A Finland Swede in Bavaria

Monday, 20 October 2014

On obsession, Monty and gender issues

The problem with Monty is his communication. For instance, he often starts his sentences with The problem with <x> is. Like last Friday, where x == women. Not a great start of any statement, and a particularly bad one for x == women. For the context, you can read Laine Campbell's blog "The Meritocracy, and a Call to Action, thanks to Monty Widenius" and for the ensuing Twitterstorm, look for #ayb14 and Monty.

I wasn't present at #ayb14 in Oxford, nor have I heard exactly what Monty said. Judging from the strong tweet by @AYBConf that "We work hard to make every AYB inclusive and were gutted by @montywi's comments earlier which were plain wrong #ayb14", it wasn't pretty. 

On the other hand, I know Monty since the late 1970s, and I am sure Monty as a father of two daughters feels bad about being called a sexist. I am certain he thinks he honestly respects women. Period. This includes respecting women as programmers.

So I called up Monty to get his side of the story. 

Kaj: Monty, have you read Laine Campbell's blog entry from the All Your Base Conf?

Monty: No, I'm outside internet reach travelling with my family near Oxford. But I expect the worst. I am afraid I failed to get my thoughts across properly during AYB. 

Kaj: Well, in a way, it could be worse. Laine says "MySQL has been my bread and butter and one of my top passions in open-source for over a decade. I am always happy to see Monty and hear from him." And moreover, she confirms she is not out to "villainize you". What really happened?

Monty: I had said MariaDB was "10 man years in front of MySQL" in the optimiser area, and Laine asked me "what about woman years". First, I used "man years" because I am a non-native speaker and express myself in a way I believe to be understood by other non-native speakers. That doesn't capture lots of nuances or diplomacy. Second, I am not the sort of person to simply give a politically correct boilerplate answer, so I went into thinking out loud about the underlying reasons for the unequal gender distribution. 

Kaj: And, in your mind, why are there more male programmers of lead open source projects?

Monty: Obsession. My personal feeling is that men get more easily obsessed with things, including computers. They forget their surroundings, neglect their families and friends. And I think that's more insulting to men than to women, who I think are better at combining work and having a life. This type of male obsession I find to be mostly a negative trait, even though single-minded focus on coding from the early teens, followed by 16 hour coding days at work, is what has enabled many successful pieces of software.

Kaj: Can't women get obsessed? Or do you think they shouldn't?

Monty: I know lots of excellent women coders. I have personally been involved in hiring several of them, and I encourage more women to become hackers, software architects, and software engineers in general. And when it comes to obsession, I'm not talking about what should be, I'm just observing what has been the case around me. When women are obsessed, I find they tend to focus on social aspects, which is why there have been much more of an uptake in women in senior management, rather than among introverted, solitary hackers.

Kaj: Is obsession a prerequisite for being a hacker?

Monty: I am not saying that one needs to be obsessed to become a good programmer or have an influence in the software industry, and most are not. However many of the extreme programmers, especially those who create big software protects by themselves, are obsessed with what they do.

Kaj: In her blog, Laine mentions you acknowledging women as having a better work-life balance, quoting you as saying "It was the men who worked too hard, had no lives, no connection to families.". But Laine's key point is this: "this is not a gender problem, and shame on you Monty for saying it was." And she goes on to say "I don’t want this to be a witch hunt. I’d love to see Monty acknowledge his generalizations, and the harm that they cause." Can you follow this?

Monty: Yes, I can. Let me relate to my own situation. There were over ten years of 16 hour coding days during the 1990s and early 2000s, when I wrote MySQL. I was in the fortunate situation to have a wife, Carola, that took care of our daughter My (as in MySQL) and our son Max, taking a two-year break from her own career. Her sacrifices are what enabled me to focus so single-mindedly on MySQL. I am very grateful to her for that. Now, do I think the role division between Carola and me in the 1990s should be the norm, in the 2010s? No. It wasn't wrong, it wasn't right, it just was. And without that role division, there wouldn't have been MySQL as we know it.

Kaj: Looking back, are you happy with that role division?

Monty: I don't think it makes much sense to regret past decisions that I can't influence any longer. And I did spend what people call "quality time" with my children, not large quantities of it, but if you ask my children, I do believe they will acknowledge that "although dad worked all the time, he was still there for us when we needed, he often made delicious food, and we spent fun times together during holidays when we travelled".



 Picture: Monty on Galapagos 2002, surrounded by children, some of which are his own

Kaj: Laine thinks you made gender generalisations and should acknowledge them as such, including the harm that they cause. Do you?

Monty: I am deeply distressed by the situation. I feel bad to be thought of as a sexist, because I believe everyone should be given equal opportunities. While I am not a social policy maker, I replied with my reflections on the root causes of why there are more male programmers in the cultures where I have been working, and I understand that wasn't wise of me. I don't want to cause women coders any harm, and as I can observe from the angry tweets after my outspoken comments at AYB, I have done just that. I am deeply sorry!

Kaj: Thank you, Monty!

Friday, 26 October 2012

Paperwork and personal accounting: Archiving financial documents


Expense reports, tax statements, invoices, insurance policies, claims, ordinary receipts: Life is full of boring bureaucracy. Most of us try to avoid spending any unnecessary time on paperwork, and I'm no different. That said, I've noticed that I am probably not really minimising my paperwork time, going about some matters in a sloppy, irrational way. Procrastination and bad habits lead to spending more time, not less -- fighting exceptions, asking for extensions, and most of all, searching for badly organised but vital information - be it on paper or in electronic form.

