I think itās definitely wrong to apply more preassure on the team. Being a product manager myself I know firsthand how hard and painful it sometimes can be to manage limited resources and push the team to the limit.
Obviosly I can only interpret this from the outside but Iāll just give it a shot and try to analyze a concrete example in the hope of contributing some productive feedback.
Letās try to take an example. Most of the people here know by now, that DC-1 is a huge thing; the main roadblocker for Alpha 3 you could say. I have the feeling, some of the above frustration came from the excitement about detailed milestones (and connected tasks) regarding DataChains (part 1).
I really love the fact, that MaidSafe is publishing their JIRA. However, I sometimes get the feeling, that itās sometimes not used for every task or that there is a another ābehind the scenesā task system. For instance, last week or so I asked why the number of ticked-off JIRA tasks hasnāt changed since 3 weeks or so. @dirvine told me, that they changed plans for the routing team and now experiment with the so called āgossipā protocol.
For me thatās absolutely cool and as a developer I can also see the benefits from such an experiment. However, when I search JIRA again today I see that the DC-1 DONE count is still unchanged for 4 weeks now and that there also isnāt a single task regarding āgossipā: https://maidsafe.atlassian.net/issues/?jql=text%20~%20"gossip"
Please, if there is a regarding task, just with a different name, forgive me. If not, I would be really interested in how JIRA is currently being used or how task management / prioritization is currently done.
From an outside perspective I would suggest to put everything which has been decided internally on JIRA and mark it correctly (which module, who works on it, etc.). If gossip is relevant for DC-1 then tag it like that. If itās not relevant and just an (important) āside projectā thatās also fine. Then I would mention in the weekly update, that work on DC-1 tasks is currently postponed as long as gossip tasks are finished.
Again, for me itās compeletely understandable, that requirements/tasks change (especially in such a complex project). But then the plan (JIRA, milestones, tasks, etc.) need to be updated too and each major change schould be communicated.
In the above example then everyone could see that new tasks appeared (āgossip integrationā) and can track the progress and understands how many resources are blocked due to this change. I think this could take out a lot of frustration potential.
Would be cool to hear some thoughts from you about this or if Iām missing something.
Good point, sometimes though we just do need to try some things and quickly. There is an issue we are working round right now for instance and for various non technical reasons we need to do it off github. Its a test of something that could help us a lot. We will of course say much more as we usually do, but you are correct. There are some libs, routing a lot, that look to go quiet when we try some things out, occasionally in parallel (LEAN) ways. Swarm is such an issue as well as intra section ordering of consensus events, kinda ordering consensus of group consensus events. Its not easy to explain and we have a few folk on that, but it is an example of where we have a large number of hangouts, test a lot of ideas and have a lot of engineers digging deeply for the best answers.
It happens in other libs as well, even with things like SOLID integrations or some simulations (i.e. a gossip sim was done about 6 months before we looked at push/pull gossip to help swarm).
It should not and we really hope will become very rare in the future, but sometimes its required to just huddle together and brainstorm like mad. So routing just now (well public-ally) you will see what looks like a pause, then perhaps there will be a mega commit, all going well, or we continue with a slightly slower thing with all the conditions in routing. We hope we can do better there, but as I say itās sometimes just not every task that is so public until we work stuff out. We never though hide stuff that works
Thank you to both you and Nick for caring enough to take the time out for people when they have doubts, your humanity is the reason you guys are the right people for this job.
Sending you guys a virtual badger (kinda appropriate given this thread I reckon).
It was pretty funny and I sense more will be coming.
Thatās why we should start a āWeekly snap threadā for āI-trust-nobody-any-moreās!ā, āwhatās-been-taking-them-so-longās?!ā and āIāve-got-money-in-this-you-knowās?ā
For therapeutic purposes. It could happen to any one of us at any moment. āI-lost-all-my-hopes-in-this-projectās!ā and āSubstratum-is-so-far-aheadāsā would land there pretty often too. Same about āMore-pressure-would-do-them-good-and-they-need-a-good-ole-kick-in-the-buttās!ā
The thing is, when it gets done, it will be hilarious to look back and read through these posts
I just created a poll where we can vote when we think beta will be released
https://forum.autonomi.community/t/poll-when-will-beta-be-released/22405
Just to check how optimistic/pessimistic everybody is
Voted. Iād like to be proven a pessimist in the nearest future possible, butā¦
hmhmmm you couldāve made it a non-public poll maybe some people donāt want to share what they voted (e.g. our devs )
ps: and i donāt see Q2 2018
Itās all open source
Not a bad idea. There could even be copy and paste-able FUD phrases for quick venting, kind of like a cathartic stress ball.
What in the name of all thatās holy is THAT??
Smirnoff + Guinness
Is it good for you?..
I have a vague recollection of drinking vodka and Guinness in my younger days. Vague because most of it was spent crawling around on the floor after my legs refused to behave.
How dare you question the health benefits of Guinness
My ol lady and I have been making an Irish stew with Guinness lately and itās really good, especially cause I get the left over Guinness.
Well it says here that itās good for you, but you have to be careful who you trust these days.
Happy to offer weekly or biweekly FUD therapy sessions.
I accept payment in MAID.
Thanks David for the detailed response. Now it makes perfectly sense. I can just try to grasp how hard it must be to be so transparent about everything so some āprivateā hacking from time to time seems just fair and productive.
Maybe a side sentence in the update about such excourses would also be enough. āOk guys, weāll delay X because we try out Y which should bring in benefit Z.ā
However, thanks for your effort and keep rocking.
Some time back, in this thread, @draw posted a YouTube video about a fascinating game called Star Citizen.
Reviving the subject because their development process and challenges parallel ours in ways that could be useful.
In common:
- Crowd-funded by passionate community
- Feverish anticipation for delivering an unprecedented product
- Product composed of many pieces that must be mostly developed to completion to work together
- Community frustration and disappointment over missed milestones and delays
Take a look at some of these for a taste of the pressure, anticipation, and frustration:
- Star Citizen Delays as a Content Creator
- Star Citizen single-player delayed indefinitely
- 8 months delay is disingenuous
- The fight over Star Citizenās production delay is getting dirty
- Delayed Again: āStar Citizenā Alpha 3 Now Wonāt Hit Until September
- Inside the Troubled Development of Star Citizen
- estimate ~400 days to complete all JIRA tasks
The amount of testing that required for the smallest of tasks is overwhelming.
When milestones are missed, some of the louder voices in the community claim malice or neglect on behalf of CGI, the company behind Star Citizen. CGI has replied by addressing the great divide between community and developers over development process.
Star Citizen has responded by creating several series of weekly shows on YouTube:
-
Bugsmashers
- recap of particularly interesting solved bugs for the week and summary of JIRA -
Around the verse
- highlights how a particular aspect of the game is developed -
The Loremaker's Guide to the Galaxy
- each episode goes deeper into the culture, lore, and history of a particular system in the galaxy -
The Next Great Starship
- series competition where amateur developers compete to develop a ship or weapon that will be included in the game -
Calling All Devs
- Live Q&A with various developers to learn about how components are developed and how they work
Thereās a lot more than that but, in general, each of these pieces of content educate the community about the something thatās very hard to convey: the development process.
One major difference between our projects is that Star Citizen has much greater and on-going funding, which allows them to pay for all the overhead of content creation and community development.
With all that said, CGI has a roadmap, milestones, and regular informative content, but Iām unsure of how to measure the effectiveness.
Misses and delays are inevitable but would people feel better about them with greater detail of development process?