Losing faith in DRK

[CHAOSiTEC]

Everybody, please take a deep breath...

The quest for information, for the sake of information, does not lead to anything, isnt it better to get information WHEN there is something to communicate, alot of stuff happening behind the scenes, that really is just plain boring to the average investor.

Also, i would like anyone that feels development is going slow, to look up "software development captain crunch", in regard to RC4.

Evan is working on this FULL time, he needs to also be able to have a life on the side, he is nobodys slave, besides, its always better to code, when your head is in the right state. He is also spread thin with all the stuff going on "coding, foundation, bug hunting etc"

As in regard to me, i am reading through the code, to get to know the "beast", working on odd jobs and trying out ideas... it takes time to get into things...

take a deep breath, get outside, enjoy a night out with your wife, just as we deserve to be able to do the same...

And remember, we are working toward Release Version 1 - all good things comes to he who waits...

This thirst for new stuff, new feature news news news is crazy, look at bitcoin, how often do they release new info... what about peercoin, vertcoin litecoin etc...

CHAOSiTEC

 

[vertoe]

Just a few notes,

• much like fernando I'm not part of the dev team.
• any comments loaded with too much emotion will be removed, please stay calm and stick to facts.

And back to topic, this is not about discussing our current problems mainly due to (missing~) communication, the problem is that this discussion is as old as the community itself and I personally wish to highlight that, and I think this is BelStar's main point: We have issues, we keep discussing them in the community, but anything never ever changes. And that's why he sais, he starts losing faith in DRK. I totally agree here. If we keep suggesting things and never get any feedback and see the same mistakes happen over and over again, then I totally understand BelStar's argumentation.

 

[vertoe]

Everybody, please take a deep breath...

The quest for information, for the sake of information, does not lead to anything, isnt it better to get information WHEN there is something to communicate, alot of stuff happening behind the scenes, that really is just plain boring to the average investor.

Also, i would like anyone that feels development is going slow, to look up "software development captain crunch", in regard to RC4.

Evan is working on this FULL time, he needs to also be able to have a life on the side, he is nobodys slave, besides, its always better to code, when your head is in the right state. He is also spread thin with all the stuff going on "coding, foundation, bug hunting etc"

As in regard to me, i am reading through the code, to get to know the "beast", working on odd jobs and trying out ideas... it takes time to get into things...

take a deep breath, get outside, enjoy a night out with your wife, just as we deserve to be able to do the same...

And remember, we are working toward Release Version 1 - all good things comes to he who waits...

This thirst for new stuff, new feature news news news is crazy, look at bitcoin, how often do they release new info... what about peercoin, vertcoin litecoin etc...

CHAOSiTEC

I think you misunderstood, this is not about releasing more and more stuff and delivering even faster and faster. The development is doing well, we now need to focus on other things as well, securing the network, enforcing masternode payments, writing up documentation, and so on...

I highly respect eduffield for working full time, but if he has no time for all the other stuff, why not delegating to the community? He cant do all on his own for the whole project..

 

[CHAOSiTEC]

in regard to enforcement, Evan has already chimed in on this, and it will happen before RC5, but tests still needs to be run to make sure that part works as it should, so that IS coming

the rest, well, right now im at my work (yes i work on a boat with lousy internet and a bandwidth cap) so when i get home i will see if i can whip something more coherent together after talking with everyone else...

 

[JGCMiner]

Which issues have been ignored?

I don't want the dev team taking time to listen to everything that gets suggested as then nothing would get done due to the inefficiency. Nor is it good if the community gets to see behind the curtain. People who want that level of involvement should just petition to join the dev team.

Furthermore, I think the team has been receptive to community input. Off the top of my head I can state that the masternode system (discussion with Anonymint on BCT) and Darksend+ (denominated change discussion on BCT) were direct results of listening to the community.

As it relates to the example given by the OP of the test rollout plan being ignored by Evan -- I actually think that was the best decision as none of the issues had to do with the RC3 to RC4 roll out but with things that would be hard to find on testnet (like the need for higher fees). We can't stay on testnet forever. Also, the wallet corruption thing was mostly just the old libdb version issue which has nothing to due with RC4. (I do feel for the one guy whose coins were unrecoverable due to RC4, but if you can't even send Evan the correct debug file so he can check the issue then who knows what other PEBKAC was involved.)

