Let us begin with a few suggested definitions.

1) The classics are the books of which we usually hear people say: “I am rereading…” and never “I am reading….”

This at least happens among those who consider themselves “very well read.” It does not hold good for young people at the age when they first encounter the world, and the classics as a part of that world.

The reiterative prefix before the verb “read” may be a small hypocrisy on the part of people ashamed to admit they have not read a famous book. To reassure them, we need only observe that, however vast any person’s basic reading may be, there still remain an enormous number of fundamental works that he has not read.

Hands up, anyone who has read the whole of Herodotus and the whole of Thucydides! And Saint-Simon? And Cardinal de Retz? But even the great nineteenth-century cycles of novels are more often talked about than read. In France they begin to read Balzac in school, and judging by the number of copies in circulation, one may suppose that they go on reading him even after that, but if a Gallup poll were taken in Italy, I’m afraid that Balzac would come in practically last. Dickens fans in Italy form a tiny elite; as soon as its members meet, they begin to chatter about characters and episodes as if they were discussing people and things of their own acquaintance. Years ago, while teaching in America, Michel Butor got fed up with being asked about Emile Zola, whom he had never read, so he made up his mind to read the entire Rougon-Macquart cycle. He found it was completely different from what he had thought: a fabulous mythological and cosmogonical family tree, which he went on to describe in a wonderful essay.

In other words, to read a great book for the first time in one’s maturity is an extraordinary pleasure, different from (though one cannot say greater or lesser than) the pleasure of having read it in one’s youth. Youth brings to reading, as to any other experience, a particular flavor and a particular sense of importance, whereas in maturity one appreciates (or ought to appreciate) many more details and levels and meanings. We may therefore attempt the next definition:

2) We use the word “classics” for those books that are treasured by those who have read and loved them; but they are treasured no less by those who have the luck to read them for the first time in the best conditions to enjoy them.

In fact, reading in youth can be rather unfruitful, owing to impatience, distraction, inexperience with the product’s “instructions for use,” and inexperience in life itself. Books read then can be (possibly at one and the same time) formative, in the sense that they give a form to future experiences, providing models, terms of comparison, schemes for classification, scales of value, exemplars of beauty—all things that continue to operate even if the book read in one’s youth is almost or totally forgotten. If we reread the book at a mature age we are likely to rediscover these constants, which by this time are part of our inner mechanisms, but whose origins we have long forgotten. A literary work can succeed in making us forget it as such, but it leaves its seed in us. The definition we can give is therefore this:

3) The classics are books that exert a peculiar influence, both when they refuse to be eradicated from the mind and when they conceal themselves in the folds of memory, camouflaging themselves as the collective or individual unconscious.

There should therefore be a time in adult life devoted to revisiting the most important books of our youth. Even if the books have remained the same (though they do change, in the light of an altered historical perspective), we have most certainly changed, and our encounter will be an entirely new thing.

Hence, whether we use the verb “read” or the verb “reread” is of little importance. Indeed, we may say:

4) Every rereading of a classic is as much a voyage of discovery as the first reading.

5) Every reading of a classic is in fact a rereading.

Definition 4 may be considered a corollary of this next one:

6) A classic is a book that has never finished saying what it has to say.

Whereas definition 5 depends on a more specific formula, such as this:

7) The classics are the books that come down to us bearing upon them the traces of readings previous to ours, and bringing in their wake the traces they themselves have left on the culture or cultures they have passed through (or, more simply, on language and customs).

All this is true both of the ancient and of the modern classics. If I read the Odyssey I read Homer’s text, but I cannot forget all that the adventures of Ulysses have come to mean in the course of the centuries, and I cannot help wondering if these meanings were implicit in the text, or whether they are incrustations or distortions or expansions. When reading Kafka, I cannot avoid approving or rejecting the legitimacy of the adjective “Kafkaesque,” which one is likely to hear every quarter of an hour, applied indiscriminately. If I read Turgenev’s Fathers and Sons or Dostoevsky’s The Possessed, I cannot help thinking how these characters have continued to be reincarnated right down to our own day.

The reading of a classic ought to give us a surprise or two vis-à-vis the notion that we had of it. For this reason I can never sufficiently highly recommend the direct reading of the text itself, leaving aside the critical biography, commentaries, and interpretations as much as possible. Schools and universities ought to help us to understand that no book that talks about a book says more than the book in question, but instead they do their level best to make us think the opposite. There is a very widespread topsyturviness of values whereby the introduction, critical apparatus, and bibliography are used as a smoke screen to hide what the text has to say, and, indeed, can say only if left to speak for itself without intermediaries who claim to know more than the text does. We may conclude that:

8) A classic does not necessarily teach us anything we did not know before. In a classic we sometimes discover something we have always known (or thought we knew), but without knowing that this author said it first, or at least is associated with it in a special way. And this, too, is a surprise that gives a lot of pleasure, such as we always gain from the discovery of an origin, a relationship, an affinity. From all this we may derive a definition of this type:

9) The classics are books that we find all the more new, fresh, and unexpected upon reading, the more we thought we knew them from hearing them talked about.

Naturally, this only happens when a classic really works as such—that is, when it establishes a personal rapport with the reader. If the spark doesn’t come, that’s a pity; but we do not read the classics out of duty or respect, but only out of love. Except at school. And school should enable you to know, either well or badly, a certain number of classics among which—or in reference to which—you can then choose your classics. School is obliged to give you the instruments needed to make a choice, but the choices that count are those that occur outside and after school.

It is only by reading without bias that you might possibly come across the book that becomes your book. I know an excellent art historian, an extraordinarily well-read man, who out of all the books there are has focused his special love on the Pickwick Papers; at every opportunity he comes up with some quip from Dickens’s book, and connects each and every event in life with some Pickwickian episode. Little by little he himself, and true philosophy, and the universe, have taken on the shape and form of the Pickwick Papers by a process of complete identification. In this way we arrive at a very lofty and demanding notion of what a classic is:

10) We use the word “classic” of a book that takes the form of an equivalent to the universe, on a level with the ancient talismans. With this definition we are approaching the idea of the “total book,” as Mallarmé conceived of it.

But a classic can establish an equally strong rapport in terms of opposition and antithesis. Everything that Jean-Jacques Rousseau thinks and does is very dear to my heart, yet everything fills me with an irrepressible desire to contradict him, to criticize him, to quarrel with him. It is a question of personal antipathy on a temperamental level, on account of which I ought to have no choice but not to read him; and yet I cannot help numbering him among my authors. I will therefore say:

11) Your classic author is the one you cannot feel indifferent to, who helps you to define yourself in relation to him, even in dispute with him.

I think I have no need to justify myself for using the word “classic” without making distinctions about age, style, or authority. What distinguishes the classic, in the argument I am making, may be only an echo effect that holds good both for an ancient work and for a modern one that has already achieved its place in a cultural continuum. We might say:

12) A classic is a book that comes before other classics; but anyone who has read the others first, and then reads this one, instantly recognizes its place in the family tree.

At this point I can no longer put off the vital problem of how to relate the reading of the classics to the reading of all the other books that are anything but classics. It is a problem connected with such questions as, Why read the classics rather than concentrate on books that enable us to understand our own times more deeply? or, Where shall we find the time and peace of mind to read the classics, overwhelmed as we are by the avalanche of current events?

We can, of course, imagine some blessed soul who devotes his reading time exclusively to Lucretius, Lucian, Montaigne, Erasmus, Quevedo, Marlowe, the Discourse on Method, Wilhelm Meister, Coleridge, Ruskin, Proust, and Valéry, with a few forays in the direction of Murasaki or the Icelandic sagas. And all this without having to write reviews of the latest publications, or papers to compete for a university chair, or articles for magazines on tight deadlines. To keep up such a diet without any contamination, this blessed soul would have to abstain from reading the newspapers, and never be tempted by the latest novel or sociological investigation. But we have to see how far such rigor would be either justified or profitable. The latest news may well be banal or mortifying, but it nonetheless remains a point at which to stand and look both backward and forward. To be able to read the classics you have to know “from where” you are reading them; otherwise both the book and the reader will be lost in a timeless cloud. This, then, is the reason why the greatest “yield” from reading the classics will be obtained by someone who knows how to alternate them with the proper dose of current affairs. And this does not necessarily imply a state of imperturbable inner calm. It can also be the fruit of nervous impatience, of a huffing-and-puffing discontent of mind.