So I decided to change things, for a more systematic approach. I'm my own all-in-one clerk, controller, CFO, and auditor. Perhaps writing about it will make a boring task more fun? Perhaps documenting my own processes will make me more organised? Compared to corporate finance and administration, I have many more degrees of freedom. Fewer legal requirements. No arbitrary corporate policies. More possibilities to make rational choices. And no boss to report to - which, in fact, can turn out to be more of an invitation to procrastinate, than a liberation from unnecessary duties.

My current task at hand is to create processes around the everyday paperwork. The goals are simple:

  • keep an overview of expenses
  • set my own expectations on upcoming expenses
  • pay valid invoices on time
  • write expense reports quickly
  • turn in tax declarations on time
  • submit reports to authorities like insurance companies on time 
  • manage account numbers, passwords and contact data
and do all of the above with minimum hassle, frustration and interruptions

Starting at the bottom-most level, I need to decide on a workflow for paperwork, from incoming documents, processing them, and archiving them. Paperwork is in my case distributed over a multitude of inboxes: snail mail to two physical addresses, and email which in turn may be mere notifications of the need to log in to accounts behind passwords to banks, credit card companies and various other service providers (all in the name of "simplification"), followed by mandatory use of yet another, particularly badly structured de-facto inbox in the form of the Downloads directory.

My belief is that I will find out the easiest process by looking at the desired end state: What documents do I want to archive? In which state, and where?

My tentative answers are:

  • In paper form, I want to long-term archive only such financial papers that legal, contractual and other formal obligations require me to retain on paper - i.e. very little
  • In electronic form, I want to retain all receipts, bank statements and paperwork which either documents past transactions or is relevant for future assets and liabilities
  • Work in process, I want to keep at a minimum, and in very good order - both in dead-tree and electronic form 
  • Concluded, archivable paperwork, I want to archive properly, in good order, by year - only in electronic form, disposing of the corresponding dead-tree instances
  • Turning to the Where part: The e-paperwork in process, I want to keep on my laptop hard disk, with a backup copy on an external hard disk
  • The e-paperwork archive, I want to keep on offline external hard disks in two locations (one at home in Germany, one in Finland)
  • Selected e-paperwork, giving me an overview of my income, costs, assets and liabilities, I want to keep on my kajPad

At my personal level of geekiness, the following step for me is to decide on a directory structure for the inboxes, the work in progress, and the files on archive. As a Mac user still happy with many usability related defaults on Mac OS X, my starting point is the names created in my home directory by the operating system: Documents, Pictures, Movies, Music, as well as the transitional directories Downloads and Desktop. Sure, I could use "Documents" or a subdirectory of it, but much like Pictures, Movies and Music, I think financial documents deserve a top-level category of their own. They're usually PDFs (and if not, they can be made into PDFs). They may be quite sizable. Like Pictures and Movies, they can usually be allocated to a particular year. So, in the name of consistency, let it be Finances (or, in my case, Finanser in Swedish - as Finder on my Mac also displays Pictures, Movies and Music as Bilder, Filmer and Musik).

On the directory level below, I'll just copy the habits I have refined over the years in using my Pictures directory. In Pictures, I have created a disciplined set of top-level subdirectories:

  • rep for Repository (which is then subdivided by year, rep/2012, rep/2011 etc.), 
  • sync for syncing pictures with mobile devices (in the desired resolution), and 
  • lib for a Library of key pictures that remain valid over the years (profile pictures, logotypes, components of web pages and other visuals). Also, Pictures has photography related temp files used by applications, such as Lightroom and Photo Booth. 

I can use the very same idea for Finances:

  • rep for a repository of year-specific receipts and financial statements
  • sync for syncing PDFs (or JPGs) with mobile devices; however, while Pictures/sync has subdirectories that iTunes syncs automatically with my my kajPad's photo albums, I will need to perform manual steps to sync PDF files in Finances/sync with iBook 
  • lib for a Library of key financial documents that remain valid over the years (contracts with banks, insurance companies, landlords)
  • scan as a transitional inbox of documents I've scanned (analogous to Photo Booth)
  • kajbook for the files used by my own financial application, written in Python and MySQL (analogous to Lightroom)

As for distinguishing between what goes to the internal and the external hard disks, and for cleaning up inbox type transitional directories, I have to adapt my Pictures directory related habits somewhat:

  • rep: in Pictures, I use the internal hard disk merely as a temporary repository when travelling; in Finances, I believe I'm better served by having the entire current year on the internal hard disk, keeping only previous years on the external hard disk
  • scan: quite like Photo Booth (which is hardly ever used on my laptop, but still) creates somewhat of an extra "inbox" of pictures which need to be deleted or moved into their proper directory in rep, also my scanner spews out files into a default place and without a descriptive file name; however, not only do the scanned documents need to be renamed and moved to their proper rep directory -- some of the scanned documents aren't financial documents at all, in which case they need to be moved elsewhere, out of the Finances tree 
  • Downloads: quite like I get pictures over email, over Skype and downloading them through a browser, I do get invoices, receipts and bank statements through the very same channels; these files need to be named properly, and brought into the rep directory; also, the cleaning up process of Downloads pertains to Pictures, Finances and all other downloaded files simultaneously, meaning that the cleaning either needs to happen for Pictures and Finances simultaneously, or the clean-up-the-Downloads-inbox step has to be performed as part of the download itself

Another difference between Pictures and Finances is that my photographic past is already sorted by year (and album). My financial past isn't, and I have to come up with a way to keep the two external HDs in sync while I'm sorting past finances. Likely, the total volume of files will remain small enough for me to be able to copy the entire Finances directory on my small 200 GB travel hard disk.

