Forum rules discussion thread

[Sokwe]

This thread is for discussing the forum rules. My intended purpose here is to determine what people expect or want from the rules, how they might be clarified, and in what ways they might be lacking. Please make sure to be extra careful to observe rule 1 in this thread, as rule discussion can be a contentious topic.

[Sokwe]

One issue with the current rules is that there seem to be different interpretations or rule 3(a) or at least a difference of opinion on how it should be enforced. The rule currently reads:

In order to keep the forums organized, please do not post multiple messages consecutively in the same topic. If your message is the most recent one in a topic, simply edit it if you have additional things to add.

I don't think there should be an absolute ban on making two consecutive posts in a single topic (this post would violate the rule in that case). I think the questions we need to answer are

• What is the intended purpose of the rule?
• Are there cases where double-posting should be allowed? If so, what are they?
• How can the rule be clarified to address the first two points?

My view that the rule is primarily intended to prevent users from making several posts consecutively in the same thread over the span of a few days. I also think it's visually more appealing and easy to read to have a single post with several edits than to have a large number of consecutive posts.

There are two cases where multi-posting should definitely be allowed: when a post reaches the character limit, and when a post with attached images reaches the attachment limit. However, other cases would be a matter of opinion. In my opinion, a double-post can be justified in the following cases:

• The second post covers a significantly different subject than the first while still being appropriate for the forum thread. I think it's better for organization and readability if substantially different subjects are in separate posts.
• The post represents a notable achievement that warrants extra attention. Adding a new post will bump the thread and cause people to read the new post, where an edit to a previous post might be easily overlooked.
• A substantial amount of time has passed since the last post in the topic. If a topic has been inactive for a long time, it will have drifted down the topics list and few people will read it or check the latest post for edits. This means potentially interesting results could get completely ignored by the community if they are added in an edit within a generally inactive topic.

It's this last point that I think might be the most contentious. One issue is that I don't know how to define "a substantial amount of time", and I think what qualifies as enough time depends on how interesting the new post is. I do however think there should be a threshold after which a double-post is allowed, regardless of how interesting that post is, assuming of course that it's interesting enough to be posted in the topic at all. Does anyone agree or disagree with this? If you agree, when do you think this threshold is?

Comments on this issue are welcome.

[dvgrn]

I don't think there should be an absolute ban on making two consecutive posts in a single topic (this post would violate the rule in that case). I think the questions we need to answer are

• What is the intended purpose of the rule?
• Are there cases where double-posting should be allowed? If so, what are they?
• How can the rule be clarified to address the first two points?

I definitely agree that all three of the cases you listed can be perfectly valid exceptions to the rule.

I've been explaining my own interpretation of the no-double-posting rule to people for years -- mostly when they start accusing other forum members of "breaking the rules" by double-posting in situations where a double post is the sensible thing to do.

Looks like the most recent round of this was at the end of October 2024. That might be one of the few times that some discussion happened, so that thread might be worth reading.

Reworking a couple of suggestions for wording changes in that thread, I get my suggestion

"Don't double-post without a good reason. If you're not sure you have a good reason, please edit your previous post instead."

and Vort's alternative based on the key idea that one person is unlikely to be able to quickly produce a large amount of useful information on multiple independent subjects that would need separate posts:

"Do not post short consecutive messages frequently in the same thread."

Before that, I think my attempts to give rules of thumb for when the no-double-posts rule is relevant and should be applied, and when it's not and should be ignored ... were probably mostly in private messages to various people (e.g., the quote at the bottom of this post from July 2024 -- moderator-board link) and on Discord (e.g., this summary message and this discussion).

I've mentioned elsewhere recently, that the confusion over rules like the no-double-posting rule is a result of the fact that a fairly substantial number of forum users have a hard time with inferring things from context. To get back to the first question above, "What is the intended purpose of the rule?" ... it seems like that's pretty clear from context. Nathaniel suggested editing a previous post "if you have additional things to add" -- seems like he was probably thinking about more minor related info getting added right away on the same subject. In context, Nathaniel's wording doesn't apply nearly so much to different subjects, notable discoveries, or new posts after a long time.

