Thomas Koch - information technologie

Free Software in companies

nospam@example.com (Thomas Koch) — Fri, 06 Aug 2010 10:46:25 +0200

At work we currently have a discussion: I'd like to develop some components as Free Software. My bosses like the idea, but the client goes nuts only thinking about it.
So as part of the discussion I thought to collect those companies who actively advertise their Free Software. I know about the big ones, but it would be even more interesting to get a list of succesful small companies that share at least part of their inhouse projects. On the other hand it would not be too interesting to list full Free Software companies like Red Hat.
It would also be fine to share this list in some wiki (FSFE?) when it grows larger. For some companies I also list very popular projects from my point of view.

big companies

Adobe
Facebook: Cassandra, Thrift, Scribe, HipHop for PHP, ...
Yahoo: YUI, Design Pattern Library, YSlow, Hadoop, ZooKeeper
Twitter
Google: Chromium Browser, Google Summer of Code, Android, Google Web Toolkit (GWT), Protocol Buffers - data interchange format, Java Collections Library, unladen-swallow - faster implementation of Python, ...
IBM: Developerworks (Tutorials), Eclipse, International Components for Unicode (ICU)

middle sized companies

Digg

small companies

101tec (now Datameer): Katta (distributed lucene on Hadoop), zkClient

Update: Need to add Danga (Gearman), Liip.ch with Okapi, Flux CMS, Jackalope, rackspace

Wohngemeinschaft in Konstanz / Kreuzlingen am Bodensee

nospam@example.com (Thomas Koch) — Thu, 05 Aug 2010 08:32:22 +0200

Vor einigen Wochen habe ich meine Freunde Zellmi und Natalie in ihrer neuen Heimat, der Kommune Waltershausen (KoWa) besucht. Es hat mir sehr gut gefallen und ich wäre am liebsten gleich eingezogen. Aber meine Frau und ich sind gerade erst an den Bodensee gezogen und auch beide beruflich gebunden. Also warum nicht so etwas ähnliches am Bodensee beginnen? Es muss ja nicht gleich eine Kommune sein; eine Wohngemeinschaft wäre ein Anfang. Die Miete ist der größte Haufen Geld, den wir jeden Monat erwirtschaften müssen. Und dann arbeiten wir so lange, dass wir den Wohnraum meist nur zum schlafen nutzen. Was für eine Verschwendung. Und dann halten wir diese Wohnung im Winter auch noch warm, obwohl keiner zu hause ist.
Wenn wir mal die Zeit haben, uns selber Essen zuzubereiten, dann tun wir das für zwei Personen. Was für eine Verschwendung von Zeit und Energie! In der (KoWa) wird immer für die ganze Gemeinschaft von über 10 Personen gekocht. Jeder kommt etwa alle 2 Wochen einmal dran.
Die Mitglieder der Kommune berichten, dass sie durch die gemeinsame Nutzung von Aufenthalts- und Sanitärräumen, Sammeleinkauf und gemeinschaftliches Essen deutlich kostengünstiger lebten. Dann sollte diese Wohnform doch gerade in einer so teuren Wohnregion wie am Bodensee die richtige Wahl sein!
Gerade lese ich, dass in Konstanz und Kreuzlingen ein riesiges Überangebot an Büroflächen existiert: Teil 1, Teil 2. Es ist also vielleicht gar nicht so utopisch mit mehreren Leuten eine geeignete Immobilie günstig zu erwerben.
Wer hat also Interesse? Einfach einen Kommentar schreiben oder Mail an thomas (at) koch punkt ro.
Vielleicht können wir ja mit Ute Hauth kooperieren. Sie und andere möchten einen Co-working space in Konstanz aufbauen. Kann man in einer geeigneten Immobilie sowohl Arbeitsbereiche als auch Wohnbereiche haben?

Update: Ein schönes Vorbild, das Rainbow Mansion

Update2: Links zum Thema: Forum Gemeinschaftliches Wohnen e.V., i3 gemeinschaftlich leben & wohnen, Wohnprojekte Portal. Auf meine Anfrage in einem Couchsurfing.org Forum hat nach wenigen Stunden bereits eine Interessentin geantwortet, die auch aktiv auf ein Projekt hinarbeitet.

