Also one quick question: Is it appropriate to post a (rather brief) answer and then expand on it over the course of an hour or so, or is that considered unnecessary bumping? I just finished an answer I first posted an hour ago and it has accrued 11 revisions in that time.
@clickbait regarding your edit here (10k only), I don't think it was deleted because of what you edited. It's simply not a fitting answer to the bug report. It's more of a feature request, that should come on its own question, not answer to related bug report.
Bringing it here anyway so others can see and maybe undelete anyway, I might be wrong.
@Feeds no one-box for pod-cast. So maybe one-cast, or pod-box?
Sometimes, particularly in small Stack Exchange sites that have relatively low activity, bountied questions get literally nothing — no answers and not even comments. Nothing useful is gained out of spending large sums of reputation for these bounties, yet future bounties must be at least twice as...
@cocomac not random, someone (can also be ordinary user) added xkcd RSS feed to this room, you can see the list here, and add feeds yourself to any room where you're owner.
The chat code "subscribes" to the RSS feed and listens for new entries.
Think you can also choose which "user" will post the message, but that's where my knowledge ends, never used it myself.
@Luuklag no, it's on us that it's still open, OP has the right to post bounty.
Dupes aren't bad, so no big issue if it stays open for a week. </2-cents>
@JNat possible network troll, based on their deleted answer on MSE at least. Highly offensive and meaningless. If did the same on the other sites, well... you're best with handling it. ;)
@clickbait well if five users think it's a duplicate (or one mod/gold hammer) it will be closed, they likely know the site better. I'm still neutral on this, and like I said if it's not closed by now it's your right to bounty it.
2 close votes for dupe of meta.stackexchange.com/questions/188481/… but that post suggests very different criteria. That post says the next min bounty shouldn’t double if no bounty was awarded; my post says it shouldn’t double if the question received 0 responses (and maybe also if it has low views).
@Luuklag Definitely worthwhile. It seems from the history that the user may be gaming the system, given that they deleted the question and the undeleted it once it became eligible for a bounty, then put a bounty on it, to evade close votes.
As to the argument of it being a duplicate: I don't think Rob's target applies, but my target definitely applies. The same arguments against the request I linked applies to this one: "[if] an initial bounty did not attract any good answers, a new bounty should be for a larger amount, to work better towards attracting answers" and "[doubling] the required bounty is a feature partly intended to limit 'bounty abuse'."
Also, per the duplicate closing policy, section Do duplicate closures work differently on meta sites?, it's OK to close a question as a duplicate if an answer to the target addresses the question, even if the questions aren't the same.
@ShadowWizardSaysNoMoreWar What they were saying in that context is, they don't clear bounties, instead waiting until after the bounty expires, to close the question, so that the author's reputation remains deducted.
@ShadowWizardSaysNoMoreWar Wouldn't be declined on the spot. Instead, they'd set a timer to remind them of the bounty expiration and they'll close it then
@JourneymanGeek If a feature existed where you could remove bounties such that the rep remains deducted instead of refunded, would you use it to remove immediately?
Also keep in mind that the privilege for users to see close votes on their own questions is gate-kept at 250 rep. Implementing such a feature would effectively allow for privilege escalation as users with 75+ and <250 rep would be able to see that their question has close votes
Solution 1: not show the option to bounty, or give an error when initially opening the bounty wizard. Privilege escalation, detailed above. Solution 2: allow proceeding through the wizard, but give an error at the last stage. Causes users to complain about lost work, especially if they took the time to compose a detailed explanation for the bounty.
Well, to be fair, such an escalation already exists for duplicate close votes: the automatic comment and the option to self-close. But it doesn't exist for other close vote reasons.
Shog hated bounties... many others do too. With enough support from community, likely SE will just sunset the bounty system and solve all the problems in one hit.
@JourneymanGeek Another feature request idea. In this case, the author deleted the question and undeleted once it got to two days old, so it wouldn't incur close votes in the meantime. Perhaps time spent deleted shouldn't count toward the 2-day requirement?
@ShadowWizardSaysNoMoreWar And reduce answer quality in one hit
@forestdistrustsStackExchange Yes, it's acceptable to post a brief answer and then expand it. But the original version has to be a valid answer in its own right. And 11 revisions in an hour is pushing it, 2 or 3 is probably ok, depending on how much traffic the site has. However, you do risk downvotes from people who hate any form of FGITW behaviour.
If Smokey devs would have accepted feature requests I'd ask them to change "Link at beginning of answer, link at end of answer" to just "Link only answer", in case of course it's answer that consists of single link. But Smokey devs aren't here and no idea how to send them a request, so meh. :D
@forestdistrustsStackExchange Oh, ok. But it can still be a bit annoying for readers who are interested in your post if you keep updating it while they're trying to read it. ;) JMG's idea of temporarily deleting can be good, especially if you need to make some drastic changes. OTOH, if you leave it too long before undeleting another person may post a new answer in the mean time that replicates some of your material, especially if they think the deletion is permanent.
