I found some helpful stuff on YouTube. One video had about 15 tweaks in windows for continuous wifi. Some were about power management issues, but others dealt with changing preferences on the wifi device itself (Device Manager/Networking/Network adapter). BAR had a great suggestion about turning off “Connection monitoring” in the FluidNC preferences tab. Baby steps. Don’t lose hope.
I would consider the transition to 1.09 to be a bug/failure.
The location of the kinematics variables (anchors, etc) changed, but when I did the full install, it didn’t show the new variables correctly in the fluidnc config tab to let me change them until I uploaded the yaml file separately
Also, when you do the next video showing full install, show what success looks like (and ideally a way to confirm that it worked without any unexpected problem) rather than just ‘click here and iyou are done”
Also, when the parameters move, rather than requiring a new upload and manual entry, it should detect that the new parameters are blank, look for the old one and create the new one (and remove/zero out the old one)
Absolutely. Testing for the ‘old’ parameters as part of any upgrade workflow (preferably within the code rather than an installer) is a great way to set up for actual migration, although I would leave the ‘old’ parameters behind to allow for easy roll-back, and instead flag them as deprecated.
Then identify which future version is intended to finally remove them as parameters, both in the config file, and in the code.
Lee H wrote:
Absolutely. Testing for the ¡old¢ parameters as part of any upgrade workflow
(preferably within the code rather than an installer) is a great way to set up
for actual migration, although I would leave the ¡old¢ parameters behind to
allow for easy roll-back, and instead flag them as deprecated.
the only reason I would remove them is that we have instructions for manually
editing the file and if those old variables are not kept in sync with the new
ones and are ignored, the confusion that could happen would be worth removing
them.
but I guess adding a comment line for each of the old parameters pointing to the
new ones could work.
Then identify which future version is intended to finally remove them as parameters, both in the config file, and in the code.
yes, but that needs to be years out (and as long as there are automated tests to
make sure the code doesn’t break, why remove it at all)
David Lang
After completion of a job, the pause in orange stays. This should revert back to the Green Arrow
when you release tension then apply tension the Z values are way out. When I started a new job without resetting it wound right up to the top and more on the Z axis. Bit nowhere near the work area.
I noticed this one too and I’m working on fixing that. Refreshing the page will fix it (obviously not a solution, but a workaround)
That is a FANTASTIC insight. I’ve been feeling like there is something weird with z going on, but I wasn’t sure where to start looking. I will track that one down.
Thanks for the reply. I rather not waste anymore time. I am currently 90% done designing a 4x8 cnc machine for under $800. I will make a videos and post to youtube as I go along. I entrusted to much of my time and money to maslow. If your interested on a better alternative, my youtube CompactCo. I have another build series of another cnc machine I fully designed and built.
Good luck in your future endeavours ![]()
Thanks! Likewise! I’ll peak in here once in a while to see how maslows issues get fixed. It’s looking like most issues might be hardware related to connectivity that creates the issues.
To be clear - I disagree with what you’ve been saying, I especially disagree with the way you’ve been saying it, I disagree with the weird fetish you seem to have for spamming every thread (to the point of ressurecting very old threads) trying to drown helpful voices in negativity, and I disagree with your attitude that you know better than everyone else, to the point of ignoring a lot of questions of people who have got it working because ‘you knows what the problem definitely is’.
But I understand you bought something that wasn’t what you thought it was because you didn’t do enough research, and that sucks. It’s clear you’ve burnt all your bridges and it’s time you move on, and ultimately I hope you can find something you can make work for you and be happy.
I also noticed this going from FW version 1.02 with index.html.gz version 0.88. It seems to work fine after uploading the yaml manually afterwards.
Hey Dave, not sure where all the negativity came from. I’m really shocked this is how people are reacting to a product being given the chance and ultimately failing for some end users. I’ve never criticized anyone individually or the community rather I’ve solely targeted the massive flaws of Maslow.
Fetishizing is strange to say, but I don’t judge others kinks or fantasies.
I believe my biggest mistake was believing in Maslow so much. The very little videos of it working looked amazing and the marketing did a great job at that. However, looking closer now, the photos of projects posted in the website are projects made with the previous versions on Maslow. How many, I’m not sure but some have the machine in the background.
Thank you for your kind words. I hope you understand that you cannot marry a product, service or idea. You have to know when it’s beyond help and spend your time and resources on something that it truly worthwhile. ![]()
Juan
I have read most of your posts. Your posts contain a lot of complaining with little to no tangible information on what problems you had or what you attempted to do to fix your issues. You barely asked for advice or help.
For someone that claims to be smart enough to create your own milling/cnc machine, you have not provided any code suggestions, solutions, or constructive ideas. You are bashing just to bash.
Anyone that is on the forums know and understand the problems that exist and how far along the project has come. It clearly works. Are there some bugs and some tweaking that need to get worked out, sure there is. If history has anything to say about it - Bar and those on this forum will work them out.
As far as marketing - this thing is only sold through the website tied to these forums and the Kickstarter. Everyone that buys one has full access to research and understand what they are getting and the current problems with the machine which are actively getting worked.
Your comment about having to buy a mobile hotspot and pay a subscription to run the machine is just ridiculous. Again, just bashing to bash.
I have no doubt that you will continue to post and respond trying to justify your comments and position with absurd reasoning.
Sell your machine, grow up, and move on.
Hey Tim
Thanks for the response.
