On Tue, Jan 6, 2009 at 4:30 PM, Freeland Abbott <gwt.team.fabb...@gmail.com> wrote: > Patch(es) not attached, so this is fairly high-level. ;-)
Doh! Attached for real now, with updates. > So long as your patch puts jarjar into tools, and you're just not including > the binary file 'cause patch can't deal and it's big anyway, that SGTM. If > you're proposing that folks can get jarjar themselves, that sounds bad. No, I just didn't want to send the actual jar around unless it seems necessary. The jar would go in the tools directory, under lib/tonicsystems. > I think it's a bit dangerous to assume that nobody doing development runs > ant (I still have old emacs/ant habits, for example, and occasional uses for > them that seem to make sense)... but I don't think that changes your > conclusions; if I had to, I'd just work off the -raw.jar or look up the > renaming. I didn't mean it as an assumption, but as a description of the most typical build setup. It's true that using the raw file directly is an additional approach that should be pretty reliable. > Note that my rules don't solve "all" of these problems; the first several > clauses are all about listing exceptions... most of them are from the JDK > (no problem), but tapestry is an outlier. Given that it seems to > dynamically build classnames, I don't see a jarjar-like way to address it... Yes, there are still some internal dependencies exposed. We can do more over time. If nothing else, we can develop a patch for tapestry to make it fit jarjar's assumptions. One step at a time, though. This patch already renames a large pile of classes. > SWT was another outlier; I take it you resolved that? No, I just used some rules of yours I found in a stray email somewhere. Indeed hosted mode is broken with the above patch, so I must have had an old set of rules. The attached patch adds SWT to the list of exceptions. > Given that we're in XML and thus can have comments, we should explain why > the exceptions are there Good point. I don't know all the reasons though! I have reasons for SWT and tapestry. Do you remember the reasons for the remaining ones? <rule pattern="org.xml.**" result="@0" /> <rule pattern="org.w3c.**" result="@0" /> <rule pattern="org.omg.**" result="@0" /> <rule pattern="org.ietf.**" result="@0" /> -Lex --~--~---------~--~----~------------~-------~--~----~ http://groups.google.com/group/Google-Web-Toolkit-Contributors -~----------~----~----~----~------~----~------~--~---
jarjar-tools-r4377.patch
Description: Binary data
jarjar-trunk2-r4391.patch
Description: Binary data