The software is not bug free, but I personally have used it on mainnet and it does WORK. Squashing bugs is a normal part of software development. Better usability and communication are coming down the road. Let them get the tech right first.

Darkcoin is 8 months old... now is time for patience as it moves out of beta -- not time for panic.

 

[georgem]

"software development captain crunch" ?

I don't know what you are referring to?

 

[elbereth]

Interesting thread and I agree with all the points of Belstar.

While I am a software developper myself and I understand developping an application is long and tedious (and this one specifically, as it is mighty complex, I really admire Satoshi's work), I don't understand how, even if we are talking about beta software (not marketed as such but let's be honest we barely reached beta status) we can go to production (mainnet) with highly important features not even tested for regression (or not at all like the enforcement). I know testnet is not big enough and not comparable to mainnet in term of what we will see, is that an excuse to not test a feature at all and rush to mainnet while bugs (and not cosmetics ones only) were still discovered on a daily basis on testnet? A bugless software is impossible (even more for distributed data like that) but still I think it could be better.

We (long time believers) know this project is long term and this is a valid argument from people saying: "why the hurry?". But I will say the same to Evan, why the hurry in putting RC1, RC2, RC3 and RC4 on mainnet? Everytime the release was not ready and what 1 or 2 weeks extra of testnet would have cost to Darkcoin?

I like some of the constructive answers here (not quite the ones dismissing or trying to diminish the valid points), and I really would like to be sure the dev team will be up to speed soon enough and that flare QA expertise can gain more momentum (JIRA was a much needed tool and is a blessing). It is good to hear that CHAOSiTEC has the code. What about the others? Hopefully DRKLord is fine and we can also see or hear progress from UI redesign.

I think what we all want is to see that things are improving, I take the post of Belstar in that way. We really don't want Darkcoin to go to waste as it looks promising despite the current flaws (we all see the vision or we would have fled already). But honestly, release after release we keep seeing the same mistakes done and that's hard (for me at least)...

To bring another note to the already long post (sorry...):
There is one thing that I am afraid of at the moment is the masternode payment (I know some masternode owners are worried also).
The feeling if you got many masternode as I do (did..), is that the payment is weird and really not random:

• Some of my nodes were having up to 3 payments every day for 4 weeks consistently
• Some others nothing AT ALL during the same period

As discussed on another thread, now that I did my block page I tried to do some payment repartition to check if there was something fishy.
Here is the graph I obtained (was done for RC4 only on 21/08/2014 and I know it should be finetuned, therefore why I didn't publish it back then):

As you can see on the graph:

• The vast majority of the masternodes received 5 or less payments since RC4 went live
• Next group had received between 6 and 9 payments
• And then some lucky masternodes are receiving 18+ payments.

How can this be caused by just variance? The payee selection seems not smooth at all or we would see a much smoother graph.

I know it is a difficult analysis to do and I am not mathematician, so bare with me if this is completly useless data.
As proposed in another thread, all data is available (I can provide with dumps of my database).

Long post already, I better stop.

 

[CHAOSiTEC]

"software development captain crunch" ?

I don't know what you are referring to?

http://cs.stanford.edu/people/eroberts/cs201/projects/crunchmode/econ-software-differences.html

 

[stonehedge]

You're doing a better job that I could at expressing my fears about masternode payments elbereth !

All I can do is point to my small group of nodes and say that over approx a month there has been a huge difference in "luck". A quick peek at here http://drk.poolhash.org/ also shows that the consecutive block payment issue is back again despite being closed off on Jira. This in itself means that it is not random variance.

I really think this needs to be a top priority investigation for the dev team. I have been crunching numbers this (GMT+1) morning and for some of my nodes, they are costing more to run on EC2 than they are bringing in. Since RC4, CPU and network utilisation has increased. Imagine what its going to be like when we get greater adoption?

