Release notes for MECCA 2000 version 4.07. Oct/31/2005 Fixed: 1. "Text along Path" keeps a copy of the path used, but did not ensure the line weight is set to zero for all cases (m2kbin). And if the path was a simple vector, typdrv may sometimes print it (typdrv). 2. "Reset to defaults" in box and text dialogs did not reset color(s) (fcplist.tcl). 3. Corrected a long standing galeyc composer mis-handling of the merge- form LEAD value. To maintain backward compatiblity (i.e., old fp files would still work), the new fp file need to add a non-zero interger as the last field of the first line (galeyc). That is, the old format first line: pos# llx lly urx ury would ignore the LEADING field value(s), and treat it as LM for the compose measure. To have the correct (as documented in MECCA User's Guide, volume 3, Chater 9, on page 6) LEADING treatment, the first line should be: pos# llx lly urx ury 3 where the '3' can be any non-zero integer. 4. Gray scale raster images did not print because of a problem introduced in v4.06 (PSbasic.ps). 5. When text-along-path/circle-text elements were using a path other than line segments, typdrv mis-handled the text and sometimes caused memory addressing error (typdrv). 6. Sometimes PS conversion resulted in "PS Error:" without any explanation. This is likely due to the input PS file produced no graphical element after conversion -- e.g. it set color and/or screening attributes, but didn't "draw" anything. For such case, "No graphical element produced." will now be the explanation (psconvert.tcl). 7. Corrected a long time problem where a filled area using a long spline as border, typdrv would core-dump when printing the job (typdrv). 8. In the output PostScript file, jobname and Title string is made up from the parts/seps description text supplied by the user. All parentheses in such data are now replaced by '-' (short dash) to prevent potential PostScript syntax error (typdrv). Changed/New: 1. In v4.06 release, level-3 shfill (shaded fill) was used when printing a graduated screen on a level-3 device. We since found GNU Ghostscript 7.0x outputs a pattern fill which ignores any custom screen (e.g. Hline) setting. So a test has been added: if being processed by Ghostscript (which claims to be level-3), tradtional level-2 graduation is performed instead (PSgraduate.ps). 2. Batch compose: column definitions and reference now permit up to 16 columns (galeyc, pmunu, gpcdump). Below, "<=" means "less than or equal to": \COLXn: 1 <= n <= 16. \CD n,n,n \END: 1 <= n <= 16. Before v4.07, the upper limit was 8 for both. 3. When using Ghostscript to produce a PDF, "ps2pdf14" which generates PDF-1.4 format, is now used if found on the system (print.tcl). Also, there appears to be a problem with Ghostscript generating PDF: if grayscale raster downsampling is enabled, sometimes Ghostscript dumps core. Grayscale image downsampling is therefore disabled for the time being. 4. If "out_pslevel3" is set in a printer tray file, or "out_pslevel: 3" is set in the plot parms file, typdrv will output a raster image in its entirety (as a currentfile data stream), instead of the old method of segmenting the image into bands (in order to comform to stack size limit when setting the pixels as string data). This will help downstream processors, such as PDF conversion, to see the whole image as one entity (typdrv, PSbasic.ps, ps3form). Please be advised that many printers (bw and color) claiming to implement "Level-3 PostScript", do not work with this new raster image data format (different Ghostscript versions also behave differently). 5. Changed bcompose to add "-g" option to generate graphic (.g) file (bcompose). 6. The psexpand program has been updated to not look for "%%Creator: Mecca" comment in the PostScript file being processed, so that other tools may be used to create the PS file -- but the specific tokens psexpand looks for, remain unchanged (psexpand).