Performance testing HTTP/1.1 vs HTTP/2 vs HTTP/2 + Server Push for REST APIs – Evert Pot

<script src=”/assets/js/request-simulator.js”/>

When building web services, a common wisdom is to try to reduce the number of
HTTP requests to improve performance.

There are a variety of benefits to this, including less total bytes being
sent, but the predominant reason is that traditionally browsers will only make
6 HTTP requests in parallel for a single domain. Before 2008, most browsers
limited this to 2.

When this limit is reached, it means that browsers will have to wait until
earlier requests are finished before starting new ones. One implication is that
the higher the latency is, the longer it will take until all requests finish.

Take a look at an example of this behavior. In the following simulation we’re
fetching a ‘main’ document. This could be the index of a website, or a some
JSON collection.

After getting the main document, the simulator grabs 99 linked items. These
could be images, scripts, or other documents from an API.

The 6 connection limit has resulted in a variety of optimization techniques.
Scripts are combined and compressed, graphics are often combined into
‘sprite maps’.

The limit and ‘cost’ of a single HTTP connection also has had an effect on web
services. Instead of creating small, specific API calls, designers of REST
(and other HTTP-) services are incentivized to pack many logical ‘entities’
in a single HTTP request/response.

For example, when an API client needs a list of ‘articles’ from an API, usually
they will get this list from a single endpoint instead of fetching each article
by its own URI.

The savings are massive. The following simulation is similar to the last,
except now we’ve combined all entities in a single request.

If an API client needs a specific (large) set
of entities from a server, in order to reduce HTTP requests, API developers will
be compelled to either build more API endpoints, each to give a result
that is tailored to the specific use-case of the client or deploy systems
that can take arbitrary queries and return all the matching entities.

The simplest form of this is perhaps a collection with many query parameters,
and a much more complex version of this is GraphQL, which effectively uses
HTTP as a pipe for its own request/response mechanism and allows for a wide range
of arbitrary queries.

Drawbacks of compounding documents

There’s a number of drawbacks to this. Systems that require compounding of
entities typically need additional complexity on both server and client.

Instead of treating a single entity as some object that has a URI, which
can be fetched with GET and subsequently cached, a new layer is required
on both server and client-side that’s responsible for teasing these entities

Re-implementing the logic HTTP already provides also has a nasty side-effect
that other features from HTTP must also be reimplemented. The most common
example is caching.

On the REST-side of things, examples of compound-documents can be seen in
virtually any standard. JSON:API, HAL and Atom all have
this notion.

If you look at most full-featured JSON:API client implementations, you will
usually see that these clients often ship with some kind of ‘entity store’,
allowing it to keep track of which entities it received, effectively
maintaining an equivalent of a HTTP cache.

Another issue is that for some of these systems, is that it’s typically
harder for clients to just request the data they need. Since they are often
combined in compound documents it’s all-or-nothing, or significant complexity
on client and server (see GraphQL).

A more lofty drawback is that API designers may have trended towards systems
that are more opaque, and are no longer part of the web of information due
to a lack that interconnectedness that linking affords.

HTTP/2 and HTTP/3

HTTP/2 is now widely available. In HTTP/2 the cost of HTTP requests is
significantly lower. Whereas with HTTP/1.1 it was required to open 1 TCP
connection per request, with HTTP/2 1 connection is opened per domain. Many
requests can flow through them in parallel, and potentially out of order.

Instead of delegating parallelism to compound documents, we can now actually
rely on the protocol itself to handle this.

Using many HTTP/2 req

Truncated by Planet PHP, read more at the original (another 333361 bytes)

Ajax live search, enhance your site’s search experience – PHP Scripts – Web Development Blog

If your website has a lot of pages, a good working search function becomes more important. Is your website based on WordPress? Then you don’t need to worry about the search function, because it’s a standard feature. But how about a great search experience? Relevant search results are one part of the game, but how […]


The release of Object Design Style Guide – Matthias Noback

Book cover

Today Manning released my latest book! It’s called “Object Design Style Guide”.