NoSQL summer at lake constance

nospam@example.com (Thomas Koch) — Thu, 17 Jun 2010 09:31:00 +0200

Have you already started to try out these new storage/database things commonly referred to as NoSQL? (For you google-bot, I'm talking about CouchDB, Cassandra, HBase, Hadoop, Hypertable, MongoDB, Tokyo Cabinet, etc). Well, it's one thing to install and use them but another to understand all the computer science background about them.
Therefor many developers around the world thought to make this summer a NoSQL summer. Local meetings are held in many cities: London, Los Angeles, New York, Paris, lake constance ...
If you like to meet for a beer and learn and discuss about some of the hottest stuff in computing, then come around! Please select the dates when you can join and the papers you'd like to discuss. You can subscribe to email announcements or ical and rss feeds at the lake constance nosql summer page.

udev ate my laptop today?

nospam@example.com (Thomas Koch) — Sat, 12 Jun 2010 11:10:57 +0200

I finally got punished today for using unstable without knowing enough about my system. Since I can't access my emails I'm hoping for help from planet-debian. Please excuse...

So the last words of my machine:

Loading, please wait...
  One or more specified logical volume(s) not found.
Unable to find LVM volume mylvm/root_crypt
  One or more specified logical volume(s) not found.
Unable to find LVM volume mylvm/swap_crypt
File descriptor 3 (/conf/conf.d/cryptroot) leaked on lvm invocation. Parent PID 
352: /bin/sh
udevd-work[77]: kernel-provided name 'dm-0' and NAME= 'mapper/mylvm-swap' 
disagree, please use SYMLINK+= or change the kernel to provide the proper name

_

Please, if you can help me, write a comment to this blogpost!
(Thank you, Hanno for borrowing me your laptop!)

update: It wasn't udev (this time). Sorry. The problem was, that I updated libdevmapper without also updating dmsetup. I could boot into an older kernel and solve this. A Bug against the LVM Debian package is already filled.

tnt is not topgit

nospam@example.com (Thomas Koch) — Sun, 16 May 2010 16:43:49 +0200

As I've already written, I'm working on an alternative to topgit. I made a first attempt in perl some weeks ago, but gave up after some frustrating hours. Yesterday I started again in python and had a very nice time putting together the groundwork and the first two commands.
It may be noted, that I've no previous programming experience in neither perl nor python!
By now, I can create a patchset branch and add a patch branch to it. There's still a lot to do. For my talk at the Debian Mini Conference in Berlin next month I'd like to be able to update patch branches, export patchsets and give a status summary.
Maybe I can already find somebody who's interested in joining me with this project? The code is in my github account, however the name will most probably change.
One reason that I've been much faster in python is the fantastic python-git library. I can only recommend it!
In other news: I'm searching a couch to surf in Berlin from june 7.-12. I prefer couchsurfing over hotels mostly to get to know nice people around the world. Please contact me, if you'd like to host me for a night or two. (thomas at koch punkt ro)

Update: Slides of my talk at the debconf are available.

Design document for a patch management system on a DVCS

nospam@example.com (Thomas Koch) — Fri, 02 Apr 2010 12:48:33 +0200

Dear friends of Debian,

this is my first post to Planet Debian. - The planet with the most geeky registration procedure in the known universe!

I proposed an alternative to topgit some days ago on the vcs-pkg.org list. Martin asked (and encouraged) me to give a better explanation of the idea, which I'll hereby try. Sorry for not giving any drawings, but I'm totally incapable of anything graphical.

Hopefully, I'll manage to come to the Debian Miniconf in Berlin. Then we could discuss the idea further and maybe even start implementing it. (Somebody would need to help me with my first steps in Perl then...)

The following text is available on github. Please help me expand it!

Design document for a patch management system on a DVCS

Requirements

The system to implement manages patchsets. A patchset is a set of patches with a tree-ish dependency graph between the patches. There's one distinct root of this dependency graph.

