Testnet June 2018 Mandatory Upgrade Testing Thread

All,

1.1.3.6 is ready for testing for all platforms.

In this version we now have the ability to add proposals.
NOTE: You must have your wallet unlocked before submitting the proposal.
After submitting the proposal you must check back on the Proposal Add page after 6 confirms to see if it went in the chain.

Lets test that.

Lets also test the ability to heat mine in 1136.

We also moved the Proposal List to an integrated tab for more professionalism.
Additionally, we made the List update whenever you switch tabs and give it focus.

Hopefully this version will become the release candidate.

All,

1.1.3.6 is ready for testing for all platforms.

In this version we now have the ability to add proposals.
NOTE: You must have your wallet unlocked before submitting the proposal.
After submitting the proposal you must check back on the Proposal Add page after 6 confirms to see if it went in the chain.

Lets test that.

Lets also test the ability to heat mine in 1136.

We also moved the Proposal List to an integrated tab for more professionalism.
Additionally, we made the List update whenever you switch tabs and give it focus.

Hopefully this version will become the release candidate.

hmm, testing heat mining, did you change something there?

I noticed .3.6 seems to take longer to load the block index. (could be cause of the forks going on currently in prod).

I also had some crashes on the windows client on 3.6 while mining.  (go back to machine and the wallet is just gone heh)

edit: crashes seem to be because of the mining issues today, may not be related to the update. so far had 3.1,3.3, and 3.6 crash

hmm, testing heat mining, did you change something there?

I noticed .3.6 seems to take longer to load the block index. (could be cause of the forks going on currently in prod).

I also had some crashes on the windows client on 3.6 while mining.  (go back to machine and the wallet is just gone heh)

edit: crashes seem to be because of the mining issues today, may not be related to the update. so far had 3.1,3.3, and 3.6 crash

Hmm lets test everything in testnet today; not a good day to intermingle testing with prod.

No, doesn't take longer to load the block index.  Lets try to only post a statement if you KNOW its true.  Meaning that you ran a baseline and have something to compare to.  And you know your hard drive cache was empty both times.

Hmm, shouldnt have any crashes if you are mining on one thread.  What are you mining with, 60 threads?  And what OS?  It would be helpful to post the tail of the log (in a code snippet of course) if you encounter a crash.  Try to reproduce it on linux with valgrind then you can give me the line #.

No, doesn't take longer to load the block index.  Lets try to only post a statement if you KNOW its true.  Meaning that you ran a baseline and have something to compare to.  And you know your hard drive cache was empty both times.

Hmm, shouldnt have any crashes if you are mining on one thread.  What are you mining with, 60 threads?  And what OS?  It would be helpful to post the tail of the log (in a code snippet of course) if you encounter a crash.  Try to reproduce it on linux with valgrind then you can give me the line #.

Loading of the index, was noticeably longer on several machines in prod(I'm talking 30sec to 5+min difference) I suspect this is due to the current issue and all the orphan blocks.   


as for mining,  I have several machines mining 4 cores...  windows of course so no valgrind.  I won't be around tomorrow, but i'll see if i can get logs.

I'm creating some proposals from the wallet now. Am I correct in asserting that the fee for the proposals are paid right from the wallet (the fee seems to be 5 BBP now, but that's probably because we are in testnet)?

Frankly, I think that is amazing!

I'm creating some proposals from the wallet now. Am I correct in asserting that the fee for the proposals are paid right from the wallet (the fee seems to be 5 BBP now, but that's probably because we are in testnet)?

Frankly, I think that is amazing!

Yes, the reason its currently 5 bbp, is we went live with that value last year.  However I feel that will end up eventually being abused (as it was in the pool at first), and also, we have been charging 2500 historically in the pool, so I made a transition point as of August 1 to switch over to 2500 BBP for the collateral fee.  So in prod, it will temporarily be 5 also until that date.

Im going to look at any outstanding test cases; we really need to ramp up testing this weekend in order to try to get a version out soon; we cant notify the exchanges of a coming mandatory unless we are sure of a release date.

