On 5/27/05, Ralf Stubner <ralf.stubner@...128...> wrote:

That looks ok in principle. The space character has a width of 377 units. The two lines with '74 304 rmoveto' and 'closepath' look strange and are unnecessary, but I would be surprised if they cause any harm. You can test this though by removing these two lines and assembling the font again with t1asm.

I added for tests a point to the space character, to see if it helps with the problem. These two lines probably describe that point.

Another wierd possibility is encoding. Maybe space is not encoded or in an unsual position in your font, and the application you are using uses code positions instead of glyphnames. So look out for /Encoding near the begining of the disassembled file. What does it say there?

/Encoding 256 array

If those ideas don't help, then i am lost. Can you make the font in question available so that others can have a look at it?

I made a new test font with the space, exported it as pfa and this time it worked. So I exported the original font again in pfa and, as previously, the space width was zero. So the program is able to generate pfa with non--zero space width, but for some reason either it can not do it for a particular font, or the font system does not interpret the particular font file correctly.

So I send the original PFA:

http://www.iitis.gliwice.pl/~arataj/test.pfa

Thanks, Artur