Hi Mary and all >> BayesIt functionality is very important, you're right. But who knows what >> else did break with the new version. BayesIt error was detected a few days >> later after the release - not after a few hours.
> I think that Natasha spotted it within a few hours--unfortunately, she put it > in a correction, posted to herself, of her earlier post to Maxim saying .31 > was running fine. Rather than directly posting the bug to Maxim. Quite so. I wanted to recant my earlier acceptance, and in so doing I made it more difficult than it should have been for Maxim to notice my report. It would have been more sensible to respond directly, quoting my original MID. BTW, anyone have a handy macro to quote one message but reply to the initiator of the thread to which it belongs? (Actually, what I'd like is something similar to the PUT macro, but for MIDs, but I don't think this can be done with QTs.) >> "Sufficient time" - that's the expression, that is interpreted in very >> different ways from all of us. ;) > I agree with you that sufficient time should mean at least one day--especially > given how the beta testers are scattered across all the time zones of the > world. Agreed. >>> You may be sure that if he doesn't, he will hear from, at least, me. I am >>> trying to look out for Maxim's and for RitLab's best interest all the time, >>> wearing both my hats, as customer and as beta tester. :) >> Nice Mary. I'll stand with you. > Yay! :girlcheer: Count me in, too. >>> 9Val is a developer, isn't he? >> It sems to me he is one. A very good and competent one. > He's my hero! I take my hat off to 9Val, Maxim and the entire crew! I've seen a few recent comments worrying about the fact that sometimes fixes to one aspect of TB can 'break' seemingly unrelated aspects. FWIW, my personal opinion is that this isn't surprising. As I think I may have rambled about at other times, the complexity of the Windows APIs (especially since TB supports multiple Windows versions) and various email standards the guys are trying to work with makes development rather complex and (extrapolating from experiences on other projects) perhaps forces them into workarounds which may later prove unstable. Which explains why what 'seems' unrelated isn't always what 'is' unrelated. The team are also working under time and (presumably) budget constraints, which means they perhaps don't always get the choice of what to fix or even how to fix it. So it's only when these instabilities arise that they can go back and repair those parts of the design which are now causing the problems. This is also why some things which may seem 'trivial' at the time are repaired before other, apparently 'more important' items are looked at: those 'trivial' items may actually be prerequisites (direct, apparent or otherwise, to repairs of the larger items. And this is where we, as the beta team, come into the picture: users add that extra level of complexity (as if there weren't already enough :) ) that can almost never be accounted for in the lab for such a complex project, no matter how hard the team may try. We use different features of TB, different mail servers and protocols, and our cultural differences give us different perspectives on what's important to us in an email program. So when Maxim asks, "Is 3.5.31 generally better than 3.5.30?", he knows RL think so, but also is aware that the diversity of beta testers makes us a much more demanding jury. He's also signalling to us that we're important to RL as well as acknowledging previous hastiness in uploading files to the public download site. I applaud him for all these reasons. >> It was just a note, that me didn't really understand the question, nor it's >> origin. > It will help all the developers who post here, and they will perhaps be able > to tell us what they want in a little bit more detail, in the future. Agreed. Detail is as valuable in questions as it is in answers. -- Groetjes Natasha The Bat! 3.5.0.31 on Windows XP Professional 5.1 Build 2600 Service Pack 2 ________________________________________________________ Current beta is 3.5.31 | 'Using TBBETA' information: http://www.silverstones.com/thebat/TBUDLInfo.html IMPORTANT: To register as a Beta tester, use this link first - http://www.ritlabs.com/en/partners/testers/
