Stuff Michael Meeks is doing

• Babies teacher's on strike, so took a deferred spring bank holiday from 30th May. Up late, into Cambridge to the Funky Fun House - babes all enjoyed it, and managed to entertain / rescue each other leaving us to talk, lovely.
• Into Cambridge on the bus to Downing, picnic on the paddock, walked to the Fitzwilliam. Sadly E. has learned to say M-use-eum, instead of the much sweeter 'I-e-am'. Admired some artwork, and some mummies. Bus to the park & ride, chat with jrb, home.
• Fed babes, pottered about with machining brass blocks into heater blocks, ground down the wire-resistor fire-cement casings a little.
• Put babes to bed, taught H. some hacking - concept of variable, two-dimensional axes etc, and got her pygame stars looking prettier.
• Chewed mail quickly, tried to draw a pretty picture of our release schedule from different user perspectives.

• Woken by N. playing the violin by my bedside. Prodded mail, off to TEAM via picking up Mike. Talks on Ecclesiology, visited Maplin to buy up their entire stock of heating resistors for repraps - both of them, ditto for 100k thermistors, again both - hmm.
• More training, back home. Steering Committee conf. call, marvelled at openoffice.org re-directing to an oracle.com apology for technical problems web-site.
• Mary Rogers kindly baby sat, so J. and I could go out in the evening, friendly pint together, then visited a newly refurbished Indian restaurant at the other end of town nice food, but some chicken raw in the middle - not ideal.

• Mail poking, chat with Kendy & Thorsten; lunch. Booked Desktop Summit flight and hotel, there from 5th to 8th Aug. Worked away at some paperwork.

• Somewhat amazed to finally notice this insight from Rob Weir (cf. reply to Maxei). In one short block of text he manages to malign The Document Foundation's governance, vendor neutrality, insinuates SUSE collusion with Microsoft (to not compete in the office space) and suggests LibreOffice will be mainly a Linux product. So lets deal with some of these:

  TDF's governance is vendor neutral

  This means that no one vendor can control TDF - we aspire to having a project people love to belong to. That should mean that people are loyal primarily to the project, and less so their company when taking decisions. Just in case that breaks down, and people act in a partisan way, we have a Conflicts of Interest section in the bylaws that limits the number of members from any one company to 30%. That applies to the Board of Directors, Engineering Steering Commitee, and Membership Committee.

  SUSE competes with Microsoft

  Clearly there is a well publicised collaboration with Microsoft around interoperability, but to suggest that there is some kind of hideous collusion to not compete in certain areas is execrable and emphatically wrong. Hard to prove the absence of that beyond affirming its absence, and/or if you could all head to the Novell LibreOffice web-page and sign up for our services and support it'd be much appreciated.

  LibreOffice downloads are mainly for Windows

  I did a rough analysis of recent downloads by platform early this month and more than three-quarters of our downloads are for Windows. Of course, many linux distributions ship with LibreOffice by default, a download may be multiple installs, etc. etc. but the picture seems pretty clear.
• Quick chat with Ralf, more paperwork. Put babes to bed, out for a run, dinner with Lydia. Back to work later.

• Mail triage; filled out Clarity, then read the mail saying this (unpleasant top-down time & motion study) is finally gone - lovely. More mail, patches etc. Lunch. Call with Kendy.
• Started booking travel: something I hate getting to - got flights for meetings in Nurnberg 13-15th of July, finally got Desktop Summit dates sorted out, only to discover that our family holiday in France is plonk right on either side of it; irksome; bother.
• Dinner, put babes to bed; tried to find flights from Angoulême to Berlin - amusingly the nearby airport is famous for having infrastructure and being "the airport that has no airline operators and no flights"; comic. The problem apparently dissolves in Bordeaux.

• Up early, kept babes quiet while J. slept. Off to NCC, communion, spoke; coffee; home. Lunch, plugged away at some PEEK rod turning it into a holding piece for an extruder nozzle.
• Played in the garden with the babes in the sun, a hot and lazy day off; barbeque for dinner, babes to bed - watched a movie, sleep.
• Restless night, de-camped to bed in the office to avoid waking J. only to be woken myself by a worried, apparently abandoned wife - asking There you are ! are you awake?, fun.

• Substantial lie in; skyped Kate & James happily. Brunch with the family. Machined some PTFE - quite fun, though the lathe is still somewhat touch and go.
• Babes watched Chitty Chitty Bang-Bang while I did some sermon prep for tomorrow, Daniel 3 - first half. Returned to play with the babes before dinner, bed, more work until late in the evening on the topic.

• Up early; prodded mail, dug away at reviewing Matus' nice patches to help enable lots of goodness in LibreOffice. Merged most of them, lunch. PEEK and PTFE rod arrived for making reprap nozzles from - lovely.
• Call with Charles & Phil, chat with Federico, managed to miss the travel booking - annoying. Forgot to link the ESC minutes from yesterday's call.

• Up, mail; conference call with Lanedo, chat with Thorsten. More mail, lunch. Team meeting, read the TDF planet and noticed Eilidh's nice work on reverse engineering Visio import being show-cased, more nice clean copy-left goodness, and (thus far) some missing strings, great progress. Pleasing to see Fridrich's Evolution on Windows built in OBS work too.
• More mail thrash, bit of hacking, up later digging out a blocker bug in 3.4.1rc2.

• Up; J. also ill - took babes to school, and car for its MOT. Chewed mail quickly, hacked happily at gtk3 locating the cloned and subset BitmapDevice modes that were defeating my damage notification logic; finally something that does client-side rendering reasonably efficiently and properly - neat. Call with Jos.
• Home, lunch. Finished Linux Format column and sent it out, more hacking here & there. Took elder babes to swimming, and put them to bed. Bed early.

