chaosbot.org sysadmin lacks supervisor skills #554
Comments
|
I don't think chaosbot.org has a chance to succeed, unfortunately. |
|
@PlasmaPower succeed at what |
|
Attracting contributors. |
|
If you're planning to host your own independent public instance, I would suggest using Docker. However, it's likely that the future of ChaosBot will stay in this repo, according to #553. |
|
@PlasmaPower it surely does. |
well, alright, why not write a PR outlining what changes you feel need to be made---in terms of code, domain, sysadmin, process---in order to position the chaosbot so that it can take maximum advantage of the contributor network that github provides access to etc... |
|
@dbpokorny I don't think @PlasmaPower meant that in a harsh way. I think whoever is left that wants to work on In any case, best of luck! |
|
@dbpokorny I was referring to making a name-alike In any case, personally I feel like my efforts would be better spent trying to get More power to you for keeping it going on your fork and the |
|
I think if there is a V 2.0 coming up, it's best not to try and split the community by encouraging people to a specific fork of V 1.0. Better to move people towards 2.0 I think... |
|
Lol ok dude 👍 |
|
Due to this PR I could DMCA the repository (and so could other contributors), as the MIT licence requires the licence to be included when the program is distributed. You should probably revert that. |
|
@PlasmaPower no, you should probably DMCA me |
|
😕 |
|
Is this part of the Chaosbot experiment ? I dont want to be salty, but who are you ? And you are asking the 81 contributors of the initial project to move toward yours ? OOS, is not a fantasy land where you can pick things and claim they are yours. |
There is no working bot at dbpokorny/chaos. The webserver runs, but not the merge-and-restart feature (because supervisor is not configured). So, even if I were to ask, "please use this repository", it isn't possible yet because the bot is not working. |
|
This is a coup? |
|
I agree with @rudehn above. Though I don't think it will be a big issue because what you have is basically a non-working fork of Chaos v1, while v2 will be an official (and original) repository. |
@eamanu it's an abdication. Look at chaosthebot.com and look at chaosbot.org. The code running at chaosthebot.com is not, as far as I know, related to chaosbot/chaos, nor has @amoffat announced any plans to continue running the bot at chaosthebot.com, nor has he indicated that he wishes to continue in his present role as sysadmin. It isn't a seizure of power because @amoffat had the power to define the reference implementation of the bot, and he held this power by running it at chaosthebot.com. He gave up that power willingly and made a public announcement indicating he would not take that power up again in the future. We have reason to believe that chaosthebot.com will not reflect the state of the code in this repo in the future, and @amoffat didn't go out of his way to hint or suggest or otherwise indicate that we should wait for chaosthebot.com to run the code in chaosbot/chaos. @andrewda As far as I know, the code in this repository does not work, so saying "dbpokorny/chaos is a non-working fork" implies that the non-working status doesn't apply to the code in this repository. There is currently one fork and one original of the bot, the fork I made that omits the auto-erase-git-history-on-crash behavior @mark-i-m contributed, and the code in this repo. Neither works in the sense of being able to perform all of the critical operations as part of a connected whole: vote counting, merging PRs, and restarting. What I've done is to arrogate veto power by saying, "if I were running this project, I would veto this change". It isn't a mutiny because @amoffat stopped running the project before I introduced the scenario, "what would happen if I had control of the repo?" If @amoffat had not abdicated, I think he would have been justified in interpreting a remark like "what would happen if I had control...?" as an attack on his leadership. I certainly would be uncomfortable if I were in a leadership position and someone questioned my leadership in this way. He abdicated and he asked for a postmortem analysis...my remarks about a "what-if" scenario are not intended as an attack on his leadership. The kernel of my postmortem analysis is that the auto-erase-git-history-on-crash behavior is worthy of a veto. So I submitted my postmortem analysis in the form of a diagnosis of what ails the project, and in my opinion, what ails the project is the auto-erase-git-history-on-crash behavior. |
|
I feel like you're the pirate organizing the mutiny that has nobody backing them up. |
|
Note before I begin - I don't intend this to be inflammatory at all :)
As far as I can tell, @amoffat hasn't abdicated. As you may not have noticed, the Chaos repo has now been mirrored to a new organisation at botchaos/Chaos
Side note: nginx, which is the "code" you describe as running at chaosthebot.com, is configured by two config files in chaosbot/Chaos, as it had been doing transparently for quite some time even before the experiment was officially declared dead in #553.
The "public announcement" you mention is, presumably, #553. I don't see where he announced that he would not be running the experiment in the future - at the end of the day, @amoffat is another member of the community just like us, so I wouldn't interpret "up to the community" as his abdication in any way. In addition, this quote:
doesn't seem to be helping your case in terms of leading the next ChaosBot.
Another side note and a slight jump: as far as I can tell, this was implemented for ease of maintenance, and now you're asking for help regarding maintenance...?
I certainly don't have any reason to believe that chaosthebot.com will not reflect the state of the code in this repo in the future, and "not going out of your way" doesn't generally imply that you are fully accepting of the pirate organizing the mutiny that has... wait, this seems familiar...
