Improving LibreOffice for Windows
0 likes790 views
The document discusses improving LibreOffice for Windows users by creating a "Windows vendor" organization within the project to support the Windows build. It also recommends developing a new Windows installer to provide a better user experience and differentiate LibreOffice from other legacy products. Additionally, it suggests playing nice with other StarOffice legacy products to avoid incompatibility issues and supporting Windows developers by improving tools for native Windows development.
Improving LibreOffice for Windows
- 1. Improving LibreOffice for Windows Jesús Corrius Llavina
- 3. Table of Contents Windows as a “vendor”
- 4. A new Windows installer
- 5. Playing nice with the StarOffice legacy ecosystem
- 7. Windows as a “vendor” I guess Microsoft will not join TDF but...
- 8. Windows as a “vendor” Major Linux distributions do: Have a team and a organization to work on LibO
- 9. Build customized and branded versions of LibreOffice
- 10. Offer support and convenient way to distribute and update it
- 11. Marketing about their product and that includes LibO
- 12. Have dedicated developers working in LibreOffice fixing bugs and add new features their customers want
- 13. This model works very well
- 14. Windows as a “vendor” Windows is an important platform but things don't work the same way
- 15. Why not? Maybe... There are no customers, no bugs or feature requests
- 16. No companies are migrating Windows users to LibO
- 17. No one makes money Why not? Probably... Market is too fragmented
- 18. No real FOSS / collaboration culture
- 19. Lack of proper channels to canalize this support (“I am a company and I want to support the Windows build, how can I do it?”)
- 20. Windows as a “vendor” Let's create a Windows “vendor” inside the project
- 21. What's that? Organized infrastructure to support the Windows build
- 22. An umbrella where small and big companies can feel at home
- 24. Marketing
- 25. A small company migrating Windows desktops can't afford to have a dedicated developer, but maybe 20 companies can.
- 26. Make it very easily for companies with no FOSS culture to join the effort
- 27. Windows as a “vendor” Why do we have to spend our time and effort on that? Windows is still a very important platform, especially for companies, but end users also
- 28. We have already won the Linux/BSD market
- 29. The StarOffice legacy ecosystem will concentrate on Windows and (probably MacOS X too)
- 30. We need to clearly show the outside world that we do really care and we are serious about it
- 31. We also need to clearly show that the other platforms are still very important and they always be
- 32. A new Windows installer... If it works why do we have to change it?
- 33. A new Windows Installer The two steps installation (NSIS/MSI) is not very friendly for Windows users
- 34. Why are we leaving temporal files in the desktop?
- 35. Not optimal for large deployments in corporations
- 36. On Windows the installer is the first impression you get of an application
- 38. But... I think there's a more compelling reason...
- 39. A new Windows Installer SHOCKING! Most Windows users don't perceive a difference between LibO and other products of the StarOffice legacy ecosystem
- 40. We need a new nice, good-looking and friendly installer to differentiate LibO from the other legacy products
- 41. We need to involve all the community and especially the design team
- 42. It will be send a hint to the outside world than we do care about the Windows users and we are working in the right way for them
- 43. Playing nice with the StarOffice legacy ecosystem Because we are not alone
- 44. Playing nice with the StarOffice legacy ecosystem None of the other projects of the StarOffice legacy ecosystems have produced installable builds yet
- 45. But there's a chance that they do
- 46. Products will not be installable side by side on Windows
- 47. There are many reasons: “Greedy” installers
- 48. Application Ids (COM+ and OLE stuff)
- 50. File associations (bad clean up code, backup, etc)
- 51. Same file names ;) In short: it's a mess
- 52. Playing nice with the StarOffice legacy ecosystem Should we play nice? I think so
- 53. Let the others be the “bad” guys
- 54. Some things are trivial to fix
- 55. Others are not: Very bad decisions when designing the OLE support can make us incompatible with older documents or documents generated with other products based on StarOffice It's fixable, but it's a lot of work
- 56. Other products will have the same problems
- 57. But.... let's be the nice guys of the ecosystem
- 58. Developers, developers, developers (No need to comment about this one)
- 60. Developers, developers, developers What would a Windows developer ask first?
- 61. “Dude, where's the Visual Studio project to compile LibreOffice?”
- 62. Cross-compilation is good, but we have to continue to support the native Windows tools
- 63. Support for latest Windows compilers and environments: Visual Studio 2010
- 64. Newest Microsoft SDKs, etc
- 65. Properly detect all the tools and dependencies, etc.
- 66. Easy way to set up a Windows development environment (msi that installs everything in the right places?) By popular demand: 64 bits version for Windows
- 67. Developers, developers, developers Full support for the free Microsoft Visual Studio Express: remove ATL dependencies
- 68. Extensions builder for Visual Studio
- 69. Known languages for Windows developers to write extensions: C#, Visual Basic, etc
- 70. Thank you … for listening to me! See you at the LibreOffice Paris Conference
- 71. Find out more at http://conference.libreoffice.org