14

Five years ago, the issue was raised as to whether questions related to problems with beta software are on-topic. The gist of the logic for why they are not is in suggested wording for a custom close reason:

Questions about bugs in pre-release/beta software or prototype hardware are off-topic. Bugs in pre-release products are common and even expected, and are often fixed before the final version is released — at which point the question becomes obsolete. Consider reporting these types of problems directly to the product manufacturer.

The same logic applies to capabilities that haven't been implemented yet, finding needed drivers or libraries, and similar questions.

We recently had a number of questions posted about ReactOS, which the developers identify as being at the alpha stage. The questions were about certain applications not yet being supported, and various error messages encountered when trying to run certain applications.

Before I vote to close a large number of questions from a single user, I wanted to get community input, especially since the closure reason isn't an official one. I don't want this to appear to be a vendetta.

The questions are:

The first three seem clearly off-topic. The last two, seem to be in a grey area. Any input on how the community views these?

4
  • I'm looking for a way out though. Why can't I be just like the other user of ReactOS who asked and self-answered a question in 2014 and wasn't harangue and helped out ~100 people a year (by view count) superuser.com/questions/815281/…? Why is that his question wasn't the source of inquisition and mine was? Why would you close that question now? It's not my intention to make the ReactOS community pay the price of our attempts to make me unwelcome. Aug 13, 2018 at 2:42
  • 1
    @EvanCarroll, I posted an answer to try to clarify some of your perceptions about the process here, get into more detail about topicality, and address some of your points.
    – fixer1234
    Aug 13, 2018 at 6:45
  • 2
    @JakeGould, I understand where you're coming from. But, the question is about closing questions on pre-release software (site policy issue) and proposed action (historical basis). The purpose isn't to to convince Evan of anything or to engage in discussion with him. That's a tangent and an outcome. The fact that he doesn't agree, and the resulting discussion was unproductive, shouldn't be a basis to close the question itself. That won't stop discussion, it will just block any new answers. The recent one by music2myear provides valuable insight. (cont'd)
    – fixer1234
    Aug 14, 2018 at 2:20
  • 3
    Actually, some of the points in Evans answer deserve consideration, and should be clarified in any guidance that gets created, even if we disagree with his conclusion. Perhaps it would be better to just move the comment threads to chat and disengage from discussion with Evan on this matter since that is unproductive.
    – fixer1234
    Aug 14, 2018 at 2:20

4 Answers 4

7

Well, ReactOS is perpetually in alpha and has been as long as I can remember it.

Darned thing is brokenish, and a good many of these are bugs that should be reported to ReactOS developers rather than asked here. While I did argue against it, there is no roadmap for when ReactOS would be in a usable state unlike Windows 8.1.

I'd say that the logic is still valid, and the close votes seem like a good idea.

As for the last one, I don't remember the developer in question but it also recommends asking their forums. This is a good idea.

So, feel free to vote to close these, and we'll follow up with our closehammers shortly.

4
  • 3
    Thanks. You saved me some decision-making on the last two questions, and probably avoided some arguments and bad feelings.
    – fixer1234
    Aug 12, 2018 at 6:46
  • 4
    Nope. I acted on pre-existing precedent which OP had brought up. I looked up previous meta questions to see what was done before. I also looked at the value of the posts being here - and whether they would be useful in a wider context, especially with the context of ReactOS's development roadmap. As for "logic" - the question links another, older question. As with many things on SE, we are relying on precident on meta and elsewhere.
    – Journeyman Geek Mod
    Aug 13, 2018 at 1:42
  • 2
    @JourneymanGeek As for the "should be reported to ROS devs rather than asked here" statement.... I would say if the dev is not responsive and the software has not changed for years (as you said), then to someone like me, this would be a "classic" example why I want to come to SuperUser to ask questions to a broader audience, share information, look for content for an answer out there already, etc. Perhaps someone using it for x years has a trick they will share to workaround the so called bug/issue whatever it may be—it hasn't changed for years... and not welcomed here... Wow!!! Aug 13, 2018 at 2:02
  • 7
    Developers are super responsive. Its just a long running project with a history of unfortunate setbacks.
    – Journeyman Geek Mod
    Aug 13, 2018 at 2:08
10

I'll post an answer because there is too much to cover in comments.