Re-reading and iterating my thoughts, I believe that the directory structure I've thought out enables a smooth work process. Now, I have motivated myself to start enforcing the Finances structure. On my internal hard disk, it should be a matter of an hour or two. Checking which files are on the external hard disk only may take a bit longer - but just like with emptying the inbox, I need something to get started. And having a proper structure for financial documents on the internal hard disk already goes a long way towards orderly finances.

Saturday, 19 March 2011

My personal web presence: Migrating to Google Sites, from blogs and emacs

This is a personal account of my effort to clean up my web presence. My web pages of personal and semi-professional nature needed weeding, and I ended up moving from sporadically maintained multi-lingual blogs and manually edited HTML pages to a set of Google Sites. Generations of cruft are being superseded by a solution based on a weighted balance of contemporary ease of use, versatility and my limited available time.

Before joining MySQL AB in 2001, I had lived through a set of personal web pages that I had set up with PHP. My mental mode was one of combining some fun PHP coding, a bit of editing of pics, and manual uploading of HTML files with scp with emacs editing on the server, that I hosted at my then-employer Polycon Ab. Once at MySQL, I continued in the same mode, but from 2005 onwards started blogging on blogs.mysql.com/kaj. With increased travel, I started an experiment of multi-lingual blogging in 2008, with blogs in over a dozen languages. I hosted them at blogs.arno.fi in a Wordpress installation on my own server, rented from the Swedish ISP Bahnhof.se. With the advent of Google Sites, I started to understand that most of what I'm doing can be accomplished in the cloud, at near-zero hosting cost, with better functionality and no technical maintenance from my part. The last part was most important. Installing, maintaining and playing with a LAMP environment may be fun, but I gave it a fairly low priority amongst my personal pastimes, as I'd rather concentrate on sports, photography and reading instead. The final straw was my server being hacked a month ago, used for sending out mails of dubious nature. I don't need that, nor do I need the quarterly payments for owning a web server I hardly use.

Time to think about what I really need. Or, given that a personal web presence isn't amongst the basic human needs, time to think about what I would like to accomplish.

The purpose of my personal web presence is to support other interests in life: to share ideas on my favourite topics, and to be social with my fellow human beings. This translates into a number of requirements, none of which are absolute, and all of which have to be weighed against the cost in time and money spent. I list time before money, as a web presence can be an endless time sink, and as the cloud gives ever better possibilities of accomplishing things without cost.

Let me start with the key resources in the state they were before the clean-up:

  1. arno.fi: A domain I had acquired in November 2005 from the Finnish governmental web authority Ficora.fi, who decides on and grants .fi domains. I had thereby obtained both the email address firstname@lastname.country and the web address http://firstname.lastname.country, in addition to the same vanity URL:s and addresses for the handful of family members and relatives who share my last name. The cost is 12 € a year.

  2. A dedicated server and DNS at Bahnhof.se: The place I was hosting arno.fi. I payed over 1000 € a year, and had to install and maintain the LAMP stack myself, in return for the freedom to affect any detail in the configuration. A complication was that the Bahnhof.se account was also the place for entering all DNS entries and CNAME records. From Ficora, I had only bought the domain, but Ficora has no DNS servers; instead, Ficora requires me to point to two DNS servers that in turn point to where arno.fi is hosted. By contrast, a company like GoDaddy would sell not just domains but also do the DNS hosting.

  3. kaj.arno.fi: a web page edited with emacs on my own server in Bahnhof.se. The index.html had various gadgets pasted in, pointing to Blogger (Google aggregation of the blogs hosted on my self-configured blogs.arno.fi), Flickr, Facebook and Delicious. Of these, the Delicious web pointers were in active use, for my most frequent links to Lufthansa, Deutsche Bahn, buses in Finland, the weather, and my favourite news sources. In the transition phase, this page is at kaj2.arno.fi.

  4. kajarno.com: A domain I had acquired in 2009, in order to at some point be able to more clearly differentiate between web presence in different languages. I thought kajarno.com could be the "international", English site.

  5. kajarno.de: A domain I had acquired in 2009 on similar grounds as kajarno.com, this time for the German site.

  6. Google Mail: Personal mail hosted on Google.


Desired improvements:

  1. Application Cloud benefits: No need for maintaining my own software. No need to kill -HUP my Apache server, to protect myself against hackers, to upgrade MySQL or Wordpress. Go hosted.

  2. Professional design: Better look-and-feel, modern gadgets, easy analytics. Use the secondary benefits of the application cloud, as an enabler of more modern technology than what I had installed.

  3. Separation into three sites: kaj.arno.fi for Swedish, kajarno.de for German, and kajarno.com for English (potentially with subsites in other languages). Part of the general cleanup.


Desired status quo:

  1. any_name@arno.fi: Possibility to easily forward emails like I have done for ages using /etc/aliases -- not only "kaj" but a number of other aliases go to my Google account. And "alexander" goes to my son's Google account, etc.

  2. Redirect 301: Possibility to easily redirect short URLs like I have done for ages using .htaccess -- for instance, "Redirect 301 /fff2010 http://picasaweb.google.com/kaj.arno.fi/FuruvikFashionFestival2010#slideshow/" used to make http://kaj.arno.fi/fff2010 point to a URL on Picasa with a name too long to remember.

