To test it, let's get two computers on the same local area network, e.g. connected to Wi-Fi on the same home modem router.
On computer B:
- find computer IP with the
ip
CLI tool. Suppose it is 192.168.1.102 - then run Ciro's
nc
HTTP test server
On computer A, run on terminal 1:
sudo tcpdump ip src 192.168.1.102 or dst 192.168.1.102
Then on terminal 2 make a test request:
curl 192.168.1.102:8000
Output on terminal 1:
TODO understand them all! Possibly correlate with Wireshark, or use
17:14:22.017001 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [S], seq 2563867413, win 64240, options [mss 1460,sackOK,TS val 303966323 ecr 0,nop,wscale 7], length 0
17:14:22.073957 IP 192.168.1.102.8000 > ciro-p14s.55798: Flags [S.], seq 1371418143, ack 2563867414, win 65160, options [mss 1460,sackOK,TS val 171832817 ecr 303966323,nop,wscale 7], length 0
17:14:22.074002 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [.], ack 1, win 502, options [nop,nop,TS val 303966380 ecr 171832817], length 0
17:14:22.074195 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [P.], seq 1:82, ack 1, win 502, options [nop,nop,TS val 303966380 ecr 171832817], length 81
17:14:22.076710 IP 192.168.1.102.8000 > ciro-p14s.55798: Flags [P.], seq 1:80, ack 1, win 510, options [nop,nop,TS val 171832821 ecr 303966380], length 79
17:14:22.076710 IP 192.168.1.102.8000 > ciro-p14s.55798: Flags [.], ack 82, win 510, options [nop,nop,TS val 171832821 ecr 303966380], length 0
17:14:22.076727 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [.], ack 80, win 502, options [nop,nop,TS val 303966383 ecr 171832821], length 0
17:14:22.077006 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [F.], seq 82, ack 80, win 502, options [nop,nop,TS val 303966383 ecr 171832821], length 0
17:14:22.077564 IP 192.168.1.102.8000 > ciro-p14s.55798: Flags [F.], seq 80, ack 82, win 510, options [nop,nop,TS val 171832821 ecr 303966380], length 0
17:14:22.077578 IP ciro-p14s.55798 > 192.168.1.102.8000: Flags [.], ack 81, win 502, options [nop,nop,TS val 303966384 ecr 171832821], length 0
17:14:22.079429 IP 192.168.1.102.8000 > ciro-p14s.55798: Flags [.], ack 83, win 510, options [nop,nop,TS val 171832824 ecr 303966383], length 0
-A
option to dump content.Amazing tool that captures packets and disassembles them. Allows you to click an interactive tree that represents Ethernet, TCP/IP and application layer like HTTP.
Start capture immediately from CLI, capture packets to/from 192.168.1.102:
sudo wireshark -f 'host 192.168.1.102' -k
Capture by instead:
sudo wireshark -f http -k
sudo wireshark -f icmp -k
Filter by both protocol and host:
sudo wireshark -f 'host 192.168.1.102 and icmp' -k
For application layer capture filtering, the best you can do is by port:
There is an
sudo wireshark -f 'tcp port 80'
http
filter but only for as a wireshark display filterSample usage:
This produces simple one liners for each request.
sudo tshark -f 'host 192.168.1.102
What you likely want is the
-V
option which fully disassembles each frame much as you can do in the GUI Wireshark:
sudo tshark -V -f 'host 192.168.1.102
TODO didn't manage to get it working with TP Link ARCHER VR2800 even though it shows DHCP as enabled and it also shows MAC addresses and corresponding hostnames in the router management interface.
For IP-level communication, askubuntu.com/questions/22835/how-to-network-two-ubuntu-computers-using-ethernet-without-a-router/116680#116680 just worked between P51 and P14s both on Ubuntu 23.10 connected with a regular Cat 5e cable.