Interpersonal

Let me start with the interpersonal issues, since the bulk of the questions were authored by Evan Carroll and there is an appearance, especially to him, that personal animosity was behind this action.

This was not personal, and I don't dislike Evan, as he suggested in his posting here. I don't recall even having any interaction with Evan prior to the TempleOS discussions. A significant number of users opposed the TempleOS threads and discussions got heated. It is natural for someone to get defensive when seemingly attacked from all sides like what transpired then. And I suspect that the TempleOS experience flavored Evan's perception of this activity regarding ReactOS.

Given the potential for this to turn into a repeat, I took the route of posting this question. There was a large number of threads, almost all his, and I didn't want the appearance that this was personal. I wanted to confirm my own perception that the threads in question were off-topic via community input before doing anything. And I posted exactly that in the question to be upfront about it, and to avoid readers interpreting it as bait to have another go at Evan and his questions.

There wasn't a conversation in chat, just a single statement that I posted, which received no response. Perhaps it was poor judgement to post that at all. It was a matter of venting frustration over what looked like the start of another run of threads that I think are a poor fit for the site. The TempleOS questions blew up and then settled down (and I pretty much stayed out of that fray except for what I thought were a couple of constructive comments). Then recently, there were four similar ReactOS questions in rapid succession that looked like the start of another run, and I vented with that statement on Root Access. Not my finest hour and I apologize for that. To be clear, all of this activity and the issues I have relate to the posts themselves, not Evan.

I'll try to expand on the issue of topicality and address some of Evan's points.

Pre-Release Software

As a preface, I couldn't find any Meta discussion on the topicality of developmental software. I actually composed a question to discuss whether it should be off-topic and laid out the arguments pro and con.

When I was done, the system picked up on some keyword in the post and found several threads where pre-release software was discussed. That's how I found what I posted in this question. The issue had already been decided years ago and looked clearly off-topic, along with a precise rationale based on existing site standards.

So I rewrote the question. Based on the prior Meta discussion, I added the proposal to close the listed questions. That was not part of my original draft.

Closure rationale

Not every question concerning pre-release software is off-topic. What is off-topic is questions related to problems caused by the developmental nature of the software. That is, issues such as:

  • bugs
  • capabilities that are not yet fully implemented
  • infrastructure (drivers, DLLs, and the like) that have not yet been provided for

These are expected holes and problems that exist only because the software is still in development. They are things that would be fixed by the time the software is released, so they are transient issues that don't have long-term value on the site.

We have 11 questions tagged ReactOS. Based on a cursory look, I identified three that I thought were clearly off-topic and two I wasn't sure about but thought probably were. The others didn't appear obviously off-topic to me. A couple of others were previously closed. It looks like as of now, a total of 9 of the questions have been deemed off-topic. Clearly, it is not an attack on Evan Carroll's questions.

Evan lists a number of well-known products that are technically pre-release and that are not seen as an issue. Things like Debian Testing and the Windows Previews are pre-release in name-only. Massive numbers of people use, or used, them on a production basis. Clearly, ReactOS is not in the same category as Debian Testing or Windows Preview.

I'm not familiar with many of the packages listed. However, mainstream packages that are industry standards, pre-release in name-only, the version everyone uses, and which are used for production of work output, aren't in the pre-release class that is a problem for the site. Further, they are close enough to "release quality" that they don't suffer from the magnitude of issues that is found in software that is still developmental.

