On Tue, 22 Jan 2002, Mike Tock wrote: > It seems I come across all these common problems for most of the 3.2 > versions of htdig. For some I have found temp fixes that allow me to > compile to the next problem.
Great. But sending about 8 messages to the list is just going to annoy people and make us want to dump your message to the bottom of the pile. Please see <http://www.htdig.org/FAQ.html#q1.5> Also please keep in mind that I don't know of anyone else who's tried compiling on Tru64 5.1, so there are bound to be some porting problems as you've seen. > It seems that there is something strange going on with the definition > of vnsprint. If you notice the configure script seems to know about > the system defined one. Yet when the make comes through the #define Is HAVE_VSNPRINTF defined in your include/htconfig.h file? > once by the system and once by the htdig code. Changing the > definitions so that they match seems to fix the problem, after looking > through the system definitions I found that off_t was defined as type > int anyway. Actually, if I remember correctly, one problem is that Digital Unix shouldn't be using the included regex.* because it causes add'l problems. We're working on a more bulletproof regex test for the configure script. You might want to try <http://www.htdig.org/FAQ.html#q5.14> as a workaround. > There also appears to be some conflict with the system defined macro > stat() in sys/stat.h which accepts only 2 params and the one defined ... > htword/WordDB.h > inline int Stat(void *sp, void *(*db_malloc)(size_t), int flags) { > if(!is_open) return DB_UNKNOWN; > return db->stat(db, sp, db_malloc, (u_int32_t) flags); > } This is very strange--it doesn't happen on any other system I've tried compiling. The sys/stat.h takes 2 parameters, but this is clearly not a call to a system library. You said you're using gcc 3.0 (and not 3.0.1 or 3.0.2)? I never actually used gcc 3.0 since by the time I could get gcc-3.x to compile on this RH box, it was already at gcc-3.0.1. > Server.cc:329: no matching function for call to `URL::URL(const URL)' > ../htcommon/URL.h:27: candidates are: URL::URL(URL&) > ../htcommon/URL.h:26: URL::URL(const String&) > Server.cc:329: initializing argument 1 of `void URLRef::SetURL(const URL&)' > from result of `URL::URL(const String&)' ... > Server.cc:331: initializing argument 1 of `void URLRef::SetReferer(const > URL&)' from result of `URL::URL(const String&)' I haven't the faintest why this is causing problems for you. The current CVS tree is shifted slightly, but judging from the error messages, this would correspond to the Server::push method: URLRef *ref = new URLRef(); ref->SetURL(path); ref->SetHopCount(hopcount); ref->SetReferer(referer); _paths.Add(ref); Now path is an argument and in my declaration of this it's const String &path, so there shouldn't be an error unless the compiler is mis-parsing this. Try changing the text of 'path' in this method to 'p1' or something like that and see if it works better. -- -Geoff Hutchison Williams Students Online http://wso.williams.edu/ _______________________________________________ htdig-general mailing list <[EMAIL PROTECTED]> To unsubscribe, send a message to <[EMAIL PROTECTED]> with a subject of unsubscribe FAQ: http://htdig.sourceforge.net/FAQ.html