Maybe the ideal thing would be to hearken to current events as we do to the din outside the window that informs us about traffic jams and sudden changes in the weather, while we listen to the voice of the classics sounding clear and articulate inside the room. But it is already a lot for most people if the presence of the classics is perceived as a distant rumble far outside a room that is swamped by the trivia of the moment, as by a television at full blast. Let us therefore add:

13) A classic is something that tends to relegate the concerns of the moment to the status of background noise, but at the same time this background noise is something we cannot do without.

14) A classic is something that persists as a background noise even when the most incompatible momentary concerns are in control of the situation.

There remains the fact that reading the classics appears to clash with our rhythm of life, which no longer affords long periods of time or the spaciousness of humanistic leisure. It also contradicts the eclecticism of our culture, which would never be capable of compiling a catalog of things classical such as would suit our needs.

These latter conditions were fully realized in the case of Leopardi, given his solitary life in his father’s house (his “paterno ostello“), his cult of Greek and Latin antiquity, and the formidable library put at his disposal by his father, Monaldo. To which we may add the entire body of Italian literature and of French literature, with the exception of novels and the “latest thing out” in general, all of which were at least swept off into the sidelines, there to comfort the leisure of his sister Paolina (“your Stendhal,” he wrote her once). Even with his intense interest in science and history, he was often willing to rely on texts that were not entirely up-to-date, taking the habits of birds from Buffon, the mummies of Frederik Ruysch from Fontanelle, the voyage of Columbus from Robertson.

In these days a classical education like the young Leopardi’s is unthinkable; above all, Count Monaldo’s library has multiplied explosively. The ranks of the old titles have been decimated, while new ones have proliferated in all modern literatures and cultures. There is nothing for it but for all of us to invent our own ideal libraries of classics. I would say that such a library ought to be composed half of books we have read and that have really counted for us, and half of books we propose to read and presume will come to count—leaving a section of empty shelves for surprises and occasional discoveries.

I realize that Leopardi is the only name I have cited from Italian literature—a result of the explosion of the library. Now I ought to rewrite the whole article to make it perfectly clear that the classics help us to understand who we are and where we stand, a purpose for which it is indispensable to compare Italians with foreigners and foreigners with Italians.

Then I ought to rewrite it yet again lest anyone believe that the classics ought to be read because they “serve any purpose” whatever. The only reason one can possibly adduce is that to read the classics is better than not to read the classics.

And if anyone objects that it is not worth taking so much trouble, then I will quote Cioran (who is not yet a classic, but will become one):

While they were preparing the hemlock, Socrates was learning a tune on the flute. “What good will it do you,” they asked, “to know this tune before you die?”

—translated by Patrick Creagh
English translation copyright © 1986 Harcourt Brace Jovanovich, Inc.

A newly deployed space telescope has struck pay dirt almost immediately, discovering a quasar – a superheated region of dust and gas around a black hole – that is releasing jets at least seventy times hotter than was thought possible.

RadioAstron is unusual among space telescopes in operating at radio wavelengths. Although the telescope itself is small compared to giant ground-based dishes (10 meters or 33 feet across), it is capable of combining with ground-based instruments operating at the same wavelengths. Together they produce images with the resolution of a single telescope as wide as the distance between them, far exceeding collaborations between dishes half a world apart.

One of the first targets for this extraordinary tool was the quasar 3C 273, the first of these enormously bright objects to be identified, and one of the most luminous. Despite being 4 trillion times as bright as the Sun, 3C 273 is hard to study, located an estimated 2.4 billion light-years away at the center of a giant elliptical galaxy.

Something that bright has to be mind-bendingly hot, and the same applies to the radiating jets 3C 273 spits out. Models suggested that it was impossible for these jets' temperatures to exceed 100 billion degrees Kelvin, at which point electrons produce radiation that should quickly cool them in what is known as the inverse Compton catastrophe. However, in The Astrophysical Journal Letters an international team has estimated the true temperature.

“We measure the effective temperature of the quasar core to be hotter than 10 trillion degrees!” said Dr. Yuri Kovalev, RadioAstron project scientist, in a statement. “This result is very challenging to explain with our current understanding of how relativistic jets of quasars radiate.”

Combining with Earth-based telescopes, including Arecibo and the Very Large Array, RadioAstron examined the radiation from 3C 373 at wavelengths of 18, 6.2, and 1.35 centimeters (7, 2.4, and 0.5 inches) providing both an overall temperature estimates that varied from 7 to 14 trillion degrees, and a view of the substructure of the quasar's jets..

^{A quasar as seen by the Hubble Telescope, but the resolution provided by space and Earth-based telescopes in combination is far greater. ESA/Hubble & NASA} 

"Only this space-Earth system could reveal this temperature, and now we have to figure out how that environment can reach such temperatures," said Kovalev in a separate statement. "This result is a significant challenge to our current understanding of quasar jets."

The resolution made possible by this collaboration was so detailed that the team was able to detect the scattering effects on their measurements of variations in the ionized interstellar medium within the Milky Way. “This is like looking through the hot, turbulent air above a candle flame," said first author Dr. Michael Johnson, of the Harvard-Smithsonian Center for Astrophysics. "We had never been able to see such distortion of an extragalactic object before.”

The authors explained that when “averaged over long timescales – days to months – the scattering blurs compact features in the image, resulting in lower apparent brightness temperatures,” part of the reason these extraordinary temperatures had not been recognized before. Over shorter periods the scattering creates the impression of bright and dark spots known as “refractive substructure.”

RadioAstron has been in space since 2011, but it has taken time to analyze the first observations. Knowing that the maximum baseline it can provide is more than double the 171,000 kilometers (106,000 miles) used in this case, astronomers can't wait to see what it discovers next.

^{Quasar 3C 273 as seen at different wavelengths showing the effect of the interference of the interstellar medium on the incoming rays. Johnson et al, The Astrophysical Journal.}

NextCloud is an awesome, secure and private alternative for proprietary platforms like Dropbox and Google Drive. Installing NextCloud can be achieved easily with pkg_add nextcloud - but I'd like to do it manually to benefit performance and stability.

General

Setting up NextCloud manually offers some room for improvements:

• PostgreSQL as the database engine. This benefits a multi-user instance when it comes to performance.
• PHP 7.0 for even more increased performance - it is way faster than PHP 5.6.
• Caching file-locking through APCu and Redis.

Caveats:

• PostgreSQL is more performant for a multi user setup. It is possible to switch database engines afterwards.
• FFS might not be the best fit for bulky data (>8TB) or when high IOPS are a requirement.
• OpenBSD doesn't support RAID stacking (eg, RAID1 and FDE) - but NextCloud offers server- and user level encryption.

Step 1: Preparations

This guide should work with the latest -stable and the most recent -current. You might prefer a clean and fresh OpenBSD installation to prevent existing configurations from interfering. Without further ado, brew yourself a cup of coffee, disable any distractions and let's go!

We're going to install NextCloud on the subdomain nc.h3artbl33d.nl and are using regular user johndoe. Take note of this and replace these values in commands and configuration files.

1.1: Login and set some defaults

After having logged in to the target machine - that is going to run your NextCloud machine - we'll first set some sane defaults.

1. Enable doas.

# echo 'permit johndoe' >> /etc/doas.conf

2. Edit /etc/ssh/sshd_config and set these values - unless you explicitly need them:

   LogLevel VERBOSE
   PermitRootLogin no
   MaxAuthTries 2
   MaxSessions 2
   AllowAgentForwarding no
   AllowTcpForwarding no
   TCPKeepAlive no
   Compression no
   ClientAliveInterval 2
   ClientAliveCountMax 2

3. Download the pf ruleset I've prepared for you and edit it (specifically the interface and remove the IPv6 rules if you are not using IPv6). When you are done, move it to the default config and check the syntax.

   $ ftp https://h3artbl33d.nl/pf-nc.txt
   $ doas mv pf-nc.txt /etc/pf.conf
   $ doas pfctl -nf /etc/pf.conf

1.2: Download/install the prerequisites

We'll need some files further on, let's make sure we have everything ready.

1. Install PostgreSQL server:

   $ doas pkg_add postgresql-server

2. Install PHP and some modules. If you are asked which PHP-version to install, answer PHP 7 each time.

   $ doas pkg_add php php-curl php-gd php-intl php-mcrypt php-pdo_pgsql php-xml php-zip

3. Enable the PHP modules.

   $ doas ln -s /etc/php-7.0.sample/* /etc/php-7.0/

4. Automatically start these services. Do not start these services yet.

   $ doas rcctl enable postgresql-server && doas rcctl enable php70_fpm && rcctl enable redis

5. Install a file so the webserver can resolve DNS queries from inside its chroot.

   $ doas mkdir /var/www/etc && doas cp /etc/resolv.conf /var/www/etc/resolv.conf && chown -R www:www /var/www/etc

Step 2: Prepare the webserver

We need to prepare the webserver. Obviously, we're using httpd(8) - see this article if you prefer nginx for some reason. We're going to kick this chapter off with httpd itself.

