Of course we're covering the game. (BTW does anybody know when it starts in EST?)

< WannaBe: This is always the fun part, <ALL> games are 15:00 Berlin... Berlin is GMT+1, as of right now, Berlin is 6 hours ahead of CG.com Server time (Eastern Time Zone), 9 hours ahead of Pacific.

So, 15:00 Berlin is 09:00 East coast, 06:00 Pacific.

However, US will move clock ahead one hour on March 11. Games starting Sunday 11th will start 10:00/07:00 (Round 2)

On March 25th, Germany will move their clocks ahead one hour, so for the games on 26th and 27th, will be back to 09:00/06:00 (Rounds 13 & 14)

Everyone got that? Test Next week, 9th of March. Anyone who aces the test will get the "Einstein" Visa to enter US.

How does laboratory algorithm work, if two people request the same game to be analyzed, but different fuel?

e.g. if I submit a request Aronian/Kramnik with Fuel level 4, before/after someone else request the same game, but with fume, or 1

Is it smart enough to tell the second person to wait, as analysis with 4 is better than fume, or 1? (In theory...)

Or if two people submit the same game, with same Fuel, will it notify the second requestor that there is one already in progress and to check back later?

Is the analysis cached? If someone request Fuel 2 full analysis, minutes later I request same game, Fuel 4, is it smart enough to use the previously done work or does it go its own merry way like a (brand) new analysis?

It tries to do "the right thing." First of all, if the game is already analyzed at a higher fuel rate, it will flatly reject the request with a statement such as "This game has already been analyzed at 30s/ply."

Now, if you request an analysis and while it is taking place I come along and request one, I will get a message, "Sorry, this game is already being analyzed at XXs/ply." It doesn't matter who requested the deeper analysis. The one in progress must finish before a new one begins.

<e.g. if I submit a request Aronian/Kramnik with Fuel level 4, before/after someone else request the same game, but with fume, or 1

Is it smart enough to tell the second person to wait, as analysis with 4 is better than fume, or 1?>

It tries to be smart about it, like described above. If somebody wanted a 30s/move analysis while somebody else ordered up a 15s/move analysis we considered making it cancel the "shallow" analysis to make room for the deep one, but that would be sort of weird. What if the quicker analysis was almost finished? And what about the poor person who has been waiting an hour or two for his analysis only to be rudely ejected from the process because somebody else did something?

We didn't want that. So the rule is, first come first served.

<Or if two people submit the same game, with same Fuel, will it notify the second requestor that there is one already in progress and to check back later?>

Exactly.

<Is the analysis cached? If someone request Fuel 2 full analysis, minutes later I request same game, Fuel 4, is it smart enough to use the previously done work or does it go its own merry way like a (brand) new analysis?>

Yes it is, and that's one of the neatest features about it. Just for example, you can do something like this:

1. Request a 15s/move analysis of a game.
2. At a critical juncture, grab the FEN and ask for 1-hour analysis of that position.
3. Go back and re-request a 15s/move analysis.

It will take only a few seconds to complete step #3 because all of the chess-engine work has already been done. What's more, it won't charge you any fuel for step #3, since the engine is not needed.

If you're still not satisfied you can do more hour-long probes of various positions in the game and continue to go back to your 15s/move analysis. It will always say "Analysis by Stockfish at 15s/ply" but the ply-depths by the heavily probed positions will tell the full story.

<And the request for new 'field', like a flask to be added, to be displayed if there is a completed game analysis, like balloon displayed next to games with kibit'n or pencil for game with annotation.>

I don't know if this message will ever get through.... I agree with <BP>. It's a total shocker.

It is time to cut back Rogoff and KCafé to just a few pages each.

https://www.ixiacom.com/company/blo...

SAN FRANCISCO--(BUSINESS WIRE)--Distributed denial of service (DDoS) attacks using domain name server (DNS) amplification increased more than 357 percent in the fourth quarter of 2017, compared to 2016, according to Nexusguard’s “Q4 2017 Threat Report.”

The quarterly report, which measured thousands of attacks from around the world, attributes the skyrocketing attacks to servers enabled with Domain Name System Security Extensions (DNSSEC), a significant new risk if not properly configured.

Although they’re intended to add integrity and security to the DNS protocol, DNSSEC-enabled servers can be deliberately targeted to reflect amplification attacks, due to the large size of the responses they generate.

Nexusguard’s quarterly distributed denial of service (DDoS) reports are based on the company’s collection of real-time data regarding threats facing enterprises and service provider networks around the world.

The company gathers data from botnet scanning, Honeypots, internet service providers (ISPs) and traffic moving between attackers and their targets to help companies identify vulnerabilities and stay informed about global attack trends.

Although the overall number of DDoS attacks fell 12 percent compared to the same period last year, a new class of powerful botnets may appear from wider DNSSEC adoption. Nexusguard warns teams to evaluate the DNSSEC response and security flaw to strengthen systems against future attacks.

“Enterprises have worked hard to patch against snooping, hijacking and other DNS abuses; however, improperly configured DNSSEC-enabled nameservers may be a new plague for unprepared teams,” said Juniman Kasman, chief technology officer for Nexusguard.

“Admins and IT teams need to check security for the entire network, as well as correctly configure DNSSEC on the domain to properly harden servers against these new attacks.”

Hackers also continue to favor multi-vector attacks, blending combinations of network time protocol (NTP), universal datagram protocol (UDP), DNS and other popular attack vectors in more than half of all botnets over the past year, according to Nexusguard’s “2017 DDoS Attack Landscape” infographic.

China and the U.S. continued to reign as the top two sources of DDoS attacks in Q4, contributing 21.8 percent and 14.3 percent of the botnets, respectively. South Korea climbed to third place, contributing nearly six percent of the global attacks, up from sixth place last quarter.

Founded in 2008, Nexusguard is the global leader in fighting malicious Internet attacks. Nexusguard protects clients against a multitude of threats, including distributed denial of service (DDoS) attacks, to ensure uninterrupted Internet service.

Nexusguard provides comprehensive, highly customized solutions for customers of all sizes, across a range of industries, and also enables turnkey anti-DDoS solutions for service providers.

Nexusguard delivers on its promise to maximize peace of mind by minimizing threats and improving uptime. Visit www.nexusguard.com for more information.>

Anyhow, the good news is today the site is a *little* sluggish, but with almost 2000 users here that's to be expected. I think we shook off those stupid memcached amplification-attackers.

There are still other issues (notably intermittent "server not responding" messages) but at least things are usable.