What about the beta versions of the big, popular packages, like the nightly releases of the major browsers, open source office suites, and the like? Those are transient, interim packages. The stable packages are available for production use. Questions not related to development issues should be on topic, but bugs and other development issues should be off-topic.

All of the discussion so far is about beta software. Alpha software is another matter. Alpha software isn't close to ready for prime time. It's unsuitable for use as production software to accomplish the purpose for which it is being developed. There are endless development-related issues that are too numerous to deal with on this site. It isn't an appropriate use of the site to replicate or supplement the developer's support resources.

Pimp Juice IT brought up the fact that ReactOS has no stable release version on the horizon, so questions about it actually would be expected to have a long life here. Aside from the fact that it is subject to major developmental revisions, which would limit the time value of questions, it isn't production software. Nobody relies on it to get their work done, and nobody has "real, practical" problems they need to solve with it to get their work done.

It is shared in its current condition for the purpose of alpha testing. People who want to dabble with it are essentially part of the alpha testing. That's part of software development, and software development is explicitly off-topic. Even if the people who want to dabble with it don't participate in the developer's alpha testing, what they can do with it is their own alpha testing.

Pimp Juice IT mentions using Super User to share workarounds, etc., among the people who want to dabble with it. The fact is that the developers already have a mechanism for reporting bugs, sharing workarounds, etc. Most of Evan's questions involved rediscovering bugs or issues that had already been reported, and then reposting the workarounds here.

Regarding other exceptions Evan cited that have unclosed questions, the system doesn't catch everything. Aside from the packages I mentioned earlier, it looks like there is a small number of questions for the others. Just because they're pre-release doesn't mean they're off topic. If someone wants to go through the questions and make a determination, more power to them. The fact that some slipped through doesn't make them on-topic if they're not.

I agree with Evan that the policy on pre-release software should be incorporated into the site guidance. Regardless of whether we agree or disagree on the value of Evan's threads, he spent substantial time preparing a lot of questions and answers in good faith. "Ignorance of the law is no excuse", but there is still some unfairness to having a significant chunk of effort closed based on policies that should have been readily accessible but were buried.

1
  • Comments are not for extended discussion; this conversation has been moved to chat.
    – DavidPostill Mod
    Aug 14, 2018 at 17:30
5

One of the main goals of SU is to be a compendium, an encyclopedia, of questions and answers.

Some questions should be put elsewhere because of the state of the technology they are about. This is first because the state of development of that technology is such that...

  • bugs are very likely to occur
  • there isn't likely to be a solution to them yet available in the broader community
  • and those working on the technology should be made aware of the bugs so they can be fixed and be made into not-bugs

Therefore, the path to the solution is, and should be, through the developers, not here.

The fact that a given piece of alpha- or beta-tech may be perpetually alpha- or -beta does not change that measurement.

However, I believe the fact that an answer exists for a question, even if the question is otherwise very much about alpha- or beta-tech and the devs should be made aware of the issue, CAN change our perspective on it, especially if the question is otherwise intelligent, well-researched, honest, etc.

With the specific questions in ... question ... here, I'd say it's possible they be left open because they are specific questions, clearly and concisely presented, and they have answers. This means that the posts COULD be helpful to others experiencing the same issue. Even if there is only likely to be a small handful of people to ever come upon this problem, that does not mean that that the questions should not be disallowed for this reason.

But, there's another "However": While it is good for knowledge, especially "complete knowledge", where a question has an answer, to exist here if it meets the criteria, there's also the fact that we cannot become the repository for ALL information, and so have to draw the line somewhere. Questions about ReactOS straddle a line:

  • they are complete questions
  • they were asked with the answer already found
  • they are about all-but-abandoned software
  • they are only likely to be helpful to a small number of people
  • they have the flavor of existing to point people to official FAQs and KBs for the technology which should have come up in a Google search for the issue
  • etc

Basically, personally I could go either way on this, but I thought it would be helpful to give the reasons I see that each way has validity within this community.