Speaking of Miller-Rabin, I did a deterministic version in Python a few years ago: math.stackexchange.com/a/1638487/207316 but of course you need a probabilistic version for big primes. miller-rabin.appspot.com has some nice sets of witnesses for deterministic M-R.
Last week, I was contemplating writing some primality certificate code in Sage / Python. Sage can certify big primes, a few different ways, but it just returns a boolean, not the certificate, which is really annoying.
It's maybe worth mentioning in that Security.SE answer that not only does it take longer to find safe primes, they can take longer to certify, too. But I guess that depends on if (p-1)/2 is safe. :)
@ShadowWizardSaysNoMoreWar I like "0.1" more than "<1" because it gives me more information with not much more space. (0.1 tells me that it's not 0.6 or 0.3).
@forestdistrustsStackExchange Sure. And Sage has all sorts of lovely elliptic curve stuff. Some of the libraries it uses produce certificates, but Sage just doesn't expose them.
Sage is built on top of a lot of powerful sophisticated stuff. sagemath.org It provides a nice unified high level framework, but it sometimes hides stuff that you wish it made visible. Often, there is a way to get at that hidden stuff, but not always.
I've grown quite accustomed to my mac over the years. Its lack of support for Excel VBA however is really a downside. Or well the lack of MS to support Mac on VBA, whichever way you look at it
WSL is a blessing and a curse for Linux folk like me.
On the one hand, it means MS has finally admitted that Linux is not going away. On the other hand, it draws some people into using Windows instead. I guess it's fine for running ELFs directly instead of firing up a VM.
It's scary to think how much of the world runs on Excel, often involving large amounts of manual data entry & manipulation, much of which could be automated...
gLinux is a Debian Testing-based Linux distribution used at Google as a workstation operating system. The Google gLinux team builds the system from source code, introducing their own changes. gLinux replaced the previously used Ubuntu-based distribution, Goobuntu.
== References ==
The problem is that ALSA used to not support playing multiple sounds at once, but the dmix plugin fixed that, but not before Pulseaudio became popular and made everything more complicated than it needs to be.
And some browser problems are due to pages created using Microsoft tools that deliberately produce HTML that doesn't properly conform to W3C specs, but of course behaves as intended on MS browsers. Although admittedly, that's not as much of an issue these days as it was in the past.
I'm certainly not going to defend every decision the company's made, but I do have a generally favorable opinion of them. And I work on the FOSS stuff :-)
And to be fair, it wasn't just MS's fault. During the height of the Browser Wars, both MS & Netscape kept piling on new features that the other side didn't handle.
@forestdistrustsStackExchange That vaguely rings a bell. We could check this on Wikipedia; it's got a fairly thorough history of GIF, IIRC. :)
en.wikipedia.org/wiki/GIF The original version of GIF was called 87a. In 1989, CompuServe released an enhanced version, called 89a, which added support for animation delays (multiple images in a stream were already supported in 87a), transparent background colors, and storage of application-specific metadata.
> By default, an animation displays the sequence of frames only once, stopping when the last frame is displayed. To enable an animation to loop, Netscape in the 1990s used the Application Extension block (intended to allow vendors to add application-specific information to the GIF file) to implement the Netscape Application Block (NAB).
> This block, placed immediately before the sequence of animation frames, specifies the number of times the sequence of frames should be played (1 to 65535 times) or that it should repeat continuously (zero indicates loop forever).
So it was looping Netscape added, not animation itself. I misremembered. Close enough, I guess.
And the ancient anim format on Amiga, developed by Electronic Arts (before they became evil) looped by default. Amiga image formats, both single frame & anims, also supported color cycling.
Fun fact here is that Gif is also a dutch word, for poison.
Once a colleague of mine tweated something, including a GIF image. Then another colleague saw some half rendered version of that tweet, that included the word GIF, and got furious.
Can we please move the "Start a bounty" link to be in the same location as all of the other action buttons? It makes no sense to me for this to be buried after the comments - It's an action on the Question. It can be very difficult to locate, especially when your question gets a lot of comment...
It's amusing how many changes SE does to their site that reduce usability, but small improvements like this which are literally one-line changes take forever. It's a shame SE doesn't have access to a large, qualified community of programmers and web developers who would be interested in helping them out. Wherever could they find such a community?
Stacks provides everything you need to quickly design, build, and ship coherent experiences across all of Stack Overflow—from the brand and product itself, down to how we send emails and write copy. stackoverflow.design
Stacks is the new rational way to build stuff on the network. I guess they're trying to escape from this kind of architecture: