Hi, sorry for your misfortune and thanks for the heads-up for other forum-members.
When planning to start from scratch, maybe you could save some time by copying the blockchain file and importing that in your fresh setup.
Syncing from the network from block#1 takes considerable time now, the blockchain has grown pretty big.
I guess the blockchain itself is not compromised and starting the sync from a somewhat recent block may save a lot of time.
I don’t know where the blockchain database is located in windows and which file you would have to copy (cetainly not the wallet-db!). Maybe some other windows user can help. Or ask the dev you’re in contact with, the least they can do to help i think…
1 Like
lolz! Thanks, but I avoid downloading block chains (security issues) and am not concerned about the sync time anywaze. I will be synced before I have gotten my first new plot started due to madmax installation and learning time. Then I have a couple months of plotting to do again, rofl! Sync time not a problem.
I think that it is in:
%userprofile%\.chia\mainnet\db\blockchain_v1_mainnet.sqlite
Save that file from your old/synced installation, and copy to the new one. I think that file is user-independent. However, maybe someone can confirm that.
I did copy it around before, but always used for the same keys/user.
The issue is with this method:
And the one above it. For performance reasons, we limited it to 20 so that it would be faster to scan for all wallets, and we assumed that nobody would have that many.
We should at least prevent users from creating more than 20, or perhaps increase the limit.
I apologize for forgetting to prevent users from doing this, I can imagine it’s stressful to lose all those plots.
If you are building from source, you can change those files yourself, and it should work, but it might be pretty slow.
2 Likes
That makes sense; madmax is a bit more efficient (depending on plotter system) than the gui plotter so hopefully not too many months…
There is a Windows fork for madmax by Stotik: Releases · stotiks/chia-plotter · GitHub.
Digital SpacePort has lots of info and youtube video’s on madmax configuration and optimization, i think he uses windows mainly.
His ‘beginners’ guide (https://www.youtube.com/watch?v=7xmEdJHYjTs&t=855s) may be helpful as a starting point.
Good luck
I guess, you still don’t understand the problem. What you are pointing to is some garbage code that was not designed right, not peer-reviewed, not testd / no QA procedures to go over that.
What we are talking is that your UI is just pure garbage that doesn’t do any checks, that doesn’t show any errors from those engineering logs. Just high school level coding. That is the main problem that somehow is so hard for Chia Net team to grasp.
Preventing more wallets is a good start, but what really needs to be addressed is the GUI.
The Chia GUI has no help or direction at all. I happily hit the create NFT button every plot with no warning or understanding that NFT was not just the name of the plot but the thing that all your plots should be pointed at. Hitting create NFT seemed perfectly sensible to me every time.
Now that I understand and have looked at the plotting screen again it makes perfect sense as it obviously did to most others, but lack of any direction or description in the interface started me on a road leading to disaster.
In addition to making the GUI more clear I would suggest that documentation and/or a manual be available alongside or even with the Chia.exe download. Lastly, it was asked in this thread for a list of the major error codes and their meanings.
Be Well.
1 Like
Just to be on the same page with that code you pointed to.
- I don’t see any comments there what each block is intending to do
- I see some magical numbers that are being repeated, instead of having descriptive manifest constants
- I see basically the same code repeated (lines 57-61 and 69-73)
- I see some lame casting there done left and right, instead of having the source providing you with proper types
And what the above means for me:
- If you don’t have any comments, there is no point of analyzing such code by anyone else (i.e., your coworkers), it takes too much time to analyze what the other engineer was too lazy to do
- When you repeat some magical numbers, you are making an opening to have them being changed in one place, but not the other, thus further degrading the code
- When you are using some magical numbers, you are the only one that understand where the fuck those came from, and for what reason - again, that is what manifest constants with proper descriptions do
- When you start casting let and right, you just opening your code for some unexpected casting done by potentially less experience engineers that are just “following” your bad example
- When you see that you need to repeat code, you start making functions. When you repeat code more than two times, it is time to get you replaced, as you have no understanding of coding
- when you need to loop on those wallets, and you put some limits, it is implied that the other part of code already screwed up the system, and you are just trying to recover from shit that should be exposed (I don’t see alarms going off that the number of wallets is over those magical numbers.