Patches are managed as branches with each branch representing a patch. Modification of a patch is done by a commit to the respective branch. A branch representing a patch as part of a patchset is called patchbranch.

The patch of a patchbranch is created as the diff between the root of the patchbranch and the head.

The most important management methods are:

Export a patchset in different formats
- quilt
- a merged commit of all patches
- a line of commits with each commit representing one patch
Update a patchset against an updated root.
Copy a patchset
Delete a patchset from direct visibility while preserving all history about it
Hide and unhide a patchset from direct visibility

Additional requirements:

The system should be implementable on top of GIT, Mercurial and eventually Bazaar.
The system must easily cope with multiple different and independent patchsets.
All information about a patchset must be encoded in one distinct branch. Publishing this one branch must be sufficient to allow somebody else to recreate the patchset with all of its patchbranches.
The system should not rely on the presence of hooks.
The system should not require the addition of management files in patch branches (like .topmsg and .topdeps in topgit)
The system must be easy to understand for a regular user of the underlying DVCS.
The implementation may allow a patchset to depend on another patchset(s).

implementation

patchset meta branch

A patchset meta branch holds all informations about one patchset. First, it holds references to the top commits of all patch branches in the form of parent references of commits. Thus pushing the patchset meta branch automatically also pushes all commits of all patch branches.

Secondly, the patchset meta branch contains meta informations about the patchset. These meta informations are:

The names of all patch branches together with the most recent commit identifier of a particular patch branch. Let's save this information in a file called branches.
A message for each patch branch that explains the patch. These messages can be saved in the file tree as msg/${PATCH-BRANCH-NAME}
References to the dependencies of the patch (other patches of the same patchset or the root of the patchset). This is also encoded in the file branches.

Since the patchset meta branch holds all this informations, it is possible, to delete all patch branches and recreate them from this informations.

Although the commits of the patchset meta branches hold references to the patch branches, its file tree does not need to contain any files from the referenced patches. This may confuse the underlying DVCS, but the patch meta branch is not ment to be directly inspected.

The branches file

A branches file for a fictive patchset could look like:

# patch branches without an explicit dependency depend on the root of the
# patchset tree
# A Root can be given as either a fix commit (seen here), a branch or a tag.
# A fixed commit or tag is useful to maintain a patchset against an older
# upstream version
ROOT: 6a8589de32d490806ab86432a3181370e65953ca
# A tag as a dependency
#ROOT: upstream/0.1.2
# A branch as a dependency
#ROOT: upstream

# A regular patch with it's name and last commit
BRANCH: debian/use-debian-jars-in-build-xml 4bab542c261ff1a1ae87151c3536f19ef02d7937

# two other regular patches
BRANCH: upstream-jira/HDFS-1234 a8e4af13106582ca1bfbbcaeb0537f73faf46d87
BRANCH: upstream-jira/MAP-REDUCE-007 e3426bcbcb2537478f851edcf6eb04b34f6c7106

# This patch depends on the above two patches
# The sha1 below the dependency patches references a merge commit of the two
# dependencies
BRANCH: upstream-jira/HDFS-008 517851aa829d77e09bc5e59985fed1b0aa339cc6
DEPENDENCIES:
  upstream-jira/HDFS-1234
  upstream-jira/MAP-REDUCE-007
    cc294f2e4773c4ff71efb83648a0e16835fca841

# A patch branch that belongs to the patch branch, but won't get exported (yet)
BRANCH: upstream-jira/HDFS-9999 74257905azgsa4689bc5e59985fed1b0aa339cc6
BRANCH-FLAGS: noexport

