![]() ![]() Suppose a system somehow says, "from now on I'm going to restrict how something operates." How will components (people, for example) react? Some would go with the flow, some would leave the system, and some would find workarounds. Seeking change and relying on what has worked in the past require a balancing act in the end. At the same time, we seek path dependence, the idea of relying on old habits. Developing NVDA itself and its add-ons is a good example of humans seeking change. ![]() But humans seek change and wants to experiment, creating new ideas and improving existing practices, and transform and redefine the idea of what is acceptable and can coexist. In our minds, the image of a screen reader code that allows add-ons from different eras to coexist without problems is a "gift" from heaven. This practice stems from the belief that add-on code can survive future NVDA changes such as Python version upgrades with little to no problems, something that was tested before (2019) and will be tested again in a few months. The issue being discussed has to do with the practice of declaring a far future date/year as last tested NVDA version in add-on manifests, which was introduced in 2019 when minimum and last tested version manifest keys were introduced. Reply to this email directly, view it on GitHub, or unsubscribe.You are receiving this because you were mentioned.Message ID: following is quite philosophical and discusses two related issues: coexistence and perfection at the system level (do let me know if what I'm writing is incoherent): That is, as long as 2024.1 can not possibly be released before January 1, 2024. This only works, as longĪs NV Access never releases a next year's version in the current year. That version 9999.1 of NVDA hadn't been released yet. ![]() That way, if the lastTested said 9999.1, but it was only 2023, NVDA would know I think the only way you could test this with validity, would be to have NVDAĬompare the current system date idea of the year, against the first part of the This is perfectly valid from the add-on's point of view. Somebody in 2025 decides to use an old version of NVDA 2023.2, with anĪdd-on which Has minimum of 2019.3.0 and lastTested of 2025.1.0. If this is expensirve, this can be closed.Enviado desde mi iPhoneEl, a las 6:55, Luke Davis How does NVDA 2023.2 know, that NVDA 2025.1 isn't available?Į.g. For flexibility and bc independent of years, nvda should consult the addondatastore transform repo. ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |