Ok, this is me griping about this rather odd missing “feature” in Chia GUI.
Flexpool just sent me a message saying my farmer is offline. I went over to my rig, checked the GUI and everything seemed fine.
Half an hour later Flex was still reporting offline so I went and checked my logs…MAJOR FUBAR, everything screwed up, lookup times 250sec, plots not found, read errors, everything.
But according to the GUI everything was fine. You are synced, you are farming all green lights.
Nice to have feature updates and new icons and fancier graphics but come on, this should not be happening a year into this thing. That’s just not having the right priorities imo.
Some forks already have nice overviews that show lookup times etc. Weeks into the launch a year ago some people where quick to slap together some simple tools to monitor farming status.
If the log is full of critical errors, the app should mention this to the user.
First of all, I think that this is the wrong audience. Sure, some of us will sympathize with what you wrote, but none of us can help.
I think this may be the biggest problem with Chia blockchain. The pre-farm belongs to Chia Net, and there is no XCH (earned monthly) to support independent devs (promoted by some sort of farmers consensus, not a corporate desire to go IPO soon).
Sure, Chia Net has some sort of “Chia Cultivate Grants” (e.g., Nucle wallet, I think), but how that is distributed is based mostly of how much sense does it make for the corporate Chia (i.e., banking-based business plan based on pre-farm). You would think that such grant could / should be extended to Gil (farmr developer). You may also wonder whether such grant went to Max (MM plotter). If those grants are real, those should be front and center part of their website, if not just to make us farmers appreciate that work, but also to guide devs to apply for them (and maybe us farmers supporting those that we feel really count).
As far as getting something done, I participated in few threads on github (either as a supporter, or as a originator) to change logs to be more explicit when catastrophic failures happen, and all those threads were basically ignored. So, that is Chia MO right now (IMO).
By the way, I am using ChiaDog to monitor my farm. I exposed my log folders on the network (read only), and have ChiaDog analyze those. ChiaDog reacts rather fast when something happens with “known problems” on harvester side. You may want to try it.
Agreed. They seem to have pretty much left the GUI development process. I’m guessing it’s because they rarely personally use it, and they don’t care, as the OG plotter works, but has few features of note helping farmers. Or at least they are not adding much of anything that they, themselves created. They did eventually hire MadMax’s developer to make a decent plotter (for normal users) to include in the GUI, But that wasn’t their effort.
My thought has been for some time… why not hire someone like the ‘farmr’ developer (or maybe someone from a pool) to put some decent reporting in the GUI? The work’s been done, it’s just not included. I’m sure there a bunches of people like yourself, relying on the GUI alone, that blindly farm away for nothing, as there setup has failed.
If I was to make an even more global opinion, Chia is more and more favoring features benefiting ‘whales’ over small farmers. Whales don’t need much, if any, handholding, and so they main intent is mass production of plots…think Bladbit and using well endowed server hardware…everything else, such as node health is assumed to be the obligation of the node owner.
Here’s 1.3.6-dev136. The updated plots view is nice and actually gives a view of the remote harvesters and plots. It won’t alert you to anything going down, but being able to see the harvesters and plots with those basic stats above is a great step forward.
I’m not sure I’d ever expect the software to also have built in monitoring capability. Perhaps Prometheus metrics exporting would be a good feature add. This would open up monitoring possibilities for things like Grafana alerts or other 3rd party tools to monitor based on the exported metrics.
But that’s the thing, they have plenty of hard cash. Just last AMA Gene said they were on budget and so not feeling any need to get additional funding because their bank is fine.
I would guess finding staff is a bigger problem.
Might give it a try, right now I am using my 100 plots on Flexpool to act as my warning system, and this time it worked like a charm
But this is kinda my point they are doing all sorts of work on the GUI and underlying software, quite regular new versions if you ask me. But if you have time to update the graphics, you should have time to do something relatively simple as adding error to the front-end. I mean the errors are there, the program is already throwing them out, they just get written to a log and that’s it. Errors like long lookup times and plots not being found should just get to the user on the front-end.
But that’s the weird thing, the whole idea of having a GUI is for the smaller, less technical user. “one-click-farming” is a major part of their roadmap and vision and this is geared toward small-time users. My guess is the team is just to nerdy They don’t think like regular computer users anymore.
From their job description page:
We have a focus on easy to use software across architectures and operating systems
If they really want to get this “one-click-farming” off the ground they need to get someone onboard who understands UI/UX a bit better. That’s not about doing more work to make everything fancier, it’s about making the right choices.