Zookeeper for web developers

nospam@example.com (Thomas Koch) — Fri, 19 Feb 2010 10:42:48 +0100

Have you ever developed any kind of distributed system? When doing so for the first time, you're very likely to fall in the trap of the Fallacies of Distributed Computing. I've done so, you'll do so too.
Now zookeeper is an application, that helps you implement many distributed protocols on top of it. The hard work of implementing fault tolerance, assuring consistency and that kind of stuff is done by zookeeper in the background. A zookeeper cluster consists of at least three servers running zookeeper (zk). A client can connect to each zk server and issue read and write requests. Zookeeper guaranties that a write either fails or is consistent and that a read will get you the most recent state.
Zookeeper exposes a filesystem like hierarchy of so called znodes. Every znode can have children but also stores data. The data stored in a znode is assumed to be small (less then 1MB). Clients can subscribe to different events on a znode and will be notified of changes in the znode itself of it's children.
Some usage examples come to my mind, which could be especially interesting for PHP developers:

save the PHP session in zk and have it therefor available to all web servers
save a shopping card
save the online status of a user (for a chat system)
synchronize configuration files
master election: many monitoring servers, but only one is active at any given time
logging system: bookkeeper is a zk contribution, that receives your logs and assures they'll be kept save
create a lock on a document in a CMS

Companies that use zk include Facebook, Yahoo, Rackspace and as rumors tell also twitter.
Now this is all kind and sweet, the only sad thing is, there's no PHP binding yet. There are bindings for Java, C, Python and Perl. So if you're desperately searching what PHP extension you should write next, just take the zk C binding and expose it to PHP!
Maybe you're a student and would like to participate in the Google Summer of Code? It should be possible to find a mentor for this project either in the PHP or zk project.
For GSOC the project could include a PHP session implementation in the extension code and the possibility of persistent zk sessions across PHP calls.
If you're in the region of eastern Switzerland: There'll be a presentation about zookeeper at the Webtreff Kreuzlingen on one of the next mondays.

Going to FOSDEM 2010

nospam@example.com (Thomas Koch) — Sat, 30 Jan 2010 20:52:08 +0100

I'm going to fosdem 2010.

These are things I'd like to attend:

time	room	topic	speaker
10:45-11:30	Janson	Promoting Open Source Methods at a Large Company	Brooks Davis
13:15-13:45	AY	Packaging Java Software for Debian	Thomas Koch
13:40-13:55	Ferrer	CiviCRM: Common goals of FOSS and Not For Profit Organisations	Xavier DUTOIT
14:00-14:45	H.1308	The Maemo Community Council: a case-study in governance	Dave Neary
14:00-15:30	Guillissen	LPI exam session 1	Klaus Behrla
14:00-14:45	Janson	What is my system doing - Full System Observability with SystemTap	Mark Wielaard
14:30-15:00	H.2213	Building Federated Social Networks on XMPP	Tuomas Koski, Simon Tennant
15:00-15:45	Janson	Ganglia: 10 years of monitoring clusters and grids	Bernard Li
15:00-15:30	H.2213	XMPP and the Social Web	Alard Weisscher, Laurent Eschenauer
15:30-16:00	H.2213	PubSub Gone Wild: Info Sharing at Mediamatic	Ralph Meijer
16:00-17:30	Guillissen	LPI exam session 2	Klaus Behrla
16:00-16:15	Ferrer	SIP Communicator: Skype-like conf calls with SIP Communicator	Emil Ivov
16:20-16:35	Ferrer	Kamailio (OpenSER) 3.0.0: redefinition of SIP server	Daniel-Constantin Mierla
16:40-16:55	Ferrer	asterisk: An introduction to Asterisk Development	Mark Michelson
17:00-17:15	Ferrer	csync: Roaming Home Directories	Andreas Schneider
17:30-18:00	H.2213	Jingle Nodes: An Open Alternative to Skype	Tiago Camargo
18:00-18:30	H.2213	Multi-User Jingle: Voice and Video Conferencing with XMPP	Dafydd Harries, Sjoerd Simons
18:15-19:00	H.1302	Spacewalk: Linux Systems Lifecycle Management	Marcus Moeller, Sandro Mathys