  3. blogs.arno.fi/blogname: Possibility to retain the existing blog URLs, such as blogs.arno.fi/isit (for English), blogs.arno.fi/fib (for Swedish), blogs.arno.fi/efib (for German) and so on.


In principle, these goals were clear to me two years ago. Yet, I hadn't progressed for all those months. Why? Let me return to the point in time, roughly two years ago, when I defined the "desired improvements" and the "desired status quo" (not in writing like now, but )

The key was Google Sites and Google Apps. An email outage had forced me to try Google Mail, with which I was happy. As part of a series of articles (in Swedish) about Social Media, I studied Google Sites. I was impressed. For the first time, the dead easy, free solution seemed to be good enough. For years, the "easy" web solutions had been severely crippled, having a condescending attitude to usability, i.e. giving hardly any degrees of freedom. But with  Google Sites, I could influence the layout, the URLS, and the navigation reasonably well, and I found I could create and edit "static" web pages faster than writing blog entries. Wow! Time to move to the Application Cloud.

Then came the obstacles. Google Sites had limitations of the very same nature as earlier "dead easy, near free" solutions.

  1. Redirection. The most irritating of those turned out to be the missing "Redirect 301" feature. How can something so easy and so functional be missing? I looked through help pages and I asked around amongst friends and colleagues. Others had similar issues, and there were "answers", but they were special cases, such as redirecting an entire domain. The best thing I've stumbled upon is a "Redirect Gadget" for Google Sites, which amounted to creating a new web page, and installing and configuring a Gadget on it, for each entry in a .htaccess file. This is the opposite of usability, and a waste of time for replacing the near-hundred lines in various .htaccess files I had.

  2. Blog URLs. Blogger and other blog hosting sites kindly allow me to use my own domain, but no subdirectories. I can do "isit.arno.fi" but not "blogs.arno.fi/isit". And that changes the URLs to all existing blog entries, which means that all external pointers to them will die. If at least I could redirect them!

  3. Email forwarding. I hadn't ever hosted my emails on arno.fi, just forwarded them with /etc/aliases. I didn't find any equivalent solution for Google Sites or even Google Apps. I ended up asking a friend for favours; he still manages his own server and added a postfix entry to his email configuration, and ssh access for me to edit the postfix equivalent of /etc/aliases. I would happily pay decent money for this type of email forwarding from a bulk commercial entity, though.

  4. DNS. Remember that I got DNS as part of my expensive Bahnhof contract for a dedicated server, which I want to get rid of. With the arno.fi domain costing nearly nothing, and Google Sites being completely free, retaining the Banhnhof contract merely for the name server pointers seemed pointless. So who is specialised in DNS hosting? Googling revealed "DynDNS", which after an agonising web experience turned out to have unacceptable constraints both in its free version and in its cheap version. Further Googling finally lead me to http://freedns.afraid.org/, which so far has worked just fine and, yes, is free. It's been reliable and it has easy editing of all the necessary CNAME, A, MX etc. records that one needs for setting up Google Sites.


But now, I've taken nearly all my intended steps into the Application Cloud:

  1. kaj.arno.fi is on Google Sites. And it has been "cleaned" of contents which is not in Swedish.

  2. kajarno.com and kajarno.de are on Google Sites. Yet, so far, it has only embryonic content.

  3. Emails to @arno.fi are independent of the Bahnhof.se server. They are forwarded through the Postfix mail transfer agent on the server of a friend of mine.

  4. DNS to @arno.fi is independent of the Bahnhof.se server. I use freedns.afraid.org.


What still remains before I can scrap my dedicated server is not much:

  1. Backing up the remaining few files from my dedicated server, downloading files such as old apache logs.

  2. Moving blogs.arno.fi/isit, blogs.arno.fi/fib and blogs.arno.fi/efib to new blog sites; probably Blogger hosted sites mapped to isit.kajarno.com, fib.arno.fi and efib.kajarno.de

  3. Moving the contents from other blogs.arno.fi blogs to static pages on kajarno.com

  4. Entering the key Delicious links from kaj2.arno.fi to the new kaj.arno.fi


None of the above is fun or creative. It's mostly boring. At least, it should be a little bit less frustrating than moving DNSes from Bahnhof.se to FreeDNS.afraid.org (which involved long moments of figuring out whether I was making mistakes or merely waiting for chicken-and-egg type changes to propagate through the internet).

Let me end by a few thank you notes, to people and web resources who have inspired and helped me in this quest towards the cloud, with everything from general encouragement to specific knowledge on Google, MTAs, DNSes, blogs, Ficora etc.: Thank you Siegfried Hirsch, Giuseppe Maxia, Jonas Sundin, Daniel Bartholomew, Michael "Monty" Widenius, and Vesa Linja-aho, plus of course FreeDNS.afraid.org and Google Sites.

Tuesday, 12 October 2010

Joining SkySQL Ab, back in the startup business

I swear, my intention was to go for a break. A year taking pictures, sharing them over the web, writing texts, running, kayaking, just being social. Honestly, ask my family and friends!

But this was not to be, in spite of what I said when I announced my resignation just days ago. Instead, I am joining SkySQL Ab, the startup that aims to become a new centre for the MySQL universe. My role will span Marketing and Engineering, and is like the title "EVP Products" inspired by Zack Urlocker's role at MySQL AB.

I would have preferred the company to have been called KajSQL, but have come to terms with the company having an extra phonetic "s" in the beginning. Quite a while ago, having eaten my favourite fish "Börjes fisk" in my country house in Nagu, Finland, my now-former colleague Giuseppe Maxia jokingly even suggested a logotype for KajSQL, then labelled "The database for community lovers".

