Steemit Retrospective for June

in #steemit5 years ago

Ever since we officially left “survival mode” we have begun performing “retrospectives.” Retrospectives are conducted at the end of a “sprint” and are a time for teams to reflect on how they have been working with the goal of continuously improving their processes. We want Steemians to have as much insight into what we are doing as possible, so today we’d like to share with you a summary of what we discussed in our most recent retrospective which covered the past month.

What went well?

  • By deciding to remove the hidden cap from ICO's we have reduced the possibility of 'scam' ICO's and also decreased our development time for SMTs.
  • In general, we made attempts to cut scope wherever possible in order to ship new products and features as efficiently as possible.
  • Made adjustments to jussi and the size of the redis that jussi uses which will decrease future AWS bills by close to $5k a month.
  • Rolled out new ad partner integration which is already generating more revenue than anything we've done before.
  • We did a lot relating to Communities including: Communication, a working group, future sprints, and a final design document was released
  • We created “Sponsored” posts
  • We've started to receive a steady stream of community PR's and features. We wanted more community contributions to Condenser and we’re getting it!
  • We updated the community contribution guidelines and communicated about it well.
  • We effectively communicated about beneficial HF21 features with the community, which resulted in a major improvement getting added: the EIP. It looks like another beneficial feature will be added as well; an increase of the per block limit on custom JSON ops. This will be great for developers.
  • Our hard work integrating with Flipside finally paid off and we received an A score.
  • While we knew we had left “survive mode,” it really feels like we're truly thriving because we’re finally working on innovative software like Communities and Smart Media Tokens (as opposed to cost cutting solutions).
  • Because SMT issues/stories were sub-par (also mentioned in the next section) we were able to turn lemons into lemonade by having a super productive SMT planning session which resulted in significant time savings.

What didn't go so well?

  • The SMT issues/stories that we had developed in the past were sub-par, which required another SMT planning session (though that ultimately delivered significant benefits)
  • We had to deploy the testnet multiple times due to various issues. Our knowledge involving the orchestration of testnets could have been better.
  • We can do more to engage with community developers on github (although we did engage more than in the past).
  • We often get bogged down due to limited resources. Engineers are finding themselves too frequently called away from important projects to address relatively minor issues.
  • We can still communicate better about additions to Hardforks. We want discussions of new additions to be more open and public so that users understand what is being proposed, who is proposing it, and the costs (in terms of risk and delays) of those proposed changes.

There were no escalations for this retro.

Your Feedback

We aren’t always a very good judge of what the community feels is needed which is why we’d love for you to leave your feedback in the comments section below. What do you think went well over the last month and what do you think we need to improve the most?

The Steemit Team

Sort:  

We can still communicate better about additions to Hardforks. We want discussions of new additions to be more open and public so that users understand what is being proposed, who is proposing it, and the costs (in terms of risk and delays) of those proposed changes.

While I think it's good to involve a community when it comes to discussing protocol changes, I also think that maybe the community at-large doesn't always understand the rationale for changes and maybe doesn't even care for the most part.

I think it would be more appropriate to at least consider that many of the protocol changes and discussions about them ought to be focused more towards investors and developers. These are the people who would likely care most about economic adjustments and underlying tech improvements - and also know the most about code changes and how they could influence behavior.

The average end user - on the other hand - who comes here to post their content and/or to read and comment on the work from other users is not likely to be technically inclined or heavily invested financially. (And I think our observations of behavior and the public discourse that has happened over the last three years pretty much confirms this.) So it makes little sense to try to convince these users that all of these economic and technical changes are needed or to go into great detail (especially using technical terminology) in order to make them feel involved. Because the reality is, most of them likely still don't. It makes even less sense to solicit their input on such technical matters just for the sake of "discussing."

Instead, the end user needs to know more about how proposed changes may affect their social experience. Yes, rewards distribution is important to understand, but it's the social impact that many people continually feel...such as fewer views, votes, engagement, etc.

So we have two different "classes" of users in that regard. It's important to understand the needs of both and it's equally important to understand how to communicate to them in a way that each group can understand and offer feedback.