In November 2018 I started working on this book. The idea for it came from a conversation I had with the friendly folks at Akeneo (Nantes) earlier that year. It turned out that, after days of high level training on web application architecture and Domain-Driven Design, there was a need for some kind of manual for low level object-oriented programming. Not as low level as the kind of programming advice people usually refer to as clean code, but general programming rules for different kinds of objects. For instance:

  • A service gets its dependencies and configuration values injected as constructor arguments.
  • A service is an immutable object.
  • An entity always has a named constructor.
  • An entity is the only type of mutable object in an application.

And so on…

While exploring the subject area I realized there would be a lot to write about, and a lot to declare as out-of-scope too. I thought it would be very important to make it a practical book, with short and clear rules for object design.

Soon enough I was able to publish a couple of chapters on Leanpub. For me publishing through Leanpub has always worked well, maybe because my writing process usually turns out to be more or less linear. This allows me to publish the first chapters, and then keep publishing new chapters over the course of several months. I absolutely love how early supporters bought this book when they could only read a few chapters. I’ve said it before, but this is an important encouragement to keep going.

Re-publishing an existing book

In January of this year I was approached by Manning. They asked if I would like to work on a book about Domain-Driven Design. I suggested working on a new edition of the Style Guide instead, and they were happy about that from the start. Manning has a rather extensive process for onboarding, and although I thought it wouldn’t be worth it because the book already existed, they still pushed me to write a profile of the “Minimum Qualified Reader” (MQR) and to create a “Weighted Table of Contents” (WTC). I’m glad they did, because it definitely improved the quality of the learning experience for readers of this book. The MQR is a description of the reader: what do I expect from them in terms of experience, knowledge, interests, etc. This helped me write the most useful book for the imagined reader, and also allowed me to skip certain basic concepts that I could assume the reader to already know about.

The WTC shows for each chapter how hard it is to read and understand. Ideally a book would start simple, with some introductory chapters, and build up towards the more complicated chapters. I think I’ve actually achieved this. The book starts with some basic concepts, building up towards an explanation about changing behavior without changing code, and separating write from read models. The final two chapters provide an overview of different types of objects and how they work together in a web application, and an overview of related topics and suggestions for further reading.

Overall, the publishing process was well managed by the team at Manning. They provided lots of advice on using diagrams, adding exercises, etc. I can recommend them to anyone who is looking for a serious process leading to a book published by a serious publisher. Just be ready to spend some more time on making everything just right. For comparison: the first edition of the book, as published on Leanpub, took me only 127 hours to write. Working with Manning, processing review suggestions, fixing issues, following the steps in their publishing process, etc. took another 132 hours.

Comparison with the PHP edition

The Manning edition of the style guide is different from the PHP edition published on Leanpub in the following ways:

  • It has more words; there are many little additions, like “asides” to answer reader questions. It is now about 290 pages in print.
  • It has better words; it has been checked for grammar, spelling and style issues.
  • It has exercises; the first chapters have exercises. Some multiple-choice questions and some open questions where you have to write a bit of code. The answers can be found at the end of each chapter. For multiple-choice questions I explain why the correct answer is the only correct answer and why the other possible answers would be wrong. For the code exercises I prov

Truncated by Planet PHP, read more at the original (another 806 bytes)

Why You Should Use Asynchronous PHP

The Difference Between Synchronous vs. Asynchronous Programming

Before we discuss the merits of asynchronous PHP, let’s quickly review the differences between the synchronous and asynchronous programming models. Synchronous code is sequential. Individual tasks must be completed before you can start another one. In asynchronous code, multiple tasks can be completed simultaneously, which can dramatically boost application performance and user experience.