Why this change of minds? For five reasons:

Reason 1: The team. Behind SkySQL Ab is a set of top people: founders, executives, and experts who all were colleagues during MySQL AB times. Our Chairman is MySQL AB angel investor Ralf Wahlsten, who found MySQL AB's first Chairman, and defined the MySQL Core Values with MySQL AB founders Michael "Monty" Widenius and David Axmark. Monty and David are investors and stakeholders, but have no operational or management roles. Our CEO is Ulf Sandberg, formerly SVP Services at MySQL AB. The co-founders include Mick Carney, formerly European Field Sales Manager for MySQL at Sun Microsystems, Patrik Backman, who worked with me on the SAP partnership and continued as Director of Engineering, and Max Mether, who set up training at MySQL AB with me, and already prior to that at Polycon Ab. Early colleagues include MySQL AB time stars such as Boel Larsen in HR, as well as Dean Ellis and Alexander "Salle" Keremidarski in Support.

Reason 2: The timing. SkySQL Ab is happening now, not when I've got my pictures sorted out. I want to support the team, and keep colleagues together, who work well together. I would hate to see an exodus of talent from the MySQL ecosystem.

Reason 3: The role. I can be back, making a difference, making decisions again. It isn't as if I wouldn't have given it a try at Oracle. I promised myself to make an attempt at influencing Oracle from the inside, explaining what the MySQL Community is, how the MySQL ecosystem works, and how it can be adapted to Oracle. Sure, like Oracle, SkySQL strives for profit, but the Oracle focus on the bottom line seemed to me to overshadow everything else, perhaps not respecting Open Source and community dynamics to the degree I would consider adequate. As listening to a Vice President from Munich wasn't high on the list in Redwood Shores, I clearly prefer having real influence in a startup than a mere title but no influence at Oracle.

Reason 4: The credibility. Like I said in the previous reason, I tried to exert influence from inside Oracle but felt I failed. What I cannot live with is conveying information I don't fully believe in myself, lending my ten years of MySQL presence to give Oracle credibility I feel it may not deserve. At SkySQL, I expect to be able to stand behind my statements for a tad longer than what happened with the ones I made in April. Referring to my resignation blog entry, The Register hints at my statements during the April 2010 keynote at the O'Reilly MySQL Conference:
The loss of another MySQL veteran in Arnö on the back of other Sun exits will counter attempts by Oracle to try and reassure users that their open-source database's future is safe on the corporate mothership.

Last year [I do think this means April 2010; my own note] Arnö had tried to convince MySQLers that nothing would change for the worse under Oracle, that their database was safe, and that concerns over the database were unfounded.

Among the facts he highlighted was a "huge talent pool" of MySQL experts inside Oracle — a talent pool that's now one member smaller. He also said that Oracle planned to be proactive in its dealings with the community and would emphasize maturing the database.

Reason 5: The opportunity. MySQL as a database hasn't seen it's apex yet, and SkySQL can contribute to the well-being of the MySQL ecosystem. The initial SkySQL Ab team collectively has over one hundred years of experience working for MySQL AB. I think this team has an excellent chance of grabbing the opportunity to build a profitable, sustainable commercial business providing MySQL related offerings, while preserving open source values.

In summary, to do the right thing, I felt I didn't have much of a choice! I continue to feel responsibility for the MySQL ecosystem and I don't want to let anyone important down.

So yes, I'll still pursue my private goals. But on top of that, I plan to have lots of fun working at SkySQL Ab, and do my best to make the company a success comparable to MySQL AB!

Sunday, 3 October 2010

Thank you, everyone behind MySQL AB!

For more than nine years, I worked for MySQL AB and its successors, Sun Microsystems and Oracle. I handed in my resignation late June, two days before Sun's German legal entity ceased to exist. Germany isn't a country where you quit HP one day and join Oracle the next, so I had a long summer with plenty of so-called Garden Leave. Last Thursday was my last day, and I'm now outside MySQL AB, outside Sun Microsystems, outside Oracle.

Like all the many former colleagues who resigned before me, I did so with mixed feelings. Leaving the colleagues, finding freedom, I think you follow. The topmost feeling I have, the one I want to highlight right now, is gratitude. I'm very grateful for what MySQL AB has meant for my career, my personal development, my life experiences, my social life. Let me mention a small subset of the people that made my life at MySQL AB, and later Sun Microsystems, a truly memorable and enjoyable one.

Back in early 2000, when MySQL AB founders Michael "Monty" Widenius and David Axmark were just beginning to see the traction for MySQL, I got a proposal from Ralf Wahlsten, an old friend of Monty's and mine: Since you've done training and consulting, and Monty hasn't, why don't you create a training program for MySQL? I'm sure Monty will help you and promote it. I followed the advice. So, in my company Polycon Ab where I was an entrepreneur for fourteen years, I started working with MySQL a good ten years ago. Ralf connected the dots! And extracted the MySQL Core Values from Monty and David, and found our first Chairman John Wattin, and became an Angel Investor in MySQL AB.

In February 2001, after a good half year of working with MySQL, it became obvious that I was experiencing something which was going to be big. My last doubts were removed when, at the outset of a boys' trip to Rio de Janeiro, I understood Mårten Mickos (whom I had known since the early 1980s and respected for his leadership and judgement) had signed on as CEO. I asked Monty (whom I had known since the late 1970s) whether he was interested in me selling out the MySQL training operations of Polycon and formally joining MySQL AB. He was. Monty was kind enough to have me, and welcomed me with open arms. And created MySQL the product and the MySQL community.