• Up early; tired. Poked mail, lots of encouragement, updated the website a little, worked on a filter issue with Marc-Andre. Lunch.
• Chat with Thorsten, then Charles. More hacking, GNOME Advisory Board call with some good news: Karen Sandler is bringing her experience to bear on the free software desktop as Executive Director of the GNOME Foundation.
• Lydia over for dinner, fell asleep early instead of working; feeling increasingly ill.

• Up early, bid Karl & Katherine goodbye; dug through mail; pleased to get a nice statement of support from RedFlag2000. More mail. Lunch.
• Calls with Kendy & Frank. Clarity, status report. Back to various gtk3 pieces, what fun - it seems the idea that anyone can render anything at any time anywhere is well entrenched in VCL, which causes some grief with a pure on-demand rendering to a provided cairo context.
• Worked late.

• Up early, off to NCC, visiting speaker preaching on God's Fatherly love for us all as favoured, inheriting sons.
• Home, pasta lunch; poked at repsnapper a bit, merging in libreprap for a much simpler build experience, fixed a few issues. Babes out to a party.
• Karl & Katherine Geller arrived from Mendelheim, for some English language immersion, good to meet them again after some decade plus; chatted, fine dinner and evening together.

• Mini lie-in, the girls seem to be getting better at sleeping nowadays. 34th birthday today; some lovely hand-made cards from the family, and a present or two.
• Set too trying to deduce the cause of the cooker's failure, unfortunately the installer didn't leave enough cable to remove it from its position. Disassembled parts of it by leaning and balancing it (a Belling 644bl), spent a long time hunting ineffectually for cracked cables causing earth leakage. Eventually isolated the problem to the top oven.
• Bruce & Anne over for lunch, with some fine gifts - a vernier caliper giving a better than 1/10th mm accuracy, and some a pair of micrometers. Out into town for a fine meal with the family at Prezzo - much fun.
• Back, Bruce helped get the laptop lathe setup nicely, with the help of some spare bolts machined a beautiful brass reprap nozzle; now I just need some PTFE and passive components.
• Raspberry & cream sponge cake, with candles; and bid 'bye to Bruce & Anne.
• Back to the cooker while J. put babes to bed; eventually discovered the top oven heater element concealed under the right hand side panel, removed it - no oven, but no RCCD trips either. Movie in the evening.

• Up early, prodded mail, read news - thrilled with Tor's great cross compiling blog post, a key piece to help LibreOffice move faster.
• Interested to read the Register's analysis of IBM's continued growth and success; clearly I need to read up on the old-world of computer rental and its market impact to understand the cloud better.
• More mail; lunch, sent out TSC minutes.
• Tim & Susan from SIL dropped by, really good to meet Susan in person, and to catch up with Tim, particularly given all the good work SIL have done with Graphite2 integration into LibreOffice.
• Poked repraps, pictures of home-made blast furnaces, oven packed in so out to buy fish and chips for dinner, admired some beautiful texts, bid them 'bye. Put babes to bed, bed early.

• Poked mail; Interested to read Brian's latest TDF entry. Unfortunately the fairly positive, accurate, firm but friendly approach of TDF is undermined by a few well-intentioned but un-restrained guys. One of these is Keith Curtis. Sometimes he makes some good points, and it can be good for the debate for them to be heard and addressed, indeed that is to be encouraged. Yet as a fully self-appointed 'un-official spokesman' he in no way speaks for TDF, or indeed for anyone but himself. Anyhow, I have been personally embarassed by his un-asked-for intervention in the recent past, heckling a (hard working) papers committee (apparently) on my behalf; disappointing. Having said that, he is a nice guy; clearly annoyed by the ASF action, that much is obvious - plenty of people are, but let no-one think that his articulation of that reflects TDF's views generally.
• More mail digging; back to more gtk3 work, I suppose one nice side-effect of the cairo-context being passed to the Draw method, is making it suitably hard to app-driven rendering - ie. the app can't spontaneously decide to re-render a certain piece of the screen and just draw there: it has to queue a re-draw of it (which is cleaner of course) and respond to that. Hopefully that fits the VCL model, lets see.
• Off to the dentist to check the teeth - apparently they are all still there (encouraging). Into town to post a laptop back to South Korea; train to Cambridge. Met up with the Gnome Contacts hack-fest guys - good to catch up with the assembled heroic hackers. Fine South Korean dinner, back late.

• Up early; more mail pokeage, chat with Florian; breakfast. Dropped babes off for their sports day, picked up Mike Geach, and into Cambridge for TEAM.
• Simon Scott for a couple of interesting hours on the Holy Spirit. Lunch - wandered over with Ben to Mackay's to buy some bulk brass, and brass tube to go with it: time to get the lathe to work turning brass into beautiful nozzles. Unfortunately delayed on return.
• Picked up babes, and stayed with them for H. and N.'s swimming lesson, home - poked mail quickly, fixed repsnapper snafu.
• Pleased to see The Document Foundation Advisory Board get announced at last, after much hard work from all concerned. It is fantastic to have strong corporate and non-profit support for our mission.

• Up early, prodded lots of encouraging mail; pushed some minor tweaks, pushed a screenshot of my prototype style preview patch (inspired by Ricardo Cruz's work), to make style selection easier - it'd be wonderful to have someone to take that further while I'm distracted elsewhere.
• Crunched some statistics, got past mail; prodded at gtk3 bits. Built on Caolan's good start to the work, and got something rendering pixels without any theming, but reasonably robust. Quite a bit more work required to target the backend at cairo, and rid ourselves of the X backend I think, particularly wrt. the ROP_XOR nastiness, either that or abandon X rendering of most things in favour of basegfx and just chuck pixels around. Pushed the prototype to master: --enable-gtk3 --with-system-cairo fun.
• Dinner, put babes to bed. Out to see X-Men First Class in Bury with some of the lads from Church: Brian, Simon & Chris - fun in some ways. Most curious to see the eugenic motif ultimately adopted by one of its initial, show-case victims - que ?. Similarly extraordinary given the film's message, and the primarily white racial representation to have the token coloured person killed early on with some failing to evolve idea: weirdly disturbing - can you make this stuff up ?
• Call with Jonathan, home late, worked, bed.