So, on my book that code is worthless, as far as production code.
It also shows that whoever was coding that was given a carte blanche to do some junk coding just to move stuff forward, no clear indication what the intended goals were (i.e., make sure that the end user is safe). Production code is mostly about making sure that no harm is done, and if there are any signs of harm, it is immediately exposed, not buried in another layer of garbage.
So, yeah the code looks to me like done in Elbonia.
1 Like
I agree, but kind of. The code example given is not production ready. It is hard to blame those engineers, as they need to gain experience somewhere. In this case that somewhere is the management. When you start doing some production code, you start putting some guidelines, some structure around the development team. You don’t let them code whatever they feel like that day.
So, as I agree 100% that the UI is garbage, my take is that it is more than that, the software team management is in need of replacement, as they have no clue what a production software group means. You can force a software engineer to change their coding style, but you cannot force an engineering VP to do stuff that he doesn’t understand, as he should be the top level of that knowledge - thus the need for replacement.
Just to be clear, in my opinion Willam Blanke (if he is still the VP Engineering) should go. It is way over his head what needs to be done. I would assume that whoever replaces him will know how to drive that team.
1 Like
The shoddy GUI is a result of the rushed and lazy coding. Chia.exe is still low github quality when it should be consumer ready with help screens and documentation.
You and many others have pointed out the bad coding and the GUI is still more like a beta test version waiting for input and improvement.
I just formatted 10 drives full of plots. 
I have a clean copy of 1.2.7 installed with a new key. I had to delete the config.yaml before installing 1.2.7 to avoid contamination by the many pool wallet references and whatever else had gotten screwed up. In spite of the new key the blockchain sync’d immediately with only the new wallet having to sync from scratch.
I ran chia wallet check and chia plotnft check to make sure the extra wallets were really gone. They are.
I installed PSChiaPlotter. Looks good. I should have madmax running 72 K32s to my 1st Chia drive tomorrow … taking the rest of today off, lolz!
1 Like
I don’t think it is like beta. On my book, it is still an engineering only prototype (useful for devs to debug, useless to guide end-users).
1 Like
Included in that is an excellent heat maps module. It only requires read-only access to your log folder (so can be run over the network for those that are paranoid about chia security - me included). I run it 24/7, and it helps when chia behavior changes. I run it with 2 seconds option, as I don’t want to wait for things getting close to bad. It is a simple tool, but rather robust (I have never had any issues with it). Here is more info:
1 Like
Long thread. First thing I’d do is leave the pool and go back in again or go to another pool.
I’d see what happens when I have no pool also.
I’d try another entire machine, because unless you completely format all your hard drives used for plotting, something might stick around - because that’s windows.
Try another OS would be good.
I really doubt you need new keys. Additional wallets thing is weird, but - once running shouldn’t matter. You just use one and ignore the others.
Short of that, I’d say you’d be looking at a hardware fault e.g. RAM, CPU or something fundamental.
If the machine is dedicated, try something like TrueNAS which would give you a non-windows option - though do check up first if it will read your NTFS drives.
If you’ve got some bodgy plots, it is really likely to be hardware. It might be disk, or it just might be RAM, or CPU. Definitely try another complete machine if you can.
That’s my 2c. 30 years of fault finding ‘instinct’ there. You just gotta try ruling out major areas and hone in on one when you isolate it.
So actually on that front. See if you can run a memtest on your machine. Or any other hardware test you can find to rule it out.
Yup, we just isolated it, garbage code, garbage UI permitting obvious (to more experienced end-user) mistakes.
That just leads to a pissing contest, as it doesn’t add anything to this discussion, but is intended to imply that you have some superior expertise comparing to others.
Sorry for that harsh response, but you need to go over most of what has been written already, not just providing general suggestions.