BigQuery: Using multiple cursors / rectangular selection in BigQuery UI – Pascal Landau

Multiple keyboard shortcuts usable in the BigQuery UI are listed in the
official documentation, though the one for
using multiple cursors is missing:

ALT + left-mouse-button-drag

Using multiple cursors in the BigQuery UI via ATL + left mouse drag


  • keep the ALT key pressed first and then click the left mouse button and drag it up or down vertically
  • the same hotkeys can be used to draw a rectangular selection (aka column selection)
    Rectangular / column selection in BigQuery
  • using ALT + LEFT and ALT + RIGHT will position one (or all) cursors at the beginning respectively end of the line

Use cases

We often deal with multiple datasets and tables that have the exact same structure, e.g. due to sharding. In those cases it’s
often required to modify different parts of the query in the exact same way so that multiple cursors come in handy.

Routing in Slim 4 – Rob Allen

Routing in Slim 4 works pretty much exactly the same as in Slim 3. They are used to map a URL that the browser requests to a specific handler that executes the code for that particular page or API endpoint. You can also attach middleware that will only be run when that route is matched.

The route in the slim4-starter looks like this:

$app->get('/[{name}]', HomePageHandler::class);

It is made up of the method, the pattern and the handler.

The method

Slim’s App object contains a number of helper methods for defining routes. The important ones are:

$app->get($pattern, $handler) For responding to an HTTP GET request
$app->post($pattern, $handler) For responding to an HTTP POST request
$app->put($pattern, $handler) For responding to an HTTP PUT request
$app->delete($pattern, $handler) For responding to an HTTP DELETE request
$app->map(['GET', 'POST'], $pattern, $handler) For responding to more than one HTTP method with the same handler

You can probably spot the naming convention!

The pattern

The $pattern parameter contains the path that you want to match. Internally, Slim uses FastRoute and the pattern syntax comes from there. This is a regex based system and has a number of ways to match:

Literal '/news' Matches the string as typed
Placeholder '/news/{id}' Use braces to create a placeholder that is a variable available to your handler
Placeholder pattern match '/news/article-{id:[0-9]+}' Only matches if the text for the placeholder matches the regular expression (digits only in this example)
Optional '/news[/{year}[/{month}]]' Use square brackets for optional segments. These can be nested.

The slim4-starter route’s pattern is ''/[{name}]'> This will match / and also any other string that does not contain a /, such as /rob.

The handler

The $handler parameter is the code that will be called when the route is matched. This can be any PHP callable or a string.

If it is a string, then it needs to be either the name of a class or the name of a key in your dependency injection container’s configuration. The class class should implement PSR-15’s RequestHandlerInterface though it can also be any class that implements __invoke().

If you are using a DIC, then Slim will try and retrieve your handler from it, so that you can ensure that any dependencies that the handler needs can be injected. If the DIC cannot provide the handler or you’re not using a DIC, then Slim will instantiate it directly.

The HomePageHandler class in slim4-starter is a PSR-15 RequestHandler and looks like this:

<?php declare(strict_types=1); namespace App\Handler; use Psr\Http\Message\ResponseInterface;
use Psr\Http\Message\ServerRequestInterface;
use Psr\Http\Server\RequestHandlerInterface;
use Psr\Log\LoggerInterface;
use Slim\Psr7\Response; class HomePageHandler implements RequestHandlerInterface
{ private $logger; public function __construct(LoggerInterface $logger) { $this->logger = $logger; } public function handle(ServerRequestInterface $request): ResponseInterface { $this->logger->info('Home page handler dispatched'); $name = $request->getAttribute('name', 'world'); $response = new Response(); $response->getBody()->write("Hello $name"); return $response; }

In this case the handle() method is called by Slim and we must return a PSR-7 Response. We’re using a dependency injection container (PHP-DI) and it injects a logger into our handler so that we can log what we’re doing when we handle a request.

The route object

When you add a route to Slim, a route object is returned. There’s a few useful things you can do with it.

Firstly you can set a name for the route:

$route = $a

Truncated by Planet PHP, read more at the original (another 971 bytes)

PHP 7.3.13 Released – PHP: Hypertext Preprocessor

The PHP development team announces the immediate availability of PHP 7.3.13. This is a security release which also contains several bug fixes.All PHP 7.3 users are encouraged to upgrade to this version.For source downloads of PHP 7.3.13 please visit our downloads page, Windows source and binaries can be found on The list of changes is recorded in the ChangeLog.