09:15-10:00	H.1302	Linux distribution for the cloud	Peter Eisentraut
10:00-10:45	Janson	RepRap - Manufacturing for the Masses	Adrian Bowyer
10:30-12:00	Guillissen	LPI exam session 3	Klaus Behrla
11:00-11:45	AW1.120	My life with HBase	Lars George
11:00-11:45	Janson	Tor: Building, Growing, and Extending Online Anonymity	Andrew Lewman
12:15-13:00	H.1302	Cross distro packaging with (top)git	Thomas Koch
12:15-13:00	H.1308	Shared libraries in Debian	Sune Vuorela
13:00-14:30	Guillissen	LPI exam session 4	Klaus Behrla
13:15-14:00	AW1.120	CouchDB, a database designed for the web and more	Benoit Chesneau
14:00-14:45	Janson	Large scale data analysis made easy - Apache Hadoop	Isabel Drost
15:00-16:30	Guillissen	LPI exam session 5	Klaus Behrla
15:30-16:15	H.1308	Continuous Packaging with Project-Builder.org	Bruno Cornec
15:30-16:15	AW1.120	Comparing the MapReduce way in CouchDB with the SQL way in a RDBMS	Stéphane Combaudon
16:15-16:45	AW1.120	Designing a scalable content management system on NoSQL technologies	Evert Arckens
16:15-17:00	H.1308	Debian Secrets: power tools for power users	Wouter Verhelst

Exploring Eclipse, PDT and Eclim

nospam@example.com (Thomas Koch) — Sat, 24 Oct 2009 16:48:24 +0200

Back in 2006 I learned, that KWrite (a simple text editor) was not enough to write PHP code with. I tried eclipse, but it seemed too complex and was way too slow on my old laptop.

So inspired by the Comfortable PHP editing with VIM series of Tobias Schlitt I learned VIM.

My blogpost about Vim as an IDE for PHP is a popular page and I often get requests about details of my setup. Please don't ask me anymore, since I intend to move away from this setup!

Some weeks ago I've been at a conference (GearConf) about team collaboration. Some of the tools presented there (like Mylyn) made me jealous of the java people.

So it was time to reconsider Eclipse. However editing text with what's called an "editor" in eclipse is a pain to say the least. But life's good and somebody already created eclim. Eclim allows you to keep using gvim for text editing but integrates with Eclipse to give you code completion, manual lookup, code projects and more.
This alone may not convince you, since I already have this integrated in vim. But there are at least three major advantages of vim+eclipse:

There's not only the integration for PHP but also Java, Javascript, Python, Ruby, C/C++, CSS, HTML and XML
Eclipse comes with a bunch of other cool plugins (like Mylin)
The code completion from eclipse is much more powerful then anything you can do with VIM. For example it also proposes you the correct classmembers after foo:: or $bar->

My collegue Jiayong just installed the bundled version of PDT with eclipse, ran the eclim installer and was ready.

However I'm much more addicted to pain, so I took the Ubuntu packages of eclipse 3.5.1, linked the pdt runtime download and all dependencies in a separate directory, checked out the GIT repository of eclim, configured the classpath in the ant buildfile and build eclim from source. The deployment of eclim can be looked up from the ant buildfile.
There are some more tweaks to do, which I can point out later, if you should dare to follow this road. You shouldn't have a need for this until you'd be interested to become famous and beloved and package PDT or eclim for Debian.

Do you already have experiences with eclim+pdt? Are there tutorials I've not found? Are you interested to come with me to explore this setup?

eZ Components Template is not slow...

nospam@example.com (Thomas Koch) — Fri, 09 Oct 2009 13:40:42 +0200

... you may just not have used it appropriately. Fabien Potencier has blogged about templating engines, compared their speed and finally concluded that his own templating engine is the fastest. Not only that, but my beloved eZ Components Template engine, which we use with great pleasure at YMC, came out as the slowest in his benchmark. (Althought he complimented it as "probably the one which have the most features".)