In May 2001, I formally joined MySQL AB, together with my Polycon colleagues Bertrand Matthelié, Max Mether, and Sylvia Arte, soon to be joined by Olivier Beutels. There was a good dozen of employees before us. My initial title was VP Training, and Mårten invited me to join the management team, together with Monty, David and others. In the years to come, I was to get a number of other roles, VP Consulting, VP Services, VP Engineering and CIO, before becoming VP Community Relations in 2005. Mårten gave me all these opportunities, trusted me, and supported me as his reportee until he left Sun. And made MySQL AB into a company, grew it from a dozen people to 500 people, created a success story, and was a role model for how to do business with Open Source.

During the many years until the Sun acquisition, I had the privilege to work with some members of the MySQL AB Board of Directors. I learned a lot from you. John Wattin, our first Chairman, successfully guided us through financing rounds and growth pain. I fondly remember John referring to me as a "fireman", given that I swapped roles so many times through the ride. I also had the pleasure to work with Fredrik Oweson of Scope Capital, Kevin Harvey of Benchmark (our second and final Chairman), Danny Rimer of Index Ventures, Bernard Liautaud (then of Business Objects), and Tim O'Reilly. Later on, at Sun, the board connections were replaced by contacts with Sun executives, where I most of all appreciated working with David DouglasRich Green and Alain Andreoli, and where all of us MySQLers got some unforgettable help from Rich Lang and Julie Ross.

MySQL AB opened the doors for learning to know many brilliant minds. Co-chairing a GPLv3 Committee with HP senior counsel Scott K. Peterson, I experienced Software Freedom Law Center's Eben Moglen first-hand. Through a combination of intelligence and diplomacy, he tamed a conference-call-ful of the seniormost US corporate counsels, who all bought into Eben's plans for the next generation of free software licenses. In 2005, I supported Florian Müller's successful efforts to (at least for a while) save the EU from the software patents. I think Eben, Florian and I all agree swpats are obsolete legal tools, used to protect incumbent players against having to innovate. Sadly, Eben and Florian have since had some disagreements. Speaking of brilliant minds, the SAP negotiations in 2002 and 2003, and MySQL AB's subsequent relationship with SAP AG, introduced us to people like Shai Agassi and Rudi Munz. Other memorable events was introducing MySQL Conference guest speakers, such as Guy Kawasaki and Mark Shuttleworth. I left the stage for Mark and his Ubuntu presentation on a MySQL conference by cheering him with "Поехали!" (Poyechale, Off we go!), a retired cosmonaut as he is.

Perhaps most rewarding was learning to know and appreciate the colleagues from nearly 30 countries. Yuri Gagarin's exclamation when he left into space was something I learned to know from Alexander Barkov and other Russian and Ukrainian colleagues, whom I've had the pleasure to work with since 2002. I learned so much about sales and customer relations from Larry Stefonic, Kerry Ancheta, Joe Pen, Mark Rubin, Mark Burton, Mick Carney, Magnus Stenberg, Richard Mason, Philip Antoniades and Ivan Zoratti. I learned about Services and Support from Ulf Sandberg, Dean Ellis, Tom Basil, Alexander "Salle" Keremidarski, and Sinisa Milivojevic. I had the privilege to work with top engineers like Serg Golubchik, Kostja Osipov, Jan Kneschke, Igor Babaev, Georg Richter, Georgi "Joro" Kodinov, Heikki Tuuri, Kent Boortz and Brian Aker. I enjoyed working with my Community Team members, such as Lenz Grimmer, Jay Pipes, Duleepa Dups Wijayawardhana, and Colin Charles, and with management team colleagues, such as Zack Urlocker, Dennis Wolf, Clint Smith, Jeff Wiss, Tomas Ulin, Hans von Bell, Maurizio Gianola, Jeffrey Pugh, and Boel Larsen. And my sanity was saved through being excellently supported by some of my longest-time reportees, Patrik Backman and Giuseppe Maxia. And now having gone out on a limb by mentioning names and thereby most certainly having omitted at least a dozen people absolutely worth mentioning, I would like to thankfully highlight Edwin Desouza for labeling me as diplomatic, although this blog post is bound to fail on that account.

What next? As I resigned from having worked for and with MySQL for nearly ten years, I decided to give myself some time to spend on matters not directly related to IT. First, I enjoy expressing myself in writing, mostly in Swedish, German, and English. Second, I appreciate the beautiful things in life, and for me, aesthetics go hand in hand with photography. Third, I think there are opportunities to combine these into some experiments in the social web. Let's see how long the break will take, before I return to more conventional duties, such as developing a startup, evangelising technology or devoting myself to Venture Capital.

What I want to do today, though, is simply to express my gratitude. Thank you, everyone behind MySQL AB!

Friday, 23 April 2010

An English language timeline of my volcanic adventure

From Sunday 11.4.2010 to Thursday 22.4.2010, I used my Twitter account @kajarno to send

  • 26 Swedish tweets

  • 41 German tweets

  • 69 English tweets


(not counting direct replies) of which a clear majority (but not all) were of a volcanic nature.

Most Swedish readers can follow German and English too, and most German readers also follow English. Hence the large number of English tweets and the larger number of German tweets.

Here's a subset of the English tweets. The dates and times are local (California 9 h different from Munich; Chicago 7 h different).

