0007393: Avoid #mousePoint to make cryptography Pharo-friendly

Notes
(0014060 - 2202 - 2610 - 2738 - 2738 - 2738 - 2738) Ron 03-10-11 14:44	Hi Daniel, Any source for Entropy is optional. The more we have the better. You will notice that there are a number of different sources for entropy. #responsdsTo: is probably what you are looking for. Adding a source that is comparable in Pharo that is missed by Squeak and leaving one that works in Squeak that is missed by Pharo is a very good solution, in my opinion. Since the methods really only collect a seed they are called very infrequently. Adding a check for respondsTo: will not harm performance in any significant way. All the best, Ron > -----Original Message----- > From: Daniel Lyons [mailto:fusion@storytotell.org] > Sent: Thursday, March 10, 2011 2:10 AM > To: Ron Teitelbaum > Subject: Re: Cryptography package > > > On Mar 9, 2011, at 9:22 PM, Ron Teitelbaum wrote: > > > Hi Daniel, > > > > The issues is probably from the changes in Pharo vs. Squeak. I would > > imagine that the change is in Random where we are gathering entropy to > > try to seed the generator. The best thing to do is file a bug on > > mantis at bugs.squeak.org. I'm not exactly sure how to tell if I'm on > > Pharo or on Squeak but it would be nice to be able to support both. > > You could upload your patch to the bug. > > > There's already an issue for it: http://bugs.squeak.org/view.php?id=7393 [^] > > The code is in RandomGenerator as you supposed. I'll see if I can figure out > how to tell which kind of image code is loaded into. Sensor cursorPoint > doesn't do the same things as Sensor mousePoint, apparently, which is a > shame because Squeak has this method too. I think the selector we actually > want in Pharo is peekMousePt, which also returns a point that changes when > I move the mouse. I don't know how to do this, but I believe it should be > possible to look for the existence of this method and not the other at > package load time and patch it. > > In thinking about it though, I'm not sure it's a good idea to rely on this > anyway, since there's not going to be a mouse position in a headless session > such as on a server anyway. Is there? Forgive me, I'm very new at Smalltalk! > > Thanks again! > > — > Daniel Lyons

(0014069 - 180 - 192 - 192 - 192 - 192 - 192) DanielLyons 03-11-11 01:41	This patch uses mousePoint if it can, or peekMousePt if that doesn't work. Otherwise it uses 0@0. That should make it reasonably portable. Tested under Pharo 1.1 and Squeak 4.2.

Issue History
Date Modified	Username	Field	Change
09-02-09 15:12	nicolas cellier	New Issue
03-10-11 14:44	Ron	Note Added: 0014060
03-11-11 01:40	DanielLyons	File Added: Cryptography-DanielLyons.37.mcz
03-11-11 01:41	DanielLyons	Note Added: 0014069

Relationships