It looks like newly created pdf file after pdf2ps && ps2pdf conversion can't be searched, and you can't copy text from it.
Hmm, well there can be many reasons for this, not least the fact that you appear to be using quite an old version of Ghostscript. I'd suggest upgrading and retrying, but if you post an example PostScript file I can try it and let you know the results. Also, if this turns out to be expected, why it happens. Note that the dual conversion really isn't a terribly good idea, as you potentially lose information on each conversion. In particular the pdf2ps script uses the pswrite (PostScript language level 1) device, rather than the much more capable ps2write device. Most likely the conversion via pswrite is discarding information. If you are unable to copy text this suggests that there is in fact no text in the document, merely bitmaps.
Created attachment 5195 [details] test_in.pdf test PDF created with OpenOffice 3.1
And now when I run `pdf2ps test_in.pdf && ps2pdf test_in.ps test_out.pdf` the test_out.pdf becomes unsearchable and uncopyable.
OK, quick work around to the problem seems to be adding '-sDEVICE=ps2write' option to the pdf2ps. But why even in the latest, http://svn.ghostscript.com/ghostscript/trunk/gs/lib/pdf2ps , version that device is not being used by default?
pswrite device converts everything to path and image objects. All text information is discarded. pswrite device is used in pdf2ps script for backward compatibility. The PostScript generated by pswrite device is rather simple and works on most printers including Level 1 ones. We have pdf2ps2 script that uses ps2write device. Even with ps2write device conversion is still lossy. PostScript has no transparency, and transparent objects are converted to images. If you want to convert PDF to PDF, you'd better do it in one step. ps2pdf src.pdf dst.pdf Yes, ps2pdf also accepts PDF input because Ghostscript auto-detects the file type. In short, Ghostscript works as designed.
all make sense now.. thanks!