2009/8/17 Miguel Enrique Cobá Martinez <[email protected]> > El lun, 17-08-2009 a las 18:13 -0300, Mariano Martinez Peck escribió: > > > > > > 2009/8/17 Miguel Enrique Cobá Martinez <[email protected]> > > El lun, 17-08-2009 a las 16:42 -0300, Mariano Martinez Peck > > escribió: > > > Hi folks: I think I found a little bug with the Author > > refactor. To > > > reproduce: > > > > > > 1) Open Monticello browser > > > 2) select a package you modified something, select a > > repository and > > > press save > > > 3) The popup asking you the full name and surename is > > opened. The > > > thing is that if you press cancel, it closes the popup but > > another one > > > is opened :( yo have no other choice that put "ok" > > > > > > > Yes, this is how the original prompt for initials worked > > before the > > rewrite, I didn't modify it. > > The thinkg is, it keeps asking until you enter something. > > > > If you could press cancel, when you do other modification, it > > will > > prompt again, so I prefer to keep the old behavior of a > > continuous > > asking for full name. > > > > But this could be changed, of course. What should be the > > behavior? > > > > The behavior must be just cancel. Close the popup and do nothing. Not > > to open the popup again. > > > > > > But this maybe can brake other parts, as the generation of signature for > methods. Supposing that the Author class let the cancel proceed. This > will necessarily to store nil or '' in the fullName inst var of Author. > the calling code (in this case, the evaluate part of the browser or the > debug windows where you modified some code) won't know that the author > hasn't stored the name correctly, so it will procedd as if were. Then > when saving the method/exporting to mc/generating the signature of the > method maybe will go on the assumption that there is a name stored. > Well you get the idea. Maybe I am exaggerating but will be necessary to > review the code using Author to see what will happen if a '' or nil is > stored there.
Ups.... You are right. Sorry, I was thinking funtioncallity, not programatically. I would love that when I press "cancel", the generation of signature and all the next code isn't called. > > Of course also can be true that nothing wrong happens (at least with a > nil stored the next time it will prompt again for a full name). > > What say the community? :) > > > > > > > > > Can anyone reproduce this ? If so, I'll open a bug ticket. > > > > > > Best, > > > > > > Mariano > > > > > _______________________________________________ > > > Pharo-project mailing list > > > [email protected] > > > > > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > -- > > > > Miguel Cobá > > http://miguel.leugim.com.mx > > > > > > > > _______________________________________________ > > Pharo-project mailing list > > [email protected] > > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > > > > _______________________________________________ > > Pharo-project mailing list > > [email protected] > > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project > -- > Miguel Cobá > http://miguel.leugim.com.mx > > > _______________________________________________ > Pharo-project mailing list > [email protected] > http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project >
_______________________________________________ Pharo-project mailing list [email protected] http://lists.gforge.inria.fr/cgi-bin/mailman/listinfo/pharo-project