On both machines, first we found the Ethernet cable interface name with the
which outputs on the P41s:
so the interface was
ip
CLI tool:
ip a
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
valid_lft forever preferred_lft forever
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: enp1s0f0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether fc:5c:ee:24:fb:b4 brd ff:ff:ff:ff:ff:ff
3: wlp2s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc noqueue state UP group default qlen 1000
link/ether 04:7b:cb:cc:1b:10 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.123/24 brd 192.168.1.255 scope global dynamic noprefixroute wlp2s0
valid_lft 61284sec preferred_lft 61284sec
inet6 fe80::3597:15d8:74ff:e112/64 scope link noprefixroute
valid_lft forever preferred_lft forever
enp1s0f0
, because wlp
is wireless and lo
is localhost.So on the P14s we assign an IP of 10.0.0.10 to the P51:
sudo ip address add 10.0.0.10/24 dev enp1s0f0
Then on the P51 analogously, giving IP of 10.0.0.20 to the P14s:
sudo ip address add 10.0.0.20/24 dev enp0s31f6
And after that, P14s can:
and P51 can:
ping 10.0.0.10
ping 10.0.0.20
TODO after a few seconds, the settings appear to be forgotten, and
ping
stops working unless you do sudo ip address add
on the local machine again. This seems to happen after a popup appears saying "Activation of network connection failed" as it fails to obtain Internet from the cable.TODO list and delete such manual assignments we've made.
This one is not generally seen by software, which mostly operates starting from OSI layer 2.
A good project to see UARTs at work in all their beauty is to connect two Raspberry Pis via UART, and then:
- type in one and see characters appear in the other: scribles.net/setting-up-uart-serial-communication-between-raspberry-pis/
- send data via a script: raspberrypi.stackexchange.com/questions/29027/how-should-i-properly-communicate-2-raspberry-pi-via-uart
Part of the beauty of this is that you can just connect both boards directly manually with a few wire-to-wire connections with simple jump wire. Its simplicity is just quite refreshing. Sure, you could do something like that for any physical layer link presumably...
Remember that you can only have one GNU screen connected at a time or else they will mess each other up: unix.stackexchange.com/questions/93892/why-is-screen-is-terminating-without-root/367549#367549
On Ubuntu 22.04 you can screen without sudo by adding yourself to the
dialout
group with:
sudo usermod -a -G dialout $USER
The frequency range of Wi-Fi, which falls in the microwave range, is likely chosen to allow faster data transfer than say, FM broadcasting, while still being relatively transparent to walls (though not as much).
There is no userland process for it, it is handled directly by the Linux kernel: unix.stackexchange.com/questions/439801/what-linux-process-is-responsible-for-responding-to-pings/768739#768739
Bibliography:
- some good interview excerpts with some of the pioneers on Glory of the Geeks
This is a standard way to embed images in HTML pages with the
img
tag.Hardcoded and unique network addresses for every single device on Earth.
Started with 48 bits (6 bytes), usually given as 01:23:45:67:89:AB but people now encouraged to use 64-bit ones.
How they are assigned: www.quora.com/How-are-MAC-addresses-assigned Basically IEEE gives out the 3 first bytes to device manufacturers that register, this is called the organizationally unique identifier, and then each manufacturer keeps their own devices unique.
As of 2023, working with DNS data is just going through a mish-mash of closed datasets/expensive APIs.
We really need some open data in that area.
- opendata.stackexchange.com/questions/1951/dataset-of-domain-names
- opendata.stackexchange.com/questions/2110/domain-name-system-record-a-database
- webmasters.stackexchange.com/questions/33395/find-the-ip-address-of-expired-domains/142751#142751
- superuser.com/questions/686195/how-to-find-the-last-ip-used-for-an-expired-domain-name/1793224#1793224
Some interesting analysis by Parth Shukla twitter.com/pparth | www.linkedin.com/in/parth-shukla-59583b20/:
Apparently most of the routers were Chinese. No surprise there.
Data format overview: opendata.stackexchange.com/questions/1951/dataset-of-domain-names/21077#21077
Some interesting usages:
The CIA really likes this registrar, e.g.:
- CIA 2010 covert communication websites
- 2014 www.newsweek.com/former-cia-officials-ready-defend-agency-after-torture-reports-release-290383
A group of former CIA officials are gearing up to defend the agency when the Senate releases its long-awaited report investigating "enhanced interrogation" tactics used on prisoners after 9/11. The highlight of their PR push will be a website, "CIASAVEDLIVES.COM," which is set to go live when the report is released on Tuesday, Foreign Policy reported.The domain was registered on November 2 under a private registration name, through DomainsByProxy, a company that guards the identity of registrants.