• Up early; packed babes off to school. Dug back through mail, merged a patch or two. Clarity. Poked at UNO component activation with Matus. Lunch.
• Chat with Caolan, sync with Kendy. Read Bruce's latest article on the (sad) decline of OpenOffice.org.
• Poked an oddness in a unit test; amazingly it seems some (internal) registry XML parsing throws an exception and bails out if we e'X'tend the thing with an new attribute. Seems like we have a new hand-coded XML pull parser too; odd. [Update] - it was pointed out that there is a nice wiki page with information on the thinking around this here.
• Worked on some analysis, poked at some nice iogrind work from Mihai. Dinner, put babes to bed. Poked at libreprap briefly - improving the re-send handling, still a somewhat hideous protocol sadly.

• Mini lie-in; off to NCC, helped out in creche with Ruth. Caught up with Wesley again; out to Elize's birthday party at the local scout camp-site. Picnic lunch (in the rain), and Marsh-mallow toasting action.
• Back home for J's birthday cake icing action, and to get the girls trained in using the mini-lathe. Then and out to the severals for a Pentecost party; sadly raining still. Chatted variously to all & sundry. Back, put babes to bed.
• Wrote notes on Gordon's second Daniel 3 sermon; and did more libreprap improvement.

• Up early, house tidying, out shopping for baguettes, back to get Barbeque going - it burned rather well, cooking both the sausages and the hair on my right arm to a fine crisp. Met up with the parents of E's friends (we accidentally over-ordered, with ~20 kids). Eventually got everyone fed, games played, things cleaned up etc.
• Left J. hard at it; set off to the Newmarket Carnival with the babes. Amused by the neat "do not drill this panel" legend inside the fire engine (presumably to protect the enclosed water-tank), makes you wonder how often it was pierced previously. H. went in a mini-Zorb (rather fun), N. and M. in some back-of-a-trailer "fun house" thing, and E. plonked for an ice-cream. Lots of wandering, and choosing, climbing etc. Much free amusement from the rather steep bank by the exit.
• Babes to bed, watched Hustle, spent a good while fixing libreprap's serial comms code; adding a few tricks to reset confused firmware when re-starting the client software; nearly back to full printing glory. Switched repsnapper master over to the 'meeks' branch of libreprap.

• Mail prodding, voted in GNOME's election; poked TDF's Membership Committee to encourage re-use of the same voting infrastructure. More mail. Backed up prototype Samsung netbook needing posting back; Lunch. Calls.
• Finally got around to re-working my style preview patch, got the basics pushed to master. Pleased to read the FSF's supportive statement on LibreOffice, hopefully that puts to bed the idea of the FSF's blessing on sub-optimal behaviour.

• Up early, read mail. Read Richard's nice article on strip mining, which has a great comment that needs to be a FAQ:

• Could Oracle have donated code to TDF ?

  It is often asserted that they could not - since we do not yet have a formal entity to assign to. In fact, with ASF's software grants, the code is licensed under a very broad license - but it is indeed to an entity. We are privileged to have two affiliated entities: the German Association, and Software in the Public Interest (SPI). However, if you read our recommendation to Oracle - it says this:

  The existing OpenOffice.org code-base should be made available under our preferred (dual) license: MPL / LGPLv3+, with Oracle retaining the copyrights. We believe that the MPL (over say an Apache license) as a copy-left license, is crucial to community growth and acceptance, and has proved itself with Mozilla.

  ie. the idea that we require an entity in place to use a copy-left license effectively is nonsensical. Furthermore, the idea that we would treat a single large player as a special case requiring them to assign code, when our other contributors continue to own their part of the franchise is odd.

• Whose move is this ?

  Well, clearly it is Oracle's choice to whom it licenses its software, as well as the software of those who assigned it their rights in good faith over many years. Having said that, this announcement - with its three-way co-ordinated IBM blogging, and high-energy advocacy from Rob appears (from the outside) to be primarily an IBM inspired, resourced and driven move.

• Weak copy-left licenses at Apache

  Reading around the topic, it appears that one approach to solving the ASF loathing of copy-left code, at least for licenses that allow binary distribution without corresponding source is to check in binaries. Perhaps this leads us back to the spectre of the olden-days, when OO.o CVS contained vast .zip files of pre-compiled mozilla binaries for handfuls of platforms and architectures; clearly not ideal but perhaps possible.

• Dissecting some advocacy

  On Sean's original Terrible Move blog there is a comment by a 'Doug', who is doubtless IBM's Doug Heintzman, a man with considerable authority, whose words should be taken seriously. So, I would like to address it, in its entirety, in sections:

  I think that you may be missing some key points. Apache represents a mature, established, and stable home for OpenOffice with a respected governance model.

  So far so sensible.

  It costs a lot of money to create and support a new foundation and maintain a high volume download site. These are not insignificant considerations.

  Indeed these are useful considerations - so lets look at these questions:

  How much does it cost to host TDF infrastructure ?

  Does this need to break the bank ? As previously discussed, TDF's income from donations and sponsors is rather less than an order of magnitude smaller than ASF's entire budget. So by inspection this seems an unlikely contention, bearing in mind that most of our download bandwidth cost (for our huge binaries) is (kindly) carried by our extensive mirror network.

