J’ai lu récemment une série d’articles publiés par le site RoughlyDrafted Magazine, intitulée Platform Death Match (The rise and fall of platforms). Cette série est constituée de 12 articles, qui passent en revue les différents systèmes d’exploitations, depuis l’époque des machines 8 bits jusqu’à maintenant. Je vous en conseille vraiment la lecture ; c’est bourré d’informations et d’anecdotes très intéressantes. Si l’histoire de l’informatique vous passionne, si vous êtes interpellés par les noms qui la parsèment (Apple, Microsoft, IBM, Commodore, Amiga, Atari, NeXT, Be), vous ne serez pas déçus.
De manière un peu paradoxale, les conseils qu’on y trouve sont applicables à plein de situations différentes, notamment à des activités de développement logiciel. Une plate-forme reste une plate-forme, et même si les échelles sont différentes, j’y ai trouvé certaines similitudes avec ce à quoi je suis confronté au quotidien.
Pour faire simple, je vais simplement citer les quelques conclusions qui parsèment les articles.
Notable platform lesson: provide an optional, burnable bridge with the past, and use it to lead existing users into a significantly better platform.
Notable platform lesson: it’s difficult to compete against free; bundle software and hardware to avoid competitors getting in the way.
Notable platform lesson: don’t release technology that is easy to copy and expect to maintain ownership, and don’t partner with cutthroats.
Notable platform lesson: sloppy, incompatible cousin platforms are not a good money maker.
Notable platform lesson: there isn’t much room to compete within a platform, and there can be only one driver.
Notable platform lesson: Successful platforms create start foundations to build upon, enabling growth in new directions for both users and developers.
Notable platform lesson: introducing a new platform is very difficult, particularly against an entrenched market.Finding an entirely new market can leverage growth towards establishing a sustainable platform.
Notable platform lessons: supporting multiple platforms is a huge drain on development resources. Backward comparability may actually work to prevent forward progress.
Notable platform lessons: supporting an integrated system is far easier than supporting a product of assembled parts from different vendors.
Notable platform lessons: an easy to copy platform will be copied. Once you give away your business model, it can’t be easily reclaimed.
Notable platform lessons: there is rarely enough room in the mass market for entirely new and incompatible platforms.
Notable platform lesson: bold risks create the prospect of big rewards, but don’t guarantee them.
Notable platform lesson: Software can be as significant of a barrier to new platform adoption as new hardware.
Notable platform lesson: Remaining the big fish is only possible if you can resist being overwhelmed by smaller parasites.
Notable platform lesson: Platform ownership makes advancements easier to deliver to users.
Notable platform lesson: The catch-22 of development support for new platforms is one of the most significant barriers for entry.
Notable platform lesson: Real artists ship!
Notable platform lessons: Backwards compatibility has to be a temporary bridge to a new platform, not a permanent crutch that preserves legacy or promotes competing platforms. Also: developers, developers, developers!
Notable platform lesson: in theory, theory and practice are the same; in practice, they are not.
Notable platform lesson: undeliverable zombie projects have to be put down or they will simply eat up all the brains in a company.
Notable platform lesson: raw technologies don’t sell; finished products do.
Notable platform lesson: it’s much easier to work with the industry rather than against it.
Notable platform lesson: Sometimes, what everyone remembers is wrong.
Notable platform lesson: Free software development is expensive to maintain.
Notable platform lesson: Integrated products help insulate from competitive attacks, and provide more value for users.