User talk:Verdana Bold

From Wikipedia, the free encyclopedia
Jump to: navigation, search


Equivalence class edits[edit]

Hi Verdana, I just wanted to thank you for your edits to the Equivalence class article. I sometimes get a bit sloppy and don't take enough time to review what I've written to pick up these grammatical errors, dot my "i"s and cross my "t"s. I do appreciate it when someone comes along and corrects these things. I notice from your user page that you intend to stick with these types of corrections, which is fine, but I do hope that, as your confidence grows, you will delve into some of the content. We need editors who understand mathematics and can write about it in a clear and forthright manner; perhaps you will be one of these. Happy editing. Bill Cherowitzo (talk) 04:09, 31 March 2014 (UTC)

No way. I've seen what happens to people who attract the attention of the admins. I've seen it several times, with several people. Arbcom is like a Salem witch hunt run by the kid who runs North Korea. I saw a published PhD expert get hauled to arbcom and banned by a know-nothing enforcing a POV. I saw another guy get banned just because someone didn't like him. No bad edits, no explanation. This isn't even a secret. Everyone at WP knows it, and nobody else cares. After the latest spectacle I observed, I told a newspaper what goes on and he said, "You want to tell me that Wikipedia admins are "a gang of schoolyard bullies with no integrity?" I'm afraid that's not news! We even did a story about it ten years ago."
Sadly, he was right. I've since come across several academic studies that revealed the blatant bias of admins in conflicts here. One admin even said (paraphrased since I really and truly don't remember the exact quote, but I swear on everything I care about that this *was* the meaning), "The people who set up Wikipedia had good intentions, but they were naïve. We have to deal with conflicts between editors the way we do because if we followed our own rules, we'd be overwhelmed by adjudication."
Thus, attempts to get them to obey their own rules are "wikilawyering" — a dismissive term used to avoid discussion of their clearly unjustifiable behavior. Supporting someone clearly wronged by a corrupt admin is being a "meatpuppet."
Wikipedia is the single most important source of information after television. It is the first resource everyone in the world uses to find the truth. What's in it matters. And it matters that some articles are "owned" by admins or their friends, causing them to be shockingly biased propaganda. We need to elect new management who will make admins obey their own rules about due process and ban the use of the word "wikilawyering" by admins.
WP has a responsibility to the citizens of the world not to run it like a medieval Machiavellian prince, totally devoid of any sense of principle, fairness, accuracy, truth, or even common decency.
So I'm basically Anne Frank, writing in Wikipedia and hoping not to attract attention from the goon squads who own articles. If I do only grammar corrections and no content editing except in math, maybe the roving gangs of Sturmabteilung and brownshirts won't up-against-the-wall me. VerdanaBøld 16:04, 5 August 2014 (UTC)

Paroxetine[edit]

Message for you on the Talk page :>) Formerly 98 (talk) 21:00, 16 May 2014 (UTC)

Side effects in the lead[edit]

Yes they do belong in the lead. The lead is to summarize the body of the text. Doc James (talk · contribs · email) (if I write on your page reply on mine) 04:00, 17 May 2014 (UTC)

Reference fixing[edit]

Hi: I've updated your edit to the hypersexuality article, and replaced your hard-coded numeric reference with a name-based reference to the cite in question. Here's the diff for my edit. Using direct numeric references is fragile, because even small edits to the article (such as the one I just made) can easily make major changes to the ordering of references in articles (in this case, reference 10 became reference 7, because its first mention moved higher up the article).

If you'd like more information about how to make this sort of name-based referencing work, just drop me a line on my talk page. -- The Anome (talk) 15:01, 28 January 2015 (UTC)

ACCK!! Thank you! --VerdanaBold 06:18, 4 March 2017 (UTC)

Negative metric?[edit]

Please look: Talk:Metric space#Questions to User:Verdana Bold. Boris Tsirelson (talk) 14:56, 25 June 2015 (UTC)

FYI, please see WP:TALKO. Don't respond in the middle of others' posts. It is disrepectful, and creates conversations that are impossible to read. Instead, please thread your replies the way others do. Sławomir Biały (talk) 17:41, 28 June 2015 (UTC)

UUID[edit]

How many and how fast do you have to generate them? In a version 1 UUID, the timestamp part is a 60-bit number representing 100 nanosecond intervals. That is 10 million per second. And the clock sequence is 14 bits (16 bits minus the 2 variant bits), which is 16384. So, just using regular version 1 UUIDs with a network card-derived MAC address is going to give you 164 billion UUIDs per second per node. What kind of node is this that can generate them that fast? What application is this that needs to generate so many? Where are you going to put all these UUIDs? Since each UUID is 16 bytes if stored in binary, you are going to be generating 2.6 terabytes of UUIDs per second. That is just for the UUIDs. What is the disk or network technology which is going to let you write them out fast enough from one node? Supposing that (1) your application actually does require this many per second; (2) that they all have to be generated on one node; and (3) you have one node that can generate them that fast, you can't possibly do it continuously for very long. In just one day you would have 160 petabytes of UUIDs, which is an order of magnitude bigger than the biggest database on the planet. All generated by one node. That is quite a node! Given that, even in the case where you have a (very short) burst where you require more than 164 billion UUIDs per second, you can use UUIDs which you have "pocketed". Assign some UUIDs you didn't use last week. Person54 (talk) 16:41, 14 March 2017 (UTC)

God damn!
What is a node (in this context)?
> 160 petabytes of UUIDs, which is an order of magnitude bigger than the biggest database on the planet.
Only ONE order of magnitude!! I assume that's base 10. It means there's a 16 PB database somewhere. Where?
And does anyone really need millions or even thousands of UUIDs per second? why? Are they generating unique keys for new database records?
Also, it's not possible to create them that fast on a PC. You'd need a supercomputer. My rig is real fast (5 GHz), If someone is generating a billion/sec, then they have to do it each one in 5 of my machine clocks.
Yet you said that someone really does need the 32,768 counter to insure uniqueness. Who is that? What kind of appalling machine do they have?
On the other hand, if you're talking about generating one UUID every 100 nanoseconds, then if they can parallellize the gen procedure, I could easily do it on my PC. My GTX 1080 graphics card can do nine trillion machine instructions per second. That's fast enough that it can execute up to 9,000 instructions to generate each UUID.
But who needs UUIDs that fast?
By pure coincidence, my PC clock fires every 200 nanoseconds, exactly half of the 64-bit clock speed.
I remember when we used to use milliseconds past midnight and keep it in 36 bits.
Also note that if it needs a million UUIDs every second, then a process somewhere is doing somethin complex a million times/sec. If they're doing it 85 million times/sec, then they're completing a task in the time it takes one bit to arrive on my 85 Mbit internet connection.
Probably the NSA. It's all the phone calls and IP packets they stole.
--VerdanaBold 20:24, 14 March 2017 (UTC)