summaryrefslogtreecommitdiff
path: root/dviware/dvi2bitmap/TODO
diff options
context:
space:
mode:
authorNorbert Preining <norbert@preining.info>2019-09-02 13:46:59 +0900
committerNorbert Preining <norbert@preining.info>2019-09-02 13:46:59 +0900
commite0c6872cf40896c7be36b11dcc744620f10adf1d (patch)
tree60335e10d2f4354b0674ec22d7b53f0f8abee672 /dviware/dvi2bitmap/TODO
Initial commit
Diffstat (limited to 'dviware/dvi2bitmap/TODO')
-rw-r--r--dviware/dvi2bitmap/TODO51
1 files changed, 51 insertions, 0 deletions
diff --git a/dviware/dvi2bitmap/TODO b/dviware/dvi2bitmap/TODO
new file mode 100644
index 0000000000..ffe4650ef0
--- /dev/null
+++ b/dviware/dvi2bitmap/TODO
@@ -0,0 +1,51 @@
+
+Implement PostscriptImage, either using PipeStream, or using libgs
+ [potentially messy, but not terribly hard]
+
+Handle multiple DVI files on the command line. Similarly, have it able
+ to read more than one DVI file from the same source: that would
+ mean that it would be possible to have dvi2bitmap sitting reading
+ from a FIFO, while a succession of TeX processes write to the
+ FIFO. [a bit harder than it might seem, as a variety of things
+ might have to be re-initialised, and the code isn't currently
+ structured with this in mind]
+
+Have dvi2bitmap able to write one of its bitmaps to the stdout. This
+ would make it possible to regard dvi2bitmap as a filter [this
+ should be pretty easy, and only depends on the output functions
+ (essentially libpng) accepting file handles rather than file
+ names, which libpng, at least, does]
+
+Support PBM graphics; support PICT graphics on the Mac. [PICT would be
+ a lot harder than PBM, because there's a lot more Mac apparatus
+ has to be learned about and built into the build procedures. This
+ is a rather low priority, I think, and as far as PICT goes, more
+ trouble than it may be worth]
+
+Perhaps have a configuration file, rather than all those options
+
+Improve antialiasing, if there is in fact a more sophisticated
+ algorithm available than the simple scale-it-up-then-
+ scale-it-down used at present. At least add a --antialiasing
+ option to abbreviate --scale + --magnification [the latter at
+ least is quite trivial]
+
+Improve BitmapImage documentation. Improve? Write it!
+
+Create a process-DVI-file main loop as a function, allowing hooks to
+ be registered for each of the possible events.
+
+----------------------------------------------------------------------
+
+Refactor runCommandPipe as PipeStream, a subclass of InputByteStream
+ Done, 3 August 2003
+
+Reimplement kpathsea as a singleton object
+ Done, 3 August 2003
+
+Have InputByteStream mmap its target file, if possible [very easy,
+ really just a matter of the right bookkeeping]
+ Done, 22 August 2003 (pretty easy in fact, barring a couple of gotchas)
+
+Implement reference points in output bitmaps, reported by --output=bitmap
+ Done 30 September 2003