[confocaloid]

I think my opinion on this can be simplified into avoiding long chains of short posts (even if there are different subjects, and even if the updates are relatively infrequent). "If you don't have a good reason to doublepost in this case, then please don't doublepost."

The second post covers a significantly different subject than the first while still being appropriate for the forum thread. I think it's better for organization and readability if substantially different subjects are in separate posts.

I think there are exceptions. In cases when the alternative choice of making a separate post for every different subject would lead to a long chain of short posts from the same forum member, I think it is better to have the information edited into a single longer post (or at least into fewer longer posts); even if the subjects are different and appropriate for the forum thread.

(For example, in the context of the forum thread viewtopic.php?f=2&t=5338 I personally don't view different SKOP oscillators as substantially different subjects. However, someone else might consider different SKOP oscillators as "different subjects". Still, I think that editing several SKOP oscillators into a single longer post is preferable to having them in several consecutive short posts by the same forum member.)

In cases when people are reluctant to doublepost and choose to edit more information into a single longer post, I believe that choice is usually a good idea.

When a single post is edited over time with consecutive additions, it can be a good idea to mention the dates of additions. (For example, "EDIT by confocaloid (2025-01-05): ..." or similar wording.)

The post represents a notable achievement that warrants extra attention. Adding a new post will bump the thread and cause people to read the new post, where an edit to a previous post might be easily overlooked.

I agree with that. In such cases, the author of the post likely will be able to add sufficient explanation to the post, so that other people can appreciate the achievement.

[...]

• A substantial amount of time has passed since the last post in the topic. If a topic has been inactive for a long time, it will have drifted down the topics list and few people will read it or check the latest post for edits. This means potentially interesting results could get completely ignored by the community if they are added in an edit within a generally inactive topic.

It's this last point that I think might be the most contentious. One issue is that I don't know how to define "a substantial amount of time", and I think what qualifies as enough time depends on how interesting the new post is. I do however think there should be a threshold after which a double-post is allowed, regardless of how interesting that post is, assuming of course that it's interesting enough to be posted in the topic at all. Does anyone agree or disagree with this? If you agree, when do you think this threshold is?

Comments on this issue are welcome.

On the one hand, I do think it would be a strange situation for someone to choose to edit a post that was made months or years ago merely because of the forum rule 3a. (But there can be other good reasons for editing an old post.)

However, I believe that kind of situation would be also affected by the forum rule 3b ("Do not post in topics that have been inactive for over a year without good reason.")

If someone indeed has a good reason to "revive" a long-inactive forum thread, then I think that reason would justify a new post regardless of whether the author already had the most recent post in that thread.

[get_Snacked]

would it be a good idea to make a thread go to the top of the topics list if the last post in it was edited?

[confocaloid]

would it be a good idea to make a thread go to the top of the topics list if the last post in it was edited?

An optional (non-default) way to sort topics by the "last activity" or "last edit" date coud be "nice to have" (unless I missed an already existing way to achieve that). Probably technically possible to implement, but might not be implemented in practice.

Note there are substantial problems with existing bump-topic functionality: viewtopic.php?p=200902#p200902 Using it damages (loses) the true date of the most recent post in the thread. I think that may be a good reason to disable the functionality entirely, if disabling it is simpler/more implementable compared to fixing it to preserve the true dates of posts unchanged.

[get_Snacked]

Using it damages (loses) the true date of the most recent post in the thread.

...how?

I think that may be a good reason to disable the functionality entirely,

if you're talking about the functionality of "when someone makes a new post in a thread, it gets bumped up to the top of the topic list", i don't see why such a functionality would need to be disabled at all. then how will the topic list be decided?

[confocaloid]

Using it damages (loses) the true date of the most recent post in the thread.

...how?

I think that may be a good reason to disable the functionality entirely,

if you're talking about the functionality of "when someone makes a new post in a thread, it gets bumped up to the top of the topic list", i don't see why such a functionality would need to be disabled at all. then how will the topic list be decided?

No, I'm not talking about someone making a new post and the thread getting bumped as a result.

