22 Commits

Author	SHA1	Message	Date
Thomas Harte	a38639d099	Eliminates the concept of an iCoordinate. ... Real-life precision appears not to support the idea of sub-sample pixel storage.	2018-09-12 20:05:39 -04:00
Thomas Harte	80b281d9f1	Switches back to whole bytes per pixel, owing to persistent precision problems at 1bpp. ... Also fixes the inaccurately-named `cycles_since_update_`.	2018-06-02 18:25:00 -04:00
Thomas Harte	0b771ce61a	Removes all instances of the copyright symbol.	2018-05-13 15:19:52 -04:00
Thomas Harte	6547560e52	Gives the CRT the ability to move iCoordinate multiplication outside of the fragment loop. ... That resolves precision issues, as were plaguing the Apple II.	2018-04-18 19:29:03 -04:00
Thomas Harte	1c605d58e3	Removes the CRT requirement for an integral relationship between cycles and samples.	2018-04-16 20:00:56 -04:00
Thomas Harte	865c47a1ac	Names the magic constants.	2018-04-08 10:35:07 -04:00
Thomas Harte	3821679efd	Switches to bit unpacking on the GPU.	2018-04-07 22:17:47 -04:00
Thomas Harte	c0b4dd65da	Mades the expected video signal usage explicit.	2018-04-04 19:01:18 -04:00
Thomas Harte	ac80d10cd8	Separates the component parts of running an audio stream: task deferral, filtering and generation. ... Walking towards improving opportunities for composition.	2017-12-17 21:26:06 -05:00
Thomas Harte	2203499215	Enables -Wreorder and corrects a few of the more trivial fixes thereby suggested.	2017-11-09 22:14:22 -05:00
Thomas Harte	ad9df4bb90	Commutes uint8_t *, uint16_t *, uint32_t *, size_t, off_t and long to functional-style casts.	2017-10-21 22:30:15 -04:00
Thomas Harte	5e3e91373a	Switches all unsigned int and double casts to functional style.	2017-10-21 19:49:04 -04:00
Thomas Harte	4abd62e62b	Standardises on const [Half]Cycles as the thing called and returned, rather than const [Half]Cycles & as it's explicitly defined to be only one int in size, so using a reference is overly weighty.	2017-07-27 22:05:29 -04:00
Thomas Harte	b82bef95f3	Decided to follow through on Cycles and HalfCycles as complete integer-alikes. Which means giving them the interesting range of operators. Also killed the implicit conversion to int as likely to lead to type confusion.	2017-07-24 20:10:05 -04:00
Thomas Harte	c2a7dffa7d	Converted the ZX80/81 video component into a ClockReceiver. As it happens, it's most convenient to take the half-cycle bus here.	2017-07-22 23:02:28 -04:00
Thomas Harte	ed28260aaf	Hardens the ZX80/81 video routines to ensure they never try to push data into the future and don't double-count time when pixels would ostensibly run into sync. You could previously see the CRT being handed negative run lengths if sync interrupted pixels or if a run of more than 320 pixels (my arbitrary buffer size) occurred, with corresponding poor behaviour given my use of unsigned numbers.	2017-07-09 19:33:05 -04:00
Thomas Harte	30e93979d2	Removed data work if sync is enabled; in that case no data is output.	2017-07-08 21:01:07 -04:00
Thomas Harte	b9dbb6bcf8	Discovered my timing error: the I/R <-> A loads should take an extra cycle. This means the ZX80 now finally takes the correct 207 cycles per line. Fixed the video output wave to be clocked at the appropriate rate.	2017-06-12 18:55:04 -04:00
Thomas Harte	ba5f34f827	Narrowed view to the centre 80% of a frame.	2017-06-11 17:24:32 -04:00
Thomas Harte	d910a4fd38	Adjusted to signal an interrupt during the refresh cycle rather than weirdly just afterwards. Which cuts video timing down by 4 cycles a line. There still might be a problem here somewhere though, as I'm getting 206 cycles/line and the internet states it should be 207. ... Also: lots of printfs have grown temporarily as I try to figure out what I'm doing so wrong as to break loading.	2017-06-11 13:32:20 -04:00
Thomas Harte	e3ee9604a5	Added comments.	2017-06-06 18:01:33 -04:00
Thomas Harte	8c66e1d99d	Factored out ZX80/81 video and rejigged to ensure it will keep ticking over irrespective of whether the machine is supplying data.	2017-06-06 17:53:23 -04:00