2.1: Enable and configure httpd

Between 6.3-stable and 6.4-current there have been some syntax changes in httpd.conf. Hence, hereby the valid configuration for both versions.

6.3-stable: use this configuration as a starting point for your /etc/httpd.conf.

server "nc.h3artbl33d.nl" {
        listen on * port 80
        root "/nextcloud"
        location "/.well-known/acme-challenge/*" {
                root { "/acme", strip 2 }
        }
}

6.4-current: use this configuration as a starting point for your /etc/httpd.conf.

server "nc.h3artbl33d.nl" {
        listen on * port 80
        root "/nextcloud"
        location "/.well-known/acme-challenge/*" {
                root { "/acme" }
                request strip 2
        }
}

Check whether the configuration is deemed valid with doas httpd -n. Moving on to preparing acme-client for the SSL certificates, courtesy of Let's Encrypt. Edit /etc/acme-client:

authority letsencrypt {
        api url "https://acme-v01.api.letsencrypt.org/directory"
        account key "/etc/acme/letsencrypt-privkey.pem"
}
authority letsencrypt-staging {
        api url "https://acme-staging.api.letsencrypt.org/directory"
        account key "/etc/acme/letsencrypt-staging-privkey.pem"
}
domain nc.h3artbl33d.nl {
        alternative names { www.nc.h3artbl33d.nl }
        domain key "/etc/ssl/private/nc.h3artbl33d.nl.key"
        domain certificate "/etc/ssl/nc.h3artbl33d.nl.crt"
        domain full chain certificate "/etc/ssl/nc.h3artbl33d.nl.pem"
        sign with letsencrypt
}

And create the corresponding directories:

$ doas mkdir -p -m 700 /etc/acme
$ doas mkdir -p -m 700 /etc/ssl/acme/private
$ doas mkdir -p -m 755 /var/www/acme

Time to fetch the certificates!

$ doas rcctl enable httpd && doas rcctl start httpd && doas acme-client -vAD nc.h3artbl33d.nl

If that went successful, grab the OCSP stapling file.

$ doas ocspcheck -N -o /etc/ssl/nc.h3artbl33d.nl.ocsp.pem /etc/ssl/nc.h3artbl33d.nl.pem

Edit the crontab to automatically renew the certificates and stapling file. Append the following in the crontab (doas crontab -e).

0 0 * * * acme-client nc.h3artbl33d.nl && rcctl reload httpd
0 * * * * ocspcheck -N -o /etc/ssl/nc.h3artbl33d.nl.ocsp.pem /etc/ssl/nc.h3artbl33d.nl.pem && rcctl reload httpd

2.2: Configuring httpd further

6.3-stable: edit /etc/httpd.conf with these values. Do not restart httpd afterwards!

server "nc.h3artbl33d.nl" {
        listen on * tls port 443
        hsts {
                preload
                subdomains
        }
        root "/nextcloud"
        directory index "index.php"
        tls {
                certificate "/etc/ssl/nc.h3artbl33d.nl.pem"
                key "/etc/ssl/private/nc.h3artbl33d.nl.key"
                ocsp "/etc/ssl/nc.h3artbl33d.nl.ocsp.pem"
                ciphers "ECDHE-RSA-CHACHA20-POLY1305:ECDHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384:ECDHE-RSA-AES256-SHA"
                dhe "auto"
                ecdhe "P-384"
        }
        connection max request body 537919488

        location "/.well-known/acme-challenge/*" {
                root { "/acme", strip 2 }
        }

        # First deny access to the specified files
        location "/db_structure.xml" { block }
        location "/.ht*"             { block }
        location "/README"           { block }
        location "/data*"            { block }
        location "/config*"          { block }

        location "/*.php*" {
                fastcgi socket "/run/php-fpm.sock"
        }
}

server "nc.h3artbl33d.nl" {
        listen on * port 80
        block return 301 "https://nc.h3artbl33d.nl$REQUEST_URI"
}

If using 6.4-current, replace the location "/.well-known/acme-challenge/*" block to

        location "/.well-known/acme-challenge/*" {
                root { "/acme" }
                request strip 2
        }

Step 3: PHP kung fu and SQL jiujitsu

In the first step, we've installed PostgreSQL-server and PHP along with some extensions. We've met the preconditions, time to get this all up and running. This is the most intense step, so you might want to grab another cup of coffee. Or tea, if that is more your thing.

3.1: Kicking PostgreSQL online

Since this is the first time we're using PostgreSQL on this machine, we'll need to initialize the database.

$ doas su - _postgresql
$ mkdir /var/postgresql/data
$ initdb -D /var/postgresql/data -U postgres -A md5 -W

Switch back to our regular user, johndoe, by typing exit twice. Now, let's start the database by issuing a single command.

$ doas rcctl enable postgresql && doas rcctl start postgresql

If you expect to house a busy NextCloud instance, you might want to do some configuration tweaking according to the instructions in cat /usr/local/share/doc/pkg-readmes/postgresql-server-{ver}.

3.2: Compiling PHP modules

Like mentioned earlier, for caching and improved performance, we are going to use the APCu and Redis extensions for PHP. We need to compile these ourself - which isn't as hard as it might sound. First, let's take care of some dependencies. When asked which autoconf to install, select the most recent version and write the version number down, minus the patch level. Eg, if you are installing autoconf-2.69p2, write down 2.69. We'll need that later on.

$ doas pkg_add install autoconf
$ doas ln -s /usr/local/bin/php-7.0 /usr/local/bin/php && doas ln -s /usr/local/bin/phpize-7.0 /usr/local/bin/phpize && doas ln -s /usr/local/bin/php-config-7.0 /usr/local/bin/php-config

Fetch both APCu and Redis for PHP.

$ ftp https://pecl.php.net/get/{apcu-5.1.12.tgz,redis-4.1.1.tgz}
$ tar zxf apcu-5.1.12.tgz && tar zxf redis-4.1.1.tgz && rm apcu-5.1.12.tgz redis-4.1.1.tgz

We'll start with APCu. Remember writing down the autoconf version? This is the part where you'll need it. In the following example, we're using 2.69.

$ cd apcu-5.1.12
$ export AUTOCONF_VERSION=2.69
$ phpize
$ ./configure
$ doas make install

That was the whole process of compiling and installing APCu. Peanuts, isn't it? Let's repeat these steps for Redis:

$ cd ../redis-4.1.1
$ export AUTOCONF_VERSION=2.69
$ phpize
$ ./configure
$ doas make install

Now we have to change the PHP configuration with some higher limits - as the default only allows uploading of files that are two MB at max. Open /etc/php-7.0.ini and edit these values.

memory_limit = 512M
max_input_time = 180
upload_max_filesize = 512M
post_max_size = 32M
opcache.enable=1
opcache.enable_cli=1
opcache.memory_consumption=128
opcache.interned_strings_buffer=8
opcache.max_accelerated_files=10000
opcache.revalidate_freq=1
opcache.save_comments=1

Next, edit /etc/php-fpm.ini.

security.limit_extensions =

And to enable both APCu and Redis, there is one more step. Create /etc/php-7.0/cache.ini with the following content.

extension=redis.so
extension=apcu.so

Step 4: Finally, installing NextCloud

We've come quite a long way. Fortunately, we're almost there! Grab the most current version of NextCloud and extract it to /var/www/nextcloud.

$ ftp https://download.nextcloud.com/server/releases/nextcloud-13.0.5.zip
$ doas unzip -d /var/www
$ doas chown -R www:www /var/www/nextcloud

Before we can use NextCloud, we need to create a database to store the data. Replace secret-password with a strong passphrase of your liking.

$ doas su - _postgresql
$ psql -d template1 -U postgres
template1=# CREATE USER nextcloud WITH PASSWORD 'secret-password';
template1=# CREATE DATABASE nextcloud;
template1=# GRANT ALL PRIVILEGES ON DATABASE nextcloud to nextcloud;
template1=# \q

Check whether you are still running as the postgresql user with whoami - if so, just give it an exit to switch back to johndoe. Start the required services. We've set these services earlier to enabled, so with a restart, the services will also start on boot time.

$ doas rcctl start httpd redis php70_fpm postgresql

Fire up your browser and head to your subdomain. You should be greeted there with the installation wizard. Select PostgreSQL as your database, using 127.0.0.1:5432 as the server, nextcloud as the user and the password you've set. Take caution to place the datafolder outside htdocs, eg, in /var/www/ncdata.