  We can look at other organisations serving web-pages, say the Wikimedia foundation to get an idea. It seems that they serve 10bn+ page views each month, at a cost of less than $10m per year (which seems to include some real staffing cost). This gives a very upper bound guesstimate of 1k page views per month per dollar. If we have 100m users that visit us once per month, that gives an outside, upper bound of $100k pa. That looks reasonably affordable, with suitable sponsorship. A wikipedia page hit I assume is some very funky database interaction, with rich media and image contents coming from hugely concurrent read/write access of tens of thousands of disparate pages. The TDF website has some quite simple static-izeable pages with typically a very simple flow through to the download page so my guess would be at some fraction of this cost. Perhaps I made a numerical blunder ? Our current spend is of the order of rather under $10k per annum.

  How much does it cost to setup a foundation ?

  Another good question, thus far, not very much. The number of excellent Free Software lawyers, willing to donate their time on a pro-bono basis is most encouraging. Our preferred jurisdiction takes some time to incorporate in, otherwise this is not (thus far) a real cost.

  Lets just code !

  Nobody wants to spend time money and energy on legal, adminisetrative, and political overhead when they could be coding.

  Personally, I love that attitude, though there is a rich seam of irony in hearing it from the political, non-coding department. Of course there is the issue of show me the code ! - so far nothing; only blogs, E-mails and hopeful looking statements. (ie. administrative and political overhead). Perhaps when we see code, more informed decision making will be possible. Having said that, being encouraged to be just a coding machine - while devolving the tricky issues of licensing and governance to other higher minds, can lead to real problems. Licensing matters.

  Won't somebody please think of the proprietary software ?

  Furthermore, There are likely interests involved who have an interest in doing real innovation in the document space involving such things as rich semantics and analytics. Think of what Watson could have done if it had access to "smarter" documents.

  Some sadly unspecified likely interests who would want to innovate in the document space - so that Watson: a bit of proprietary software (albeit built using a lot of open source) from a single vendor can do better. At least, this possibility is not something that motivates me to work harder, or to contribute code under a non-copy-left license. I for one, would not like to live in a star-trek future if it was built on proprietary software.

  The powers that want to invest in and evolve OpenOffice will be much more comfortable in Apache, as will their customers. The various parties including Libre office and the Linux distos need to seize upon this opportunity to reconcile and drive value creation from a collective investment. With the Oracle developer subsidy gone, to both both OpenOffice and LibreOffice, the remaining parties need to work together and make a concerted effort to recruit new parties with new ideas. This could be a great outcome!

  More hidden powers that want to invest, but just can't if they cannot sit inside Apache (or is it proprietise?) - show me the powers! might be a good riposte. It is hard to make a balanced judgement on the basis of likely hidden things. As for working together - I can only applaud the idea. Step one for working together is, I would hope not starting and supporting a competing project. As for recruiting new parties with new ideas - who can argue with that ? It is the old ideas such as demanding the right not to contribute in a timely and constructive way that stick in my throat. Indeed, arguably a major motivation here is to get the politics around contribution, its extent and timing out of the project, permanently.

  In conclusion

  Since Doug is a very sharp and personable chap, I imagine that the concerns about cost and overhead are rather secondary to the real issue of the ability to proprietise: which has to be near the crux of this, rather than just a detail.
• Team meeting, appreciated this write-up from Monday I missed. Boggled at E-mail variously, chased an intermittent bug, fixed a nasty OSX build issue.
• Dinner, babes to bed, poked mail.

• Up early, prodded mail quickly; drove Mike & myself into Cambridge for TEAM. Enjoyed another day of great teaching, and helpful feedback.
• Home, fed babes dinner, paid the piano tuner (now finally getting to grips with the higher octaves), read stories to babies. Poor J. under the weather, snoozed a while. Call with Mum, pottered around getting my newest laptop (a lathe) partially re-wired - scary to see the degree of perishing of the original plastic cabling.

• Up extremely early; worked until breakfast, filed updated e-d-s deadlock. Slept, awoke, massaged patch to fix deadlock, more mail. Pleased to see Lanedo's Support for LibreOffice, a tad late for the link I admit.
• Read Mark's essay on balancing power. I tend to think that blessing a single market segmentation is unfortunate. If QA and documentation are weak, and a project has no-one that wants to do that (even for cash), why not out-source that work (at zero cost) to a 'distribution' company ? they can then aggregate customer needs to the point that transaction inefficiencies become small and help fund fixing them. Of course there is some duplication there, but that is the sad downside of competition surely. Having said that, I'd love our ecosystem to have more, viable, well funded companies, and more contracting-out of support, QA, documentation services etc. moving in that direction would no doubt increase flexibility and provide incentives to attract more business and investment around technologies. Happily some of these companies exist and thrive today, yet apparently do not require aggregating ownership of contributions (eg. EnterpriseDB, or PGX or ... ). Finally, having bazillions of small, fragile foundations is indeed a bad idea, things like the Software Freedom Conservancy can help adopt and aggregate established projects giving them some of the advantages, without the overheads.
• Got fed up of writing text, so dug into a fun crasher bug instead, satisfying to actually improve the code a little; wrote and sent status report for the week.
• The rest of the family returned in emotional pieces, after a somewhat disasterous attempt at a combined music lesson, hmm. Created a couple of new freedesktop accounts.
• Lydia and Janice over for dinner in the evening, spent some time riveting the barbeque together (in its increasingly dillapidated state), poked mail late, bed: off at TEAM tomorrow all day.