I appreciate that the price of DRK will be key to ROI on masternodes so I'm not going to close them in the short term but masternodes are so critical to Darkcoin that I think these concerns need to stop being dismissed as random variance and properly investigated.

I hope I'm not expressing a very ignorant view here but isn't clarity and stability on the masternode network the most important thing for DRK at the moment in terms of core functionality?

 

[JGCMiner]

While I am a software developper myself and I understand developping an application is long and tedious (and this one specifically, as it is mighty complex, I really admire Satoshi's work), I don't understand how, even if we are talking about beta software (not marketed as such but let's be honest we barely reached beta status) we can go to production (mainnet) with highly important features not even tested for regression (or not at all like the enforcement). I know testnet is not big enough and not comparable to mainnet in term of what we will see, is that an excuse to not test a feature at all and rush to mainnet while bugs (and not cosmetics ones only) were still discovered on a daily basis on testnet? A bugless software is impossible (even more for distributed data like that) but still I think it could be better.

What was not tested? Enforcement wasn't but that is also not "on" on mainnet. DS+ was tested for a month and there still were new issues noticed in a few hours on mainnet. Would another week have made huge difference? What exactly should have been different or tested longer?

Maybe if specific incidents where referenced with a clear statement of how it should have been better then I could understand, but a general statement like "testing needs to be longer" or "we need better teamwork"' is easy to make and doesn't really help.

P.S. The issues with the RC4 MN payments have been acknowledged and Evan has already confirmed on Jira that the old RC3 system will be used for RC5.

 

[elbereth]

P.S. The issues with the RC4 MN payments have been acknowledged and Evan has already confirmed on Jira that the old RC3 system will be used for RC5.

I also did a graph with RC3 payments and the problem was also present (if not worse), so going back to RC3 logic will not fix the problems. That's the whole issue, RC4 was supposed to fix the problems we had with RC3 MN payments.

Edit: More answering:

Maybe if specific incidents where referenced with a clear statement of how it should have been better then I could understand, but a general statement like "testing needs to be longer" or "we need better teamwork"' is easy to make and doesn't really help.

Well then masternode payments... Never officially tested (focused on DS+ exclusively) but the problem was already visible in testnet (I never stopped my masternodes in testnet and even with only 15 MNs it is visible, some have 30k tDRK while others have 3x less if not worse).

 

[stonehedge]

P.S. The issues with the RC4 MN payments have been acknowledged and Evan has already confirmed on Jira that the old RC3 system will be used for RC5.

Evan also confirmed that the consecutive block payment issue was resolved but as of just now, 2 out of the last 8 paid masternodes were double paid.

Edit: It was actually 2 out of the last six.

 

[elbereth]

Evan also confirmed that the consecutive block payment issue was resolved but as of just now, 2 out of the last 8 paid masternodes were double paid.

Edit: It was actually 2 out of the last six.

For this Jira bug, I think Evan said we will rollback to RC3 to fix the problem but as of RC5.

 

[JGCMiner]

I also did a graph with RC3 payments and the problem was also present (if not worse), so going back to RC3 logic will not fix the problems. That's the whole issue, RC4 was supposed to fix the problems we had with RC3 MN payments.

Edit: More answering:



Well then masternode payments... Never officially tested (focused on DS+ exclusively) but the problem was already visible in testnet (I never stopped my masternodes in testnet and even with only 15 MNs it is visible, some have 30k tDRK while others have 3x less if not worse).

Good point about not testing the masternode payments and the odd variance. But in fairness to Evan, was that bug report on Jira before he went live with RC4? If so, then this is a legit complaint and something that shouldn't happen in the future.

 

[elbereth]

Good point about not testing the masternode payments and the odd variance. But in fairness to Evan, was that bug report that on Jira before he went live with RC4? If so, then this is a legit complaint and something that shouldn't happen in the future.

Unfortunately the Jira is only up from middle/end of testing RC4, there was nothing before to keep track of problems. We were several people pushing (mostly behind the scenes) to have something to log bugs and flare did it.

 

[stonehedge]

For this Jira bug, I think Evan said we will rollback to RC3 to fix the problem but as of RC5.

