home *** CD-ROM | disk | FTP | other *** search
Text File | 1995-07-31 | 6.1 KB | 120 lines | [TEXT/R*ch] |
- HTML Markup Version History
- ---------------------------
-
-
- HTML Markup ALPHA versions (5/9/95)
- ==========================
- The alphas implemented some of the basic filters like use first line as
- <TITLE>. They had a 3 meg memory partition for safety because they had
- no solid memory management technique. Unlike more recent versions, they
- did not estimate the size of files and allocate memory appropriately.
- Instead, they used the formula of adding a small buffer to the size of
- the original file. It worked surprisingly well and was fast, but
- unpredictable for situations where the formula didn't work. Very little
- of the ISO 8859-1 character set was implemented so while some characters
- like < and > would be converted, other international characters would
- not. These versions expired on 6/1/95.
-
-
- HTML Markup 1.0b1 (5/11/95)
- =================
- Version 1.0b1 was almost identical to 1.0a2. There were some minor bug
- fixes, interface changes, and documentation changes. This version expired
- on 6/1/95.
-
-
- HTML Markup 1.0b2 (5/22/95)
- =================
- Version 1.0b2 fixed a few minor bugs. The biggest change was that it was
- the first PowerPC-native version of HTML Markup. Version 1.0b2 and all
- subsequent versions were distributed as Fat binary applications. Speed
- improvements on the PowerPC were only about 15% with a large number of
- files. I attribute this to the fact that the file system is not native,
- so in theory Copland should bring a big speedup.
-
-
- HTML Markup 1.0b3 (6/3/95)
- =================
- Version 1.0b3 fixed a few minor bugs, once again. There was a problem with
- memory allocation that caused the shareware message to appear more than
- once in the output. Also, words separated by a carriage return were squished
- together. This was fixed. The memory allocation algorithm was modified to
- double the size of the original file and then add a 10k buffer, making it safer
- but still unpedictable in strange circumstances and incredibly inefficient
- for others.
-
-
- HTML Markup 1.0b4 (6/24/95)
- =================
- Version 1.0b4 was the first really stable and powerful version. On the
- technical side, it implemented an all new estimation algorithm to replace
- the old memory allocation algorithm. This new technique would run through
- the conversion twice. The first time, it would figure out how much memory
- it needed for the output file. The second time, it would actually write
- the file. I added more escape code conversions, implementing all the
- essentials from the ISO 8859-1 character set. I added a "smart list"
- algorithm that replaced the old list conversion function. With the
- old algorithm, any word with a dash in it would be converted into a
- list. The new one only treats lines that start with a dash as list
- items. Version 1.0b4 introduced the output ticket button. The output
- ticket combined all the output options from the main job ticket with
- a few extras like add .html to the name, and setting the creator code.
- Two new options on the main job ticket allowed converting dashed lines
- to <HR> flags and keeping inline HTML flags.
-
-
- HTML Markup 1.0b5 (7/24/95)
- =================
- The output ticket now allowed you to convert spaces in the names of output
- files into underscores and/or convert the file names to lower case.
- If you were a paid user of HTML Markup, you received instructions on
- how to enable the registered features. You could automatically convert
- any standard URL's in your document into clickable links in version 1.0b5.
- A new option turned lines beginning with # (pound sign) characters into
- items of an <OL> (numbered list). A new option created an index.html file
- with links to all of the files that Markup created. You could now have
- two-level lists. If you simply indent a list within another, Markup would
- make it into a nested <UL>/ The job ticket was rearranged so it would fit
- on smaller screens and have some sort of order to it. A dialog box was
- added that tells users how to properly run HTML Markup when the double
- click the icon instead of dropping files on it. I set default items for
- all of the dialog boxes to "ok". As the conversion algorithms for Markup
- grew more and more complex, I began to wonder if the algorithm that
- estimates the final file size would hold up. It appears that it did, but
- just to be sure, I increased the buffer size from 500 bytes to 5 kilobytes.
- If you used the add .html option in 1.0b4, there would be problems with
- files that already had long times. Markup will now truncate it.
- Pressing cancel on the job ticket will now quit immediately instead of
- showing the final report. In fact, the button has been renamed "quit".
- If the end part of your source file was a list and you had Markup convert
- lists to <UL>'s, it would cause the shareware message to be indented
- because the list was never ended. Markup will now end all lists at the
- end of the file. If you had special characters like ¢ or ß in the title
- of your document, they would not be converted to the proper codes in
- earlier versions. They now will.
-
-
- HTML Markup 1.0 (8/1/95)
- ===============
- For the registered version, there is now a popup menu that allows you
- to select the colors. Previously, you had to know the RGB values. There
- is a tradeoff though. With this method, you only have a selection of 11
- colors. There is also a popup menu in the output dialog for selecting
- the file creator. You can still enter it manually, but you can also choose
- from Netscape, BBEdit, Microsoft Word, Jot, or SimpleText. I added <BR>
- tags to the index.html file. A bug in earlier versions caused all the
- links to be on one line! One user reported that he used URL's within
- sentences and when Markup converted them to links, it included the trailing
- periods. I recommend that you always enclose your URL's with < and > but
- if you use them in sentences, Markup will now recongnize that and treat
- it appropriately (i.e. it will ignore the period). I got rid of some
- extra /r (carriage returns) that I had originally put in for debugging
- purposes. This makes the output look nicer, especially for lists,
- embedded lists, and numbered lists. I also added support for two more
- required AppleEvents, namely OpenApp and QuitApp. It's pretty pointless
- to add PrintDoc because Markup can't print anyways.
-
-
- --Scott J. Kleper
- 8/1/95
-