Poo-tee-weet

OSS projects at Liip

Lukas Kahwe Smith — Thu, 02 Jul 2009 15:07:04 +0200

Yesterday we had an entire day at Liip dedicated to various talks, workshops and hanging out together. It really reminded me again what an amazing company Liip is. Its really what I wished m company back in Berlin should have been. Fun, smart, successful, productive, agile, good. At any rate I wanted to mention the stuff we are working on in terms of OSS. Obviously Chregu and myself are active on PHP. Jordi is also an active member of the PHP.net community. But we have other less known stuff that should also get noted, like Jackalope, Mahara, Picok, GottaGo and Okapi.

Jackalope is an implementation of the client side JCR standard for PHP. This will enable us making happy agile frontends that are managed with all the various Enterprise features like versioning, ACL's etc. Liip created this project but is looking for others to join. We are already sharing the interfaces with the typo3 guys.

Mahara is a social networking platform that has spawned out of the Moodle e-learning community. Penny recently joined Liip and she was actually one of the original founders of the project.

Picok is a one of my babies. Its sort of an OSS iGoogle. So it allows you to easily create new so called portlets, that can aggregate content from various internal and external data sources and present them in a user customizable interface. Originally created for one of our clients the Raiffeisen Bank here in Switzerland. They were kind enough to let us open source the code. In return we won them a couple prizes at the Swiss Web Awards :)

GottaGo actually took the Swiss Web Awards by storm. It won the grand prize along with a few other category prizes. Its an iPhone app that basically finds the fasted route via public transportation from point A to B here in Switzerland. Oh and I should stop referring to it as GottaGo, since its now called "Transport".

Okapi is our super lightweight framework. Recently Chregu and I have started working on version 2. Basically we took various Symfony2 components and integrated them into our code. Currently its still in proof of concept stage, but it looks quite promising. I am really seeing a lot of potential in the service container and event dispatcher. I also think it could become a very useful addition to the Symfony eco-system as it seems quite feasible to migrate from Okapi2 to Symfony2 or vice versa. Meaning it will be able to choose if you want something simple and lean or you need something full features, without having to need to learn two totally different frameworks. Chregu has ported liip.to to this proof of concept, so check it out!

Its done!

Lukas Kahwe Smith — Tue, 30 Jun 2009 14:54:52 +0200

W00t! 5.3.0 stable is out. Never expected it to take this long, but its a lesson to keep faithful to the "release early, release often mantra" in the future. At some point a branch can just get too big, that it becomes close to impossible to release it. While I do appreciate the various thank you (or in the case of backslash haters "drop dead") emails, I want to highlight that all I really did is take over the tasks that require at most a bit of patience and certainly no technical skill (well I did have to update a few websites, but even there I needed Hannes to fix all my typos).

So Johannes was the one who was especially in the last few weeks laboring over tons of patches and bug reports and sleeping less and less as we got closer to the actual release date. Pierre was of course also a huge help in this process .. and of course the countless developers that worked on the actual code that you all can now download and play with.

Now I am not saying that 5.3.0 hasn't eaten up my share of time. I am very thankful to my employer Liip that let me spend time on all of this. A bit of free time also went into 5.3.0, which I hope to now direct towards my music career a bit more :)

PHP 5.3.0 stable almost released :)

Lukas Kahwe Smith — Thu, 25 Jun 2009 08:55:32 +0200

It almost happened, but it didn't for now. Originally we planned to release today. But again a few issues came up, even with Johannes deciding that sleep is for the weak, it just seemed unwise to announce the release today. So we pushed things back a few days, so the new date is June 30th (meaning it will be a Tuesday release). This also gives the documentation team, who have been expanding the 5.3 docs like crazy, a few more days to beef things up even more. Now is really high time to ensure that you have a PHP 5.3 compatible release of your software ..

A new coding standard for the PHP world?

Lukas Kahwe Smith — Tue, 02 Jun 2009 22:29:47 +0200

There is currently a heated debate spawned with the creation and subsequent announcement of a more or less final decision made by a select group of people at php|tek. Now I am a friend of openness, but I have a hard time remembering any significant existing project within the PHP community that actually changed its coding guidelines. For all I know the main CS in the PHP world originated in the Horde project, but adopted and expanded by PEAR, got adopted and expanded by Zend Framework. See the pattern?

Now PEAR is actually trying to do what hasn't been done before, change the CS in an existing community. From being heavily involved in PEAR years ago, I can tell you that all CS discussions on the mailinglist where useless. Of course there are bad CS and there are very good CS decisions. I personally have a few things to say here. But for the most part its about defining a standard and sticking to it. So handing over the bulk of the deciding to the PEAR group (not to be confused with the PHP group that has no real active responsibilities beyond being the copyright holders) was the only sane thing to do. Now they took things one step further and got a bunch more frameworks together. Awesome.

The mistake they made was to publish their findings on a mailinglist on php.net called php standards. It should have been published on the PEAR developer list. Or maybe some new mailinglist, just not one that makes it sound like anything more than a standard to be adopted by the named list of frameworks. Of course they should promote the standard, adjust the standard if needed, but keeping the actual deciding in a small circle again is the only sane approach for this. But making it sound bigger than it actually is, is not the way to go. Let it grow naturally and all will be good.

New MySQL development approach?

Lukas Kahwe Smith — Sat, 23 May 2009 21:39:04 +0200

Looks like there is hope yet for the owners of the MySQL copyright to bring around their development. There is a proposal (actually it seems its already decided to ahead with it) up on their public wiki that details a development approach that could finally become compatible with the real world, rather than managers checking off feature lists. Maybe its too late already, but maybe its just in time to "fend off" the increasing competition for who provides the best MySQL distribution.