There is a separate bump-the-topic functionality which is sometimes available under certain conditions.
https://www.phpbb.com/community/viewtopic.php?t=2514671

To see an example of the problem with "bump the topic" functionality, look first at this archived copy of a forum thread (November 2024):
https://web.archive.org/web/20241118164 ... =11&t=6623
Note in particular the last (at the time) forum post, which is by Para, and made in July 2024.

Now compare this with the current view of the same forum thread, archived copy (January 2025):
https://web.archive.org/web/20250105210 ... =11&t=6623
Note that the exact same forum post by Para is now displayed as if it was made in January 2025.

The displayed date of the post was changed to an incorrect one.
The reason why it changed is because someone else recently bumped the thread, and the date stamp of the most recent (at the time) post in the thread was damaged as a consequence.
It was someone else, because the author of that post wasn't recently active on the forums.

The bottom line is that bumping a topic damages the date of the most recent post, and shouldn't be used even when available, unless the underlying problem is fixed in some way.

[get_Snacked]

There is a separate bump-the-topic functionality which is sometimes available under certain conditions.

where is the functionality found?

[confocaloid]

This thread is for discussing the forum rules. My intended purpose here is to determine what people expect or want from the rules, how they might be clarified, and in what ways they might be lacking. [...]

I think the forum rules need to warn against posting very large [patterns that appear to be] pseudorandom soups.
The forum rule 4 already lists "Some specific examples of discoveries that won't be of interest to this community" and large random-looking patterns may need to be mentioned there too.

Sometimes someone posts a big pattern (hundreds cells by hundreds cells) that appears to be a soup. Sometimes the description says or implies it is indeed a soup; other times, there may be no description.
In most such cases, it appears to be unnecessary to post that particular soup. Posting such patterns is problematic, because the resulting forum post takes lots of space and becomes hard to quote partially (a large volume of data to be trimmed from the quote), and also because the random-looking pattern by itself doesn't contribute to the existing knowledge or understanding (one needs additional explanations, and then those additional explanations can replace the big random-looking pattern, avoiding the need to post it explicitly).

Note that LifeViewer has scripting commands which can autogenerate pseudorandom soups, and can avoid the need to post them explicitly. (The scripting commands also clearly communicate to other people both the intent to show a soup and the specific way of generating the soup.)

Code: Select all

#C [[ RANDOMIZE RANDWIDTH 256 RANDHEIGHT 192 RANDFILL 33 RANDSEED 31415926 ]]
x = 1, y = 1, rule = B3/S23
o!

[hotdogPi]

You're assuming it's truly random. What if someone created a 300×300 soup that contains a rare object and there's no obvious way to reduce it?

[confocaloid]

You're assuming it's truly random. What if someone created a 300×300 soup that contains a rare object and there's no obvious way to reduce it?

No, I'm just guessing it's pseudorandom (in most cases there is no sufficient description to explain the need to post a random-looking pattern).

If the final pattern contains a rare object that is a reason to post the soup, then the forum post needs to clearly explain that. If the forum post doesn't explain anything and just gives a random-looking pattern, readers will have to guess. Which is a problem.

[...] Sometimes the description says or implies it is indeed a soup; other times, there may be no description. [...]

[b-engine]

Suppose I found something from a soup generated by a seed, should I post all the commands or just the seed?

Found a toad in this soup:

Code: Select all

#C [[ RANDOMIZE RANDSEED 0 ]]
x = 0, y = 0, rule = B3/S23
!

Found a toad in seed 0.

[confocaloid]

Found a toad in seed 0.

Which one?

(That is, it's better to post the complete scripting command, to tell how the pattern was generated. Even better would be to isolate the relevant reaction that produces the object of interest, and post only that reaction / a small predecessor, with additional comments.)

Code: Select all

#C [[ RANDOMIZE RANDWIDTH 16 RANDHEIGHT 16 RANDFILL 33 RANDSEED 0 ]]
x = 1, y = 1, rule = B3/S23
o!

Code: Select all

#C [[ RANDOMIZE RANDWIDTH 32 RANDHEIGHT 32 RANDFILL 33 RANDSEED 0 ]]
x = 1, y = 1, rule = B3/S23
o!