• Up; pleased to read a sympathetic groklaw take. Adjusted my blog license to public domain. Some encouraging comments too. Re-compiled chen's e-d-s patch for more testing.
• Spent a little while updating my 3.4.0 release blog with an updated contribution graph for good measure.
• This little article on Microsoft's developer platform story is worth a read. It is interesting to see GNOME 3.0 going in a similar direction with its Clutter + Javascript world and shipping rather before Windows 8, I guess the QML world is very java-scripty too. Personally, I find Javascript pretty loathesome as a language for serious work, but doubtless that's just compressed crustiness.
• Lots of mail thrash, Clarity. Lunch. More mail, Christian Lohmaier did a fantastic analysis of the completeness of the code dropped on ASF judging by the filenames at least. Some highlights: 40k of 69k files to be licensed to ASF ie. 40% of files missing, by his (clever) file movement analysis, the snapshot was taken at m103 ie. before 2011-03-21, ie. old and missing lots of branches integrated since then. Then a lot of details: no icons but then no toolbar XML descriptions either, no translations, no templates, no filter configuration, the list goes on. Anyhow, onto the constructive bit: it would be much better, cheaper, safer and more complete for Oracle to license all rights that it owns in the code in the complete set of mercurial repositories to ASF, and to let them clean this up, absent that - many more man weeks of delay, and many more expensive legal round-trips will be required I suspect.
• Tragically didn't manage to get my Monday inbox processed to the point of productivity after a whole day at it: not good.
• Miriam over for dinner, lovely to hear of her wedding plans, and flip through our old wedding albums (ten years next year) together. Back to work before bed.

• Fine lie-in, and off to St Peters on the level, an HTB church plant. Lots of bounce, and friendliness - great kids work, well delivered helpful sermon, sadly rather de-coupled from the text.
• Back, for a roast beef blow-out; tried to read the good bits from Schneier's Applied Cryptography - (of Thomas') at least got the basics of RSA and DES under the belt - much less scary than expected a rather helpful practical book.
• Tea, packed, bid 'bye to the parents; drove home via an M25 congested Long Delays in both directions, J. did some great off-piste map reading around it.
• Quick tea, watched the sixth episode of Hustle, and bed.

• Mini lie-in; breakfast, climbed trees in the parents' garden with a succession of small people - some getting 25+ feet into the air, nice.
• Knocked up a packed lunch, and out to Worthing's Brooklands park / lake. Paddling in the sun, picnic lunch with the babes, lots of play-ground / rushing around; had a break in the sun.
• Back home for afternoon tea. Babes watched a movie, prodded at mail, did some research until dinner.
• In the evening pottered around, poked at a lap-lathe we had lying around (still in good order - modulo perished electrics it seems) - hopefully ideal for reprap nozzle creation if the concentricity is still ok. Dis-assembled a bread-cutter to access a broken gear (seemingly using it for meat introduced grease to attack the plastic). Spent a while modelling a replacement gear in openscad, while catching up with Tom & Becky.
• Pleased with J.'s progress at solving the rubicks cube - most impressive (I couldn't stick it). Bed late.

• Up early, breakfast, somehow missed an 8am meeting; Encouraging sync with Ralf. Insightful article from Brian on the IBM move, and an even more insightful comment from Marbux doing some of the spade-work on expanding the history. Marbux is right - this has all happened before: IBM trying to fork, based on an old code-base (under a non-copy-left license), against the existing developer community's wishes as/when the non-copy-left license was withdrawn last time. Nice to have that Doug Heintzman interview from 2007:

  [three+ years ago]: IBM's contributions will include 35 dedicated programmers as well as editing, accessibility, and other code ... We've been investing in creating many complementary things in Notes that aren't in the core OpenOffice code, such as the SmartScreen filters and iAccessibility2 ...

  While in fact I applaud IBM's lack of engagement with the (highly unreasonable) Sun regime that was OO.o - it is interesting to note the exaggerated claims of dedicated developer numbers never really materialised (or perhaps there was just an effectiveness problem). Also IAccessible2 integration is still one of the (great) promised distinctive new features they plan to finally contribute for Apache Office (much improving accessibility for the impaired on Windows, analagous to SUSE & Sun's integrated work on atk integration for Linux - an unfortunate feature to politic with). The code was re-promised to the community in 2008 in an IBM keynote, and eventually dumped vs. a very, very obsolete branch (1.1.5) and left; and is still not integrated to this day, hence re-appearing as a new value-add. Of course times change, and strategies morph and some commitments are un-sustainable, no doubt it happens to us all.
• Poked at mail, wrote up some thoughts on the 3.4.0 release, and the encouraging progress there; with so much good work by so many, I wanted to highlight some of the behind-the-scenes work too.

• Are the LibreOffice statistics just made up ?

  I was amused last night with Rob's attempt to dispute the statistics we publish, he was promoting the idea that LibreOffice had only 54 developers in the last six months, most of them Novell. It seems (on a benign assumption) that this is likely to stem from a rather simple error, counting just 'bootstrap'. Admittedly we do the rather evil thing of downloading more code when you run 'make' (can be forced with ./download after configure) so perhaps understandable. It raises the question of how we could possibly publish graphs with a straight face that show the opposite. I quickly re-generated the numbers last night and posted them. Of course we all make mistakes, perhaps we have - we use Jonathan Corbett (and GregKH's) gitdm which is quite well established, and all the code and patches are published. They show for the last six months:

  Employers with the most hackers (total 214)
  (Unknown)                  138 (64.5%)
  Oracle                      45 (21.0%)
  Novell                      18 (8.4%)
  Known contributors           7 (3.3%)
  Canonical                    4 (1.9%)
  Redhat                       2 (0.9%)
  

  Of course - the numbers show nothing about volume of contribution, only discrete numbers of affiliated contributors: so RedHat did a much larger volume of work than Canonical - who donated a more diverse set of changes. Cue argument about the usefulness of lots of volunteers contributing small changes (personally I love and want to encourage that - I love 'Unknown' affiliated hackers they sound spicy and interesting).

• Is TDF going to 'fall over in the next storm' ?

