On 05/01/2013 06:47 PM, Rick Jones wrote: > On 05/01/2013 03:52 AM, Stefano Lattarini wrote: >> Hi Rick. >> >> On 05/01/2013 01:18 AM, Rick Jones wrote: >>> I have been asked if I was willing to eschew creating the .ps version >>> of the netperf manual. >>> >> Why? What harm do the existence of the "DVI -> PostScript" recipe do to you? >> Just avoid generating the PostScript file if you are not interested in it >> (I do that all the time ;-). But if somebody else wants to later create such >> a PostScript for his own reasons, he still can. The best of both worlds. > > Why is a good question. I am the messenger in this case. I'm not sure > that the .ps actually makes successfully for the netperf manual, but I > do not know that to be the reason for the request. > Ah, OK. We don't shoot the messenger ;-)
> I have been told there has to be some special handling for the .ps in > the Debian package building of netperf. That is all Dark Magic to me > though. > Sounds the same to me as well. > Indeed, a "make netperf.ps" fails - there are a boatload of messages > about over and underful hboxes, and the end is: > > Transcript written on netperf.log. > /usr/bin/texi2dvi: no such file or directory: netperf.dvi > make: *** [netperf.dvi] Error 1 > Stop, this is saying that the building of the *DVI* itself has failed. (Then of course one won't be able to build the PostScript: that is generated from the DVI...). My guess is that this isn't an automake issue, but a texi2dvi one --- I don't know whether due to a bug in texi2dvi or in the netperf's texinfo files. And there is another thing to consider: Automake 1.13 has removed support for any makeinfo/texi2dvi older than 4.9. If you are using a Makefile generated by Automake 1.13.x with a texi2dvi < 4.9, you might be out of luck. HTH, Stefano