UPDATE: Edits responding to Evan Carroll's request that I revisit his answer

Regarding bugs and developers hearing about them:

  • I am a long-time It professional and am very familiar with, and remain frustrated by, the propensity for users to not inform me of problems. They assume I'll just know, or that someone else reported. Yes, it can be annoying to get 50 reports of the same problem, but FAR better that than no reports and a load of unexpected frustration when the problem isn't fixed quickly enough. When a developer has an existing bug tracking process, it both falls under the "minimal necessary research" rule here on SU, and an unwritten "do the right thing" rule that we direct people to the correct bug-reporting process.
  • Next, we don't want to build a shadow support system. If the developers have chosen to use a given bug-tracking system, and we step in as a 800lb gorilla and say, even through basic inference, that we will take their bug requests over here, we are not being a good internet neighbor, and there's the very good chance answers found here will be valid for less than the usual amount of time as an alpha or beta product iterates more quickly in its development cycle than a published and finished product does.
9
  • Thank you for at least openly grasping with these issues and trying to extrapolate a policy on it without making it about me. To clarify ReactOS isn't "all-but-abandoned" in fact, there are 150 developers/users right now on IRC (Freenode/#ReactOS). That's 10% the size of Debian. This month it became self-hosting and was written about in The Register (which turned me onto it) theregister.co.uk/2018/07/24/reactos_gets_stability, Packt Hub hub.packtpub.com/… and OSNew osnews.com/story/30605/ReactOS_0_4_9_released Aug 13, 2018 at 21:24
  • I'd like you to revisit my answer and clarify your thoughts bugs and arbitrary designations about releases. Bugs are always very likely to occur, and nothing magical happens after or before beta designation. Developers are always fixing them. I think we should focus on the quality of the question and whether or not it can be answered, and we should strive to be impartial to everything about the software. Some people use buggy software. If they're using it, it's not so buggy that it's useless -- why would anything else matter? Aug 13, 2018 at 21:28
  • @EvanCarroll I've updated my post with comments responding. Aug 13, 2018 at 21:39
  • But isn't that true about all bugs and issues for all software? Why should software designated beta be different? For exactly, if I find a bug in Chrome that is reproducible and interferes with my workflow and I want a workaround is it on topic here if there is or is not an issue on the official tracker? And, isn't the same true about beta software? You should always submit bugs to devs, regardless of the beta or alpha designation. I'm not sure why we should treat one differently than the other. Aug 13, 2018 at 21:43
  • 3
    Excellent answer. One observation: in this case, the questions were from rediscovery of known problems, and the answers were reposting of solutions found at the developers' site; the threads were duplication of content that already existed where it should. The question becomes what is the value of duplicating it here. We often say that this site isn't intended to replace or compete with Google when basic information already exists elsewhere and we're not adding value to it here. Evan's ROS questions don't contribute new knowledge, they just house it at an additional location.
    – fixer1234
    Aug 13, 2018 at 23:03
  • 4
    Nearly all posts here are duplicating information found elsewhere. I don't take that aspect into consideration by itself. Aug 13, 2018 at 23:05
  • @music2myear nor should anyone. There is no rule here that contributions must be unique material and primary sourced. I had a problem. I couldn't find the solution. I submitted the problem here and oftentimes found the solution after I submitted it -- (though sometimes they're self-answered on creation, and that's fine -- don't upvote them if you don't like that). My goal is only to help others and the community of ReactOS users which exists even on this site (like it or not). Aug 14, 2018 at 1:18
  • It may be worth nothing how the objection to my material has now against changed. From the software being "alpha" to the answers themselves existing elsewhere on the internet. At what point will others conclude that Fixer1234 is simply not being nice and targetting me? He did this too with TempleOS, which was merely not "well used" (which also isn't against the rules). Interestingly, there the answers were totally unique to the site. I wrote them all myself. Aug 14, 2018 at 1:19
  • 6
    @EvanCarroll, I'm with JakeGould on this one. The reason hasn't changed, and even if it had, the answers here have mostly noted several reasons. You are seeing bugbears and villains where I (and others who have commented and answered here) don't, and that is making it difficult to engage constructively. The consensus of the community appears to be rather solidly that, in THIS case, you are incorrect. The specific rules which the questions run afoul have been clearly noted and re-noted. Aug 14, 2018 at 16:21
-4

