I think the first step would be to understand the changes involved, as to
be able to assess the effort & risks vs. the advantages of doing such a
late change, just before release of FreeMind.
Dimitry Polivaev said:
> Hello everybody,
> this mail is addressed to Dan and Chris first, but I also ask Eric and
> Ryan to respond.
> I applied the SimplyHTML patch submitted by Dan. Dan has done a good job
> removing some shortcomings from SimplyHTML. I would like to integrate
> his patch into the head branch and even change the next SimplyHTML
> version from currently 0.12.5 to 0.13.1 because I think the changes are
> that significant.
> Dan wrote that his changes are well tested because he used the patched
> version for a long time. I could however see that his changes removed my
> last workaround for pasting corrupted html under linux. I restored the
> work around, but therefore I can not be sure that there are no new bugs
> introduced into the program. On the other side I can not easily localize
> changes and review them because the most changes are just simply
> refactoring like renaming of variables, and the most important changes
> can not be easily localized among them and are also not always evident
> for me.
> Therefore I created a new branch called shtml_2009090128_dan and
> submitted the patched code there and not to the head. Some source files
> where only trivial refactoring was done are committed directly to the
> HEAD, but all the files with less trivial changes are committed to the
> new branch. (Obviously the branch should be called shtml_20090128_dan ,
> but I typed "09" twice and noticed it a bit late. Anyway it is not a
> problem, it is just a mistake).
> And therefore I want to propose some different ways what we could do
> next for safe integration of the Dan's code into the main branch.
> 1. If Dan could provide a short description of his changes I could
> review them in the code side. Ideally Dan could take the latest version
> from shtml_2009090128_dan, insert comments in code describing his
> changes and mark them by e.g. // DAN or somehow else.
> If Dan would like to do it I would give give him a direct access to CVS
> of SimplyHTML so that he can enter his comments directly there.
> 2. If somebody else wants to test the new version I could compile it and
> create an unstable release in project SimplyHTML. And if Dan could
> describe what has been changed from the user point of view (change log)
> it could help to see what should be tested in more depth.
> 3. Creation of new unstable freemind version with the test version of
> SHTML is another option.
> And before I do any furhter steps I would like to ask your opinion.
> Best regards,
>> Hello Dimitry,
>> in my private version of SHTML, I have corrected a group of SHTML bugs.
>> Some of the most notable correction is the one of pasting of HTML copied
>> from external applications such as web browsers into tables and list of
>> SHTML; before the correction, pasting into a table or a list breaks the
>> table or list. There are many other corrections, including corrections
>> of issues with backspace from a non-list into a list, and issues with
>> cursor movement within tables. Also included are a few new functions
>> like moving of rows up and down within tables; these are well isolated
>> in their individudal methods, and can be turned off by removing them
>> from menus.
>> Are you interested in me creating a patch for these corrections? If yes,
>> I will create the patch against the current version of SHTML in CVS.
>> The corrections are well tested, as I have used my version productively
>> many hours a day for several months.
>> Best regards,
>> Keywords: Simply HTML
Eric de France, d'Allemagne et de Navarre