# We need to talk about the Sandbox

Warning: this post will contain things that people don't want to hear. But it needs to happen.

Now that that's out of the way, let's get down to business.

The Sandbox was a brilliant idea. It allowed people to see what others thought of their challenges before actually posting them. It was a great place to redirect new users to in order to improve their challenges.

But not anymore.

I don't have access to any site stats but it's obvious without them. The Sandbox is dying.

Based off of my experience, the only person who regularly reviews challenges there is Peter Taylor, for which I would like to thank him. Everyone else pops in to either post a new proposal and then reads one other or just skim over challenges to check for any mistakes.

I've been meaning to make this post for a while now but my latest challenge pushed me over the edge, in particular these two comments.

<begin_anecdote>

I had sandboxed this challenge on the 19th of May, 1 week ago today and it had 2 upvotes and no unresolved issues for a week. By the Sandbox's own definition, it was ready for posting.

So I decide to post it, edit the Sandbox post and went to sleep, confident that all issues had been filtered out and corrected. Imagine my surprise when I wake up and find 20+ items in my inbox asking for clarification!

Two of them were even asking me to post this in the Sandbox! When I read these comments, I wanted to reach inside of my computer, crumple the Sandbox into a piece of paper and throw it in the bin. I even wrote (but didn't post) a comment that was all £&@%^*+ symbols, I was that annoyed.

<end_anecdote>

This has to stop. Something needs to be done about the Sandbox. And I'm going to be honest with you here, I have no idea how to fix it. This post isn't to tell you how to fix it or what needs to be done to fix it. This is to tell you that it needs to be fixed. And why can't I tell you how? Because it is so broken that I can't think of even beginning to correct it.

But maybe you do. Please, answer this question with your ideas for what we could do about the Sandbox. Whatever it is, it is better than the current system.

Now, as ever, there will be people who disagree with me, who say that the Sandbox is working. If you are one of these people, think about this:

• A good challenge is posted to the Sandbox and quickly receives 3 upvotes.
• While reading through it, you prove a couple of mistakes/contradictions and notify the OP.
• They correct the mistake and, in your eyes, looks good for posting. They give it a week or two to gain attention but nothing changes
• They then think that it is ready for posting and post it.
• Immediately, someone downvotes and asks for a clarification. Soon after, you post an answer that you've been working on since you saw it in the Sandbox.
• The OP updates the question and invalidates your answer. But by this time, you've moved on to other questions and have forgotten about it. You gather downvotes as your answer doesn't work.
• All this, simply because no-one else spotted a simple ambiguity while in the Sandbox phase, giving both you and the OP the negative feeling you get when someone downvotes your post.

Has this ever happened to you? It probably has and we can stop it. Why stand around, letting this happen again and again, when we can simply resolve this issue?

I say it again, the Sandbox is dying. And we need to save it. But how?

• Your last example (everything after if you disagree with me, think about this) is purely hypothetical, and does not prove that the sandbox is broken. And even in this example, the same problems would have happened if it was posted directly to main; the issue in common here is the ambiguities in the challenge, not the sandbox. – James May 26 '17 at 23:11
• That's an awful lot of inflammatory language, and very little in the way of a clear statement of a problem. – Steve Bennett May 28 '17 at 10:31
• @SteveBennett sometimes the truth is harsh. If you ask me, saying it as it is rather than sugar coating it will get better results. Maybe I didn't make my point clear. The Sandbox no longer helps to develop challenges before posting – caird coinheringaahing May 28 '17 at 10:35
• Try re-reading your post. 80% of it is waffle, meta-commentary on your own post and tangents. The only paragraph in which you describe the problem is "Based off...mistakes". I'm all for harsh truth, but let's have more facts and less drivel. – Steve Bennett May 28 '17 at 11:10
• – user62131 May 28 '17 at 15:56
• And here I sit looking at my second sandbox entry which has received zero attention and desperately needs it. I'm pretty sure there's a good puzzle or challenge in the idea somewhere, but I need help identifying it. – Draco18s no longer trusts SE May 28 '17 at 16:03
• @Ilikemydog the one about combination locks. (The other one went to main, almost got closed, but survived and had some good answers) – Draco18s no longer trusts SE May 28 '17 at 16:04
• I get how this might have been frustrating to you, and I don't mean to discourage the community from pursuing a better solution, but honestly, it strikes me as absurd that you considered 2 upvotes as an indication of "consensus". – Cody Gray May 30 '17 at 8:06
• On regards to your comment about my comment on your question, I hadn't seen your challenge in the sandbox. I also don't spend too much time browsing it so that's probably why. The reason I said that is there was a lot of things needing clarification (from my POV) so it seemed as if you hadn't used it. I did have an extra note in the comment about how some things might still need clarifying after a sandbox post but removed it. – TheLethalCoder Jun 1 '17 at 8:36
• I wonder if part of the problem is that the question rate on main is much higher than it used to be. When the question rate was low, people could run out of challenges to answer and go and browse the sandbox instead. With more challenges available now, maybe that means fewer people finding spare time for the sandbox. – trichoplax Jun 4 '17 at 10:51
• Same thing happened to me in Music with pi and eulers number. In the sandbox I got 0 comments, and in the main site I got 2 comments notifying 2 main mistakes! I am thankful to god that Kevin Cruijssen didn't downvote my question and first asked for clarification. – math Jul 11 at 13:37

# There are several common types of problems with challenges; the Sandbox is good at catching some but not others

## Things the Sandbox is good at

Where the Sandbox really shines is in circumstances where a challenge breaks a challenge design principle that most experienced PPCG users are aware of. For example, it's fairly easy to go through the recent Sandbox entries and comment on all the ones that require large fixed strings in their output, but are ostensibly about something other than . There are a number of similar issues which are almost always quickly caught in the Sandbox.

Another situation in which the Sandbox works well is in cases where the whole concept behind a challenge is poorly conceived. Most such challenges immediately receive negative feedback, and even though a couple of attempts are often made to fix them, it becomes clear that the challenge could never work. As such, these challenges tend not to reach the main site at all, meaning that the Sandbox is doing a good cleanup job in these cases. This might not seem encouraging as a question asker – the only real benefit they get from it is a slight reduction in the (very low on PPCG) chance of a question ban, when the challenge inevitably gets closed – but it certainly helps out the question answerers.

The final way in which I'm aware of the Sandbox helping well is in checking for duplicates. Duplicates are often almost impossible to search for on PPCG, so you typically have to rely on someone who saw the earlier challenge to remember that something was up. (My favourite example of this is this comment, in which a PPCG user remembered competing on an earlier duplicate challenge and already had a solution to it in a file on their own computer, and yet couldn't find the challenge itself. We found it in the end, but it was way harder than it should have been.) It's often possible to get a suspicion that a challenge might be a duplicate, and Sandboxing it with an explicit request for a dupe check often produces useful results.

## Things the sandbox is bad at

Now, what the Sandbox is worst at is finding ambiguities in the specification (i.e. the situation for which you're trying to use it); there's definitely a problem here, because this is something that is commonly perceived as something that the Sandbox does/should do. Although I don't know for certain, I suspect there are two reasons for this.

One is that finding an ambiguity often needs a lot of knowledge of specific programming languages; I can see a challenge that says "in situation X, return Boolean false; your program must not error" and immediately realise "there are programming languages in which Boolean false is a type of error", but it's not a fact that the programming community is generally aware of, so most people wouldn't be able to catch that specific issue in the Sandbox. Likewise, there are a large number of potential issues which could only be caught by specific other people. In other words, the issues that I mention that the Sandbox is good at can each be caught by a large proportion of PPCG users; however, ambiguities can often only be noticed by a much smaller subset, meaning that in order to remove all ambiguities from a Sandboxed challenge, you'd probably need around half of PPCG to review the challenge, and even then you'd run the risk of missing some. (Note that posting to main isn't much better at catching this sort of ambiguity, but if nobody catches it on main, the ambiguity turned out not to actually matter.)

The other issue is that some ambiguities only become clear when trying to actually solve the problem, something you aren't really meant to do with Sandboxed problems. (I rarely try to solve even my own problems before they're posted on main, to keep things fair; it's not uncommon for a challenge to change between the sandboxed and main version anyway, so working on an answer in advance is inadvisable as it is. The main exceptions here are solving Sandboxed challenges in Jelly in my head in order to determine if they're too trivial – if I can do it, the challenge probably needs to be harder – and situations in which I write a program I'm particularly proud of and post a challenge to PPCG simply as an excuse to be able to post the program in question as an answer.) Actually writing code can give you a better appreciation for where the corner cases are than looking at the spec.

## Ways in which we can improve the situation

So, what can we do about this? First, we probably shouldn't be downvoting challenges unless they're actually downvotable according to SE rules ("this question does not show any research effort; it is unclear or not useful"). Sandboxing questions correctly shows effort into making a good question, and that the question is not immediately unclear; lack of utility (in PPCG terms, being uninteresting to compete on) is also something that you'd hope would at least be commented on in the Sandbox. Incidentally, you can help deal with this issue both as a voter and as a challenge poster; voters can simply not downvote challenges unless the challenge deserves it; and challenge posters can use the trick of linking to the Sandbox post from the main post (even though doing so really shouldn't matter), in order to give a subtle hint of "if you want to complain about this, you should have done so while it was Sandboxed" (which in my experience actually helps to avoid downvotes). For what it's worth, I've Sandboxed a few challenges even though I knew they'd get no useful feedback, purely to be able to link to the Sandbox and say "see, I Sandboxed this!". (Much of the time, though, I am expecting the Sandbox to be able to improve the post.)

That basically just fixes the issue with downvotes, though; it doesn't fix the issues with the amount of exposure that some posts need to resolve ambiguities being lower than the amount that the Sandbox can reasonably give it. I know that I suggested sandboxing posts as on-hold posts on main a while back; the suggestion was generally well-received on average, but controversial. I've since come to view it as a suggestion that would probably be beneficial overall, but which is fairly drastic (and which would likely have significant UI issues), and therefore should be avoided unless necessary; however, if we view the Sandbox as being on the point of collapse, we might have to go through with it. (It's not like the reopen queue is at all struggling at the moment.) That said, there seems to be a rough, but far from unanimous, consensus that the Sandbox is in fact functioning mostly correctly at the moment.

• Overall, I agree with this post. But there is one thing I can't believe. You can make a Jelly program in your head? I don't even understand them with an explanation! – caird coinheringaahing May 28 '17 at 16:01
• Jelly programs are so short that it's one of the easier languages to do in your head; if you try to do a Java full program in your head, you've typically forgotten the start before you reach the end. Note that the challenge has to be very trivial to be mentally solvable in almost any language; for example, if I recognise the challenge as a simple combination of two Jelly builtins, it's both very easy in Jelly, and probably too trivial for main (except possibly as a catalog). – user62131 May 28 '17 at 16:06
• The biggest problem I see with ambiguities in sandboxed questions is neither of the ones you mention: it's that some questions have so many, or the OP doesn't understand what the problem is, that they need several rounds of revisions. A sandboxed question will almost certainly be looked at once by a few people, and maybe twice by fewer, but after two rounds of revision it takes more effort to see what's changed and lots of patience to keep re-critiquing it. – Peter Taylor May 29 '17 at 9:38
• @cairdcoinheringaahing If you know the language well enough it's actually quite easy to make trivial programs in your head. – Erik the Outgolfer Jun 5 '17 at 6:53

# Replace the sandbox with a Main site tag

I think this suggestion is doable without any changes to SE.

Introduce a special tag on the Main site which has a description along the lines of:

A proposed challenge. This tag indicates that the challenge is not ready to receive answers, but rather has been posted to receive feedback in the form of comments. When ambiguities, mistakes, contradictions, and other issues have been resolved, the author removes the tag, thus signalling that answers may be posted.

Users can then post sandboxed challenges exactly like they would post open challenges, except they would include the tag.

Behaviour around sandboxed challenges:

1. If a user answers a sandboxed challenge, we kindly remind/inform them of its status.
2. If a (new) user posts a challenge which obviously needs sandboxing, we can edit in the tag and kindly inform the poster.

If it is at all customizable, we could encourage the use of the sandbox tag by including it in the Tags field placeholder text:

at least one tag (sandbox code-golf string), max 5 tags

# Upsides

1. Sandboxed challenges are as visible as open challenges, thus encouraging peer review and hinting the sandboxing system to newcomers.
2. No need to move posts from the sandbox to main; the tag is just removed in order to "post".
3. No need to include [tag:code-golf] etc. in sandboxed posts; proposed challenges can have actual tags (although only four instead of five) with the normal auto-completion and tag description system in effect.
4. The system for detection of duplicates will work.
5. Anyone can see whether a challenge was sandboxed just by looking at the edit history.
6. No more of the Sandbox's "This question has more than 30 answers already." pop-up.
7. We can use the FEATURED ON META for something productive.
8. Users looking to review sandboxed challenges can easily access them.
9. No more lag for users on mobile or with the Infinite Scrolling user script.
10. It will still be possible to have New Sandboxed Posts in The Nineteenth Byte either as part of the regular feed, or as a dedicated one, using filters in the RSS feed.
11. When the challenge is opened (by removing the sandbox tag), it is automatically bumped to the top of the Active list (but unfortunately not the "Newest" view) due to the edit.
12. Becomes easy to see if a language was created or modified after the post was first proposed.
13. Challenges which are posted after a sandbox period do not clutter anywhere. (In the current Sandbox, high-rep users see all the deleted posts.)

# Downsides

1. Removing the tag bumps to the top in the activity view, but not in the newest view, so there wouldn't be an easy way of finding questions which have recently lost the tag.
2. The HNQ algorithm heavily uses activity since post date, so we'd lose a lot of HNQ questions because they'd have a spike in activity a day after they were posted, instead of a minute after.
3. Nothing stops people from answering these challenges anyway.
4. There are a lot of sandboxed questions, which would dilute the actual challenges on main.
5. It would be much harder to see when there's actually a new challenge to be answered, and the main question feed would become a lot more useless.
6. Votes on drafts actually affect reputation.

# Questions

1. Sandbox posts are often just vague ideas to gauge initial interest... do we accept that for draft challenges on main?
2. Is there some minimum amount of detail that needs to go into these?
3. How do we prevent answers on sandboxed challenges?
4. Should we require all posts to have either or similar to how Meta requires one of the special tags, in order to make it easier to find open challenges?
5. Is it really so terrible if answers are posted to sandboxed challenges? Worst case, they'll be invalidated by changes to the OP.
• Another benefit: well-received sandbox ideas keep their reputation when going "live". Possible downside: sandbox ideas going live will likely need a comment flush. – BradC Jun 6 '17 at 22:19
• @BradC I remove my sandbox comments when the post is fixed accordingly, but it is easy for mods to clean up anyway. I'm not sure if keeping the reputation is a benefit per se. It is a side-effect. Why do you think it is specifically a good thing? – Adám Jun 6 '17 at 22:21
• @BradC I have always wanted to get rep from the sandbox. However, if a proposal is met poorly, in the current sandbox would receive no rep changes but in this system would get -2 rep from that. So if someone isn't particularly clear, they get downvoted by a drive by who then doesn't come back to reverse that when the challenge is improved – caird coinheringaahing Jun 6 '17 at 22:23
• @cairdcoinheringaahing Hmm, do people downvote sandbox ideas for "this still needs some clarification", vs "this is a bad idea for a contest"? – BradC Jun 6 '17 at 22:24
• @BradC I think the latter, which means carrying over the downvotes is actually a good thing. – Adám Jun 6 '17 at 22:25
• Removed the tag bumps to the top in the activity view, but not in the newest view, so there wouldn't be an easy way of finding questions which have recently lost the tag. – Peter Taylor Jun 7 '17 at 6:18
• @PeterTaylor Ah, I didn't know. That's a pity. Maybe we should have a different tag open so that open challenges are findable. We could then require one of those two tags just like one of the special tags is required on meta. – Adám Jun 7 '17 at 6:45
• @PeterTaylor Better? – Adám Jun 7 '17 at 6:47
• Worse. Aside from the fact that finding a tag's newest question list takes about three times as much effort as finding the overall newest question list, having a tag dichotomy like that on main is a horrible abuse of the tag system. – Peter Taylor Jun 7 '17 at 8:18
• @PeterTaylor How is the sandbox/open system any worse than than bug/feature-request/discussion/support on Meta? Especially if we can have a system in place which forces users to use one of the two (or three; we could have tips too). – Adám Jun 7 '17 at 8:23
• If it were a forced choice that would mitigate the problem, but I don't think that's going to make the priority list for implementation in less than six to eight months. It would also require retagging every existing question. – Peter Taylor Jun 7 '17 at 10:51
• @PeterTaylor Retagging existing questions should be relatively easy, as they are all open. We only need manual treatment for those that have five tags. – Adám Jun 7 '17 at 10:52
• Couldn't we use [on hold] status rather than tags (like I suggested earlier)? Given that such questions shouldn't be answered, it has both the right semantics, and the right name. Also, it's worth pointing out that a mass retag (adding open to every question) would bump every question's activity, causing chaos with respect to sort order (and making vandalism during the retag almost impossible to spot as a consequence). – user62131 Jun 7 '17 at 23:18
• This is a moderation nightmare. I predict people will answer drafts all the time, we'll get even more comment flags than we do now (it's the most common type of flag), moderators will be informed about a large number of deleted questions by the same user (which is not only bothersome, it also means we won't know if there's a problem), people will get question-banned for poorly received sandboxed challenges, vote tallies will rarely reflect the final quality of the post, etc. – Dennis Jun 8 '17 at 14:51
• @Dennis What is a comment flag? – Adám Jun 8 '17 at 15:06

# Add the Sandbox to the Review Queue

Now, I don't know if this is possible, and I'm assuming it's not, because SE would have to do something about it. However, this makes the most sense: the people who like fixing up Sandbox questions can use the review queue to find all of the newest ones, and they can even get badges for it. People who aren't sure if they can help or not can skip challenges they have no comment on. Presumably, this will bring at least more eyes to sandbox challenges.

# Encourage more people to use the Sandbox Viewer

While all of the details are given in the link above, I'll pick out the relevant things:

• There is a sandbox button added to the header. When a new sandbox post is added, it changes colour

• It shows latest activity in a sidebar

Basically, it acts as a review queue for the sandbox

Something which could be kept in mind is Level River St.'s comment below:

I think the right way to go about this would be to have a SE developer do this officially, so that people dont have to install the viewer. Similar to the review queues for posts on the main site.

See here

• Relying on third-party tools should only be a backup solution in case we can't find a good solution with SE's tools, because not everyone will use those third-party tools. – user45941 May 28 '17 at 3:19
• @Mego That's why I say encourage. It's a handy tool and I think more people could benefit from it – Beta Decay May 28 '17 at 6:55
• I have intentionally disabled the Sandbox Viewer, it was too buggy for me. – Erik the Outgolfer May 31 '17 at 8:22
• @ErikTheOutgolfer In which way? – Beta Decay May 31 '17 at 8:24
• I think that some things like commenting didn't work. Also, the "See in Sandbox" button doesn't work either. – Erik the Outgolfer May 31 '17 at 8:26
• @ErikTheOutgolfer I see. I never used the Sandbox Viewer. It seemed like the worse option when browsing the sandbox – Beta Decay May 31 '17 at 8:27
• @BetaDecay Now that I'm testing again, yes, it's still that buggy. – Erik the Outgolfer May 31 '17 at 8:30
• I'd be interested to know if the downvotes are because this is an external solution, or because it's buggy. If there are people who like the idea but can't use it because of bugs, then it would be good to have a measure of that, as a motivation to get people working on the bugs. – trichoplax Jun 4 '17 at 10:54
• @trichoplax I'm guessing that it's more to do with Mego's comment about being third party. – Beta Decay Jun 4 '17 at 11:16
• I'm upvoting this because I think the concept is good. But I had no idea the sandbox viewer existed, and I am sure I am not alone. I think the right way to go about this would be to have a SE developer do this officially, so that people dont have to install the viewer. Similar to the review queues for posts on the main site. Whether SE would implement it is another matter. – Level River St Jun 4 '17 at 12:00
• This thing doesn't work for me. – Corsaka Nov 20 '19 at 10:04
• @Corsaka It's worth noting that this post is two years old, so there may be a newer version – Beta Decay Nov 22 '19 at 21:48

# Add an Option To Hide Deleted Posts

Most reviewers sort Sandbox posts by activity and, as deleted posts are included in that sort, they cause a lot of clutter - it's far too easy to scroll past a short challenge sandwiched between a few deleted ones, for example.

I propose an option that would easily allow for the toggling of the visibility of deleted posts. A userscript would probably be the way to go, rather than expecting SE to implement this for us.

This proposal obviously doesn't address any of the larger issues being discussed here but should help reduce some of the "noise" in the Sandbox.

• Admittedly this only affects 2k+ users rather than the new users who will be using the sandbox much more – caird coinheringaahing Jun 10 '17 at 15:41
• @cairdcoinheringaahing: Indeed. My thinking was this would primarily benefit reviewers (I'll edit to clarify), who, from what I've seen so far, are mostly north of or fast approaching 2k. – Shaggy Jun 10 '17 at 19:53
• This can be done in most browsers by opening in private browsing mode or equivalents. – jimmy23013 May 18 '19 at 17:42

So how would you respond to something like this...

Upon asking a new challenge, the owner closes the question as

On hold for final clarifications to be resolved before opening up for answers. Ask owner for clarifications in <Sandbox_chatroom_link>.