Code: Select all

#C [[ RANDOMIZE RANDWIDTH 32 RANDHEIGHT 32 RANDFILL 50 RANDSEED 0 ]]
x = 1, y = 1, rule = B3/S23
o!

Code: Select all

#C [[ RANDOMIZE RANDWIDTH 64 RANDHEIGHT 48 RANDFILL 50 RANDSEED 0 ]]
x = 1, y = 1, rule = B3/S23
o!

[hotcrystal0]

According to iNoMed on Discord, it is possible to break all forum rules except 3a in one post:

Seeing as an act of doxxing (1b) in the form of a rude (1a) and profane (2b) purely foreign-language (2a) unnecessary necropost (3b) that overuses quotes (3d) and text formatting (3c) with an improperly LV-tagged (5a) misleading auto-invoking RLE (5b) is also non-academic (4), it's safe to say a single post can break everything except rule 3a as that distinctly requires multi-posting to break

[dvgrn]

According to iNoMed on Discord, it is possible to break all forum rules except 3a in one post...

Not to give anyone ideas, but rule 3a can also be broken by the same post. It just requires that the same person has the most recent post in the same thread. If the previous post didn't break any rules, then it's only the new post that breaks rule 3a -- and that one single post would break all the rules.

Please, everyone, resist the temptation to actually attempt this ... Speaking of multiposts, confocaloid made some suggestions a while back, that I'm still thinking about how to work into the official forum rules. Right now I'm leaning towards something like

"Lots of short messages in quick succession from the same person are not appropriate on the forums. Please don't double-post without a good reason. Good reasons include, for example,
- a significant change of topic (that is still appropriate to the thread)
- a week or more having elapsed since the previous post
- the previous post being too long due to multiple previous edits, or due to hitting the post size limitation."

This still seems to be one of the most mis-understood and mis-applied forum rules. It would be nice if people could be sure that they didn't have to apologize for double-posting, in lots of cases where the rule was absolutely never intended to apply.

The current wording says "If your message is the most recent one in a topic, simply edit it if you have additional things to add." -- somewhat implying additional things to add [immediately] [about the subject of the previous post]. Read that way, even the current wording doesn't really forbid all double-posting -- but we can definitely make it clearer somehow!

[hotcrystal0]

Not to give anyone ideas, but rule 3a can also be broken by the same post. It just requires that the same person has the most recent post in the same thread. If the previous post didn't break any rules, then it's only the new post that breaks rule 3a -- and that one single post would break all the rules.

If rule 3a is broken (which requires posts in quick succession), then the post can’t break rule 3b, which is about necroposting.
Also, I think this “how many forums rules can you break with one post” discussion should be moved to Moderator Discussion to prevent people from “getting ideas”, but not deleted because it could be a resource for moderators.

[dvgrn]

Not to give anyone ideas, but rule 3a can also be broken by the same post. It just requires that the same person has the most recent post in the same thread. If the previous post didn't break any rules, then it's only the new post that breaks rule 3a -- and that one single post would break all the rules.

If rule 3a is broken (which requires posts in quick succession), then the post can’t break rule 3b, which is about necroposting.
Also, I think this “how many forums rules can you break with one post” discussion should be moved to Moderator Discussion to prevent people from “getting ideas”, but not deleted because it could be a resource for moderators.

Heh, instead I will hereby officially warn everyone not to get any ideas from any of this. If anyone actually attempts to break all the rules at once, I'll definitely resort to some sort of unpleasant punishment (which I won't describe in any detail -- just please assume it will be highly unpleasant).

Re: whether you can break all the rules at once: rule 3a currently does not say anything about "in quick succession". That's exactly why I think rule 3a is worth fixing up somehow! People are still posting messages that start out "sorry for doubleposting but..." in situations where weeks have gone by since the last post, so the rule shouldn't apply, so no apology should be necessary

-- as well as cases where the topic is completely different, so the rule shouldn't apply, so no apology should be necessary.