After the installation, edit /var/www/nextcloud/config/config.php and add these lines before the closing );.

  'memcache.local' => '\OC\Memcache\APCu',
  'memcache.locking' => '\OC\Memcache\Redis',
  'redis' => array(
    'host' => '127.0.0.1',
    'port' => 6379,
  ),

Step 5: Troubleshooting

If something is amiss, there are two logs you should check first and foremost: /var/www/logs/error_log and nextcloud.log in your NextCloud data directory.

Feel free to reach out to me on Mastodon and via e-mail (hello@h3b.nl if you require additional help.

FAQ - Frequently Asked Questions (completely made-up by the author)

Q: The error log mentiones "the Redis server went away"

A: somewhat cryptic error message, I agree. Luckily, the Redis server didn't go on holiday nor went anywhere else, out of your reach. This message merely means that the Redis server is most likely offline. Check with rcctl check redis.

Q: I got a nasty 500/520 error while trying to access NextCloud

A: Most likely, PHP went belly up or it won't start due to an error in the configuration. Check the daemon status with rcctl check php70fpm or try to restart it with rcctl restart php70fpm.

Q: How about crypto, how safe is my data?

A: OpenBSD is the most secure OS out there. Having said that, nothing is one hundred percent secure. If someone says otherwise, they are flat-out lying to you. This guide follows best practice methods. There is an additional step you could take, that this guide doesn't cover: using encryption from within the NextCloud webinterface. Be cautious, losing either your passphrase or your keys will result in inaccessible data. There is a recovery method, but each user has to allow that in the individual profile. It's disabled by default - being the sanest default.

Q: Why go to all this trouble to get NextCloud up and running - when there is a NextCloud package available?

A: As mentioned before, purely for performance reasons. Furthermore, the 6.3-stable package uses PHP 5.6 by default. Having said that, you could always use the package. If you are running -current, almost everything is available through packages - thanks to the awesome Gonzalo - except for APCu.

Q: Are you mad? Spending this much time on a tutorial?

A: Whether I am mad is arguable but off-topic. I like sharing knowledge, writing and am honored that at least someone got this far. Any and all feedback is appreciated but completely optional.

Q: Can you write a tutorial about X?

A: I can always check whether it matches with my interest and whether I deem it worth my time. Feel free to drop me a message.

ON NOVEMBER 12th a video called “Slaughterbots” was uploaded to YouTube. It is the brainchild of Stuart Russell, a professor of artificial intelligence at the University of California, Berkeley, and was paid for by the Future of Life Institute (FLI), a group of concerned scientists and technologists that includes Elon Musk, Stephen Hawking and Martin Rees, Britain’s Astronomer Royal. It is set in a near-future in which small drones fitted with face-recognition systems and shaped explosive charges can be programmed to seek out and kill known individuals or classes of individuals (those wearing a particular uniform, for example). In one scene, the drones are shown collaborating with each other to gain entrance to a building. One acts as a petard, blasting through a wall to grant access to the others.

“Slaughterbots” is fiction. The question Dr Russell poses is, “how long will it remain so?” For military laboratories around the planet are busy developing small, autonomous robots for use in warfare, both conventional and unconventional. In America, in particular, a programme called MAST (Micro Autonomous Systems and Technology), which has been run by the US Army Research Laboratory in Maryland, is wrapping up this month after ten successful years. MAST co-ordinated and paid for research by a consortium of established laboratories, notably at the University of Maryland, Texas A&M University and Berkeley (the work at Berkeley is unrelated to Dr Russell’s). Its successor, the Distributed and Collaborative Intelligent Systems and Technology (DCIST) programme, which began earlier this year, is now getting into its stride.

In 2008, when MAST began, a spy drone that you could hold in the palm of your hand was an idea from science fiction. Such drones are now commonplace. Along with flying drones, MAST’s researchers have been developing pocket-sized battlefield scouts that can hop or crawl ahead of soldiers. DCIST’s purpose is to take these autonomous robots and make them co-operate. The result, if the project succeeds, will be swarms of devices that can take co-ordinated action to achieve a joint goal.

A hop, skip and jump away

At the moment, America’s defence department is committed to keeping such swarms under human control, so that the decision to pull a trigger will always be taken by a person rather than a machine. The Pentagon is as alarmed by the prospect of freebooting killer robots as the FLI is. But, as someone said of nuclear weapons after the first one was detonated, the only secret worth keeping is now out: the damn things work. If swarms of small robots can be made to collaborate autonomously, someone, somewhere will do it.

Existing small drones are usually polycopters—helicopters that have a set of rotors (generally four or six) arranged at the vertices of a regular polygon, rather than a single one above their centre of gravity. Some MAST researchers, however, think they have alighted on something better.

Their proposed replacement is the cyclocopter. This resembles an airborne paddle steamer. Though the idea of cyclocopters has been around for a while, the strong, lightweight materials needed to make them have hitherto been unavailable and the computing tools needed to design them have only recently been created. Now that those materials and tools do exist, things are advancing rapidly. Over the course of the MAST project the researchers have shrunk cyclocopters from being behemoths weighing half a kilogram to svelte devices that tip the scales at less than 30 grams. Such machines can outperform polycopters.

Cyclocopter aerodynamics is more like that of insects than of conventional aircraft, in that lift is generated by stirring the air into vortices rather than relying on its flow over aerofoils. For small cyclocopters this helps. Vortex effects become proportionately more powerful as an aircraft shrinks, but, in the case of conventional craft, including polycopters, that makes things worse, by decreasing stability. Cyclocopters get better as they get smaller.

They are also quieter. As Moble Benedict of Texas A&M, one of the leaders of the cyclocopter project, observes, “aerodynamic noise is a strong function of the blade-tip speed”—hence the whup-whup-whup of helicopters. The blade-tip speeds of cyclocopters are much lower. That makes them ideal for spying. They also have better manoeuvrability, and are less disturbed by gusts of wind.

Dr Benedict reckons cyclocopters are about two years away from commercial production. Once that happens they could displace polycopters in many roles, not just military ones. But they are not the only novel technology in which MAST has been involved. The programme has also worked on robots that hop.

One of the most advanced is Salto, developed by the Biomimetic Millisystems Laboratory at the University of California, Berkeley. Salto (pictured) is a monopod weighing 98 grams that has a rotating tail and side-thrusters. These let it stabilise itself and reorient in mid-leap. That gives it the agility to bounce over uneven surfaces and also to climb staircases.

Salto’s speed (almost two metres a second) puts huge demands on its single leg. Ron Fearing, one of the electrical engineers developing it, puts things thus: “imagine a cheetah running at top speed using only one leg, and then cut the amount of time that leg spends on the ground in half.” As with cyclocopters, the materials and processing power needed to do this have only recently come into existence.

Dr Fearing says Salto and its kin are quieter than aerial drones and can operate in confined spaces where flying robots would be disturbed by turbulence reflected from the walls. They can also travel over terrain, such as collapsed buildings, that is off-limits to wheeled vehicles. Salto still needs work. In particular, it needs to be able to cling more effectively to what it lands on. Dr Fearing uses the analogy of a squirrel leaping from branch to branch. Arriving at the next branch is only half the battle. The other half is staying there. Once that is solved, though, which it should be in the next year or two, small non-flying robots that can go where their wheeled, or even track-laying, brethren cannot should become available for practical use.

Bouncing over the rubble of a collapsed building is not the only way to explore it. Another is to weave through the spaces between the debris. Researchers at the Biomimetic Millisystems lab are working on that, too. Their solution resembles a cockroach. Its body is broad and flat, which gives it stability but also permits it to crawl through narrow spaces—if necessary by going up on one side. Should it tip over whilst attempting this, it has wing-like extensions it can use to flip itself upright again.

Getting into a building, whether collapsed or intact, is one thing. Navigating around it without human assistance is quite another. For this purpose MAST has been feeding its results to the Defence Advanced Research Projects Agency (DARPA), America’s main federal military-research organisation. According to Brett Piekarski, who led MAST and is now in charge of DCIST, the Fast Lightweight Autonomy (FLA) programme at DARPA will continue MAST’s work with the aim of developing small drones that can “ingress and egress into buildings and navigate within those buildings at high speeds”. Some of that has already been done. In June DARPA reported that polycopters souped up by the FLA programme were able to slalom through woodlands, swerve around obstacles in a hangar and report back to their starting-point, all by themselves.

