From koffice-devel Fri Oct 28 19:48:15 2005 From: Tomas Mecir Date: Fri, 28 Oct 2005 19:48:15 +0000 To: koffice-devel Subject: Re: OpenOffice.org bloat article Message-Id: <492258b10510281248l44330829occ2ebca9bc787491 () mail ! gmail ! com> X-MARC-Message: https://marc.info/?l=koffice-devel&m=113052891008395 On 10/28/05, Robert Knight wrote: > > There are so many things we can do, if only people actually started doing > it.. > > But what things exactly and how would they be useful? This would need > some careful planning to be worthwhile. > > > Save is already implemented by a serilalized XML writer though, if I > > understand it correctly - so loading shouldn't be that hard ... We > > just need to analyze whether we need some information that gets saved > > later on than where we need it - if not, then load should be quite > > straightforward; and even if we do, we could make two passes or use > > some other trick, still better than DOM-based approach. > > So if I understand correctly, KSpread currently loads the entire > contents of the XML files into various memory structures before it > parses them? Ouch! Exactly. It was already discussed in the past, but nobody has done anything with it yet, because it's quite a big job. But it surely is something that needs to be rewritten. There's some bug report about it too in the bugzilla. / Tomas _______________________________________________ koffice-devel mailing list koffice-devel@kde.org https://mail.kde.org/mailman/listinfo/koffice-devel