  A great allegation from Rob Weir:

  I'd be absolutely giddy with joy if LibreOffice developers would come over to Apache and run their project under the Apache 2.0 license under the Apache process. I'd even be open to calling it "LibreOffice". But this is much more an issue of organizational capabilities than it is the rather narrow gulf between the current OpenOffice and LibreOffice source codes. I want an organization that will last, not something that will fall over in the next storm.

  Ignoring the red-herring about naming; the question is: is Apache really that vast, stable, storm-proofed and well funded as an organisation ?

  First I should say, that I think the Apache governance rocks in many ways, and this is no critique of them. They are a voluntary organisation, and they kick some serious ass. More amazingly than that they achieve so much on a fairly limited budget If I read it right:

  Attachment AE: Proposed 2010-2011 Budget
  
  ASF budget for FY May 1 2010 - Apr 30 2011
  -------------------
  ...
  INCOME Total                : $541,200.00
  

  With the majority of that coming from the sponsorship program. This makes me wonder, whether financially and structurally they are really playing in a different league to TDF or whether volunteer Free Software projects like this, doing amazing things on a shoe-string are the norm. Clearly age, experience, and being incorporated are helpful predictors of future stability. Another example would be GNOME's financials bringing in $187,000 last year, KDE ev reports are in German, but Die Finanzsituation im Kalenderjahr 2008: Einnahmen: 235.406,15 EUR seems of a similar magnitude. Clearly, financially TDF's raising of Eur 100k in a couple of months (from a mix of generous donors), seems quite adequate to finance the project. My conclusion - There is no need for TDF to fall over in any storm - not even a man-made one. But of course, we are eager for more contributors, members and stability.
• One of the nice things about the Attachmate acquisition is being part of SUSE; working with management with a deeply informed grasp of Free Software, the importance of the community, and how working well with them will always, in the end, be best for our customers; anyhow enough raving - here is our SUSE Perspective on Oracle's Contribution of OpenOffice.org to Apache.

Today we released 3.4.0, there is a great list of new features, specific to LibreOffice available (expertly assembled by Marc Pare and others). Each should also be credited so some of the depth of the (code) developer community is apparent, this is of course in addition to our extensive credits page (kept up to date by a volunteer of course). This is the first major release containing code from many of our new volunteers which is exciting. Of course, there are some great improvements there, I like the named range / data-pilot work that makes it easy to extend the data range you're working on without manually editing perhaps ten data-pilots depending on it but there are a load more. Some of the changes are invisible, and/or behind the scenes - so I thought I'd expand on them here.

The incredible shrinking codebase

First - ridding ourself of sillies - there is lots of good work in this area, eg. big cleanups of dead, and unreachable code, dropping export support from our (deprecated for a decade) binary filters and more. I'd like to highlight one invisible area: icons. Lots of volunteers worked on this, at least: Joseph Powers, Ace Dent, Joachim Tremouroux and Matus Kukan. The problem is that previously OO.o had simply tons of duplication, of icons everywhere: it had around one hundred and fifty (duplicate) 'missing icon' icons as an example. It also duplicated each icon for a 'high contrast' version in each theme (in place of a simple, separate high contrast icon theme), and it also propagated this effective bool highcontrast all across the code bloating things. All of that nonsense is now gone, and we have a great framework for handling eg. low-contrast disabilities consistently. This also reduces run-time memory consumption (we have to cache the .zip theme's directory), and of course binary installation and download size shrinks (we ship several themes to taste) - here is a pretty picture:

The incredible shrinking footprint on Windows

When we first released LibreOffice, in place of the individual install set per language - duplicating the code, artwork, templates, etc. many tens of times on each server we switched to bundling lots of languages (and also the run-time adapted BrOffice branding) into a single package. That shrunk our mirror load from 76Gb down to 11Gb for 3.3, which is now for 3.4 down to 7.6Gb a handy win (of ~70Gb) for mirror admins, making us more agile, and appreciated by our fantastic mirror network I hope. To achieve this, in consultation with the l10n team, Kendy worked hard to split out our help to the wiki - so you can browse it on-line at http://help.libreoffice.org. That of course helps Linux users' space-constrained live-CDs more useful too.

With that work in-place, we managed to cram the top fifty languages into only 225Mb on release, which (sadly) left the remaining languages in a rather larger 265Mb download. In 3.4.0 down to improvements such as the them work above, sharing templates, dropping the BrOffice brand (at the Brazilians request) and more importantly Kami's gret msi packing improvements we've managed to pack all our languages, and many more dictionaries, and more functionality into a single download at 197Mb. That is clearly still too big, but smaller than a MS Office service pack. We are still some 40Mb larger than the original single-lang packaging (which it is our goal to match), but there is plenty more room for improvement (eg. gutting the megabytes of pointless ICU code we ship), and we'd love more help, (cf. scale offset of 150Mb):

Better translation infrastructure

I'm not an expert here, but our fantastic l10n team, swelled to 200 contributors is doing some great work on getting the latest translations into the product. They're using pootle for that, but better (thanks to Andras) we've switched from storing our translations in SDF format, instead using native gettext .po files compiled to various odd other formats during the build. Bjoern has a prototype patch coming to allow run-time .po file translation, to allow post-ship translation and better integration with Launchpad.

Heroic merging

One of the huge, invisible tasks for 3.4.0 was the merging of Oracle's code changes. Luckily, of course we use git - and several split repositories, such that merging should be easy. Then again, we have done lots of widespread changes, many hundreds of thousands, if not millions of lines of diff, stripping dead code, translating thousands of comments, touching every single file, as well as more substantial API and code changes.