Unity is strength

The next challenge—the one that people like Dr Russell particularly worry about—is getting the robots to swarm and co-ordinate their behaviour effectively. Under the aegis of MAST, a group from the General Robotics, Automation, Sensing & Perception (GRASP) laboratory at the University of Pennsylvania did indeed manage to make drones fly together in co-ordinated formations without hitting each other. They look good when doing so—but, to some extent, what is seen is an illusion. The drones are not, as members of a swarm of bees or a flock of birds would be, relying on sensory information they have gathered themselves. Instead, GRASP’s drone swarms employ ground-based sensors to track individual drones around, and a central controller to stop them colliding.

That is starting to change. A farewell demonstration by MAST, in August, showed three robots (two on the ground and one in the air) keeping station with each other using only hardware that was on board the robots themselves. This opens the way for larger flocks of robots to co-ordinate without outside intervention.

Moreover, as that demonstration showed, when drones and other robots can routinely flock together in this way, they will not necessarily be birds of a feather. “Heterogeneous group control” is a new discipline that aims to tackle the thorny problem of managing units that consist of various robots—some as small as a postage stamp, others as large as a jeep—as well as human team members. Swarms will also need to be able to break up into sub-units to search a building and then recombine once they have done so, all in a hostile environment.

Such things are the goals of DCIST. The first tranche of grants to these ends, some $27m of them, has already been awarded to the University of Pennsylvania, the Massachusetts Institute of Technology, the Georgia Institute of Technology and the University of California, Berkeley. When DCIST itself wraps up, probably in 2022, the idea of Slaughterbots may seem a lot less fictional than it does now.

Shizuoka University and contractor Obayashi aim to launch two small (10 sq cm) satellites connected by a 10m steel cable from the International Space Station.

Containers on the cable will move forward and back recorded by a camera.

 Obayashi envisages a space elevator using six oval-shaped cars, each measuring 18m x 7.2m holding 30 people, connected by a cable from a platform on the sea to a satellite at 36,000 kilometers above Earth.

The  elevator would be powered by an electric motor pulley.

The cars would travel at  up to 200kph and arrive at the space station eight days after departure from Earth.

The total length of a cable to be used for the vehicle will be 96,000 kilometers, and the total cost is estimated at $9 billion.

The cost of transport is expected to be about one-hundredth of that of the space shuttle.

Carbon nanotube is the most likely material to be used for the cables.

The work of making parchment is unglamorous, and sometimes it smells like the inside of a boxing glove: like cheese and sweat and hard work. There is only one firm of parchment makers left in the UK. There are places elsewhere in the world where parchment is produced, but the process is partly mechanised. At William Cowley’s – located somewhat improbably near Milton Keynes – everything is still done by hand. What happens there is probably not much different from what was done in the Hellenistic city of Pergamon during the reign of Eumenes II (197-159 BCE). Eumenes was an avid bibliophile and built a library to rival that of Alexandria; at its peak it contained 200,000 volumes. According to Pliny, Ptolemy of Egypt was so enraged by his neighbour’s acquisitive habits that he banned the sale of papyrus. Eumenes instructed his subjects to find an alternative writing material, and parchment was born. Where plant-based papyrus was fibrous, brittle and liable to break, parchment was flexible, durable and milky smooth. The material gave its name to the city: pergamenum is the Latin word for parchment.

It was clear when Ouster started developing the OS-1 three years ago that deep learning research for cameras was outpacing lidar research. Lidar data has incredible benefits — rich spatial information and lighting agnostic sensing to name a couple — but it lacks the raw resolution and efficient array structure of camera images, and 3D point clouds are still more difficult to encode in a neural net or process with hardware acceleration.

With the tradeoffs between both sensing modalities in mind, we set out to bring the best aspects of lidars and cameras together in a single device from the very beginning. Today we’re releasing a firmware upgrade and update to our open source driver that deliver on that goal. Our OS-1 lidar now outputs fixed resolution depth images, signal images, and ambient images in real time, all without a camera. The data layers are perfectly spatially correlated, with zero temporal mismatch or shutter effects, and have 16 bits per pixel and linear photo response. Check it out:

The OS-1’s optical system has a larger aperture than most DSLRs, and the photon counting ASIC we developed has extreme low light sensitivity so we’re able to collect ambient imagery even in low light conditions. The OS-1 captures both the signal and ambient data in the near infrared, so the data closely resembles visible light images of the same scenes, which gives the data a natural look and a high chance that algorithms developed for cameras translate well to the data. In the future we’ll work to remove fixed pattern noise from these ambient images, but we wanted to let customers get their hands on the data in the meantime!

We’ve also updated our open source driver to output these data layers as fixed resolution 360° panoramic frames so customers can begin using the new features immediately, and we’re providing a new cross-platform visualization tool built on VTK for viewing, recording, and playback of both the imagery and the point clouds side by side on Linux, Mac, and Windows. The data output from the sensor requires no post processing to achieve this functionality — the magic is in the hardware and the driver simply assembles streaming data packets into image frames.

Customers who have been given early access to the update have been blown away, and we encourage anyone curious about the OS-1 to watch our unedited videos online, or download our raw data and play it back yourself with the visualizer.

Firmware update page: https://www.ouster.io/downloads

Github & sample data: www.github.com/ouster-LIDAR

This Is Not a Gimmick.

We’ve seen multiple lidar companies market a lidar/camera fusion solution by co-mounting a separate camera with a lidar, performing a shoddy extrinsic calibration, and putting out a press release for what ultimately is a useless product. We didn’t do that, and to prove it we want to share some examples of how powerful the OS-1 sensor data can be, which brings us back to deep learning.

Because the sensor outputs fixed resolution image frames with depth, signal, and ambient data at each pixel, we’re able to feed these images directly into deep learning algorithms that were originally developed for cameras. We encode the depth, intensity, and ambient information in a vector much like a network for color images would encode the red, green, and blue channels at the input layer. The networks we’ve trained have generalized extremely well to the new lidar data types.

As one example, we trained a per pixel semantic classifier to identify driveable road, vehicles, pedestrians, and bicyclists on a series of depth and intensity frames from around San Francisco. We’re able to run the resultant network on an NVIDIA GTX 1060 in real time and achieved encouraging results, especially considering this is the first implementation we’ve attempted. Take a look:

Because all data is provided per pixel, we’re able to seamlessly translate the 2D masks into the 3D frame for additional real time processing like bounding box estimation and tracking.

In other cases, we’ve elected to keep the depth, signal, and ambient images separate and pass them through the same network independently. As an example, we took the pre-trained network from DeTone et al.’s SuperPoint project [link] and ran it directly on our intensity and depth images. The network is trained on a large number of generic RGB images, and has never seen depth/lidar data, but the results for both the intensity and depth images are stunning:

On careful inspection, it’s clear that the network is picking up different key points in each image. Anyone who has worked on lidar and visual odometry will grasp the value of the redundancy embodied in this result. Lidar odometry struggles in geometrically uniform environments like tunnels and highways, while visual odometry struggles with textureless and poorly lit environments. The OS-1’s camera/lidar fusion provides a multi-modal solution to this long standing problem.

It’s results like these that make us confident that well fused lidar and camera data are much more than the sum of their parts, and we expect further convergence between lidars and cameras in the future.

Look out for more exciting developments from Ouster in the future!

LINKS

1. Videos:

Sunny drive with ambient data:

Pixelwise semantic segmentation: https://www.youtube.com/watch?v=JxR9MasA9Yc

Superpoint: https://www.youtube.com/watch?v=igsJxrbaejw

2. Firmware update page: https://www.ouster.io/downloads

3. Github & sample data: www.github.com/ouster-LIDAR

Every now and then I see a page that hasn’t jumped on the bandwagon; we just can’t have that so lets run through how to design for the modern web.

These practices are immutable, you must follow them because I’m a developer advocate. They’re also in effect on quite a few of the top websites as ranked by Alexa, but most importantly, developer advocate.

Let Users Know About Your Mobile Application

The very first thing you must do when a user visits your website is to show them a big modal dialog telling them that they should just install the mobile application instead.

A neat little trick is to make the link to close the modal small and put it very close to the link that lets the user install the application. This will make users much more happy as it’s so easy to install the application and not accidentally close the dialog!

Let Users Know About Cookies

Should the user continue and insist on using the web version we need to let them know that we use cookies to track them.

Let Users Know They Can Sign Up

Sometimes a link is not enough, a modal can be pretty helpful here in letting the user know they can actually sign up for your website.

Research has shown that modals that cannot be closed have the best conversion rate.

Block Users from Europe

Europeans nowadays have these pesky laws which muddy the waters on what we are allowed to track. The best practice here is just to do nothing, continue tracking everything about your users and just block Europeans from accessing the site.

