Lightning Talk : PyRoom

PyRoom appears to be a monochrome full screen editor to write novels; Full screen is not distracting you from your work. Writeroom vs PyRoom : Commercial, costs money vs Free The troll alert slide was terrific ! So far, I am not too excited... but the guy is making up for the somewhat deep void of the subject (but a good thing is that  a lightning talk lasts 15 minutes, so my suffering will not be too long lasting).

MySQL Room – Roadmap, Q&A, Recent Developments at Sun

Kaj Arnö, VP Community is going to shed some light on the subject. I was hoping to get a talk about social networks and MySQL but the title is not very much in line... Kaj has a number of blogs etc (but since the sucky connectivity of FOSDEM, I've lost my post and switched to the 3G link). So news:
  • Monty and Marten are gone.
  • Karen Tegen Padir is the new boss.
So, what to expect ? Roadmap ? So, MySQL Strennghts:
  1. FOSS
  2. TCO is low
  3. Stability
  4. Performance
  5. Ease of use
And weaknesses:
  1. Feature set
  2. Gfx tools buggy
  3. Slow release cycle (I tend to agree !)
  4. Contribution process (but developers who contributed were hired !)
Kaj presents MySQL opportunities: Enterprise (including some non-opensource software), Cluster, Support Team, ... Kaj's presentation is fun to look at since he uses his own face with funny expressions on the left of the slides in a black & white style. It's cool, I'll give the idea a shot in a presentation of mine. Kaj explains the business model of MySQL (1000:1) / put something out and then ask "will you buy?". This doesn't work, we need intermediate steps. So, whitepapers help the business audience. Webinars attract the technical guys and do register. They help in finding the pain points, the real buyers... Also a good way to get leads in a steady flow. MySQL Threats:
  1. Founder Headaches (he says that they've got rid of them, so no headaches any more).
  2. "It's just a toy" statement
  3. Software engineering process not improving
Some questions: "Why did Sun buy MySQL" Similar culture, different customer bases, sell more HW, sell more MySQL S "What are we not so good at ?"
  • half-baked features : e.g. stored procs exceptions lingering for years with nobody taking care about that.
  • create index is slow, needs lots of space
  • interoperability with other products is weak
  • convincing large accounts that MySQL is production grade, enterprise ready. Kaj answers that they should market MySQL more as "the on-line database" instead of "the open source database". And actually, 20 of the world's largest sites do run on MySQL.
  • the fact that there are various storage engines is both a strength and a curse. Performance may be killed by choosing the wrong one. Something is done these days about that.
Ok, now to the roadmap ! 5.1 is now GA. "What should happen in MySQL 5.1+ ? more support for third party storage engines. more regular and faster releases price policy: bad for cluster. Answer: we've not sales here ! "Monitoring support for the server" Mostly done with tweaks & hacks for monitoring inside the server. But the general feeling is that it is quite stable. "What do you think about the ease of use of MySQL?" Lots of things easier than with other RDBMSes When coming into specific storage engine cases, that's becoming quite hard So, lots of shos of hands at the moment. My battery is running out, I guess I'll have to swap it with a new one! Post over for now.

Attending a Drupal 7 Session – Multisite

FOSDEM  entrance After having discussed a bit about TRIM with Marc Laporte yesterday evening, I was curious to see how Drupal handled the multi-site management. Emma Jane Hogbin is going to tell us more 🙂 BTW, she's working on a book about Drupal Front End theming. Interesting concerns: - why to use it and which clients are best suited to use these features. Goal: make technology more accessible for clients. "The 100 Mile Client Roster" : 10K a year. Motto: "Reduce, Reuse, Recycle" Reduce: Security updates Reuse: Downloaded themes Documentation Answer questions once "Drupal night" Invoice less frequently for larger amounts Assumptions: * low traffic sites and unique content per domain name. Common functionality. Brochureware. Editable themes. Overnight, little work are key objectives. has the technical details. Lots of sharing: shared code base, IP, modules directory segregated databases all common modules shall go into /sites/all/modules Google Analytics and google account created for clients so that they can access stats (when they do realize they want them, maybe months later when they do realize it exists !). Drupal has a sexy installer, that's a fact. Presentations are here: irc: emmajane mail: A dynamic talk but not too deep on tech. General impression: Drupal can work for lots of small sites. Q&A touches on backup (Emma Jane uses the backup module). Weekly backups. The backup module can be automated through cron. People usually forget their password. Having a feeling of control over their site leads users to commit more money for larger projects.