I misunderstood. Thanks for correction.

 

[LonnieDRR]

I too get impatient and frustrated with the price of DRK. I bought in at 0.017 and Im down a LOT of money. The thing we have to remember is that the coin is still in development. Only at RC4 and RC5 is coming. Haven't even begun to talk about Release Version 1 yet. Its a ways away. Give Darkcoin time to grow naturally. The last thing we need is another PnD coin right? This is the next evolution of Bitcoin that we are building here. Takes time.

 

[JGCMiner]

Unfortunately the Jira is only up from middle/end of testing RC4, there was nothing before to keep track of problems. We were several people pushing (mostly behind the scenes) to have something to log bugs and flare did it.

Alright then. Given the activity that Evan has had in Jira recently I expect that all reported bugs will be fixed before RC5. The question remains how many bugs will be actually found on testnet, but that a different issue entirely.

I have no issue with testing for a good amount of time and now that we have Jira making sure that all reported bugs are gone before you go live on mainnet. But once that hurdle is cleared then have to bite the bullet... you can't just arbitrarily hang out on testnet for another month for "good measure".

 

[LonnieDRR]

I wish I knew how to code so I could help out.

 

[stonehedge]

Ultimately I think the majority of the concerns that some members of the community have could be mitigated by better communications.

Its all very well knowing that Evan and his team are busy but if they aren't reporting back to us early adopters (even with a no news update) on a regular basis then perhaps the team isn't complete yet. A no news update is worthwhile. It makes people feel that they are being included.

Take one of the projects that I am working on for one of my private clients at the moment. I'm currently deploying ZENworks Asset & Patch management in an organisation with nearly 50,000 endpoints and Solarwinds to a 1,000VM virtual environment and across the Cisco Nexxus core network for capacity management. They are existing products that have been around for years so this work is nowhere near as major and complex as Darkcoin. Most of the work I am doing is behind the scenes and due to political, supplier and (human) resource issues there are long periods where I have very little progress to report despite working my ass off. I'm still required to provide a highlight report to my board weekly even if there is no progress. It allows them to understand what I have been working on, where my problems sit and what my next priorities are. It takes me 20 minutes to produce this report for each of my projects.

If there was some way that somebody close to the core team could pull together a very brief weekly summary including highlight RAIDs and publish it on here I think some of the concerns will immediately be dispelled.

Nobody is doubting how much work is going in but people are worried that the problems that they perceive (from imaginary to major!) aren't on the dev team radar. It doesn't take much to help fix this and communication is key.

Jira is a good step foreward but I don't think it is being used to its full potential yet.

 

[eduffield]

I think it's good to remember that when the RC2 issues happened I was working completely alone. That means that all bug fixes, reporting, communication and literally everything else was coming from only me. Since then we've gained a loyal team of people helping out part time with the project. We've made huge leaps in progress from then to now including nearly finishing Darksend (I'm open sourcing it in a couple weeks), getting a foundation started and getting a better vision of what we want Darkcoin to be in the future.

There's also talk about core architecture needing to be completely redone? That's definitely not true at all. Darksend is pretty much in the form that it will be released in. I have a couple of security updates to make, some commenting and we're finished and ready to go.

In short I'm not even sure what the problem is. The project is trending up and we're making huge progress. The only thing lagging is the price. I would guess that's what's causing the nervousness in the community. Because everything else is solid. The price will eventually follow the progress, it just takes time for one to start influencing the other.

 

[moli]

I think it's good to remember that when the RC2 issues happened I was working completely alone. That means that all bug fixes, reporting, communication and literally everything else was coming from only me. Since then we've gained a loyal team of people helping out part time with the project. We've made huge leaps in progress from then to now including nearly finishing Darksend (I'm open sourcing it in a couple weeks), getting a foundation started and getting a better vision of what we want Darkcoin to be in the future.

There's also talk about core architecture needing to be completely redone? That's definitely not true at all. Darksend is pretty much in the form that it will be released in. I have a couple of security updates to make, some commenting and we're finished and ready to go.

In short I'm not even sure what the problem is. The project is trending up and we're making huge progress. The only thing lagging is the price. I would guess that's what's causing the nervousness in the community. Because everything else is solid. The price will eventually follow the progress, it just takes time for one to start influencing the other.

Hello Evan!

My only concerns are:

1. The windows wallets v.32 still freeze after encryption is put in. They do take quite a bit of CPU during loading the blockchain.
2. I tested Darksend on testnet with 1000 drk for each round and each time it worked fast from round 2 to round 4, beginning round 5 it started to get longer to anonymize (over an hour), but the other rounds worked fast amazingly, i was very happy to see. But DS still has some other problems: it doesn't anonymize the amount you want after you change the amount. For example, if you start out anonymizing 1000 drk, get some coins in the wallet, 999 got anonymized quickly. Then you change the amount to be anonymized to 5000, it won't happen. Also, if you change to a different round, doesn't happen either. Maybe I was doing something wrong?
3. I would love to see documentation with clear explanation how DS works and what we are supposed to do to make it work like it should.
4. I am concerned about the security of Darksend once it's open sourced. I guess i'm not sure what I'm concerned about but is DS ready to be open sourced when it's still not working properly on wallets?

Thank you, Evan, for all your hard work. We love to hear from you from time to time. Your posts mean a lot to us.

 

[stonehedge]

I think it's good to remember that when the RC2 issues happened I was working completely alone. That means that all bug fixes, reporting, communication and literally everything else was coming from only me. Since then we've gained a loyal team of people helping out part time with the project. We've made huge leaps in progress from then to now including nearly finishing Darksend (I'm open sourcing it in a couple weeks), getting a foundation started and getting a better vision of what we want Darkcoin to be in the future.

There's also talk about core architecture needing to be completely redone? That's definitely not true at all. Darksend is pretty much in the form that it will be released in. I have a couple of security updates to make, some commenting and we're finished and ready to go.

In short I'm not even sure what the problem is. The project is trending up and we're making huge progress. The only thing lagging is the price. I would guess that's what's causing the nervousness in the community. Because everything else is solid. The price will eventually follow the progress, it just takes time for one to start influencing the other.

Hi Evan,

I think those with the best understanding of the project aren't necessarily concerned about the price and I would hate to think that you took anything said in this thread personally.

I don't know whether you agree with anything I said about communication up-thread but I think that the whole process can seem to be a mystery to those who aren't in IT or new to alt-coins. Just dropping in to here now and again to share what you're up to would probably make a bit of a difference in-between releases.

You've set really high standards compared to any coin and you have an army of people who believe in the medium to long term future of what you (and we!) are doing. Just don't forget that not everybody understands what a release candidate or code review actually is. Because you've developed something with great potential, you're already attracting people who might not understand how projects like these have to be structured and prioritised. These people need to be included somehow.

Dan

 

[stonehedge]

Hello Evan!

My only concerns are:

1. The windows wallets v.32 still freeze after encryption is put in. They do take quite a bit of CPU during loading the blockchain.
2. I tested Darksend on testnet with 1000 drk for each round and each time it worked fast from round 2 to round 4, beginning round 5 it started to get longer to anonymize (over an hour), but the other rounds worked fast amazingly, i was very happy to see. But DS still has some other problems: it doesn't anonymize the amount you want after you change the amount. For example, if you start out anonymizing 1000 drk, get some coins in the wallet, 999 got anonymized quickly. Then you change the amount to be anonymized to 5000, it won't happen. Also, if you change to a different round, doesn't happen either. Maybe I was doing something wrong?
3. I would love to see documentation with clear explanation how DS works and what we are supposed to do to make it work like it should.
4. I am concerned about the security of Darksend once it's open sourced. I guess i'm not sure what I'm concerned about but is DS ready to be open sourced when it's still not working properly on wallets?

Thank you, Evan, for all your hard work. We love to hear from you from time to time. Your posts mean a lot to us.

Moli: Items 3 and 4 are probably covered by the work that Kristov Atlas is doing. I think so anyway.

 

[eduffield]

Hello Evan!