Some cool ones:
- playinside.me
Archive example: web.archive.org/web/20130726224338/http://librarianhelper.com/
As of 2021, Ciro Santilli feels strongly that Amazon originals are so much sillier compared to Netflix ones in average.
Of course, everything pales in comparison to The Criterion Collection.
Jeff has spoken a lot in public about Amazon, perhaps even more than other comparable founder, see e.g. history of Amazon. Kudos for that.
Her neck is huge! She also redid her teeth at some point apparently. Some good photos at: www.irishtimes.com/life-and-style/people/mackenzie-scott-how-the-former-mrs-bezos-became-a-philanthropist-like-no-other-1.4850049
MacKenzie Bezos' new husband after she divorced Bezos.
Science teacher at the Lakeside School in Seattle.
www.dailymail.co.uk/femail/article-9338723/Who-billionaire-Mackenzie-Scotts-new-husband-Dan-Jewett.html Who IS billionaire Mackenzie Scott's new husband Dan Jewett?
MacKenzie Bezos went on to marry a science teacher who taught their children.
The contrast with Bezos's girlfriend is simply comical. MacKenzie married the idealistic morally upright science teacher, while Bezos went for a silly sex bomb. Ah, bruta flor, do querer!
MacKenzie Bezos's charity instrument.
www.irishtimes.com/life-and-style/people/mackenzie-scott-how-the-former-mrs-bezos-became-a-philanthropist-like-no-other-1.4850049 MacKenzie Scott: How the former Mrs Bezos became a philanthropist like no other (2020) has some good mentions:
But as Scott's fame for giving away money has grown, so too has the deluge of appeals for gifts from strangers and old friends alike. That clamour may have driven Scott's already discreet operation further underground, with recent philanthropic announcements akin to sudden lightning bolts for unsuspecting recipients.
The name of the organization is a reference to the old man lost his horse.
I wonder where the spray painted sign went: twitter.com/profgalloway/status/1229952158667288576/photo/1. As mentioned at officechai.com/startups/amazon-first-office/ and elsewhere, Jeff did all he could to save money, e.g. he made the desks himself from pieces of wood. Mentioned e.g. at youtu.be/J2xGBlT0cqY?t=345 from Video 4. "Jeff Bezos presentation at MIT (2002)".
Bibliography:
- archive.ph/ucSHN This is what it was like to work at Amazon 20 years ago (2015). Good annecdotes from the first offices.
Apparently posted to
ba.jobs.offered
Usenet newsgroup?Jeff's email was
bezos@netcom.com
at the time.First Amazon hire, wrote and led the team that wrote v1.
He looks like an older and more experienced dude compared to Bezos at the time.
Bibliography:
. www.geekwire.com/2011/meet-shel-kaphan-amazoncom-employee-1/2/ also mentions that unlike California, there's no sales tax in the state of Washington, which is important for selling books.
- a few mentions at: Video "Jeff Bezos presentation at MIT (2002)"
Amazon is apparently notorious for having bought off many competitors, many of them just to kill off the competition and clear the way, not to actually reuse them.
youtu.be/tfAhTtBlb2Q?t=849 from Video "Jeff Bezos Revealed by Bloomberg (2015)" clearly shows Tim O'Reilly saying that very clearly about Bezos.
I do know of a number of cases in which he [Bezos] has acquired companies in order to take out competitors, potential future competitors. Rather than because he actually wants that business to continue.Perhaps O'Reilly who is the bookselling business is not the greatest fan of Jeff. But still. My God.
www.yalelawjournal.org/pdf/e.710.Khan.805_zuvfyyeh.pdf Amazon's Antitrust Paradox by Lina M . Khan from The Yale Law Journal raises this incredible issue.
Like Google custom silicon, Amazon server operations are so large that with the slowdown of Moore's law, it started being worth it for them to develop custom in-house silicon to serve as a competitive advantage, not to be sold for external companies. Can you imagine the scale required to justify silicon development investment that is not sold externally!
