You Convey To Know : X-Principles
I was reading nearly the history of X11 on Wikipedia piece trying to create Linux bugs in addition to I stumbled across the principles of X. There are 3 I'd similar to focus on, ii of which I experience apply rattling good to X-Plane in addition to i which does not.
The solely affair worse than generalizing from i instance is generalizing from no examples at all.
This is true, in addition to inward price of X-Plane, I retrieve this comes downward to "don't add together features that aren't needed." When you lot hold off at the "cost" of X-Plane inward price of how difficult it is to write novel code (and inward to a greater extent than or less ways you lot could order Laminar Research does concern yesteryear adding novel code to X-Plane) our toll goes upwards amongst the number of existing features already inward house that nosotros guide maintain to overstep away on working piece nosotros alter things.
If a job is non completely understood, it is in all probability best to furnish no solution at all.
This also is rattling truthful for X-Plane. We desire to furnish long-term compatibility for X-Plane every bit a platform for airplanes, scenery in addition to plugins. The adventure of putting inward solutions to problems that nosotros don't sympathise is that ii months afterward we'll overstep away "ah drat, our APIs aren't quite right" simply yesteryear that fourth dimension tertiary parties volition hold upwards depending on the existing broken behavior.
Provide machinery rather than policy. In particular, house user interface policy inward the clients' hands.
I am critical of X, simply you lot could deflect my criticism yesteryear pointing out that it stems from how X is used, which mayhap goes against its master copy intention. My issues, all stemming from the work of X every bit the lowest score of window administration are:
The minute criticism isn't fifty-fifty remotely fair at all - X isn't a consummate solution, in addition to it's meant to guide maintain materials on superlative of it. I suppose my number hither is: X11 is pretty ubiquitous every bit the base of operations layer inward the Unix world, simply what goes on superlative is opened upwards for debate.
And this goes to this tertiary principle. If X11 is solely the "bottom half" of a consummate UI framework, it makes feel to furnish machinery in addition to omit policy. But....is the bottom one-half of a consummate UI framework useful to clients? It's the lack of policy, encoded inward a simple, ubiquitous API that makes uncomplicated things slowly in addition to difficult things possible, that frustrates me nearly coding on Linux.
So when it comes to X-Plane, I guide maintain a rattling dissimilar approach - I process all of my SDK code every bit policy work. That is, nosotros endeavor to move inward slowly to exercise things nosotros desire authors to exercise in addition to impossible to exercise things nosotros don't desire authors to do. In my experience this is the solely reasonably option...given an opened upwards platform, users volition do, well, everything possible.
Is it fair to hold upwards then heavy handed? Call me a techno-fascist simply I retrieve it's essential. At the halt of the day, tertiary parties count on us to overstep away on the platform stable then that they don't guide maintain to waste product fourth dimension updating previous finished products each fourth dimension nosotros post service a novel land to X-Plane. The wider the array of weird demeanour that nosotros tolerate, the wider the array of weird demeanour nosotros guide maintain to process every bit backward compatibility cases later. By using technical limitations to require authors to adapt to specs the kickoff time, nosotros tin avoid compatibility breaks later.
The solely affair worse than generalizing from i instance is generalizing from no examples at all.
This is true, in addition to inward price of X-Plane, I retrieve this comes downward to "don't add together features that aren't needed." When you lot hold off at the "cost" of X-Plane inward price of how difficult it is to write novel code (and inward to a greater extent than or less ways you lot could order Laminar Research does concern yesteryear adding novel code to X-Plane) our toll goes upwards amongst the number of existing features already inward house that nosotros guide maintain to overstep away on working piece nosotros alter things.
If a job is non completely understood, it is in all probability best to furnish no solution at all.
This also is rattling truthful for X-Plane. We desire to furnish long-term compatibility for X-Plane every bit a platform for airplanes, scenery in addition to plugins. The adventure of putting inward solutions to problems that nosotros don't sympathise is that ii months afterward we'll overstep away "ah drat, our APIs aren't quite right" simply yesteryear that fourth dimension tertiary parties volition hold upwards depending on the existing broken behavior.
Provide machinery rather than policy. In particular, house user interface policy inward the clients' hands.
I am critical of X, simply you lot could deflect my criticism yesteryear pointing out that it stems from how X is used, which mayhap goes against its master copy intention. My issues, all stemming from the work of X every bit the lowest score of window administration are:
- By making remote-desktop capabilities a kernel purpose of the spec, X11 introduces a complexity toll to desktop UI programming that isn't necessary a lot of the time.
- X is non a consummate solution to a desktop surroundings layer.
The minute criticism isn't fifty-fifty remotely fair at all - X isn't a consummate solution, in addition to it's meant to guide maintain materials on superlative of it. I suppose my number hither is: X11 is pretty ubiquitous every bit the base of operations layer inward the Unix world, simply what goes on superlative is opened upwards for debate.
And this goes to this tertiary principle. If X11 is solely the "bottom half" of a consummate UI framework, it makes feel to furnish machinery in addition to omit policy. But....is the bottom one-half of a consummate UI framework useful to clients? It's the lack of policy, encoded inward a simple, ubiquitous API that makes uncomplicated things slowly in addition to difficult things possible, that frustrates me nearly coding on Linux.
So when it comes to X-Plane, I guide maintain a rattling dissimilar approach - I process all of my SDK code every bit policy work. That is, nosotros endeavor to move inward slowly to exercise things nosotros desire authors to exercise in addition to impossible to exercise things nosotros don't desire authors to do. In my experience this is the solely reasonably option...given an opened upwards platform, users volition do, well, everything possible.
Is it fair to hold upwards then heavy handed? Call me a techno-fascist simply I retrieve it's essential. At the halt of the day, tertiary parties count on us to overstep away on the platform stable then that they don't guide maintain to waste product fourth dimension updating previous finished products each fourth dimension nosotros post service a novel land to X-Plane. The wider the array of weird demeanour that nosotros tolerate, the wider the array of weird demeanour nosotros guide maintain to process every bit backward compatibility cases later. By using technical limitations to require authors to adapt to specs the kickoff time, nosotros tin avoid compatibility breaks later.