My only concerns are:

1. The windows wallets v.32 still freeze after encryption is put in. They do take quite a bit of CPU during loading the blockchain.
2. I tested Darksend on testnet with 1000 drk for each round and each time it worked fast from round 2 to round 4, beginning round 5 it started to get longer to anonymize (over an hour), but the other rounds worked fast amazingly, i was very happy to see. But DS still has some other problems: it doesn't anonymize the amount you want after you change the amount. For example, if you start out anonymizing 1000 drk, get some coins in the wallet, 999 got anonymized quickly. Then you change the amount to be anonymized to 5000, it won't happen. Also, if you change to a different round, doesn't happen either. Maybe I was doing something wrong?
3. I would love to see documentation with clear explanation how DS works and what we are supposed to do to make it work like it should.
4. I am concerned about the security of Darksend once it's open sourced. I guess i'm not sure what I'm concerned about but is DS ready to be open sourced when it's still not working properly on wallets?

Thank you, Evan, for all your hard work. We love to hear from you from time to time. Your posts mean a lot to us.

1.) These are the types of things I'm working on right now for RC5. I'll make a note and see if I can reproduce the encryption freezing issue
2.) This is a known issue, http://jira.darkcoin.qa/browse/DRK-40 . It's probably the next thing I'll fix
3.) I think I need to focus on documentation, videos and the like to get the word out that we have a working methodology and product. Maybe that's what I'll do after RC5, it should help with attracting users.
4.) I'm not too concerned about this . The crypto I use in DS is very well understood, and the processes are robust and secure. Also it's worth noting that the masternode code in it's entirety has been open source for a few months now with no exploits. DS+ uses a lot of the same logic to secure it's transactions.

 

[Stealth923]

Thanks for the update Evan - One of the biggest issues in my opinion from extensive testing in testnet was the collateral bug that you subsequently removed for RC4. This is important for network security and is critical for RC5.

have you had a chance to look/remediate at this?

 

[DrkMiner]

I compared CPU load during loading the blockchain of several alt coins (Doge, Namecoin, lite).

They all have some kind of load during loading of blockchain; however DRK had the hight load of 15% (not by much).
(other alts had around 12%)

The only alt coin that had similar load to DRK was Namecoin.

Wallets tested: 9.12.31 , 10.12.31

Tested on windows 7 64bit
CPU: i7 3770K
RAM: 12 GB

CPU load Doesn't look like an issue, but on an old PC it might slow it down. Slow CPU/old PC would have problems with any coin and not DRK in particular.

 

[JGCMiner]

Thanks for the update Evan - One of the biggest issues in my opinion from extensive testing in testnet was the collateral bug that you subsequently removed for RC4. This is important for network security and is critical for RC5.

have you had a chance to look/remediate at this?

Was just going to ask about the collateral payment... won't be good if Darksend can be DOS'd for free.

Also Evan do you have a comment on the claims of uneven payments even in the RC3 protocol as stated by Elbereth? I would expect the payouts to look like a bell chart, but Elbereth is stating otherwise.

I guess what I am asking is if the plan is to go back to RC3 payments without making any tweaks?

 

[eduffield]

Thanks for the update Evan - One of the biggest issues in my opinion from extensive testing in testnet was the collateral bug that you subsequently removed for RC4. This is important for network security and is critical for RC5.

have you had a chance to look/remediate at this?

I must have live test data to fix that one, so that's going to be one of the last things before launch. We'll start RC5 testing soon.

 

[eduffield]

Was just going to ask about the collateral payment... won't be good if Darksend can be DOS'd for free.

Also Evan do you have a comment on the claims of uneven payments even in the RC3 protocol as stated by Elbereth? I would expect the payouts to look like a bell chart, but Elbereth is stating otherwise.

I guess what I am asking is if the plan is to go back to RC3 payments without making any tweaks?

I would expect the bell curve after about 3 solid months of data and there's many considerations of the methodology used to correct the amount of expected payments for the time a masternode is active.

From my investigations, I think the RC3 payment system is fine for now. I'd love to implement a round robin system if I can figure out a way to do that decentralized.