Page contains a good summary of their hardware to date. They seem to still be the centerpiece of silicon development. There are still however people outside of Israel doing it, e.g.: www.linkedin.com/in/laurasharpless/ says as of 2021:
My team develops software for our next-generation Machine Learning accelerators: HAL, firmware, and SoC models.
2021: networking chip reports emerge: www.theverge.com/circuitbreaker/2021/3/30/22358633/amazon-reportedly-custom-network-switch-silicon-aws, presumably contesting with the likes of Cisco?
2018 onwards: Amazon AI accelerator silicon.
ARM-based servers.
One of the least evil of the big tech companies of the early 21st century, partly because Sergey Brin's parents fled from the Soviet Union and so he is anti censorship, although they have been tempted by it.
Google only succeeds at highly algorithmic tasks or at giving infinite storage to users to then mine their data.
It is incapable however of adding any obvious useful end user features to most of its products, most of which get terminated and cannot be relied on:
This also seems to extend to business-to-business: twitter.com/MohapatraHemant/status/1343969802080030720 ex-Googler tells how they lost the cloud to Amazon.
More mentions of that:
- world.hey.com/dhh/google-suffers-from-a-digital-petro-curse-908e919a "Google suffers from a digital petro curse" by David Heinemeier Hansson (2021), the creator of Ruby on Rails
- killedbygoogle.com/ dedicated website, source on GitHub: github.com/codyogden/killedbygoogle
Too many genius engineers. They need some dumber people like Ciro Santilli who need to write documentation to learn stuff.
Ciro Santilli actually attempted two interviews to work at Google in the early 2010's but very quickly failed both on the first phase, because you have to be a fast well trained coding machine to pass that interview.
Ciro later felt better about himself by fantasizing how he would actually do more important things outside of Google and that they would beg to buy him instead.
He was also happy that he wouldn't have to use Google crazy internal tools: someone once said that Google's tools make easy tasks middle hard, and they also make impossible tasks middle hard. TODO source.
But whatever the case: Ciro will not, ever, spend his time drilling programmer competition problems to join a company.
www.wired.com/story/google-shakes-up-its-tgif-and-ends-its-culture-of-openness/ "GOOGLE TGIF 1999 video". TGIF is the weekly all hands meeting abolished in 2019: www.wired.com/story/google-shakes-up-its-tgif-and-ends-its-culture-of-openness/
The 1997 Wayback Machine archives are just priceless: web.archive.org/web/19971210065425/http://backrub.stanford.edu/backrub.html. I'm so glad that website exists and started so early. It is just another university research project demo website like any other. Priceless.
Craig Silverstein was the first employee hired, in 1998: www.newyorker.com/magazine/2018/12/10/the-friendship-that-made-google-huge
In August 1998 they had an their first investment of $100,000 from Andy Bechtolsheim, Sun Microsystems co-founder. Some sources say September 1998. This was an event of legend, the dude dropped by, tested the website for a few minutes, said I like it, and dropped a 100$ check with no paperwork. Google wasn't even incorporated, they had to incorporate to cash the check. They were apparently introduced by one of the teachers, TODO which. Some sources say he had to rush off to another meeting afterwards:
Tried to sell it for 1 million in early 1999... OMG the way the world is. It would be good to learn more about that story, and when they noticed it was fuckup.
One of Google's most interesting stories is how their startup garage owner became an important figure inside Google, and how Sergei married her sister. These were the best garage tenants ever!
Bibliography:
- Video "Anne Wojcicki interview by Talks at Google (2018)" has a few mentions, e.g. youtu.be/pDoALM0q1LA?t=173
- www.theverge.com/2019/12/4/20994361/google-alphabet-larry-page-sergey-brin-sundar-pichai-co-founders-ceo-timeline The rise, disappearance, and retirement of Google co-founders Larry Page and Sergey Brin. Good timeline!
One wonders if this name has some influence from the LGBT culture in San Francisco!!!
The guy who coded the initial the BackRub, but left before the company formed. TODO how did he meet Largey Brage? Why did he leave Google?
He co0founded eGroups in 1997 with Carl Victor Page, Jr., Larry Page's brother, and sold it to Yahoo! in 2000 for $432m.