Loading of the index, was noticeably longer on several machines in prod(I'm talking 30sec to 5+min difference) I suspect this is due to the current issue and all the orphan blocks.   


as for mining,  I have several machines mining 4 cores...  windows of course so no valgrind.  I won't be around tomorrow, but i'll see if i can get logs.

Testing this morning on one home linux and one VM block index seems to load fast, I think it might have been an anomaly.

Yes, the reason its currently 5 bbp, is we went live with that value last year.  However I feel that will end up eventually being abused (as it was in the pool at first), and also, we have been charging 2500 historically in the pool, so I made a transition point as of August 1 to switch over to 2500 BBP for the collateral fee.  So in prod, it will temporarily be 5 also until that date.

Im going to look at any outstanding test cases; we really need to ramp up testing this weekend in order to try to get a version out soon; we cant notify the exchanges of a coming mandatory unless we are sure of a release date.

I feel that the 5 BBP will definitely make it easy to abuse the function. 2500BBP I think works well, as you can ask the 2500BBP back in your proposal. If you have a good proposal you have nothing to fear.

This Sunday will be my 'BiblePay day', so I will test a lot and try a lot. A tiny thing that I have for now is that you can still also access the proposal-tab from the top menu. That's probably redundant now.

I feel that the 5 BBP will definitely make it easy to abuse the function. 2500BBP I think works well, as you can ask the 2500BBP back in your proposal. If you have a good proposal you have nothing to fear.

This Sunday will be my 'BiblePay day', so I will test a lot and try a lot. A tiny thing that I have for now is that you can still also access the proposal-tab from the top menu. That's probably redundant now.

Yes, Im in total agreement, and I have it set to where we can move the Proposal-Add to the top menu, and just see the Proposals from the left menu; I figured I would leave it like this until Anton gets it as he is working on some things that affect the menu right now.

Updated to v1.1.3.6

New tabbed Proposals List
Proposal Add
and "exec totalrac"
 are AWESOME!

Hey, sorry TheSnat, I finally accidentally encountered the long load time for the block index on one of my windows machines (IE > 5 mins, compared to normally 20 seconds).

I think this has something to do with the loading of the prayers in the background...  It makes me nervous; will need to look into it today.

At the worst, I feel its safer if we let the prayers load serially like we used to-and delay the start by 10 seconds... Ill take a look.  I hate to delay the release much longer.

Quick comment/note about the new proposal view (which is really nice to see!) - it would be great if the spacing on the amounts was done the same way as in the rest of the wallet.

For instance, instead of:
8000000

It read:
8 000 000

Not a major thing by any means, but something to consider for a later leisure.

Testing this morning on one home linux and one VM block index seems to load fast, I think it might have been an anomaly.

I deleted and re-created the chain and let it re-sync.

I think it was due to all the forks and the client processing through them(memory was clocking in at 2-2.5gb on a few machines that were online in prod throughout the last few days)

Hey, sorry TheSnat, I finally accidentally encountered the long load time for the block index on one of my windows machines (IE > 5 mins, compared to normally 20 seconds).

I think this has something to do with the loading of the prayers in the background...  It makes me nervous; will need to look into it today.

At the worst, I feel its safer if we let the prayers load serially like we used to-and delay the start by 10 seconds... Ill take a look.  I hate to delay the release much longer.

I appreciate the apology, I tend to run things harder than most
I posted separately I think I understand why my clients were acting up,  (mining slowed to 1/4 the speed with the large mem usage as well)

I don't know if it has been resolved, but there was an issue in the older clients if you delete the block data, and let the client re-sync it won't pick up the CPID for mining until you restart. Easy workaround i'll see if it still happens here. (Edit: still there i'll throw it as an issue on the github later)


I started working on the UI notifications, 

Quick comment/note about the new proposal view (which is really nice to see!) - it would be great if the spacing on the amounts was done the same way as in the rest of the wallet.

For instance, instead of:
8000000

It read:
8 000 000

Not a major thing by any means, but something to consider for a later leisure.

Thanks for the heads up, Yes, we'll add this to a coming leisure.
If you dont see it in the next version after the mandatory please add an issue to github.