Thursday 15.4.2010

  • 13:12 Kaj Arnö wonders whether LH459 will leave at all MUC-SFO tonight 21:00, and if so, which route #iceland #volcao #ash

  • 20:18 Damn, flight was cancelled. #lh #iceland #volcano SFO-MUC LH459

  • 21:06 Even though it isn't a #Lufthansa owned #volcano they provided me with a free hotel. Danke! :) #Island


Friday 16.4.2010

  • 07:56 John Cleese took the taxi from Oslo to Brussels. I doubt taxi is an option from San Francisco to Munich. Stranded in Oslo, Cleese takes taxi to Brussels

  • 14:56 Whoever put the soap in the Eyjafjallajökull volcano, don't add it to Katla #ashtag - http://is.gd/bw7fz great funny 2min video


Saturday 17.4.2010

  • 7:34 If many MySQLers meet, disruption happens. 2008: Sun buys MySQL. 2009: ORCL buys Sun. 2010: Eyjafjall erupts #ashtag #mysqlconf (via JWiss)

  • 10:51 Kaj Arnö wonders if a competent lawyer could send a cease and desist letter to @Eyjafjall #ashtag

  • 11:15 Under Icelandic influence, Europe has turned to the most xenophobic continent of them all #ashtag @Eyjafjall

  • 15:42 No LH459 SFO-MUC tonight either. Later meeting with fellow stranded MySQLers for dinner in Redwood Shores. #ashtag #mysqlconf


Sunday 18.4.2010

  • 10:48 Sorry Twitter followers and FB friends for erupting so many msgs but being stuck on the wrong continent is no fun #ashtag #getmehome

  • 11:38 Kaj Arnö just extended his involuntary stay at Sofitel until Wed, like his fellow LH passengers (good hotel, wrong continent) #ashtag #getmehome #SFO

  • 11:41 Kaj Arnö sees that today's LH459 is cancelled, too -- tomorrow (for which I was rescheduled) is still on schedule, though #getmehome #ashtag #SFO

  • 13:25 Lava all, nerve all? RT @Eyjafjalla: To my 1125 followers: I lava you all! #ashtag

  • 14:40 Progress rpt: Now @Lufthansa_DE follows me and @Eyjafjalla laughs at my jokes. If only I could connect the two. #ashtag #getmehome


Monday 19.4.2010

  • 00:42 Sun's travel agency found me a flight Monday morning 9:40-15:45 SFO-ORD (=Chicago) +16:40-07:40+1 ORD-MAD. If it flies. I'll try Madrid!

  • 00:44 Issues: Little time in Chicago. And perhaps ORD-MAD will be cancelled. But worth a try!

  • 06:25 *If* I get home to Europe over Madrid, here are my trains: 16-18:38 M-Barcelona 19:38-5:45 B-Geneve 6:14-8:56 G-Zürich 9:16-13:28 Z-München

  • 17:33 Boarded! Madrid, here we come! And there I will stay, my wife predicted. Trains and buses told to be full until Monday. #ashtag #getmehome

  • 17:57 Iberia Airbus A340/600 is moving, due East! Bueno, señorita stewardess, I'll turn off the phone. To be continued. #ashtag


Tuesday 20.4.201

  • 08:18 Kaj Arnö greets Europe and thinks about his fellow European #mysqlconf attendees still captured in the US #ashtag

  • 12:32 There's an ocean of difference between waiting in Madrid and waiting in the US. #relief #ashtag

  • 12:53 Done: SFO-ORD 2956 km. Done: ORD-MAD 6763 km. Todo: MAD-MUC 1481 km. That's a mere 13% or < 1/7 #ashtag http://is.gd/bAuXU

  • 19:24 Kaj Arnö is on his 8th day of handluggage only business travel and still finds a spare pair of unused socks #happy #frugal #greedy #ashtag


Wednesday 21.4.2010

  • 08:27 Kaj Arnö will make an attempt at the MAD airport to go to MUC #ashtag 8:55-11:20 IB3534 http://bit.ly/b9Wv4a

  • 09:22 Sign of air travel normalisation: frustration over MAD "security" process temporarily won over joy of returning home #ashtag

  • 09:37 A jetful of passengers are still queuing for boarding. Hope or desperation? 8:55 IB3534 Madrid-München #ashtag

  • 11:43 Landed in Munich! Although I am now a released Eyjafjalla hostage, most #mysqlconf Europeans still aren't. #hashtag


Thursday 22.4.2010

  • 08:46 Fun science! RT @stewartsmith: Best giggle for today: #boobquake - also, a world of awesome.

Thursday, 22 April 2010

Twelve conclusions by a released volcano hostage

Five days of uncertainty, trapped far away from home by an Icelandic volcano, provide plenty of time to think. Here is my attempt at drawing conclusions the day after returning home: partly on a personal level, partly on a general human level, partly on a societal level.

To recap what has happened: Ten minutes before the intended boarding time of LH459 from San Francisco home to Munich, I got an SMS about the flight being cancelled. This was Thursday 15.4.2010 at 20:20. Thereafter I lived in uncertainty in airport hotels, most of the time 40 km south of San Francisco, to finally arrive home five days later than plan on Wednesday 21.4.2010 at 13:00..

Personal conclusion


Conclusion 1: Uncertainty and losing freedom are hard to cope with. In retrospect, I never was in any real trouble. No physical suffering, no bodily harm, hardly any material loss. Still, the experience was amongst the strongest ones in my life. If that's the case, what do I really have to complain about? Losing freedom for an indefinite period of time. Unmet expectations.