Married a Vietnamese Chick called Allison Huynh from university in 2001. Was unfaithful, and now does not want to split the cash? www.cnbctv18.com/technology/who-is-scott-hassan-the-google-founder-accused-of-divorce-terrorism-10543641.htm Bro, be a man.
That article does mention that he has 13 B in Google shares he bought before IPO, but a net worth of only 1 B. He must have made some insane losses somewhere! It does feel like they gave him a privileged deal because of his early contributions, having that much for just 800 USD sounds unlikely. Hassan likely sold a good part of the shares too early.
www.dailymail.co.uk/news/article-9912929/Billionaire-investor-helped-launch-Google-accused-divorce-terrorism-bitter-break-up.html has even better information. He tried to strike a post-nuptial after google went public in 2004, which she declined. So things were already not perfect then. It mentions that the shares would be worth 13 B today, not that he holds them necessarily. He must have sold early.
To be fair, he did work on a lot of cool stuff, not the least the company that crated the Robot Operating System, which is a cool sounding project.
The fact that he does not have a Wikipedia page as of 2022 is mind blowing, especially after divorce details. Maybe Ciro Santilli will create it one day. Just no patience now. OK, done it: en.wikipedia.org/wiki/Scott_Hassan let's see if it lasts. The page lasted but an anonymous user with IP from California removed divorce details and google share ownership details in December 2022, both of which had a New York Times source: www.nytimes.com/2021/08/20/technology/Scott-Hassan-Allison-Huynh-divorce.html. This is the first time Ciro Santilli ever goes into an Edit war: en.wikipedia.org/wiki/Talk:Scott_Hassan#Divorce_details_removed_as_%22poorly_sourced_material%22_by_anonymous_user_even_though_they_had_a_source_from_the_New_York_Times It feels exactly like the type of thing Scott would have done himself. And he possibly unadvertedly exposed his real IP in doing so: 24.6.226.102. It is pingable, but Nmap analysis shows nothing of interest.
Company co-founded by Scott Hassan, early Google programmer at Stanford University, and Carl Victor Page, Jr., Larry Page's older brother.
They were an email list management website, and became Yahoo! Groups after the aquisition.
The company was sold to Yahoo! in August 2000 for $432m and became Yahoo! Groups. They managed to miraculously dodge the Dot-com bubble. After the acquisition, Yahoo started to redirect them to: groups.yahoo.com as can be seen on the Wayback Machine: web.archive.org/web/20000401000000*/egroups.com The first archive of groups.yahoo.com is from February 2001: web.archive.org/web/20010202055100/http://groups.yahoo.com/ and it unsurprisingly looks basically exactly like eGroups.
Has some good mentions, but often leaves you wanting more details of how certain things happened, especially the early days stuff.
Does however paint a good picture of several notable employees, and non-search projects from the early 2000's including:
- the cook dude
- porn cookie guy
- the unusual IPO process
Paints a very positive picture of the founders. It is likely true. They gave shares generously to early employees. Tried to allow the more general public to buy from IPO, by using a bidding scheme, rather than focusing on the big bankers as was usual.
The introduction mentions that Google is very interested in molecular biology and mining genetics data, much like Ciro Santilli! Can't find external references however...
Two of the most compelling areas that Google and its founders are quietly working on are the promising fields of molecular biology and genetics. Millions of genes in combination with massive amounts of biological and scientific data are an excellent match for the Google search engine, the tremendous database the company has in place, and its immense computing power. Already, Google has downloaded a map of the human genome and is working closely with biologist Dr. Craig Venter and other leaders in genetics on scientific projects that may lead to important breakthroughs in science, medicine, and health. In other words, we may be heading toward a time when people can google their own genes.
The book gives good highlight as to why Google became big: search was just an incredibly computationally intensive task. From very early days, Largey were already making up their own somewhat custom compute systems from very early days, which naturally led into Google custom hardware later on. Google just managed to pull ahead on the reinvest revenue into hardware loop, and no one ever caught them back. This feels more the case than e.g. with Amazon, which notoriously had to buy off dozens of competitors to clear the way.
Their UI is horrendous as of 2020:
I need this for OurBigBook.com!! webmasters.stackexchange.com/questions/104241/can-google-analytics-track-url-fragment-in-url
They removed referrer path completely in GA4... OMG WTF are they doing!