Dear Fabien, I think it's not fair to blog a benchmark without publishing the code you used to do the benchmark! But even without the code and without doing my own benchmark, I can offer some explanations, why eZ Components may have been the slowest in your benchmark and why you've compared apples with oranges. I take the compiled template code from the Twig site:

CODE:

/* Hello {{ name }} */ class __TwigTemplate_1121b6f109fe93ebe8c6e22e3712bceb extends Twig_Template { public function display($context) { $this->env->initRuntime(); // line 1 echo "Hello "; echo (isset($context['name']) ? $context['name'] : null); } }

Thinks to keep in mind from the Twig compiled template:

There's direct output via echo (which is captured by Twig and not directly sent out)
The template variables are sent in the array $context
There's no error check, whether the required template variable 'name' is actually sent to the template
The content of the variables is not escaped

Now let's compare this with the compiled code produced by eZ Components Template. First the template:

CODE:

{use $name} {use $items} Hello {$name} {foreach $items as $item} * {$item} {/foreach}

And the resulting code:

CODE:

<?php // Generated PHP file from template code. // If you modify this file your changes will be lost when it is regenerated. $this->checkRequirements(1,array("disableCache" => false)); $i_output = ""; if ( !isset($this->send->name)) { throw new ezcTemplateRuntimeException( sprintf("The external (use) variable '%s' is not set in template: %s and called from %s", 'name', $this->template->stream, ( sizeof($this->template->streamStack) >= 2 ? $this->template->streamStack[sizeof($this->template->streamStack) - 2] : 'the application code') ) ); } if ( !isset($this->send->items)) { throw new ezcTemplateRuntimeException( sprintf("The external (use) variable '%s' is not set in template: %s and called from %s", 'items', $this->template->stream, ( sizeof($this->template->streamStack) >= 2 ? $this->template->streamStack[sizeof($this->template->streamStack) - 2] : 'the application code') ) ); } $i_output .= " \nHello "; $i_output .= htmlspecialchars($this->send->name); $i_output .= "\n \n"; foreach ($this->send->items as $t_item) { $i_output .= "* "; $i_output .= htmlspecialchars($t_item); $i_output .= "\n"; } return $i_output; ?>

The differences are:

The output is collected in a variable, not with PHP's output buffering. I don't know whether this has a performance impact, but in any case it allows for much greater flexibility in the implementation of the template engine. Thus it's possible for example to
The template variables do not come in a simple array but as an object of type ezcTemplateVariableCollection. This is surely slower as an array, but also allows to do great things.
My favorite eZC Template feature: Every variable needs to be declared at the top of the template! I wish PHP would be that strict.
And now the point which may be responsible for the observed performance impact: eZC Template by default escapes all variables with htmlspecialchars. This is of course configurable and so I could compare template execution time with escaping turned on and of: For an array of thousand elements it more then doubles the time.

So in conclusion: Please post the benchmark code, so that we can debug it and please compare the template engines with the same features turned on or off!

Timeline of PHP quality assurance tools

nospam@example.com (Thomas Koch) — Sat, 26 Sep 2009 16:58:46 +0200

I'll be giving a talk on PHP quality assurance tools next week at the gearconf in Düsseldorf. The event is focused on team collaboration and most of the audience will come from other languages then PHP. Since I'm not a contributor to any QA tool I'll center around the user perspective and the process of establishing these tools in a small to middle web company like YMC.
As an outline for the first part I thought to give an historical view on QA in PHP and therefor made a timeline of PHP QA milestones:
The dates as text:

~1997 JUnit
6.4.2002 PHPUnit 0.1
28.8.2003 PHPUnit 1.0.0-alpha1 Release
13.7.2004 PHP5.0.0
14.7.2004 release PHPUnit2
30.1.2006 eZ Components 1.0
~sept 2006 ZF uses PHPUnit (issue #374)
4.10.2006 first integration of Selenium in PHPUnit
1.7.2007 Zend Framework 1.0
18.11.2007 first release of PHPUnderControl
19.11.2007 Project Mess Detection in PHPUnit
6.2.2008 first release PHP_Depend
march 2009 qualityassuranceinphpprojects.com, thePHP.cc

main sources:

Did I get all this right? Are there things I should have added?

