Hi, I am setting up a set of test files to check if there is some problem with pdftex. Those test files should be run after each change made to pdftex, hopefully to detect some errors before they slip to end users. Each file from the set is tested as follows: - recompile the test file - check whether the pdf is ok, using as many as possible available tools. At the moment the tools being used are listed in the script checkpdf I posted, plus pdftool from fitz. The output of such a check would look like: ,-------- | Checking microtype.pdf using pdfvalidator... | PDF library evaluation version, created Dec 2006 | Checking microtype.pdf using Multivalent Validate tool... | Checking microtype.pdf using Ghostscript... | Checking microtype.pdf using pdftoppm (xpdf)... | Checking microtype.pdf using pdftool (fitz)... | Printing microtype.pdf to microtype.ps using Acrobat Reader... | Checking microtype.ps using Ghostscript... | Printing microtype.pdf to microtype.ps using pdftops (xpdf)... | Checking microtype.ps using Ghostscript... | done. `-------- [Next version of xpdf will allow checking pdf using xpdf in "batch mode", ie without user interaction. This should catch more errors than pdftoppm] - then the pdf file is compared against another (verified) pdf to see if there are any differences. The comparison is done by: - using pdftoppm to print all pages from both pdf to images and compare the image (using diff) - if the images in the above step differ, use the 'compare' tool from ImageMagick to compare the images. If the difference is "big enough", the pages will be considered as "different". - at the moment the set consists of the following files: ,-------- | encguide - the latex encoding guide | microtype - the microtype manual | MnSymbol - the MnSymbol manual | mscore-sample - sample of mscore (TrueType) fonts | survey - the free math font survey | survey-vn - Vietnamese version of the survey `-------- Your comments are welcome. Thanh PS: while testing those files, I found one issue with pdf inclusion: if the included pdf contains any type3 font, the result will fail to print to PS by acroread (type3 fonts in XOject forms are not exported to PS). Maybe we need to write some warning if a type3 font is found during pdf inclusion.