[confocaloid]

Multiple short messages in quick succession can be annoying or otherwise problematic, even when each next post would count as "a significant change of topic".
Multiple short messages can be undesirable (editing the last post can be objectively better than double-posting), even when days or weeks passed between posts.
Very long posts (which might hit the post size limitation) may mean there are unnecessarily big RLEs (such as someone posting an extremely large soup instead of just extracting the relevant reaction or interaction), or other unnecessary high-volume content.

In principle there can be good valid reasons for multiposting. In principle there can be good valid reasons for very long posts. However, I think in practice most multiposts were made without a good reason (and would go better as edits to the most recent post), and in practice many long posts are unnecessarily long and could be significantly shortened without losing anything on-topic.

The "RLE copy/paste thread" in the sandbox contains some examples of unnecessary multiposting. Not linking any specific examples, it is easy to find them. Even in the Sandbox, it can be very annoying to see multiple posts from the same person in a quick succession without an (either obvious or explained) reason for that.

[dvgrn]

...

I agree with pretty much all of this. There might be some situations where I'm going to prioritize participation over perfection, though, in relation to

"...in practice many long posts are unnecessarily long and could be significantly shortened without losing anything on-topic."

For example, if someone wants to post a series of discoveries more or less as they happen, I'll certainly want to remind them to edit a previous post instead of multiposting, whenever that reminder seems to be needed. But I won't demand, or even suggest, that a set of discoveries in the same rule, edited in as a series of small RLEs in a single long post, should necessarily be re-worked into a single stamp-collection RLE. That's a lot of work, and it doesn't seem like necessary work, especially when it's probably not known at any given time whether the chain of discoveries is complete or not.

Anyway, we seem to be substantially in agreement about the general goal here. What do you think of my suggested rework of the rule about multiposting?

If we try to write a rule that draws a bright line between every possible valid reason for multiposting and every possible unwanted multipost, we'll have to write a book-length rule (and we'll still miss some edge cases, no doubt). So I'm not trying for perfection, just for an incremental improvement over the current rule -- since the current rule definitely seems to confuse people quite regularly.

[confocaloid]

I think the existing wording of the forum rule 3a is objectively better than the proposed replacement. The suggested change would introduce more confusion.

Currently the forum rule 3a says

Do not post multiple messages consecutively in the same topic. If your message is the most recent one in a topic, simply edit it if you have additional things to add.

I think that is concise and reasonably clear. (It tells what to not do. It tells what to do instead. People thinking about that forum rule and being reluctant to double-post is an evidence of the forum rule being sufficiently clear as it is.)

I think your suggestions are significantly longer, less clear and more confusing.

The existing wording (essentially) directly tells
> "Do not X. If Y, do Z instead".

Your suggestion instead (essentially) tells
> "X is not appropriate on the forums. Please do not Y without a good reason. Good reasons include, for example, ..."
which is harder to comprehend in comparison, and where the list of "good reasons" includes
> "a week or more having elapsed..."
(making it harder to understand how the forum rule 3b "Do not post in topics that have been inactive for over a year without good reason." affects the situation), and also includes
> "the previous post being too long..."
which sounds more like a problem in itself rather than a good reason to double-post.

Further, X /= Y, the term 'double-post' may fail to be self-explanatory to a newcomer.

[...] Right now I'm leaning towards something like

"Lots of short messages in quick succession from the same person are not appropriate on the forums. Please don't double-post without a good reason. Good reasons include, for example,
- a significant change of topic (that is still appropriate to the thread)
- a week or more having elapsed since the previous post
- the previous post being too long due to multiple previous edits, or due to hitting the post size limitation."
[...]

[dvgrn]

I think your suggestions are significantly longer, less clear and more confusing.

Heh, we are in agreement again! In my response to Sokwe when this issue first came up, I wrote

The current form ... has the great virtue of simplicity. Every time I try to re-word it, I end up wandering off into discussions of special cases that get so long that I even bore myself -- and then the rewording doesn't look like an improvement over the original, so then I just give up on changing it.

However, Sokwe started this discussion because the current wording of rule 3a seems to be not working quite right.

The existing wording (essentially) directly tells
> "Do not X. If Y, do Z instead".

That part seems very easy to fix. How about this variant?