talking on basic income at google open source jam

nospam@example.com (Thomas Koch) — Wed, 12 Aug 2009 16:42:25 +0200

Tomorrow evening I'll be going to google's open source jam in Zürich. I want to explain, why free software and the idea of basic income do play together well. The slides are already prepared:
google_jam_talk_basic_income.pdf

openruhr.de übernimmt ruhrtux Idee

nospam@example.com (Thomas Koch) — Fri, 07 Aug 2009 10:06:31 +0200

Im Herbst 2007 hatte ich die Idee, die Initiative Ruhrtux zu starten. Es ging darum, Freie Software und Freie Kultur als Bestandteil der Kulturhauptstadt 2010, dem Ruhrgebiet, zu vertreten. Leider habe ich für diese Idee keine Unterstützer gefunden, weder das "Linux und Open Source Kompetenzzentrum NRW" noch die "Ruhr Networker" oder das Linux Hotel in Gestalt von Herrn Wiesemann zeigten Interesse sich in irgendeiner Weise für die Idee zu engagieren.

Inzwischen ist die Webseite ruhrtux.de wieder offline und man findet nur noch Spuren davon im Netz.

Heute sehe ich, dass Herr Wiesemann die Initiative openruhr.de startet. - Vor allem anscheinend als Werbung für sein Linux Hotel. Na dann viel Glück. Ich habe meine Lektion gelernt.

PHP SAPI to run gearman workers

nospam@example.com (Thomas Koch) — Mon, 20 Jul 2009 09:17:44 +0200

Gearman is a system to run jobs on multiple servers. We at YMC are working know to integrate gearman in our cluster solution. Since we're a PHP shop, the jobs should be written in PHP.
There already is a solution with the gearman PHP extension by James M. Luedke. But I wanted to run multiple workers on one machine and having them use an opcode cache. The solution is a dedicated PHP binary (SAPI), which runs many worker threads in one process. The code (see /sapi/gearman_worker) in branch gearman_worker is already usable and I'd be happy to receive comments on it. After all I'm not (yet) a C expert.

TODOS:

graceful stop on SIGTERM
include the possibility to run cron-like scripts in a loop
pause, if available memory is too low
documentation
get it into upstream

eZ Components: The Good The Bad and the Ugly

nospam@example.com (Thomas Koch) — Wed, 17 Jun 2009 11:07:21 +0200

I've been nominated again for the eZ Components Award. Thanks to eZ Systems for honour and to me employer YMC for the food!
So this is a good occasion to assemble a list I've had in mind for a longer time. In a big project like eZ Components, which has been programmed by many contributors over a long time it is natural, that there are some better and some worse parts. So which components are the good, the bad and the ugly? Maybe this kind of overview could also be done by other users of other libraries too. Then we could assemble a big comparison table of framework or library components and one could pick the best parts of each.
You'll notice that even if I consider myself a power user of eZ Components, I won't give a rating for every component, because there are still a few, I've not intensively used by now. So let's complete this table in a wiki somewhere together!

So IMHO:

Component	rating	justification
Authentication	works	the component should better not handle $_SESSION itself
Cache	not used yet
Configuration	not used yet
ConsoleTools	great
Database	great
DatabaseSchema	bad	You want to have more control over your database schema.
Document	great	first ReST parser in PHP? Also many other formats!
EventLog	great
Feed	great
Graph	great
ImageAnalysis	not used yet
ImageConversion	not used yet
Mail	great
MvcTools	works	too few documentation of best practices
PersistentObject	great
Search	ugly	sorry, but it's not flexible enough for more then the most basic needs
SignalSlot	bad	haven't seen any use case that convinces me
SystemInformation	bad	doesn't comply to ezc coding guidelines
Template	great	stable, fast, easily extensible
Translation	works	Misses the important XLIFF format, but can be extended
Tree	not used yet
Url	bad
UserInput	ugly	see my attempt for a new form component
Webdav	not used yet	but looks very promissing!
Workflow	great

The numbers:

not used yet: 6/25
works: 3/25
great: 10/25
bad: 4/25
ugly: 2/25