Giacomo wrote: > Gianugo Rabellino wrote: > > giacomo wrote: > > > On Mon, 26 Nov 2001, Giacomo Pati wrote: > > >>Quoting Gianugo Rabellino <[EMAIL PROTECTED]>: > > >>>Gianugo Rabellino wrote: > > >>> > > >>> > > >>>>>>I just did a fresh checkout. This is Linux Mandrake with Blackdown > > >>>>>>1.3.1. I'm getting plenty of entity resolution problems and I'm > > >>>>>>investigating them. More on this later. > > >>>>>> > > >>>>>> > > >>>>>Cool, thanks. > > >>>>> > > >>>>>BTW: You don't get any exceptions during the docs build? > > >>>>> > > >>>> > > >>>> > > >>>>Ah-ehm... I was in a hurry and didn't realize that I was building the > > >>>>HEAD branch version :) I just installed a fresh Linux and I'm getting > > >>>>the correct repo right now. I'll send the result in a short while but > > >>>>anyway yes, what I got on the HEAD branch were entity related > > >>>> > > >>>exceptions > > >>> > > >>>>(the last one was about ISOnum.pen not found). > > >>>> > > >>>>I'll send my results in a while. > > >>>> > > >>>> > > >>> > > >>>OK, looks like I have good news.cocoon_20_branch is building fine on my > > >>>Linux box. I'm running Mandrake 8.1, kernel 2.4.8, Blackdown 1.3.1. I > > >>>get no exceptions at all, and everything seems to be built just fine. > > >>>Now the problem is to understand why yours is not building. Here are my > > >>>relevant data: > > >>> > > > > > > So, here is my results. > > > > > > 2.0 BRANCH: > > > > > > 1. fresh checkout > > > 2. done a 'build docs' on this vergin checkout > > > 3. lots of 'Failed to create InputSource:...' > > > > > > Same here, but it looks like a simple warning. > > Ok. > > > > > > > > 4. but at the end it reports a 'INFO 10067 [ ] (): Done' :) > > > > > > So, there seems to be happened something to my old repository which I > > > don't know ATM. > > > > > > HEAD BRANCH: > > > > > > 1. checkout a fresh HEAD branch > > > 2. done a 'build docs' > > > 3. lots of 'Failed to create InputSource:...' > > > 4. but also a 'ERROR 10067 [ ] (): Could not read resource > > > file:/home/giacomo/cvs/xml-cocoon-21/build/cocoon/documentation/xdocs/plan/index.xml' > > > > > > I went through this: the plan/index.xml file has an incorrect DTD > > definition (it should read ../dtd and not ../../dtd).
Yes, this is incorrect default systemId (i will fix it). However, this is one purpose of the catalog entity resolver - to determine alternate system identifiers of centralised DTD copies. > > I was about to > > patch it but I was struck with a lot of errors regarding ISOpen.num. That is just an artefact of the previous problems. The entity resolver (if it was working during build docs) would find ISOnum.pen > > I realized that I was going to mess up woith David's work so I just gave > > up with the idea of letting him investigate about the whole issue. > > Any comments here, David? I had already documented into Bugzilla everything that i knew about the issue. All that i see happening with the discussion so far, is that people are discovering anew and reporting stuff that has already been documented. Today i did a stack more work on this issue. I think that i have found the reason for the problems. See the separate thread: "found the cause of Bug #5060 build docs". --David > > Anyway I'd focus on the 20_branch as of now: I think we can solve the > > HEAD problem later on... The same issue is in both HEAD and cocoon_20_branch. > Yes, let get this beast out RSN. > > Giacomo --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, email: [EMAIL PROTECTED]
