Agree. I like ESP ota support, they make it easy.
For the $7k board though, I don’t have full context, or know enough to know if this was a dev/design fail.
Or, if this was by design because of sales/biz incentive…
As a Customer, assuming I was following supported upgrade scenarios correctly, I’d be pissed about having to pay to replace component that failed, in an unrecoverable way, just because wireless update was interrupted.
No one accused you of doing anything wrong. You voiced an opinion. So did we. That’s what makes this a forum.
You, however, are the only one that took offense to someone else’s opinion and resorted to childish responses.
No one said you compared a hobby machine to an industrial one, but you used “how industrial machines are made” and an instance where an industrial board got bricked to support your reasoning of why WiFi is such a terrible thing and you can’t trust it with your hobby machine, even though no one is pushing you to use it.
My response is so that others can also see that I don’t agree that those 2 things are relevant. If you do, then that’s fine.
Many of us use WiFi here just fine. Is it perfect? no. But it serves our needs, and so it is within my rights on this forum to voice that to the OP in contradiction to your opinion
Okay, look. Everyone here is trying to help. Everyone is providing their opinion, with some facts sprinkled in. No one is pissed. Emotion rarely comes across correctly in text, assume everyone is smiling while they type. I will not let this become a hostile place people are worried about sharing their CNC related opinions.
I can assure you no one here prefers wifi…if anything everyone is pissed I use it and it is biased to a hardwired connection. I built a wireless board and the first thing everyone want is a dam hardwired connection. I can not win.
Now for the facts. If you do not like WIFI AP or STA don’t use it. The jackpot can be hardwired, you can use a pendant or USB, or just use the SKR.
If you want a little more information. The jackpot was made for people that want full control wirelessly and have spare devices. Most any wifi enabled device has full control, ability to update and have a terminal wirelessly. This is nothing like the previous machine you have used. You can connect to the jackpot directly, it makes its own network (if you are super fancy you can connect it to your current network). The only way that network goes down is if you kill the power. If you kill the power and that somehow stops if from booting just reload it. Takes 2 minutes. The board is completely self-contained if your wifi connected devices die the machine will just keep running. All the devices need to do it send a message to start your file, after that they do not ever need to talk again.
If you want wires, you can have them. I myself prefer to have wireless control, the beauty is I know people prefer one or the other so I offer both, and I support both. No one will push you one way or the other, they will try to correct you when they do not think you understand how the boards we offer function.
Personally prefer wifi. In fact, I added the wifi module out of frustration with my long USB cable being unreliable. Ran into Repetier disconnects that caused jobs to halt and fail. I could maybe have fixed with good strain relief. But there’s other benefits to controlling via wifi…
Web UI has improved since sharing my Dec 2022 opinion/experience :
Fluidnc with a pendant is the best of both worlds IMO. It’s not perfect but I’m really digging it.
I like having both Wi-Fi connection and the wired pendant.
To think i was just complimenting your video demonstrating the pendant…
Excellent video by the way.
In my experience, the more you pay for something, the more likely you are going to have these weird workarounds that you have to avoid. We paid $20k/seat for vxworks and had an OS that was significantly harder to use than $0/seat Linux. These boutique B2B solutions just don’t have to be pleasant.
That is the right conclusion though. That PLC is to blame. That is awful design.
All that said, wifi is unreliable. Which is why the design decisions on Fluid are to interface over wifi and then play gcode from the sdcard. Once you start sending gcode, any 802.11 is not good enough IMO. I wouldn’t even want it streamed over Ethernet. Somehow USB is solid enough. But running from a local sdcard is the most reliable, and if you can populate that card over wifi, it is a win win.
The pendant is awesome. But it isn’t turn key, and it isn’t stable. I would be very careful recommending it to anyone yet.
I apologize for my contribution to the discussion that got it started down a wifi vs. ethernet rabbit hole. I have had issues in the past (on machines completely unrelated to V1e) with the stability of wifi, so I always prefer a hardwired connection when that is a reasonable option. I particularly agree that wifi is prone to dropping out at a bad time and that I don’t trust it (but I still use it when necessary for a multitude of reasons).
However, I am new to this community and did not realize that this has been a controversial topic in the past. I am new enough that I did not even know how wifi plays into the interactions with a Jackpot, but was instead referencing other situations where I use wifi (thank you folks for explaining what the wifi is used for). I build custom ESP32 devices at home for wireless connection to a home assistant server and have fun with them and appreciate the quick wireless setup, despite the fact that I wish I could cheaply and easily link some of them to a wired connection. I meant no insult that you chose to make the Jackpot a wireless board.