/essay |
😭 😢 💔 😳 |
|
Thanks @md678685 @dbpokorny, I appreciate your energy for the project, and I wish you all the luck in your fork. However, using this project's issue tracker as advertisement is completely inappropriate so I'll ask directly that you please stop with that line of contribution. Making the case that I "abdicated" is completely unfounded, so let me clear the air on that right now: my role in this project is one of guidance to facilitate the decisions that the community makes and remove stumbling blocks from the path. It is not a dictatorship from which I can abdicate. If you have a different vision on how to lead your fork, again, I wish you all the luck. Right now chaosbot is at a reflecting stage, and it's ok to sit for a moment and collect our thoughts on the direction it can take moving forwards. V1 experienced a massive growth of contributions and input from everyone in a very short period of time, so much so that the weaknesses in the design made themselves very painful. Taking a step back and thinking through how to proceed is the best course of action. There's no rush :) |
Who chooses the chaosthebot.com sysadmin? |
|
@dbpokorny the community, by choosing which version of chaosbot they contribute to. |
|
@PlasmaPower I don't understand, can you explain that in more detail? |
|
@dbpokorny It's the same way any open source project works. At the beginning, the creator has control of the project. If people get fed up, someone well known in the community will fork it and the community will move to the new fork, leaving the old maintainer behind. In this case though, most people don't have a problem with @amoffat as the leader/sysadmin. There's no reason to move. However, if someone creates v2, they will likely be the next leader of the project by virtue of creating it. |
|
@PlasmaPower I asked, "who chooses the chaosthebot.com sysadmin?" You remarked that you believe "most people don't have a problem with @amoffat as the leader/sysadmin" and furthermore that "any open source project works" in some way you describe in vague terms with brief remarks on a succession plan. If this isn't clear enough already: you answered a question I didn't ask. |
|
|
@Redmega Whoever chooses what the nameservers resolve "chaosthebot.com" to has a choice of sysadmin as well. Do you agree? Perhaps I should have asked, "who picks the host, and who picks the host's sysadmin?" These are actually two separate, but dependent questions. In order to know who the host's sysadmin is, the identity of the host must be known. |
|
I thought you meant "who chooses the chaosbot sysadmin". Since you asked "who chooses the chaosthebot.com sysadmin", the answer is your DNS provider. |
|
Yep pretty much @dbpokorny
There is never one true host, and therefore never a true host sysadmin. You can't pick any one fork of chaosbot and say it's the "true" chaosbot fork. |
|
@dbpokorny Let me put @PlasmaPower 's words more simply: chaosthebot.com is not the "one true ChaosBot". It is just one domain that resolves to one instance of the bot, and so the bot that the domain points to is decided by whoever manages the DNS. At the present, this is @amoffat, who has chosen to point it to the instance he maintains. However, chaosthebot.com operating on chaosbot/Chaos has been the ChaosBot project accepted by the majority of the community, and therefore it is the dominant ChaosBot project. If the majority of the community decides they don't like @amoffat, they can fork it and decide to move to "entropybot.com" at Entropy/EntropyBot, for example, and then that would become the dominant bot in the community. Even after this, there is still no "one true ChaosBot". So, to answer your question by reiterating @Redmega, who reiterated @PlasmaPower, who reiterated common sense:
|
|
To enforce @md678685 talk, each instance of Chaosbot is a valid one, A lot of contributor have one running somewhere. If you think @amoffat is a despot, it could really be a bad one, cause I never seen him inpose onself in any debate nor the most important one. edit : I hoped we could choose to close this one democraticaly, but our child is not awake yet 😕 |
|
Also, you removed the license that all contributors contributed under, and now any future contributors you're hoping to attract may be discouraged by the fact that they are contributing to an implicitly-"all rights reserved" project. |
|
Does anyone believe the supervisor interface is an important part of the project? I'm starting to think that it is doing more harm than good. If you think it's important, I want to know why. |
|
Yeah. If your bot crashes, Surpervisor is what restarts it. |
|
I think @dbpokorny thought supervisor was for our despot leader to finagle with the bot on the server without our consent! Down with the botriarchy! |
...or accepts a PR and tries to restart, which is core functionality for the bot. |
|
That said, I would rather have a bot which brings up its new self and then kills itself -- without the need for supervisor. |
|
chaosbot.org now runs the code at the repo chaosbotgroup/chaos and can restart successfully after merging a PR. |
Hi,
I am the sysadmin for chaosbot.org and there is a running bot, which runs fine until it has to restart after merging a PR, at which point it does some process and then attempts to invoke supervisorctl, which fails. I wasn't using supervisor to run the bot in the first place (because I don't know how to) and running it with "python3 chaos.py" causes the supervisor error after merging a PR.
I can
Learn how to run the bot in supervisor (it is installed, but it isn't clear why the bot can't invoke it correctly after merging a PR)
Remove the code that depends on supervisor
Until 1 or 2 happens, I will watch powerlessly while the bot stumbles along, having to restart it manually after each PR merge, which crashes it.
The text was updated successfully, but these errors were encountered: