@XanderHenderson Regarding this question, everyone interacting with the question seems to assume that $n$ and $N$ are meant to be the same. That is, they are interpreting the question as "Is $\gcd(n^2-1, n) = 1$ for all $n$?" I think that this is the right interpretation, but the question, as written, is unclear.
@Makyen Got the script running, thanks. The annotations are useful. But the tags are almost unreadable on my monitor, there is not enough contrast between the dark grey background and text. Might be related to my monitor being wide gamut (but uncalibrated). But I've never seen anything else so unreadable so I think it has more to do with design. Any way to tweak that?
@BillDubuque Yeah, the tags in this room are quite poor wrt. contrast. Unfortunately, I forgot to adjust them in the first release I made of the URRS. I was working on some bug fixes and improvements yesterday and today. I should have another release later today. One of the changes is to implement an adjustment to the tag's CSS. With the next release of the URRS, they currently look like the following:
I may make some additional changes to the meta tags, as I just noticed that they aren't very distinctive against the white background of a oneboxed Meta post, like those above. The meta tags are OK in a message that's not oneboxed. I tried to maintain more-or-less the scheme which was already in use, but just improve the contrast. I'm also open to suggestions, if you, or anyone, wants something different.
Keep in mind that only people that are running the URRS will see this change. Unfortunately, there's nothing that can be done by us to change the stock tag format.
@Makyen Either works for me (much better than current). But would the former work better for greyed out items (does that signify completed?). Currently I can't read the greyed out tags because they are too light, which might be exacerbated by the 2nd scheme. It would be useful to be able to read them w/o having to hover.
@BillDubuque That's a good question. Yes, greyed out does signify the request is complete. One one the main goals for the "completed" style was for the message to still be readable, without the need to hover (although, with some additional effort). So, definitely they should be readable in that state. Another main goal for the the right-side status information to be clearly out of vertical alignment with the incomplete ones, enabling the user to scan down the right side to see incomplete requests.
So, here are the two styles, but using the "complete" faded format:
@stressedout Thank you for mentioning such answers in this room. Bad posts may not be seen by users with the rep required to delete it, so they may get many downvotes from other users but still not be deleted. You were correct in flagging it for moderator attention, but note that each flag needs only one moderator to handle, so typically the outcome relies on the judgement of only one moderator. Posting in this room is likely to get a much larger number of users evaluating it. It's now gone. =)
By the way, in my viewpoint (which others may not share), there are two main reasons for deleting answers that merely duplicate WA's output:
(1) As hardmath stated, I take questions on Math SE to be mathematical inquiries, not question on how to press buttons on a calculator.
(2) WA often gives the Wrong Answer. I am virtually certain that even 10 years from now I will still be able to easily give it a limit expression that it will get wrong although any mathematician will be able to get the right answer.
Anyway, I think we can't blame the moderators for sometimes declining our not-an-answer flags, because their delete-vote has instant and irreversible effect (normally you need 3 normal users to delete a post).
@BillDubuque Lol at the accepted answer. The question should have used a modulus that was on the order of 10^15 to prevent such answers. The answer is up for deletion. =)
@BillDubuque The same rules apply; it must have negative score to be up for deletion, and if it gets 3 delete votes it will be deleted (and unaccepted).
The only exception is that the answer author cannot delete it if it is accepted.
@user21820 Strangely, I see the "d:(2)" annotation on your msg, but not on my prior msg (it shows only "An:(-1)"). Anyone know why? Maybe mine didin't match any regexp?
@BillDubuque I have no idea; I'm using GreaseMonkey and not intending to switch, so I can't use the script. But if I were to guess, it could be because the script runs once per message and you might not have refreshed the page?
@user21820 The URRS works fine under Greasemonkey. The only script, of those mentioned here, that has a compatibility issue with Greasemonkey 4 is the alpha version of the Request Generator. The release version of the Request Generator should work fine for you, as well as the URRS. The alpha version of the Request Generator doesn't work with Greasemonkey 4, due to relying on userscript APIs which Greasemonkey has fundamentally changed from GM3 to GM4.
Any userscript which uses the userscript APIs must be rewritten to work with GM4. This will cause ongoing compatibility issues with scripts that have not been rewritten. It is possible to write a script such that it's compatible with both versions of the userscript APIs, but it has to be intentionally done, and may require rearchitecting the logic in the script. From a user's perspective, random scripts will fail to work in GM4.
@BillDubuque @XanderHenderson @user21820 The difference is that the SE API does not provide the number of delete-votes cast on answers, but does provide them on questions. Thus, the script doesn't have number of delete-vote information available for answers.
@BillDubuque It would be possible to scrape the information from the question page, for those users with >20k rep. However, A) scripts scraping pages that the user isn't actively viewing is something SE explicitly doesn't want to have happen; and B) requesting pages or from the SE internal real-time endpoints can rapidly run into hard rate-limits which SE imposes for access to pages and the realtime endpoints.
If the hard rate-limit is tripped, then the IP the user is on is blocked from all access to SE for a period of time, usually starting at 15 minutes. That's rather inconvenient (I've had it happen a few times due to development). There are also no indications leading up to it happening, so it's not something where you can program the script to backoff when it gets close.
Those rate-limits also take into account all traffic from the IP address, so how much is possible for a script to do will depend on what else the user is doing, even in background tabs, or what other users who share the IP are doing.
Hrm... @Makyen the script seems to have added the phrase "Seeking personal advice" into my close request. I wrote "This question uses obfuscatory notation in order to create a 'puzzle'."
Is there a way to fool the overly aggressive regex?