It would be completely voluntary, but

• Getting posted to the main site means relevant people would see it
• Reviewing in chat room shouldn't raise community user flags
• On hold will prevent answers until the owner is sure it is ready.

And hopefully people won't down vote until the question is reopened.

• My favourite response on this thread is Adám's where he suggests having a sandbox tag. If this happened, your situation would be very easily remedied. Simply add the tag and people would (hopefully) treat it as a proposal rather than a question. – caird coinheringaahing Jun 8 '17 at 19:32
• @cairdcoinheringaahing As Dennis pointed out, that would cause a moderation nightmare. I expect that this would have similar results. – user45941 Jun 9 '17 at 0:53

## Do nothing

I see no problem with the sandbox at the moment. Sure, it is stupidly big and crashes the mobile app, but use the mobile site instead, for sandboxing.

I, and I'm sure countless many others have never experienced a problem with the sandbox because it does what it's supposed to (but inevitably, a few answers have slipped through the net)

• This does nothing to help the problem – Redwolf Programs Apr 17 '18 at 17:31
• @Redwolf That's because I felt that there wasn't a problem. I'm not sure what the sandbox is like now though – Beta Decay Apr 18 '18 at 18:49
• Not to be pedantic, but I feel like ... others have ... is missing a comma. – Jonathan Frech Aug 28 '18 at 21:02
• @BetaDecay I've posted at least three sandbox posts with 0 votes, 0 comments for weeks, months, or more – Redwolf Programs May 5 '19 at 2:56
• @RedwoldPrograms This post is over a year old and I have not been active on this site for a long time. For all I know the may have been new issues that have arisen with the sandbox. All I know is that back in 2017 there wasn't a problem in my eyes. – Beta Decay May 12 '19 at 10:15

# Reward People for Helping in the Sandbox

First off, this is probably a horrible idea.

We make a post on main called "Sandbox Rewards" or something like that. Whenever someone posts a constructive comment to a sandbox challenge, they also post an answer to the main post. Then, the person who posted the sandbox challenge then upvotes the person on the main post, giving the main poster rep in exchange for their help.

This has a lot of problems. There is no guarantee that the sandbox-challenge-poster will upvote on main, and this might be confusing for new users. And the main post would technically be off-topic. And there are probably a bunch of problems that I haven't thought of.

• This seems like a good idea but the problem is it's execution. If we gave rep for comment upvotes and encouraged people to upvote helpful comments, then it could work. – caird coinheringaahing Jun 5 '17 at 17:32
• This isn't going to work IMO – Beta Decay Jun 5 '17 at 20:32
• @RedwolfPrograms Downvotes signal disagreement on meta. It's perfectly acceptable to downvote anything you disagree with, despite any disclaimers in the post. – Dennis May 10 '18 at 14:41

I say Sandbox it seems to me too much big...

I say it would be prohibited to post heavy content [only text objects not photo, embedded film etc] and organize all the posts in a way each one of us has at max 1 post in sandbox even with more question in his only post.

Or break the sandbox in 3 pieces: The one for math numeric questions, the one for strings type questions, the one for all other type questions.

• There are challenges involving images, etc. so they, obviously, need to include that media within their body. There are also other challenges that are best explained with an image or that benefit from having an image to assist with the explanation. To your next suggestion (leaving aside the enforcing of such a policy), multiple challenges per post would make the review process much more difficult with confusion in the comments with different challenges being discussed simultaneously. – Shaggy Jun 10 '17 at 19:08
• It would also create difficulties with voting; if there were 2 challenges in a post and one was good enough to upvote but the other was bad enough to downvote, what then? Your last suggestion does, in my opinion, carry some merit (I could have sworn it had already been proposed) but I would suggest individual Sandboxes for each winning criteria, rather than each tag. – Shaggy Jun 10 '17 at 19:11
• The right photo or heavy contents (MBs I think or java cpu consume in the page) I think can slow down every type of page – user58988 Jun 10 '17 at 19:17
• Ah, you were referring to file size, only. Sorry, I misinterpreted. In that case, yes, I agree with you on that point but it wouldn't seem, to me, to be an issue specific to the Sandbox. Although, to be fair, in my ~2 months here I think I've only encountered the problem once. – Shaggy Jun 10 '17 at 19:41
• I don't understand why this has been downvoted. It's 2019, the issue still stands, and I think this (at least, breaking the sandbox up) is a very good suggestion. – Geza Kerecsenyi Jul 9 '19 at 19:18