One of the things that made the merge more tricky, was that (after decades of inactivity), some Oracle developers decided now (post fork) would be a great time to do some long overdue large-scale code cleanups. One of those was replacing every instance of TRUE with sal_True, same for FALSE, all instances of legacy types like BOOL to sal_Bool, for each of UINT16, ULONG, LONG etc. etc. Unfortunately, not an easy 'sed' either, as witness to some of the bogus "sal_True" type strings that popped out of the works. While a valuable, and much needed cleanup, this results in the kind of multi-million-line patch (touching all the dead code too) that has the effect of obfuscating their other valuable changes, and takes a certain determination to merge and reconcile.

The effort, spearheaded by Norbert with many straight days of mindless grunt work from myself, and able assistance from Thorsten, Kendy, Kohei, Noel and others - hopefully highlights a triumph of determination and survival against the odds, the tedium and some RSI. Unfortunately, due to some comic, transient technical hitches (that resulted in having to do much of the work twice) this merge rather delayed the code freeze and our first betas for 3.4.

Tinderboxes / rapid building & QA

With the rush to get 3.3 out, and the stabilisation releases after it, as well as the one-shot fun of migrating many Linux distributions over to use LibreOffice there was not enough spare bandwidth to get many tinderboxes or build-bots building. Unfortunately, this had quite an impact on the QA teams particularly after the merge completed, which was sad. The good news is, that we have mostly fixed this now, and have much more recent (the aim is daily) install-sets for most platforms available for testing - a great way to get involved is to help with re-testing old bugs against the latest stable releases.

A chunk of this is down to better tooling and scripts to drive tinderboxes, though we still struggle with unreliable Windows builds, cygwin's sh.exe, or perl.exe or dmake like to wedge intermittently some hours into the build. On the Mac front, Norbert had a breakthrough to shrink the build time. An all-lang Mac build used to take 15 hours (13 inside helpcontent2) - he managed to get this down to under 30 minutes using a ramdisk - substantially improving our agility and ability to turn around builds quickly: getting the latest code to QA fast. One of many great build performance improvements - with other much appreciated packaging acceleration wins in the tangle of badly written perl by guys like Jan Darmochwal, Julien Nabet and Steve Butler.

Then of course some QA stars like Rainer, Alex, Cor, Andre and many others have been working hard at finding, cleanly filing, triaging, prioritising and marking bugs, critical work.

Time based releases

One of the key features of LibreOffice, from my perspective, is its time based release process. Italo has done a great write-up with several nice pictures of this, one of these is this:

To me - having a predictable, and time-based release is such a key concept, that shipping 3.4.0 as a carefully labeled, point-zero release on time is more important than shipping an incredibly bug-free product, at a future, undefined point. This avoids a deeply political process of deciding when and if to release, and whose pet bug is worth waiting for and why, and why is his worse than mine, and oh ! I just found another one, lets delay another week. The alternatives to a time based release seem to have lead only to long (multi-month) slips.

Clearly, we have learned a lot this cycle, and improved our processes to make future releases even better. Obviously our succession of point releases: 3.4.1, 3.4.2 etc, will incrementally improve quality: indeed we are confident of that, since we already have a lot of fixes in-place for 3.4.1, however the fact remains that 3.4.0 is buggy (in fact all software is, but it is more so than usual, and we found a lot of bugs rather late). The bright side, is that our future point-zero releases, build on our improved infrastructure will be better in future. This is why we are continuing to advertise 3.3.2 (and the soon to be available 3.3.3) as the primary download build for now (thanks Christian for all your hard Javascripty work to make that fly). Please do - check-out 3.4.0 and get stuck into helping us make the Free Software Office world even more fun, fast-moving, and exciting. A great way to start as a developer is to get stuck into an Easy Hack.

Contributors

Of course, lots of people got stuck in already, and continue to do so. We like to graph statistics of commits by affiliation - charting how many people of what affiliation commit at least one patch per month, that gives us a pretty graph like this:

Which (I hope) shows the strength and diversity of the contribitor base, as well as its extraordinary growth since we launched. Sadly, some like to denigrate and despise contributors that only come up with small changes, personally I started off in Free Software as a contributor like that - and we love to help encourage, and mentor contributors from small things - to greater ones.

Other highlights

Well, of course so many others have been involved in making LibreOffice the success it is today, tirelesss work by many Steering Committee members, like Italo and Florian writing and co-ordinating multiple press briefings concurrently, and many helping translate and promote the software. The design team, being patient as we get the basics right before getting too stuck into fixing the UI (while producing some beautiful, much improved artwork), and no-doubt many others I forgot. And please take note - this is just some of the features you can't see, there are plenty that you can"

• Up early, prodded mail, quick reality check. Set off to drive south, worked in the car - somehow the light was kind.
• Recalled happily the first days of the announcement of The Document Foundation, press-releases combined with real, building code, preview binaries for end-users, that you could build yourself, backed by a real, flesh-and-blood team of developers empowered with commit access, united and working away at our first official release even while volunteer spokes-people briefed the media.

• It seems that IBM and the ASF are encouraging individuals from the LibreOffice world to sign on as 'Initial Committers' to their new project. I have a few thoughts on that. It seems that there is a gulf between where the Apache Software Foundation is and where we, as a community are. Bridging that may prove difficult, and the main stumbling block is non-copy-left licensing.

  One thing that I think is clear, is that sticking together is more important now than ever before. While we are a young team of hackers I hope we're a friendly one, and have built some kindred spirit. We are not (to use Rob Weir's reply to Jeremy Allison) "something that will fall over in the next storm"). I would hate to see us divided which is a real risk. Consider a question you can ask a group of people, that seems reasonable: eg.

  "Can everyone who was -not- bullied as a child, please raise their hand ?"

  Unfortunately this, if acted upon, has the effect of isolating and identifying all those who sadly were, yet without their consent. The best course of action is to reject the question, and the second best is to not raise your hand.