As an example:

HF21 has protocols that will considerably impact both the economic and social behaviors on the platform. The official communications from Steemitblog and many of our witnesses about the hard fork don't adequately detail how things will potentially or even likely improve for the average social media end user. There have been details and some rationale for the EIP, but it still seems that the average end user is confused.

The 50/50 protocol, the non-linear curve, the reverse auction time...these are the "inside baseball" details that really have no impact on attracting curious bloggers looking to join Steem or the common "quality" content creator that wants to continue sharing their content and engaging with the community. Throwing out all of the technical terms and talking about "40% reductions" in content creator payouts just scares a lot of these people unnecessarily when the fact of the matter is, if the EIP works as expected, most of these "quality" content creators will probably see insignificant changes to their overall rewards...some will probably even see their payouts improve.

So sure, we can inform the community about the proposals, their impact, and discussions taking place, but I don't think that these always need to be the focus for the social media users. They mostly just want to share and view content. And as long as the platform is easy to use and has common social media functionality (which most of our interfaces are not and do not have), people will use it.

End users shouldn't focus so much on the technical details. They should be working on improving their craft and networking, just like they would be doing on any other social media site. It's our job as "leaders" on Steem to remind them of that and make the platform more suitable for achieving those ends...which requires a lot of improvement from interface owners and the Steemit, Inc. communication team.

I agree with the fact that technical stuff need to be explained in a way that everybody understand. But I also think that if we compare the level of communications, posts, discusions and chats that haven been produced during this HF21 preparation, we are in a much better position that we were ever before. @steemitblog have improved drastically and there are also many really good and understable posts written about the matter. By the way, I specially liked yours from a couple of days ago and have been using your way of telling the story to explain (in spanish) a lot of things to not tech users wanting to understand.

This is really great feedback. Thanks!

I am sure there are many technicallly minded users who are not privy to discussions that take place but do appreciate a technical explanation of the changes - even if they are not about to be changed.

I've asked questions on a previous post from @steemitblog which haven't been answered. I was first thanked by @steemit on Twitter for asking those questions but this thank you tweet has since been removed.

I'm reposting some of those questions here for visibility as they are of interest for the community and in the hope that these questions can be answered leaving everyone better off.

  1. Where did Steemit explain why linear rewards weren't flawed before implementing them?
  2. Why did Steemit Inc implement the flawed linear reward curve or never explained why they weren't flawed? (Who took the decision to support them, how was that decision taken?)
  3. Should current investors solely vote for themselves to avoid scammers to receive "free money for nothing" aka free Steem? (loophole which has been admitted by Steemit)
  4. What are the benefits, if any, for investors to not solely vote for themselves?

Many of those questions came about because in February 2017 Steemit Inc re-highlithed why linear rewards are fundamentally flawed, yet they never retracted their statement before pushing for linear rewards to be implemented.

In a world with honest people who don't vote on themselves to get "free money for nothing", a simple linear curve, aka n would produce a 1 share 1 vote proportional payout. This is the blue line and shows the ideal situation.

Unfortunately, we live in a world where people will attempt to game the system by voting for themselves. If everyone voted for themselves then the result would be simple interest payments and have no economic impact.

To clarify, I was never in favor of a move back to n^2 per say but I have been in favor of a move away from linear reward at least since November 2017 and I now support the new EIP.

Def do love the communication.

I do believe the key issues are not addressed. The reward incentives of steem are broken and dont seem to be addressed and avoided altogether.

Doing things well is not enough we also need to do the right things.

I very much appreciate the improvements that have been made and the progress being shown, great work on that.

One thing I think that is still an obstacle for us as a community is communicating with our user base in a way that is not only understandable to them, but embraces them. I know I’ve mentioned this a few times but after recent frustration brought to me by some of the community I feel like I need to bring it up again.

Much of the community feels ignored, talked over, frustrated and isolated due to most of the communication not really geared to the average end user. Things are confusing, scattered all over and generally are written for individuals who already understand the tech and logistics behind the discussion, which leaves out quite a large group of our user base.