First, it would be my position that all software that is commonly available and used by end-users should be on topic. This specific piece of software has always been declared "alpha" by the creators. That's partly because they're so far from their own goal. The software saw its initial release in 1998. Over 20 years ago. ReactOS isn't "all-but-abandoned" in fact, there are 150 developers/users right now on IRC (Freenode/#ReactOS). That's 10% the size of Debian. The project has hundreds of committers on Github. This month it became self-hosting and was written about in The Register (which turned me onto it), and other publications

This seems like an somewhat eerie distinction because it's entirely subjective.

"Beta" (definition)

I don't believe people here have an understand of what "beta" means. It doesn't mean anything concrete. Most software is actually on a time-release which means a release only means "having support." It is not at all a statement on bugs, capability, or infrastructure. So for example, when you release software every 6-months the functionality you get from the bi-yearly release is only guaranteed to be "supported". It's not magically less-buggy or more capable, future releases often even future beta releases will usually be less buggy, more capable, etc.

Here is some software that is always in beta (unsupported) and never can be released such as

Clearly, none of this is any different than ReactOS. Or, MAME. MAME adopts a different definition of "beta" from the article from Wikipedia,

The developers of MAME do not intend to release a version 1.0 of their emulator program[28] because there will always be more arcade games to emulate and thus the project will never be truly "finished." Accordingly, version 0.99 was simply followed by version 0.100.

We still have 16 questions on MAME

So for MAME the goal is total emulation of Arcade games, and so long as any games can't be emulated the software is in "Beta".

Pre 1.0?

And the list of pre-1.0 software on my computer alone (list generated with dpkg -l | awk '{print $3, '\t', $2}' | grep '^0' | less) includes,

  • camora
  • dash
  • i2p
  • xfburn
  • telnet & Linux NetKit (which includes telnet)
  • ufraw

Not-niceness

Second, this is clearly an action made with prejudice. I was never made aware of even the conversation. After only 1 hour and one submitted answer the user (who doesn't like me) chose the mod's answer for supporting the close-hammer. For those watching this is the conversation in the chat that is/was allowed. Didn't want to put forward the effort to tag me, but did mention the conversation in chat!

enter image description here

This kind of stuff does not make for a "nice" community. This is about a specific user that doesn't like me.

Let's also look at this,

Before I vote to close a large number of questions from a single user, I wanted to get community input, especially since the closure reason isn't an official one. I don't want this to appear to be a vendetta.

There are 11 total questions on ReactOS. Most of which are still not closed. We even have a tag that was created years ago!

Let's actually consider a policy here.

Thanks!

History

enter image description here

The last time this became a coordinated attack, Journeyman had to stand up and end the drama. And, it did end. My contributions even got more upvotes since then! We need the same statement again. This isn't productive and there is no rule against "beta" or "alpha" software. I am actually using and solving problems with ReactOS, and acting in good faith.

3
  • Comments are not for extended discussion; this conversation has been moved to chat.
    – DavidPostill Mod
    Aug 14, 2018 at 17:29
  • 4
    ReactOS is not beta software. It's alpha software. So it's "pre-release" in an entirely different sense and the comparisons aren't valid. But there is no point in trying to discuss this further, it wouldn't be constructive.
    – fixer1234
    Aug 25, 2018 at 6:41
  • 2
    Please do not link to content that is abusive.
    – DavidPostill Mod
    Aug 27, 2018 at 12:35

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .