... Inf-HTML v0.8b ...

History of the Inf-HTML package
0.1a
First closed alpha
- Support for Table of Contents display
- Support for basic displaying of articles with no support for styles,
links, margins, etc. Pretty shocking.
- Real rough coding
0.2a
Decision was made for a rewrite to be done.
- Much better error control and reporting. Structured in such
a was that any errors will display a message, free all allocated memory,
and exit.
- More modularisation of code. Full commenting. Better structure.
- More comprehensive support for control codes. Styles implemented - bold,
underlined, italic, and monospaced examples.
- Kludging introduced for left margin.
- Graphic buttons introduced.
- All support files now must be located in the /inf-html/ directory.
- Debug mode seperated from main code. Cleaned up and improved.
0.5b 10/12/94
First public release... in initial beta testing
- Name change from html-inf to inf-html.
- Kludged support for inline graphics inclusion.
- Look ahead on word wrap to prevent large entries sticking
out and looking bad.
- Improved layout control to stop line breaks occuring when they shouldn't
have and to be there when they should.
- Support for really large articles that are larger that one slot. This
has not been extensively tested... I added 3 lines of code to do this and
it works fine on the 2 articles I tested it on.
- Major bug fix - incorrect titles on articles and links failing.
This has improved inf-html to the stage where I am prepared to release it
for public beta.
- Basic documentation included / being written (grin).
- Support for another type of in-line linking added.
0.6b 16/12/94
Second public release... beta testing continues..
- Changed the way Table of Contents displays. (all in plain text rather
than using heading levels which looked un-nice with WE 0.92mt onward)
- Included graphics are now transparent - including Inf-Html, Team Os/2,
and Dos inline graphics.
- Document title changed to reflect content instead of always
file - x:\...\myfile.inf
If displaying a Table of contents heading is
Inf-HTML: (doc title)
If displaying an article heading is
Inf-HTML: (article heading)
This produces much better WebMaps.
- Changed some (hopefully all) references to inline images to include
alternate text for text-mode WWW browsers.
- A change from WE 0.9 to WE 0.93 meant that IBM Extended ASCII characters
used in line-diagrams (for example the Os/2 CmdRef.inf) would no longer
display properly. Changes to code mean that (hopefully) &, <,
and > symbols should display. IBM Ext ASCII (from 169 to 218) has
been remapped to +, -, or |.
- Dictionary debugging added to articles using
readinf.cmd?<filename>+text+<article#>+2
This will show when character remapping has occured (from what to what).
- Index's supported if in the file. This hasn't been extensivly tested but
it looks like it works.
0.7 Not Publicly Released
- BUGFIX... due to my pathetically shallow knowledge of batch programming I
had the file
readinf.cmd
(when I bothered to include it!)
containing a line as such...
x:\path\inf-html.exe %1 %2 %3 %4 %5 %6 %7 %8 %9 %10 %11
for which arguments 10 and 11 are interpreted as 1 with a 0 after it...
I searched the code, using a debugger, for hours to figure out why some TOC's
were not being displayed properly... arrr...
- Code rehaul / cleanup to try and move from Borland C to ANSI C for
portability reasons.
- A "Conversion" executable was developed using precompiler
instructions. This changed a Inf/Hlp file into a series of Html files.
- Table of Contents display with unordered lists.
- Due to ongoing problems with the way lynx interperets <pre>
(ie. adds a carriage return) a decision was made to rewrite the article
display code to be fully monospaced.
0.8b 26/12/94
Third public beta release.... onward, upward, forward!
- A complete re-write and overhaul from the ground up. Expect some new
bugs due to this.
- Now makes extensive use of pre-compiler statements to support the
generation of INF-FILE, and DEBUG, from the same engine. Seperated into
multiple source files for better management.
- A semi-proper command line parser was added to stop really dumb things
causing SYS Errors.
- Displays all articles with mono-spaced. A much better (although more
complex, probably slower, and maybe bug filled) word wrap engine developed
(on Christmas Eve) which was a BIG code headache!
- Centered graphics are displayed with <center> that is supported
by WebExplorer and NetScape.
- No more little arrow graphics, or [to].
- Tested for display with IBM WebExplorer v0.95, NetScape v1.0 for
Windows, and Lynx.
- BUGFIX - Multi-TOC articles did not reset "space=true" which
caused some interesting results. This should not be reset if in the
middle of a preformatted example.
- BUGFIX (hopefully) - If two spaces occured after a fullstop at the
end of a line the next line was indented a space.
Future releases / wish list
- A search engine might be on its way... I have to find 2 minutes to
look at the search tables in the files.
- Support for centered text, right justified text, would be
1) nice, 2) a challenge!
- I was thinking another executable that did "to TeX" or
"to LaTeX" conversions would be nice. Still in though stage.
- It would be nice to have something to pull all the graphics out of
a Inf/Hlp file and turn them into GIF's. (Anyone volunteer?).
Known bugs/problems
- Some of these aren't really bugs 'cause I never said it would work with
all Inf/Hlp files and it doesn't ;-(
- The Os/2 Personal Magazine v1.01a won't go online.... You can sort
of kludge this by accessing it with
op101a.inf+text+0
- GBU v1.06 has some display problems.(Fixed... due to multi-toc bug)
- Any Hlp/Inf file that relies on IBM Ext ASCII is in trouble. Hacks
included from 0.6 onward to handle line diagrams.
Later, Peter Childs
INTERNET:
pjchilds@apanix.apana.org.au
FIDONET :
Peter Childs 3:800/352 (South Australian Country Club)