Denmark is a failed socialist cup cake state anyway, its users are not worth the effort.

Let Users Know About Your Mobile Application

It’s good practice to show the user the dialog again when a user clicks on any link, most likely they just accidentally clicked away the first time.

Again we need to help our users out here, so we make the cancel link very small to minimize the chances of it accidentally being tapped. Happy users everywhere!

Allow Users To Opt-Out

Now it’s very important that we are not intrusive to the user, so we must allow them to opt-out of the mobile application modals.

Best practice is to put it somewhere the user will easily spot it, like inside one of the account preferences pages.

Advertise Your Application

If the user should opt out of the prompt’s for the mobile application, we can till get them in the long run. Advertising the mobile application somewhere on the website works great.

Eventually the users will give in and convert!

Always Bet on JavaScript

These modals obviously require JavaScript, and of course it’s important to have endless scrolling but make sure you future proof yourself by using the latest framework. You may think “oh it’s only a couple of modals” today. But in the future, it may be many many more modals and oh boy! When that happens you’ll regret you did not make an isomorphic application with React and code splitting!

Conclusion

Now that you know these best practices for modern web development, make sure you apply them everywhere. You are now also certified and ready to apply for top ranking sites like Reddit and Medium as long as you remember these simple rules during the interview process.

Don’t know anything about web development at all? Don’t worry you can just take a week long bootcamp!

Already a web developer? Buy the C programming language book here and get out while you still can.

House Majority Leader Kevin McCarthyKevin Owen McCarthyTrump: Google, Facebook, Amazon may be in a 'very antitrust situation'Bail bond industry mobilizes against Calif. law eliminating cash bailWashington's fall agenda: GOP, tech industry clash over bias claimsMORE is leading the charge against President TrumpDonald John Trump7 subtle jabs speakers took at Trump during John McCain’s funeralGraham invited Ivanka Trump to McCain's funeral: reportTrump blasts trade talks with Canada: We shouldn't have to buy our friendsMORE’s new favorite punching bag: big tech.

The California Republican, who hopes to replace Speaker Paul RyanPaul Davis RyanDems vow rules overhaul to empower members if House flipsTrump says he will 'take a look' at increases for federal workersCongress gives McCain the highest honorMORE (R-Wis.) next year, has been aggressively promoting a campaign to “stop the bias,” referring to what critics say is a pattern of discrimination against conservative voices on social media.

To that end, Twitter CEO Jack Dorsey is scheduled to testify on Capitol Hill this week at a hearing requested by McCarthy.

“I’ve had many conversations with the president about how we have to stop this bias,” McCarthy, one of Trump’s closest allies on Capitol Hill, told Fox News on Thursday.

“I’ve spoken to Jack Dorsey throughout the month,” McCarthy added. “He and I philosophically disagree, but we do agree on one thing: We believe in the First Amendment. But we also believe in transparency and accountability.”

The majority leader’s recent crusade against social media comes at a time when Trump has ramped up his own rhetoric against the tech industry. His eagerness to champion the cause could endear him to both the White House and conservative lawmakers -- two constituencies that could be crucial to securing the Speakership.

Trump recently asserted that Google and other platforms are “rigged” against him, an accusation that Google rejected, and one that came on the heels of allegations from conservatives that Twitter has been “shadow banning” certain Republicans so that their accounts are less visible to users.

That controversy started after prominent conservatives aligned with Trump, including Rep. Matt GaetzMatthew (Matt) GaetzHouse Republicans say Ohr interview escalates surveillance concernsOvernight Defense: Duncan Hunter refusing to step down from committees | Trump awards Medal of Honor to widow of airman | Pentagon names pick for Mideast commanderTrump awards posthumous Medal of Honor to family of fallen Air Force sergeant MORE (R-Fla.), House Freedom Caucus ChairmanMark MeadowsMark Randall MeadowsDems vow rules overhaul to empower members if House flipsTrump calls North Carolina redistricting ruling ‘unfair’Steele told DOJ official Russia thought it had 'Trump over a barrel': reportMORE (R-N.C.) and Rep.Jim JordanJames (Jim) Daniel JordanThe Hill's Morning Report: Trump shifts campaign focus from Senate to HouseHouse Republicans say Ohr interview escalates surveillance concernsGOP’s McMorris Rodgers on Jordan for Speaker: 'I am staying open'MORE (R-Ohio) failed to appear on Twitter’s auto-populated drop-down search box when users typed in their names.

The company said that it does not “shadow ban” according to political ideology, but acknowledged that its attempts to crack down on hate speech have unintentionally reduced search results for lawmakers from both parties.

McCarthy was quick to take up the mantle on the issue, tweeting more than a dozen times last month with the hashtag “stop the bias.” He also made the media rounds to step up pressure on Dorsey to publicly testify on Capitol Hill.

But not all of McCarthy’s efforts on that front were successful. One of his tweets attempting to demonstrate censorship of conservatives drew criticism for being misleading.

After sharing a screenshot of a tweet from Fox News host Laura Ingraham that was covered by language warning of “potentially sensitive content,” Twitter users were quick to point out that Ingraham’s tweet was covered up due to settings in McCarthy’s own Twitter account, not because of a company campaign to silence conservative voices.

Still, the crusade to keep the issue of alleged anti-conservative bias in the spotlight could earn McCarthy some political capital. One of the lawmakers who Republicans say has been targeted by Twitter’s “shadow banning” practice is Jordan, a leader of the Freedom Caucus who is also running for Speaker.

While Jordan may struggle to secure the 218 votes needed to win the gavel, his far-right group has the power to veto any Speaker hopeful if Republicans retain control of the House in the midterm elections.

The House Energy and Commerce Committee on Wednesday will hear testimony from Dorsey, but it’s unclear whether Congress or the Trump administration will take any action against tech and social media companies, which lawmakers on both sides of the aisle have been reluctant to regulate further.

Proposed steps for regulation, however, could include making Facebook and other social media platforms a public utility, forcing Google to be more transparent about its algorithms and making it easier for individuals to sue technology companies.

McCarthy says all options are on the table.

“Congress is going to look at everything, because of how powerful they have become,” he told Fox News.

Why it matters: Judge rules that copyright trolls need more than just an IP address if they want to go after copyright infringement. An IP is not enough proof to tie a person to a crime.

In a win for privacy advocates and pirates, the Ninth Circuit Court of Appeals ruled that an IP address alone is not enough to go after someone for alleged copyright infringement. They ruled that being the registered subscriber of an infringing IP address does not create a reasonable inference that the subscriber is also the infringer.

The case began back to 2016 and has been playing out in the legal system ever since. The creators of the film 'The Cobbler' alleged that Thomas Gonzales had illegally downloaded their movie and sued him for it.

Gonzales was a Comcast subscriber and had set up his network with an open Wi-Fi access point. At some point, someone had used his network to download the movie and the film creators captured Gonzales's IP address.

The judge stated that in order for a proper case, the copyright owners would need more than just an IP address. This is often difficult to provide since it is challenging to prove who was connected to what and when. This case is made even more challenging since Gonzales's network was open and anyone could have downloaded the movie.

The new ruling upholds a previous ruling by a lower district court on the same case. The Appeals Court issued the following statements:

In this copyright action, we consider whether a bare allegation that a defendant is the registered subscriber of an Internet Protocol (‘IP’) address associated with infringing activity is sufficient to state a claim for direct or contributory infringement. We conclude that it is not.

The direct infringement claim fails because Gonzales’s status as the registered subscriber of an infringing IP address, standing alone, does not create a reasonable inference that he is also the infringer. Because multiple devices and individuals may be able to connect via an IP address, simply identifying the IP subscriber solves only part of the puzzle. A plaintiff must allege something more to create a reasonable inference that a subscriber is also an infringer.

In addition to direct infringement, the copyright owner also attempted an indirect infringement claim. They alleged that Gonzales had encouraged users of his network to download the movie but this failed as well since they were unable to provide any proof. Finally, the judge ordered Cobbler Nevada LLC, the copyright holder, to pay more than $17,000 in legal fees for Gonzales.

On the April morning of 4/20, the “weed holiday” when crowds toke up in public places, Nick wrote a Grade 12 history test on the French Revolution while stoned: “I got an 82,” he says, “which is a pretty good mark.”

He’s offering this detail as evidence that, after three years of experimenting with pot in high school, his memory is intact, his brain is not fried, the life-destroying danger warned of in drug-education assemblies has not, in his experience, come true.