Conclusion 2: Europe is my home continent. Emotionally, it was terrible to be stuck on the wrong continent, 9474 km away from home (SFO to MUC). Last time Eyjafjalla erupted (1821-23), it took one and a half years, and knowing this instilled a feeling of potentially being a hostage for countless weeks, away from family and friends. How could anyone know for how long intercontinental air traffic would remain closed? Sure, the relief when returning home was enormous, but the real euphoria happened when I saw Europe again in Madrid. I was rescued! Sure, I was still 1481 km from home, but that distance I can make by train, bus, car or even bike. Fellow sufferers amongst MySQLers have shared my feeling that although it's fantastic to be home, the real pressure was off already upon "merely" returning to Europe.

Conclusion 3: Humour is a strong weapon. From the start, we who got stuck kept a good mood through joking. Some of the humour was black enough not to merit being captured in print. "Send cash not ash" was the first successful joke on the web, but the self-made jokes provided more solace. Most jokes were somehow language dependent and I shared them with the two biggest language groups amongst the stranded colleagues and friends: the Swedish speakers and the German speakers. Language jokes seldom translate but the illustration on the right is 1:1 transferrable between Swedish and German. The word for "surprised" in both languages is literally translatable to "overrashed", so leaving out one "r" means we were "overashed". One particularly fun web link is to the three-minute Icelandic terrorist video, where Icelanders keen on keeping their Icesave money threaten to put soap into Eyjafjalla's big sister Katla, so we won't have a summer for years (a Katla eruption in 1783 caused hunger and poverty across Europe, believed by some to have been a key trigger for the French revolution six years later).

Conclusion 4: Language creates closeness and strengthens a sense of kinship. This one may be impossible for monolinguals to understand, though. Basically, sharing raw information about my situation would have been technically doable in English, making Twitter and Facebook texts understandable for (nearly) all. Yet, the whole point with the social media communication for me was to shorten the emotional distance and increase the mental well-being of those at home, of fellow sufferers and of myself. For this to happen, there needs to be a en emotional closeness to the language itself. I need to communicate in the same language as I would use in person. Hence, I used mostly Swedish and quite a bit of German, with English only as a third choice. It wouldn't have felt right to use English only.

General conclusions


Conclusion 5: Humans are social creatures, and social media give true consolation in times of hardship. Facebook and Twitter became lifelines for us volcano hostages. I got more Facebook events than ever. On my way from Chicago to Madrid I had 51 events, and from Madrid to Munich 48 events. It felt great for me to know that someone cared. And to judge from the content of the comments, people really did care: "I hardly get time to work - having to follow the exciting serial story about 'will-Kaj-get-home-or-not-and-when-and-where-and-how". Reply: "That's how I feel, too! Is this Reality Entertainment? (Sorry Kaj - perhaps not from your perspective...)" whereas a third fellow hostage pointed out that "this is a bad reality show, as no matter how hard I try, nobody votes me home...".

Conclusion 6: Individual characteristics, good and bad, are underlined by the exceptional situation. Stoic peace could be seen amongst those who stay calm under normal circumstances. Systematic work and concentration amongst those not easily distracted. That I found consolation in humour, in writing and communicating with others was no coincidence. Sadly, I regressed into an old bad habit of being too keen on following superficial web news, instead of just carrying on as if nothing had erupted.

Conclusion 7: The possibility and ability to make the best of a bad situation varies according to attitude, character and luck. Myself, I had set my mind to returning home quickly. I have visited to California quite often and (picky as I am) have problems in the US to find food I enjoy, particularly for breakfast. I dislike driving cars; lazy as I am, I prefer public transport. I was in no mood to spend time somehow getting the 40 km to the city centre of San Francisco, which I have visited many times. For others stuck, the situation was quite the opposite, and they were happy that San Francisco was the place fate had chosen, if they had to get stuck. For me, the good part was that I had my running gear, books and notes, as well as car-borne colleagues with whom I could have a run and share dinner.

Conclusion 8: Wishful thinking is more common than panic reactions. "How may hours will you be delayed?" was a typical question in the beginning. As hours grew to days due to closed airspace, the questions became "Have you got closer to home?" or "Do you know when you'll be home?", although nobody had a clue when air traffic would be allowed again.

Societal conclusions


Conclusion 9: Flying, especially for tourists, will become less frequent. Both rational and irrational reasons speak for travelling less. Personally, I will evaluate the need for travel outside Europe more stringently than before. And I believe many will even avoid intra-European travel.

Conclusion 10: The value of insurance will be questioned. The idea with insurance is to minimise exposure to risk if something unexpected happens. Financial Times Deutschland points out that insurers exclude many unexpected events from the list of what you're protected against. Consequently, customers may see less value in being insured in the first place.

Conclusion 11: Social media grow ever more in importance. Although Lufthansa were bad at communicating during the Eyjafjalla eruption, it turns out KLM already has done what I have nagged for Lufthansa to start doing. KLM actively communicates in Twitter, has a good Facebook page (I became the 38704th one to like it), and an impressive YouTube video with their CEO. And the sense of belonging created amongst private users during times of crisis won't fade off as quickly as the ash cloud.

Conclusion 12: Risk assessments may become more rational. BBC's question whether it's more dangerous to drive a car than flying through ash clouds stresses the unethical aspect in the populist politician statement of "security above all", since we irrational human beings readily accept highly dangerous activities such as high speed limits for cars, smoking, alcohol and unhealthy food, all of which cost considerably more lives than air traffic. And even airport "security" control (de facto a form of state sanctioned systematic mobbing) will over time get easier to question, as we travelers use our newly gained vulcanic experiences and contemplations to become less gullible.