This leaves them confused, and confusion leads to misinformation, fud and an overall low morale. The may not have any interest or desire to know the algorithms of a curve, as an example, they could just get a explanation of what these changes mean for them and how it is beneficial for the longevity of the platform and therefore them.

We have a very diverse community with all different levels of tech knowledge. This is what we want, and we need to embrace that by offering different forms of communicating imo.

We need end users for all these wonderful things we are developing, we need “normies” and we need to communicate to them as well as give them what they need to understand and thrive here.

I truly believe most of the frustration and outrage from the most recent hardfork discussions is due to how it has been communicated and therefore is resulting in low morale.

It is pretty common practice for any company to have some sort of community manager or liaison, as it’s such a simple role that prevents issues as well as ensuring voices are heard. Even block.one is hiring a few as they see the need.

A strong community that is in love with a platform is about the best peer to peer marketing a company can ask for. Everyone becomes walking advertisements.. and I believe the lack of communication to our current user base as well as the future user base we want to attract is holding us back.

That would be my feedback on what needs to be worked on. To me it’s such a small thing that could have a huge impact.

Thanks and keep up with the forward momentum!

Justine

I don't think that steemits user base exists in such solidity to form it self as community. It is too scattered and wide bunch of people. I think "communities update" is going to help make more clear who wants and what here. Currently it depends on post topic whetever steemit inc is going to be bashed or not. It is that those dissatisfied are most eager to leave comments overall.

There was one problem with communication though. The way that the EIP was originally represented could have been better. Rewards curve change was especially one that I think didn't go to well in terms of communication. It should have been mentioned from the beginning that the non-linear part was targeted to very low amounts of steem-power hodling accounts. That caused alot of angst by defenders and opposers of the "EIP".

I understand what you mean and that diversity is pretty special as well. I guess I think of it as we have this “Steem community” followed by all these sub communities that thrive as well.

The communities feature will really take this to a whole other level and help lift and empower communities of all types while giving them the tools they need to thrive. I really look forward to that and do think many issues will be solved when they are implemented.

Completely agree with you about the EIP aspects, many things were/are just not understood which leads to unneeded backlash and frustration.

Communication is actually pretty powerful and can have huge impacts on many things we don’t even consider.

I don't consider myself technical but I have understood every piece of communication Steemit, Inc put out. There will always be people who get lost... It's the internet! some people on here don't understand English very well, so that's a big barrier.

But Steemit Inc needs to scale before hiring marketing people and people who can communicate in different languages. All the community seems to want is a company that communicates with 0 engineer behind. Think practically. Best ROI for Steemit, Inc now is to hire devs.

They have devs, some pretty amazing ones.. and the upcoming SPS will help pay for project work of community devs as well. I don’t believe I implied they should be a company without engineers, as I’m pretty aware where we are.

The fact of the matter is that how things are being communicated with the user base as well as the outside world is holding us back.

I’m not suggesting they fire all their devs to hire a translator, that would be some crazy internet antics...

I was just giving feedback on what I feel they could still improve on, which is what the post asked for.

Thanks for your reply!

Here is my feedback. If you are truly out of survival mode, you should be working on getting STEEM listed on exchanges to support the success of SMTs, but there is no mention of this. What are you doing to address the lack of STEEM on exchanges. When exchanges don't have the node set up to support STEEM and SBD, they will not be able to support SMTs.

More exchanges are nice but it's not really that important to have to be repeated every time I see your comments. Especially not considering we're on the literally biggest exchange and there's barely any volume there.

Very much agree.
They should focus on blockchain things and on getting better on communication... and they're making progress on both.

WOW somebody finally called him out on this :)

Its like listening to a broken record honestly.

It is refreshing to see continued, effective, and regular progress updates. Please keep this up. It is helpful for users, developers and passive investors alike.

Thanks for these communication posts.

I'd really love a graphical roadmap with some optimistic dates - can anyone direct me to a website or something with this? or can I suggest it should be on the 'to do' list.

Our knowledge involving the orchestration of testnets could have been better.

I know or think I know HF21 has been passed off to the witnesses for evaluation. That was what? 3 weeks ago. I took a look at a few discord chat rooms, and a couple were mentioning issues with the testnet. Is it resolved, are the witnesses now testing, and is there ever going to be an update on testing progress.