"Lots of short messages in quick succession from the same person are not appropriate on the forums. Please don't post multiple consecutive messages without a good reason. Good reasons may include, for example,
- a significant change of topic (that is still appropriate to the thread)
- a week or more having elapsed since the previous post, with the new addition being interesting enough to draw fresh attention to the thread
- the previous post being too long due to multiple previous edits, or due to hitting the post size or attachment limitation."

There might still be ways to make that a lot shorter. But I still think there's an improvement to rule 3a waiting to be found out there. Sokwe and I have both documented some significant issues with the current rule -- hence this discussion.

I'm personally not terribly worried about the last two bullet points causing new problems, especially with the addition of may above.

There can be good reasons for double-posting.
There can be good reasons for posting in a thread after significant time has passed.
There can be good reasons for posts hitting size or attachment limits.

I think we can "assume the best" about people being able to figure out if they have good reasons or not -- and deal with the exceptions as they arise, just as we currently do, mostly by politely explaining the nuances of the rules to newcomers until we get the key ideas across.

Candidate shorter version:

"Multiple short messages on the same thread on the same day from the same person are not appropriate on the forums. Please don't post consecutive messages without a good reason. Good reasons may include, for example,
- a significant change of topic
- the thread having gone inactive for a week or more, so that an interesting update to a previous post would likely go unnoticed
- the previous post having reached post size or attachment limitations."

For the third item, honestly, the right idea might really be just "the previous post is already long enough". Rule 3a is just trying to advise against chains of short posts. There are examples all over the forums of threads like this one where rule 3a is completely ignored -- perfectly reasonably, because it absolutely is not supposed to apply. Each post is a good length already; editing to make fewer longer posts wouldn't have been nearly as good an option.

[confocaloid]

incomplete reply

[...] Rule 3a is just trying to advise against chains of short posts. There are examples all over the forums of threads like this one where rule 3a is completely ignored -- perfectly reasonably, because it absolutely is not supposed to apply. Each post is a good length already; editing to make fewer longer posts wouldn't have been nearly as good an option.

I disagree with your claim that the rule 3a is "completely ignored" in the examples you mentioned.
I think in many cases people are aware of the existing forum rule 3a, and they are taking it into account when posting.

The forums are supposed to be a place for people to discuss Life/CA-related topics.
A single person posting multiple consecutive posts doesn't lead to the situation "people discussing Life/CA-related topics".
This is an objective problem. It is reasonable to expect a discussion of some topic, with contributions from different people.

It is relatively insignificant whether the posts are long or short (as long as they are neither terribly short and "lazy" in an obvious way, nor terribly long and high-volume and again "lazy" in another sense in an obvious way).

It is relatively insignificant whether the posts are frequent or infrequent, as long as they aren't terribly frequent and chatlike/noisy.

This applies to many examples of forum threads where someone posts many messages over time. One person posting multiple posts doesn't make a discussion. Threads with a single most active contributor risk devolving into "bubbles" where someone is talking to themselves.

This also includes the thread you linked to. I tried to communicate what I find interesting about the topic. I tried to communicate what I believe to be useful to others who would like to better understand the topic. I know there are readers, but I don't know whether there will be posts from other people there. Hopefully someone else will reply with further interesting ideas or questions in that thread.

Do not encourage people to stay in their own bubbles. Communication is hard but that doesn't mean one can simply ignore the need for it. The forums exist for communication.

[dvgrn]

incomplete reply...

Yup -- I'm quite close to agreeing with everything you've said so far in this post, too.

However, what I'm trying to do is find a way to adjust rule 3a so that it's clearer than it is now. People keep reading it as if it is insisting that nobody should ever make two consecutive posts -- and yes, literally, that is what it currently says.

Among other things, the current wording reguarly leads to people worrying about or apologizing for double-posting, in situations where there's absolutely no need to worry. A forum search on the word "sorry" turns up an awful lot of these.

So I'm looking for specific suggestions about how to improve the wording of rule 3a. For example, does anyone have any further feedback on suggestions laid out above -- e.g., the "candidate shorter version"?

[hotdogPi]

How do we handle the proliferation of game forks?