  Whatever we do, needs to be done together as a community. Perhaps you personally, have no concerns about licensing all of your code to corporations who have no intention to reciprocate, and who care nothing for working constructively (some with us with a track record to match). If so, can I ask you to consider your actions in the wider context: of the rest of the team you've enjoyed working alongside, particularly those who may not feel the same.

  All of which is to say, I would strongly prefer to see either all of us as initial committers, or none at all, and that is a decision we need to make collectively; clearly I have a strong personal preference for the latter option.

• Arrived home, lunch. Prodded around in OO.o some more. One of the more amusing finds is that dmake (still substantially required to build the majority of OO.o) is GPL software, not acceptable to Apache. Prodded at xalan-c, I was unaware that Rob Weir started his apache career here - an initial checkin, and six commits over less than a month, a decade ago (in the svn history if I read it right), then nothing. "I was a committer in the Apache Xalan project many years ago (1999-2000)" sounds right, apparently from mid Dec 99, to mid Jan 00. Still no doubt useful to have that thing out there somewhere, and perhaps he has commits under another name to other xalan pieces (not trunk of java or test apparently).
• Worked, unfeasibly late.

• Up early, mail prod, misc. calls. Poked at some encouraging iogrind related progress. Poked at a packagekit / zypp backend issue. Lunch.
• Interested to see that the widely trailed move of donating OO.o to the Apache Software Foundation actually happened today. TDF have a simple, friendly response, and I have a number of thoughts:
• Engaging with community members (IBM), and having a commitment to the developer and open-source communities (Oracle) are laudable goals. I can only applaud the sentiment. Unfortunately, starting that process only after finalizing a license incompatible with the communities existing work, and at a different home to the one the developers chose themselves seems an odd way to engage, and commit.
• Unfortunately, there is a problem with Free Software developers, firstly - they often don't wear suits, and (get this) some have beards: which just shows you the kind of schmucks they are. But worse - they have odd, meritocratic, collaborative decision making processes, that don't come up with suitably corporate answers. One example is jurisdiction: the community (after all is said and done) wanted to found itself in Germany. Professional, serious, serial, corporate body founders prefer to go elsewhere (US, UK) - yet, is it really that bad to compromise on the issue ? Community decision making - but only if you like the outcome is a tad unfortunate.
• Worse - Free Software hackers tend to be free-spirited, and they often believe in reciprocity: if I give you my work, surely you should give me yours ? ie. the spirit of the copy-left. Unfortunately, that is not the Apache way, which has some merits no doubt, but is alien to the existing developer community that commitment is made to. OpenOffice has traditionally included plenty of copy-left code, some of which I highlighted before. Coercing developers to do the bidding of big companies is not something they react well to (usually).
• This event highlights some of the great work that has been done as part of the GPLv3 process, and also the MPLv2 work (done by my friend: Luis Villa). These happy lawyers out there have laboured long and hard to make their new licenses compatible with Apache 2.0 - such that code under that license can be re-licensed under their terms. An example of doing this is here. Without their labours, it would not be possible to integrate the Oracle code, and the eight months of existing work by the community into a single beautiful whole. Clearly there is no rush to actually do that work, perhaps it can be done on schedule for LibreOffice 3.5. By a happy coincidence, we have a slightly longer cycle this time as we sync. up our six-monthly time-based release schedule with that of Linux distributions and desktop.
• Apparently this is a somewhat divisive attempt by an exiting Oracle, along with IBM to sideline the existing developer community, their governance, their aspirations, membership, licensing choice (explicitly adapted to meet IBM's needs incidentally), bylaws, and so on. All of this despite a profound, frequently stated open-ness to including new (particularly large) corporate contributors inside TDF, and taking their advice seriously.
• Thankfully, license compatibility lets us turn this from a closed, and finished chapter of long, sad story - into the beginning of a happy one - where everyone, regardless of size and Dilbert-ness can join together around TDF's code-base and contribute on their own merits. So, next time you meet a Free Software lawyer, please - shake their hand.
• We also have Rob Weir enthusing about the joys of his preferred outcome. It all sounds wonderful, but sadly is not what the substantial, existing developer (and marketing, and QA and ...) community chose. Luckily of course, it is not a final choice as/when the code is released they are free to choose to join TDF and engage. Still, I look forward to reading Rob's code - it'd be great to hack with him.
• Got around to reading Luke's mail and associated odf proposal That document itself has some great quoteable material:

  Both Oracle and ASF agree that the OpenOffice.org development community, previously fragmented, would re-unite under ASF to ensure a stable and long term future for OpenOffice.org.

  or how about

  The initial set of committers include people from the community of OpenOffice.org Technology projects .... The initial group of developers will be employed by IBM, Linux distribution companies, and likely public sector agencies. Localization resources are expected to gravitate to the new project, as well. Ensuring the long term stability of OpenOffice.org is a major reason for establishing the project at Apache.

  Amazing to see Andrew Rist and Rob Weir as the initial committers - I'm unaware that they have ever committed a single line of code to the codebase before: but ... there is always a first line; a whole new sense of initial committer perhaps. I was encouraged to read somewhere that: "The first step along the road leading to committership is to become a developer".
• Sean M. Kerner writes a nice piece here on the rational What a Terrible move.
• Fun quote from Benson Margulies posting on the Apache projects list:

  Instead of IBM loving and leaving ASF in an attempt to outmaneuver Oracle, we have IBM, with help from Oracle, proposing to love ASF to outmaneuver an open source foundation. It all looks like fox-hunting to me. (The unspeakable in pursuit of the inedible.)

• Call with JP. Pleased to see Christian restoring some Korean comments from the history whose encoding got mangled - LibreOffice: not just a legacy of German comments, but some in Korean too - help appreciated translating them, they're: here and nearby.
• Pleased to see Bradley's thoughtful piece on the move. Played some violin to relax, nothing like Kreutzer to free up the hands & mind.