On a Sunday afternoon of the May long weekend, with his parents away, he’s been playing the video game Fortnite in his Toronto-area bedroom, listening to the rapper Rich the Kid, and now, on his smartphone, he’s making the case for why he thinks the adults need to chill out about marijuana. He smoked a joint an hour before the interview, which might explain the mellow amble into the merits of Fortnite. Otherwise, given his rapid-fire recall of pot research, it’s hard to tell.

“Did you know,” he asks, “that you’d have to smoke 1,500 pounds of marijuana for like, 30 minutes, to die? That’s physically impossible.” (I check: He’s quoting a 1988 ruling from the U.S. Drug Enforcement Administration, almost verbatim: The time stated is actually 15 minutes.) Yeah, he says, “I’ve done a lot of reading.”

Nick would prefer his last name not be used – he doesn’t advertise his pot smoking to his parents. When we first spoke a couple of years ago, Nick was conducting his own single-subject marijuana trial. If he binged for a few days, he’d take a week or two off. “To make sure I don’t go overboard,” he said then. Now, in his last semester of high school, he’s smoking more than ever – three or four times a week, mostly on weekends. “I’ve been paying close attention to see if there are any faults in my memory, and the truth is I haven’t seen any.”

In fact, he says, he’s getting the best grades of his high-school career. He’s made the honour roll. Until recently, he was swimming competitively. He’s holding down a part-time job. And he was accepted into both his top choices for university.

As one scientist observed, Nick might be doing even better if he weren’t getting high so often. But for a teenager with a relatively heavy drug habit, he comes across as a successful, functional 17-year-old.

And when it comes to pot research, he’s still a puzzle.

Last year, a committee of experts for the National Academies of Sciences, Engineering, and Medicine in the United States released a detailed analysis of the published studies on marijuana. They looked at more than 10,000 papers, selected the ones they deemed to be of highest quality, and assessed the strength of the science. The report is 500 pages of carefully fudged conclusions that highlight the gaping holes in cannabis research. For instance: Does pot cause lung cancer? “There is moderate evidence of no statistical association,” the reports says. Does it make you more likely to die prematurely, or from an overdose? “There is no or insufficient evidence to refute or support a statistical association.” On the possible long-term consequences that cause parents the most worry – pot’s association with psychosis, brain damage and addiction – the report also offers more caveats than conclusions.

The studies that have rolled out since have been equally confounding. A meta-analysis published in April in JAMA Psychiatry, for example, found that once teenaged pot users went off the drug – even for as little as 72 hours – their scores on cognitive tests were not significantly different from non-drug users.

Other studies have also found the “stoner” deficit to be temporary. Scientists believe that the adolescent brain is at higher risk because it’s developing rapidly in the same areas affected by the “high” from pot. But brain scans showing structural changes are often based on smaller samples of chronic, frequent users. Since they don’t have a before-picture of those same brains, researchers can’t say for certain whether marijuana is the cause, or a consequence that happens along the way. James MacKillop, co-director of the Michael G. DeGroote Centre for Medicinal Cannabis Research at McMaster University, compares the problem to a crime-scene photograph. “Pot may be present, but is it the culprit?”

One issue is that the subjects in studies tend to be people using a lot of marijuana for a long time. But it’s much more common for teenagers to dabble in pot for a while, and wean themselves when adult responsibilities arrive. And researchers know even less about this second group.

As Canada heads into its own real-world pot experiment, the best they can say is that cannabis will likely turn out to be neither devastating for most teens, nor benign for all of them.

In other words, most Nicks will probably end up just fine.

But not everyone is a Nick.

The cautionary tale

A week before Nick’s pot-hazed history exam, Jonathan Kropman, 23, went to rehab. Driving to a treatment facility in Port Hope, Ont., run by the Canadian Centre for Addictions, he made his father stop five times so he could smoke pot on the side of the road. At the centre, he reluctantly parted with his last three joints, after his dad persuaded him not to sneak them in.

It was a long journey to this point. In the spring, Mr. Kropman flew home a week early from a Mexican vacation with his family – booking the flight without his parents’ knowledge – because he felt he couldn’t go another seven days without pot. The previous April, he had woken up in a jail cell at Toronto’s Old City Hall with a foggy memory of the night before. He needed to be told that, while intoxicated and smoking weed at the corner of Yonge and Dundas, he’d swung a punch at the police officer who stopped to question him. This wasn’t his first pot-induced trouble – in Grade 8, he was suspended from school for possessing a joint. But it was adding up: “I really have to change my life,” he decided. For one thing, he said later, from rehab, “no woman is going to want to be with an addict.”

Mr. Kropman took his first haul on a joint in a playground after school in Grade 6, when he was 11 years old. This alone makes him an outlier: According to a 2017 Ontario student drug survey, only about 2 per cent of teens have tried pot by Grade 8, a number that rises to 37 per cent in Grade 12. But Mr. Kropman played sports with older kids, and weed was always around. “It was the cool thing to hang out with the bad crowd. I didn’t know what I was getting into,” he recalls. “From the first toke, I knew – I was addicted to the feeling. It was euphoric, an escape.”

Even before then, Mr. Kropman says he was “definitely a trouble-maker” and a risk-taker. He was sent home in Grade 4 for fighting in school. He had a special worker who sat with him in class to help him with work and monitor his behaviour. By junior high, he was using pot all the time, and drinking as well, although he preferred weed to alcohol. Pot, he explains, “is a lot easier to hide. You just put [eye drops] in your eyes.”

The warnings that he was causing damage to his brain, he says, “went in one ear and out the other.” His mother was firm: She didn’t want him using. His father was less strict – growing up, Mr. Kropman says, his dad also smoked a lot of weed. As they became more aware of his habit, they tried to talk him into cutting back: “I wasn’t listening.”

Instead, his drug use escalated – he got weed from friends with medical marijuana licences and, eventually, his own prescription, he says, given to him for help with sleeping. He couldn’t cope without a ready supply. Working construction, after high school, he smoked weed to stay focused during the tedium of painting houses. Eventually, he was drinking, smoking and, finally, using cocaine, which he tried one night when he was stumbling drunk. (Pot’s reputation as a gateway drug was also cast into doubt by the National Academies report, which stated that the examined studies “did not provide compelling evidence.”) Mr. Kropman insists it was alcohol – not cannabis – that paved his path to the hard stuff; stoned, he says, he would have been too nervous to try it.

At the time we speak, Mr. Kropman had completed more than a month of treatment, cold turkey. “This place definitely saved my life.” The physical cravings are gone, he explained, but the mental struggle is just beginning. “I still crave it, absolutely.” On a Saturday in May, he left rehab for a sober-living house, where he can stay for three months as long as he passes regular drug tests. As of this Thursday, he’s been “clean and sober” for 78 days. “I am going to give it my best shot.”

Of all the pot-induced problems that adults worry about, addiction should be at the top of the list, says Jean-Sébastien Fallu, a psychologist at the University of Montreal who has edited a book on cannabis research coming out this fall. About 9 per cent of adults who start smoking pot will become addicted, Mr. Fallu says. Among teenagers, he says, the rate is 16 per cent – and nine times more likely, for instance, than the risk of a pot-smoking teenager developing schizophrenia.

Detective work

So why is Nick on the honour roll, and Jonathan going through rehab?

That’s the real detective work required to unravel the role that pot plays in Mr. MacKillop’s crime-scene photo. Are teens who start using cannabis at young ages different from their peers? How much do certain risk factors matter – an impulsive personality, a deviant peer group, childhood trauma – and in what combination? Is pot a supporting character in a tale of unfortunate events, or the villain that turns the tide?

Risk factors appear to be at least a partial explanation for many of the problems associated with cannabis. Teenagers who begin using early, and increase their use over time – as Mr. Kropman did – are more likely to become addicted; having a history of addiction in the family is also a contributing factor.

In some studies, accounting for risk factors significantly reduces – and even eliminates – the negative consequences associated with cannabis. A 2012 longitudinal study– one of the most well-regarded cannabis studies, and hence often cited as proof of the drug’s negative side effects – found that IQ fell among long-term, heavy users. But twin studies that compared pot-using teens with their non-using siblings didn’t find cognitive differences. When researchers adjust for criteria such as family background, parental education, intelligence and personality, the link between negative consequences, such as high-school dropouts and poor grades, often collapses. How often teenagers use pot also appears to be important: a June 2017 study found that young people who used twice a week or less performed as well as – and even slightly better – than non-users on tests measuring memory and executive control. And the National Academies report found “limited to no evidence” that, once people stopped using marijuana, there were sustained effects on learning, memory or attention.

“There is a nuance to the findings that need to be appreciated,” says Cobb Scott, a psychology professor at the University of Pennsylvania and lead author of the June 2017 study, who studies the effect of dose and abstinence on cognitive skills. “Our research suggests that these effects may be smaller than people were worried about and it’s possible they don’t reflect long-term damage to brain networks that support cognitive functioning.”