It seems to me there were issues during HF20 with the testnet, yes we eventually got explanations, that does not matter and is not really the point of my comment.

As has been mentioned in a couple of comments, there are varied users and users abilities to understand the technical side of things. Right now as far as I know there has been no testing on HF21, no word that is easily found on the status of HF21, just still a lot of talk about what is in it, what is great about it, and no talk about the progress of the testing.

Three weeks of testing is not an indication that everything is hunky dory to a non technical individual. It is an indication that something is amiss, at least to me.

what do you think we need to improve the most?

I don't really expect a response and have seen countless people ask important questions or make good points and never get answered, but... Might as well try at least once.

I just read a thread from a friend of mine on here, who said steemit is not working for him. His reasons were primarily low interaction on his posts, and this is something I've noticed as well. Not just on my own page with over 3500 followers but on most of the pages I see, there's very few comments and human interaction that I can see.

I think the most important thing, or at least one of the most important things that could be done is to fix the algorithms.

I'm not sure what they are set at but they feel similar to what Facebook uses which is probably 1% or less according to the research I've seen. Even with such a low percentage of feeds that an individuals content is shown to Facebook still is providing much more human interaction because they have many more users, however... Back when Facebook had its algorithms at around 17% it was a lot of fun posting over there!

After they started reducing the percentages, it just got worse and worse to the point where many people quit because they were basically talking to themselves and had very little interaction, even on huge pages with many thousands of followers.

It seems clear they did this to me because they were trying to encourage people to pay for advertising to get their posts seen more, but it truly destroyed the experience for many people.

I see similar here on steemit, most pages I look at people are basically talking to themselves and get almost zero human interaction. I think many people wouldn't be nearly as concerned about the value of steem in terms of it's crypto coin if they at the very least had people reading and commenting on their pages, yet combine the low value and the "ghost town" atmosphere and I can understand why so many people quit and go back to Facebook or where ever, at least people interact with them on other social media sites.

If the algorithms were put up to around 17% like Facebook had when it was still fun to use, then I bet that would make a massive difference here. Otherwise, I see most of the smaller less popular people getting tired of talking to themselves and eventually quitting cause almost no one interacts with their posts.

This is social media right? Then the idea should be to get people socializing and not feeling like they are alone and talking to themselves.

Another thing you could look at is all the dead accounts, that may be another reason why peoples posts aren't being seen as much because out of the 1% or however many people a person's post gets sent to, many of them are probably dead accounts owned by people who quit, if perhaps you could make it so that new posts aren't sent to dead accounts and only active ones that might help a lot as well, not sure if that's possible or not... But to me it seems like the main problem is just that most of the people I see are basically talking to themselves, even many bigger accounts get very few comments.

I know there is another proposal about rewards and such that may or may not help in regards to more human interaction and comments, but I think the algorithms can play a major role since I saw how fun Facebook used to be before they gimped theirs and how downhill it went and how many pages and people's experiences it ruined after they reduced the percentage that people's posts get sent to other people's feeds. That one thing could make such a huge difference in my opinion. Thanks for reading if you did.

This post has been super informative and elucidating in terms of how the team is doing, what issues/challenges you guys are identifying, how you assess not just your own work but the context you're operating within. Thank you for this.

One thing that is not very clear to me is why there isn't much more bugfixing and development of features done by other developers (outside of Steemit, Inc.). I see that you have identified that:

We often get bogged down due to limited resources. Engineers are finding themselves too frequently called away from important projects to address relatively minor issues.

Are there things you think can be done to help with this? One potential strategy might be to announce (with the help of Utopian or directly through @steemitblog) that there is this opportunity for help with developing certain features or fixing certain bugs. Or improving the reliability of certain systems that often cause problems.

Taking it further, could some of the work for Communities be announced as open to developers/designers to do it as per the provided specs?

Coin Marketplace

STEEM 0.18
TRX 0.16
JST 0.030
BTC 62852.17
ETH 2463.87
USDT 1.00
SBD 2.64