You’re correct, rather than publish the almost identical problems and errors everyone else is getting, I attempted to solve it through following others advice and videos posted for about two months.
As I’ve said to others, we do not have access to change the component of the main board, which I suspect is the problem. Why? They’ve have a defect before and could possibly have to do with the connectivity issues. It terms of the coding side, that’s honestly only possible but the original author. It’s similar to large scale social medias like twitter. Twitters engineers have come out and have said that they themselves to not know how’d they fix if a core structure is corrupted; they only add to the existing structures.
Yes, there’s been people who have been able to make it work, others have not. I’m not sure what the ratio is from working and non working, but I am in the nonworking category.
Your correct, marketing does not have to be ads on social media sights, marketing is anything tethered to a project that gets shared and reposted but others; it’s organic marketing. Research can only be done to an extent. It’s like any other product, the videos look promising and seem to work. As a retail buyer we aren’t given the products flaws because it just wouldn’t sell.
I am unaware of an internet provided giving good, strong mobile hotspots for free (not cellphone hotspots, very limiting and weak). If you are aware of a free WiFi hotspot please let me know.
I have yet to see anything that I have said being absurd. In the contrary, I’ve been pretty targeted with my criticism and with experience (2 months at this point) of trouble shooting and following recommendations and videos. I’d concede if I tried for a few days of or week, but it’s been two months.
I’m trying to sell it, but I don’t think anyone is going to buy it. If growing up is spending hard earned money for something that doesn’t function, I’d prefer not to. I’m just peaking in here to reply to other upset users. Moving on is different for many, I just reply as I get the notifications or check to see the latest problems with Maslow.
Like I said to another user, try not to marry a product, community, or idea. True growth is able to know when something just doesn’t work and finding something worthwhile.
Juan,
I’m not aware of anyone else needing to get a hotspot through their cell provider for this tool, and the fact that you keep coming back to this lead me to wonder yes maybe this project is not for you from a basic technical perspective. We all use our WiFi or our laptop/phones directly as needed to control it.
Mine works well, and if someone who has a lot of experience in IT support I know that people don’t typically call in to tell a company that things are working great. Because of that, you might be getting an inaccurate impression from the ratio of working/not working post here because in general,people like me tend to only write in when we need to troubleshoot something.
If anything, I really should share more of the projects that I’ve done with this tool. I’ve had great results, and I’ve also had some perplexing problems, that’s just the nature of the game.
It is unfortunate that your expectations were so different from the community in general. I believe most of us took to heart the very frank advice upfront at every step of the way that this was not a “out of the box“ consumer solution, but is instead a project aimed at peoplefamiliar with quirky new technology who are willing to invest time and love and occasionally a little bit of blood to make these things work.
The project is also priced very appropriately to this expectation.
It is unfortunate that you are having a negative experience and I also wish you had entered this prepared and willing to share the data needed to properly troubleshoot issues. On the occasions that I have, I’ve received tremendous assistant from open hearts and minded volunteers as well as the primary developer themselves and I have very positive association with this project and the device that I’ve built and I hope you can find something for yourself in the future that does the same for you.
Hey Thunder
Thanks for the comment.
The hotspot suggested was given by a user in Maslow Reddit. They mentioned to attempt to connect to a home network WiFi rather than the Maslow self made network. I thought it wasn’t a good suggestion just like you especially since bar stated it’d be best practice to connect directly with Maslow which I’ve been doing. Are you saying you connect to your cellphone hotspots and connect your pc and Maslow to it? I’m getting conflicting suggestions by the community and others who’ve posted on here.
I understand what you’re saying. That is true, most publish things when things are not working. I hope users begin posting more completed project and posting YouTube videos, it’ll really reinforce the success for future potential buyers.
I’d be great to see what you’ve made. It’d really show what the machine could do! I’ve delve into the community builds and most are from the oldest design, there’s where my support for this new one came from.
No it’s ok, I should have understood the context open source was being used here. I believed open source was similar to the likes of PrintNC or Prusa or Lowrider but it’s a bit of a different definition here. I know now to be weary of open source projects.
You’re correct $500+$200=$700 is low for a cnc but not cheap in general terms especially the economical climate here in the US (not sure where you’re located but hope you’re doing better than us here).
I didn’t want to repeat the same errors and problems everyone else was having. In a way to not overload the system type of thinking. I read through other posts and suggestions and did as advised as well as watching all the video in the Maslow Guides and Maslow YouTube page.
I don’t doubt the people in here, they’re great and very helpful; I doubt the product.
Nevertheless, you’re right, should have stuck to my true passion of designing and fabricating my own cnc’s rather than attempt to use someone else’s. That truly was my mistake.
Thanks Thunder for the dialogue, have fun with Maslow! I wish you good luck and health for the future!
That’s fantastic, I will check it out.
This is firmware 1.10.1 which I would love feedback on. The big change here is that in horizontal mode both apply tension and the first point in the anchor point locating process now pull all four belts tight at the same time.
The eliminates the need for the “tug to simulate gravity” that was throwing folks off before.
It seems to be working well for me at this point, but I would love to hear from others if it’s working for you too.
No need to mess with the maslow.yaml file on this one.
index.html.gz (122.0 KB)
firmware.bin (1.9 MB)
Will test this morning.
Is it possible to put the version numbers in as eg V1.10.1 so we can be sure the files are loaded correctly with these Beta versions.