If damage does occur, however, the research suggests that early exposure is likely a key factor. A University of Montreal study published last December that tracked boys in Montreal from age six into their 30s found that, among those who smoked pot after age 17, there was no overall difference in cognitive skills – intelligence, executive function, decision-making – compared with their peers who hadn’t smoked. Those boys who started using cannabis before age 14 performed more poorly, especially in verbal cognitive tests. But even so, says Natalie Castellanos-Ryan, a psychology professor who co-authored the Montreal study, “we can’t assume it is the toxic effect of cannabis.” In many cases, other high risk factors were also present before marijuana, including poor grades and delinquent behaviour.

And even when studies do find a statistically significant negative association with cannabis use, Ms. Castellanos-Ryan says that the actual effect on an individual’s life may be small. Bottom line, she concludes: “There is a lot of missing research, and many of us are trying to fill in the gaps.”

One area where the science appears to be closer to an answer is the association between cannabis and schizophrenia. Rather than cause the serious mental illness, Ms. Castellanos-Ryan says, current researchsuggests that marijuana works as “a trigger” for young people already predisposed to psychosis. (In one study of pot users, for example, only those with a family history of schizophrenia developed the disease.)

It is still not clear whether not using cannabis would have prevented the disease in high-risk individuals. But those teenaged exposures appear to be key: Steven Laviolette, a biologist at the University of Western Ontario, says that while dosing juvenile rats with THC produces memory issues and psychotic behaviour, those changes are not seen in adult rats given the same dose.

Since schizophrenia develops during adolescence, and teenagers may not know their family history or whether they have a genetic sensitivity, Mr. Laviolette says it’s another argument for waiting until adulthood to use cannabis. “We haven’t figured it out yet,” he said. “But we can conclusively say it is a risky thing to do.”

Pot smokers are complicated study subjects – many of the same teens also drink alcohol and smoke cigarettes, both of which may bring even more significant harms than cannabis. (The National Academies report, for instance, found stronger evidence for tobacco as the first gateway drug than pot.) Comparisons between studies are tricky, since researchers may not ask about the same risk factors, or record data the same way.

Most studies also rely on self-reports, which aren’t particularly reliable, even for subjects not doing drugs. Aside from being present in real time, researchers can’t determine the potency of pot being ingested, or precisely how much.

Yet, one of the most important outstanding questions is whether stronger versions of cannabis – as well as different ways of using it – will produce different levels of harm. Today’s street pot is not Woodstock weed. The amount of THC, the mind-altering chemical in marijuana, has roughly tripled since the 1970s, and new concentrated versions, known by names such as shatter, wax and budder, may be 80-per-cent THC, or more than six times the level of the modern-day plant version.

“We certainly do not have strong evidence that low frequency use is associated with a lot of negative outcomes,” says Mr. MacKillop of the DeGroote Centre . But at the same time, he adds: “What was somewhat a murky science to start with is already playing catch-up to the brave new world of cannabis consumption.”

Science may eventually conclude that what makes some teenagers more susceptible to negative outcomes when they use cannabis is a complicated mix of interwoven conditions and life circumstances, both inherited and acquired.

“That is not an unlikely scenario,” says Susan Weiss, a researcher at the National Institute on Drug Abuse in the United States. “But it doesn’t mean that cannabis isn’t adding to the problem.”

Ms. Weiss is part of a groundbreaking, new study trying to get closer to the answer by following more than 10,000 American nine- and 10-year-olds through their teenage years, using regular cognitive tests, psychosocial surveys and brain scans to capture a before-and-after picture of drug use. But it won’t have findings for many years, long after pot is legal in Canada.

So while scientists are figuring things out, how do Canadian parents talk so their teens will listen?

High on the honour roll

“I figure live fast, die young,” explained Alex, a 17-year-old in St. Catharines, Ont., who smokes pot, he says, five or six times a week but, as with Nick, can also boast about being on the honour roll, with a university acceptance letter.

When it comes to the risk of marijuana, he says, “I worry about it the same way that you worry about eating at McDonald’s. It might give you a heart attack or make you fat, but people eat at McDonald’s all the time because it tastes so good.”

Would Alex really want to eat Big Macs every day? “No, probably not,” he admits. But it’s not a bad analogy. Even with the science still to be determined, the researchers interviewed for this story were unanimous in their best advice: Teenagers should wait until their brains are more developed (ideally, age 25), use infrequently, take breaks and avoid weed with high levels of THC. Adolescents with known risk factors – such as a history of schizophrenia or addiction in the family – should take special care. At the very least, like fast food, marijuana is safest consumed in small, spare doses. In the best-case scenario, researchers agreed, teenagers would never use pot.

On the positive side, says Mr. Fallu in Montreal, the fear that legalization will launch a tsunami of teenaged potheads has been grossly overstated, according to existing research. That’s not happened to date in states such as Washington and Colorado, where recreational cannabis has already been legalized.

In a 2017 Ontario student drug survey of more than 11,000 teens, 10 per cent of high-school students said they intended to try pot after it was legalized, and 4.9 per cent said they planned to use more often. By comparison, 53 per cent said they didn’t plan to use it either way.

Canadian teens already use pot at relatively high rates – a 2014 report by the World Health Organization placed Canada second behind France for 15-year-olds who said they’d used pot in the past 30 days. (Unlike many other countries, girls use at roughly the same rates as boys.) But the Ontario student study also found that overall drug use among teenagers – including cannabis – was the lowest recorded since the survey began in 1977, especially among younger grades.

And it’s not as if weed is hard to find now, as even teenagers not using pot told The Globe and Mail. Ellen, an Alberta high-school student, said she’d only have to wander across the street at lunch hour to the stoner kids smoking outside the 7-Eleven.

As legalization approaches, researchers such as Jean-Sébastien Fallu are calling for balance, so adults don’t discredit themselves with teenagers, who check everything on Google anyway. It is particularly frustrating, Mr. Fallu says, that parents often exaggerate the risks of cannabis, while minimizing the danger of alcohol, even though there’s research to suggest that binge drinking, in particular, may be even more harmful to physical health and the teenage brain.

Mr. Fallu suggests that a rational, science-based message about cannabis should present the risks, admit what is yet unknown and make clear that using drugs when you’re young is a gamble with potentially high stakes.

“Human beings want to have black-and-white facts, but the truth is rarely that clear, and not just in drug use,” he says. “There is uncertainty, but we can educate around that.”

1. There is a lot scientists don’t know. But researchers believe that starting before age 15 and using heavily through your teenage years is associated with the highest risk of harm. The government is legalizing marijuana not because it is completely safe but to regulate it and, ideally, make the street version of the drug less available to young people.

2. If you are going to use, take breaks. Pot may not be as addictive as other drugs, but the risk of addiction is still higher for teenagers. Taking a break is also good for your developing brain (and your lungs), and a way to see how marijuana might be negatively affecting your life and your relationships.

3. Just like alcohol, there are different potencies of pot. Again, it’s not certain, but the science suggests that higher levels of THC, the mind-altering chemical in the drug, may cause more damage to a young brain. The highest level of THC – as high as 80 per cent –is found in butane hash oil extractions, called shatter, wax or budder.

4. Be careful with edibles, which are the main cause of emergency-room visits for pot. All the THC in that pot brownie will be absorbed into your body. And the effect takes longer, so people may take too much before they realize it.

5. Marijuana may not cause schizophrenia but scientists believe it is a potential trigger for teens with risk factors, such as family history. If you have a severe reaction to pot – hallucinations, for instance, or frightening paranoia – that may be a warning sign.

6. Some scientists will say that alcohol is worse than cannabis. But the harms of each drug are different. Using any drug when you’re young is more risky than when you’re an adult. Using them together appears to increase those risks.

7. Don’t drive if you have used marijuana, especially if you have been drinking too. Some people may say they drive better stoned, but research suggests otherwise – when people are high they react more slowly and think less clearly. Researchers who analyzed the existing data found a higher risk of car accidents when people drove while stoned.

8. Be skeptical of any headline that suggests a study has found the answer. Science gets misused on both sides, and legalization has also given cannabis companies a vested interest in overselling the drug’s positive effects.

9. Studies usually group a lot of people together to produce average findings. How each individual reacts to marijuana may be very different. But one thing is true for everyone: Your brain is developing until about age 25. Just like getting good sleep and eating well, avoiding alcohol and marijuana as much and for as long as possible helps ensure you’ll get the best one you can.

– Erin Anderssen