The key is the center of the issue and the additional wallets matter so much that with too many you make your key and it’s many wallets so unwieldy that the harvester and the GUI start to fail.
Your key has one Wallet. Every time you create a new NFT it creates a new, “pool wallet”. All wallets are part of the blockchain. They cannot be deleted or severed from your key. Each wallet has it’s own pk and sk so the more wallets you have the more pk sk combinations the system has to handle.
In error, I hit the create NFT button every time I plotted 3 parallel plots. Unknown to me at the time I was creating a new wallet and a new self-pool with three plots in it every time I did so. By the time I was done I had 53 pool wallets and my main wallet. The Chia GUI is only set to handle 20 wallets and even that is probably too many. This caused my harvester and the Chia GUI to choke, even though the GUI appeared to be working well.
Note to self and others: You should have ONE pool wallet unless you have some specific mission in mind. I cannot think of a need for more than one.
I use farmr and recommend it to all. If not for farmr I would still be happily looking at my machine that was just pretending to farm. As I plotted more NFTs farmr told me I was getting harvester failures. I checked my logs and found the (at that point rare) pk sk error. I went into full search and plea mode but got no answers for weeks until I finally put my questions to support at keybase.
By the time I understood the problem it was far too late.
As there is no way to delete or sever the wallets from the key the ONLY solution is to start with a new key. Since you cannot use your dysfunctional and unfixable key, you cannot farm any plots you made with it either. Key and all plots associated with it flushed down the toilet.
At least re-plotting using madmax will take some of the pain away, lolz! . 
2 Likes
Sorry but your rant is getting a bit old now…suggest you find yourself your own $60mln worth of investors, launch your own designed blockchain and see if you do any better.
Either that or learn how to be a critic in a more civilized way
2 Likes
I think that you are missing the point. We are all here the strongest promoters for Chia. We all invested thousands in the equipment believing that all this will work. We all would like to see it succeed.
So, no need to shill for Chia, as they can just step in and let us know that things will be taken care off. They don’t do it for whatever reason. Also, sending an engineer to show his code for end-user focused thread doesn’t help.
Again, if you want to help, help @Aspy68 to recover faster, or let Chia know that bad things are just being repeated, and we are really frustrated.
LOL I’d say more likely what you said above would lead to that. In fact you’ve been quite derrogatory, including the sentence above, so might I suggest you keep your opinions to yourself to avoid the exact thing you’re saying you want to avoid.
1 Like
Cool glad you’re on the home stretch!
My brain just burst out with a way to jury rig a key with too many wallets so it will work again.
If you created too many wallets, Chia and the harvester will continue to choke even if you remove the associated plots because the list of wallets for the plots is still in your config.yaml. Your key will always have too many keys but there is no reason for the Chia GUI to know that.
All the NFTs and their plots created in self-pools except the first NFT plot you produced still need to be deleted, but all OGs can be retained. After deleting all but the first NFT plot you produced, get your remaining NFT’s public key.
Edit config.yaml. Under pool_list you will have a long list of pools. The first pool group entry should have the same public key as the first one you produced. Confirm that this is the case. If so, delete all other pool group entries below the first one.
After re-starting the Chia GUI it will now see and be dealing with only your main wallet and one pool wallet. Confirm that you have kept the correct connection/combination of NFT what you left in config.yaml by taking your OG farm directories offline and farming the remaining NFT plot by itself. Check the log to make sure you do not have ERRORs and to confirm that the plot is actually correctly harvesting.
If good, put the rest of your OGs back online and you can start plotting again, only now, making sure that you plot only to the one NFT and do not continue to make more, lolz! 
Support at keybase Chia team confirmed that this solution should work. It was pointed out to me that the config.yaml would repopulate the extra wallets if you re-synced your wallet from zero so you need to make a wallet db backup to avoid this.
Unfortunately this did not come to me until after I was finished my last response telling marshalleq that the ONLY way to fix the problem was to abandon the key and delete the plots and that I had just deleted mine, rofl!
Be well!
1 Like