02/2018 08:24PM by Bill Clark.

Quote
LoboCNC
I, for one, would favor a slower rate of change if it means more stable, well documented and easy to use products.

Speaking as someone who didn't have any problems with the networking on the Duet...

For a slower rate of change, there seems to be no shortage of available RAMPS 1.4 boards, and products like the MKS GEN boards which are compatible with them, and pretend to be an Arduino with one attached. Certainly there are other solutions which are less of a leap.

I intend to use my Duet pretty much as I got it. Once configured, I will probably not be doing large firmware updates, nor will I be making hardware changes to the printer that it runs for purposes other than repairs. I expect that in time, this version of the hardware will no longer be supported for new firmware, or that other advances will leave this behind. For some people, examples are never going to be enough, unless directly from the example will work for their specific situation. How do you tell someone who doesn't understand networking to find an available address on their subnet? Picking an arbitrary one that (probably) works in most cases (Which 192.168.1.14/24 is) is probably not a bad plan, but you'd be just as screwed if your router is at 192.168.0.1. Of course, if the user understands networking enough to spot that configuration quickly while looking through the sample configuration files, then there's no problem at all. (I changed mine on that first pass through on the SD card to an address that wouldn't be a problem, but it got a DHCP address assigned anyway.)

I thought that the network configuration was pretty well documented in the configuration file, but then personally, all I needed were some general hints.

What was much harder for me were some of the definitions for things like end stops. I didn't know for sure what the stops that I had were wired as, and had to do a lot more looking to know that I needed to re-wire the connectors in order to get the expected operation. That took some digging for me. I'm sure that as I finish development of the printer that I'll have more learning curve to go through, and it's really not going to be long-term applicable for me, since I will probably not be re-configuring RRF for anything else any time soon. Just like I forgot almost everything that I'd learned about configuring Marlin once my I3 was set up the first time, and once I had to do it again, so much had changed that I might as well have been starting from scratch. I'm pretty sure that I'll get it figured out well enough to get my printer up and running, but unless I start mass producing printers based on Duet controllers, I'm not ever going to need to know much about how to fine-tune the firmware. (Even then, unless I'm making those printers in a variety of configurations, I'll get the settings figured out for what I'm doing, specifically.) This is going to be a common issue with something like 3D printers, which come in a near infinite variety, and even ones built the same can have slightly different settings tuned to slightly different uses.

Quote
SupraGuy

Quote
LoboCNC
I, for one, would favor a slower rate of change if it means more stable, well documented and easy to use products.

Speaking as someone who didn't have any problems with the networking on the Duet...

For a slower rate of change, there seems to be no shortage of available RAMPS 1.4 boards, and products like the MKS GEN boards which are compatible with them, and pretend to be an Arduino with one attached. Certainly there are other solutions which are less of a leap.

I intend to use my Duet pretty much as I got it. Once configured, I will probably not be doing large firmware updates, nor will I be making hardware changes to the printer that it runs for purposes other than repairs. I expect that in time, this version of the hardware will no longer be supported for new firmware, or that other advances will leave this behind. For some people, examples are never going to be enough, unless directly from the example will work for their specific situation. How do you tell someone who doesn't understand networking to find an available address on their subnet? Picking an arbitrary one that (probably) works in most cases (Which 192.168.1.14/24 is) is probably not a bad plan, but you'd be just as screwed if your router is at 192.168.0.1. Of course, if the user understands networking enough to spot that configuration quickly while looking through the sample configuration files, then there's no problem at all. (I changed mine on that first pass through on the SD card to an address that wouldn't be a problem, but it got a DHCP address assigned anyway.)

I thought that the network configuration was pretty well documented in the configuration file, but then personally, all I needed were some general hints.

What was much harder for me were some of the definitions for things like end stops. I didn't know for sure what the stops that I had were wired as, and had to do a lot more looking to know that I needed to re-wire the connectors in order to get the expected operation. That took some digging for me. I'm sure that as I finish development of the printer that I'll have more learning curve to go through, and it's really not going to be long-term applicable for me, since I will probably not be re-configuring RRF for anything else any time soon. Just like I forgot almost everything that I'd learned about configuring Marlin once my I3 was set up the first time, and once I had to do it again, so much had changed that I might as well have been starting from scratch. I'm pretty sure that I'll get it figured out well enough to get my printer up and running, but unless I start mass producing printers based on Duet controllers, I'm not ever going to need to know much about how to fine-tune the firmware. (Even then, unless I'm making those printers in a variety of configurations, I'll get the settings figured out for what I'm doing, specifically.) This is going to be a common issue with something like 3D printers, which come in a near infinite variety, and even ones built the same can have slightly different settings tuned to slightly different uses.

You provide a great example - you breezed through the network setup because all you needed were general hints, but were flummoxed by the end stops. I'd be the exact opposite - flailing with the network setup but pretty comfortable configuring end stops. All of these products, even the venerable Ramps 1.4, are quite complicated and realistically could use much more comprehensive documentation, including a variety of examples for each of the different aspects of the setup.