On Mon, 12 Jul 2021 at 17:01, Gavin Smith wrote: > > On Mon, Jul 12, 2021 at 4:04 PM Jonathan Wakely via Gcc <gcc@gcc.gnu.org> > wrote: > > GNU Hello has the same problem with its docs: > > https://www.gnu.org/software/hello/manual/hello.html#index-_002dg > > That URL is garbage because of the URL-encoded %2d character, and the > > fact it links to the wrong place (the description of the option, not > > the option itself). The former is no longer an issue for GCC (it was > > for many years) but the latter is still a problem. > > > > If you don't know where to find it yourself, the source is visible here: > > https://github.com/yugui/example/blob/master/doc/hello.texi#L208 > > I downloaded the source for the "hello" manual and recreated it with > Texinfo 6.8 (running " texi2any --html hello.texi --no-split"). I've > attached the results. The current output doesn't exhibit the problem > with the scrolling being at the wrong place - this problem has > evidently resolved itself since the time when the online "hello" > manual was generated. (I don't remember many complaints about it on > the mailing list, though: if we don't know about problems, we can't > fix them.)
The "copyable link" does work as I would expect. The #index-_002dg anchor still seems to be in the "wrong" place, i.e. in the <dd> element not the <dt> element. But the addition of the copyable link nicely solves the problem of needing to easily obtain a link to the right position. > The URL is mangled because index entries can have more characters in > them than what is suitable for a URL. A space character becomes a "-", > so a "-" has to become something else. Yes, I understand the reason.
