got_get_repo_commits iterates over the commits and build a queue for later display. My previous memleak fix moved the TAILQ_INSERT_TAIL *before* the logic to exit the loop and so it broke the pagination. This fixes it by delaying the insertion into the queue, but also by moving the check for the "previous" button after we've opened the commit. ok tracey

This adds a new helper function `gotweb_link' that outputs an HTML link with the parameter correctly escaped. ok stsp@

reported by the W3C validator.

due to an error in the order of the arguments to fcgi_printf that I've introduced last time I touched gotweb_render_header the title is now "/" (or SCRIPT_NAME) and one assets (the safari pinned tab svg) is a funny URL :)

This guards against {mis,ab}use such as `tog < /dev/null`. ok plus fix from op@

keeping GOTWEBD_REPO_CACHESIZE * sizeof(struct cache_repo) inside the struct server makes it too large for imsg on some platforms (linux at least.) Instead, store a pointer and allocate the array when it's received on the child processes. ok stsp@

Avoid crashing by corrupting the pointer instead of writing to it and fix how sym is generated: it shouldn't include the '=' sign. ok / improvement tb@

suggested by op@

ok op@

We fail to release the memory for the nodes. To fix it however, we some consumer of the commit graph iterator need to be corrected: the returned pointer is safe to be used only up until the next iter_next call; save a copy it if it's needed afterwards too. ok stsp@