Thanks again for the awesome CNC tools you have been making. I hope to someday have a good enough excuse, the time, and some free cash to buy the LR3 kit and build myself one at home.
Same experience. At work, we flew multiple different versions of VxWorks in many different spacecraft. We were always finding stupid bugs, and getting completely crappy support.
The one I’m flying now uses an open source RTOS and NASA open source flight software with our own custom apps on top. So will the next one that we’re designing.
I hope when you are ready to do that, you share your build experience here. The community will help you as you go.
On my pendant, I did a poor solder job that caused intermittent disconnect, but after I fixed that, it’s been rock solid. Its firmware has been released (now considered ready for public hobby maker use). I’m using it to run cut jobs.
You just posted yesterday that you are on a brand new topic branch. That is not my idea of stable software. Stable doesn’t mean it has no bugs, it means it isn’t changing. If there are 10 people that have the current version and they have been running it for a week, the chances of a new user running into a novel bug are pretty high. Compare that to the firmware on the skr pro, which has been on thousands (if not tens or thousands) of boards for months and it is based on firmware that has a similar pedigree from the months before that.
We don’t have to wait to get Marlin levels of stability. But IMHO, the pendant needs a month or so of settled use before we convince someone it is more stable than Marlin with a TFT.
It is also a project in its own. Which I think you have communicated well with the videos. No one would think it is as simple as the TFT.
That’s true, and I don’t disagree. My perception of that new branch is the same as Marlin (or whatever) making gradual improvements/slight changes over time. The new branch has only a tiny handful of minor changes. I’m not so much trying to convince people that a month’s worth of waiting has passed, as I am simply saying the month’s worth of use testing started several days ago.
Is there possibly some confusion over the difference between “stable development” versus “stable running” item? I understood “it isn’t stable” to mean “the pendant disconnects a lot” or “hangs up and has to be rebooted” (which in mine these statements are not the case - at least after redoing a poor solder joint). In hindsight, I’m thinking you likely meant “the developers are still tinkering.” ??
Fluid does not have a bugfix or RC branch so I am holding the “tested” label back in the instructions a few releases.
I do start testing whatever the new release is within a couple days of when they release it but without a DEV branch of any sort I think it is best for everyone to stay back a couple releases.
This is my feeling on it, they are actively messing with it, very actively, so I won’t take a deeper look until it is more of a fine-tuning not major changes.
Same goes with the V3 of the UI, several of us have been testing it for more than a month, getting it pulled in to fluid means everything was changed so now it needs to be tested for a long while again. I want everyone on V3, it is better in a ton of ways, but I am not pushing anyone to it until I feel it is stable.
I am coming from marlin, all I know about production releases come from that so I do not know how this would normally work. I have made the Marlin bugfix our “stable release” a few times just to get some features, but typically the stable release is nearly 6 months behind more.
Bleeding edge is good if you know what you are doing but any newcomer should not face any randomness that we can not explain. Bugs are expected, known and understood bugs are fine.
This is what I was trying to describe. I believe that it works perfectly for you. But if there is still a lot of developer churn, and not a lot of testing on a (mostly) unchanged version of the software, then it isn’t ready for someone who doesn’t know what they are getting into.
To be clear, I love that you’re using it, developing informative and entertaining videos about it, and building up knowledge about it.
I just get worried when someone asks about a screen and one of the suggestions is the pendant. UGS is stable. CNCjs is stable (even if it has some well known bugs). WebUI is probably 3/4 stable (but not everyone likes to mess with wifi). If someone wants to cut from an sdcard and a screen, the least effort and least risky way to do that is an skr pro with tft screen. That may be completely different in 3 months, looking back. But that’s my opinion right now (and it isn’t worth a lot since I don’t have one).
That is definitely less work than sourcing and assembling a pendant, which could involve either crimping or soldering for the buttons, etc. I don’t mean to imply otherwise. Just wanted to mention that at least one “wired” screen option does exist for the Jackpot, even if more work initially.
We’re on the same page. I wasn’t trying to chastise you. Just add a contrary opinion. Sorry if it was condescending.
Yes, I think we are on the same page — after my initial misunderstanding of what “stable” meant.
I think there’s an old saying something like, “There’s what you meant, then there’s what you said, then there’s what they thought you said, and finally there’s what they thought that meant.“
Communication can be… interesting.
This thread seems to have ended up turning more into an argument than a solution.
I bought an unbuilt v3 kit that came with an skr.
The cost of buying a jackpot and binning the skr is not really an issue for me if the jackpot is better.
Ultimately is there a better board? I tend to think there is a reason everyone seems to be switching to it.