Scott, Barry, et al.,
Just a few comments.
First, I am very happy to see the return of the minor version release. While this might not have been intentional, it is a welcome site since that means that smaller issues are being addressed outside of the more major releases. I hope that following the bumps in the road on the way to getting a broadly stable 2.0.x release out the door that this can continue for the purpose of fixing bugs (leaving new non-emergency functionality for major version releases.
I also greatly appreciate knowing what the issues are. This was something that was nice about the old way that Declude worked. While Scott wasn't necessarily forthcoming with documentation and disclosure, with prodding and by communicating with others on this list, that information seemed to surface when it was vital to know. That has changed.
It would still be nice to know about all of the issues in slightly more detail. The release notes for IMail for instance are a good example of this. I learn a lot about the issues with IMail by reading them and it allows me to be more proactive in managing my software. Even if they don't affect me, it does disclose information that may help me down the line. As far as I am concerned, knowing about fixed bugs, even minor ones, is as important as knowing about new features. I can understand that from a marketing perspective new features are perceived as the most important by non-administrators, but truth be told, many of us are more concerned about avoiding and fixing bugs so marketing that information to us is just as important if not more. It would also be nice to know what the major changes were even if they are supposed to be transparent. For instance, this release obviously re-worked some facets of the software, and I would really like to know how extensive that work was so that I can better determine when I am comfortable in upgrading. Personally, I'm still not comfortable, and I probably won't be at least until there is a determination as to what was deleting the files on Andy's system along with the memory leak obviously being addressed that was exposed by this condition.
One other thing that also still concerns me is the new licensing scheme. It seems like this might have been one of the issues, any while you might be confident that it is working fine now and is supportable, I really want to know how it works from a flow-chart perspective. For instance, I don't want to run into issues during an outage where I am recovering from a backup onto a different server only to find that things aren't working because of some licensing mechanism. I know that this might trip Windows Server 2003 to require re-register with them, but it will work for 30 days without that. I have no idea if Declude needs to be immediately re-registered, how you might choose to block such things, and otherwise what ever else I might have to look out for such as server outages on your end. A recent experience of my own with BackupExec doubled the time of an outage because of a simple default setting in their software that didn't allow me to catalog imported backup media properly without changing. Something as simple as that is such a big deal that I decided that I can no longer trust BackupExec as a disaster recovery solution since this is definitely not the only quirk that I have come across and I don't have the time to be any more proactive in identifying these things ahead of time. I also really, really hope that you created a backdoor registration work-around so that in the event of something happening to Declude that this information could be distributed in order for us to continue using the software. Even TiVo is reported to have a backdoor that is closely held and so far secure that they can distribute to keep their units from becoming completely useless without the required updates (otherwise they could only be used to play previously recorded content).
So please don't be fearful in releasing minor revisions, in fact this is greatly desired and a sign of what I would consider to be responsible development of an app that is involved in a highly important service (E-mail). And please do start more fully documenting every release by identifying all changes, whether or not they were intended to be transparent.
Thanks,
Matt
R. Scott Perry wrote:
We've gone back to 1.82 as well.
We'll wait again until 2.0 is proven stable. Declude hasn't been like what
has been in the past.
Just to let people know a bit about this -- the source of the crash was identified pretty quickly. And a change could have been made almost as quickly to prevent the crash.
However, in this case, the D*.SMD files (the ones containing the E-mail body) were disappearing -- a situation that should (in theory) never happen. There are causes for this (such as an on-access virus scanner), but they aren't very common. So my advice was that rather than just fix the crash, further investigation should be done to determine why those files were disappearing. That way, we can have a new release that fixes the crash without running the risk of people noticing a new problem (that they weren't seeing simply because the crash occurred).
With software, there are often minor issues that come up that don't get addressed because they seem so minor or aren't being reported. Yet many times when this happens, a bigger bug appears later that would have been fixed if the minor issue had been dealt with right away. I've seen this a number of times with software I've worked on that nobody but myself runs. One out of a hundred times, something that isn't quite right will appear. The thought process is "Gee, it would be a good idea to look into that to see why it happened, but it would be a lot of work tracking it down; I'll deal with that later." With a program that only I run, that's fine. But for software that 1,000s of people are running, most of whom consider E-mail to be mission critical, I think it is best to wait and have it done right.
And, to take credit away from where it should be taken (or whatever the opposite of "giving credit where it is due" is), the crash is occurring in code that I wrote.
-Scott
---
Declude JunkMail: The advanced anti-spam solution for IMail mailservers since 2000.
Declude Virus: Ultra reliable virus detection and the leader in mailserver vulnerability detection.
Find out what you've been missing: Ask for a free 30-day evaluation.
---
[This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]
--- This E-mail came from the Declude.JunkMail mailing list. To unsubscribe, just send an E-mail to [EMAIL PROTECTED], and type "unsubscribe Declude.JunkMail". The archives can be found at http://www.mail-archive.com.
-- ===================================================== MailPure custom filters for Declude JunkMail Pro. http://www.mailpure.com/software/ =====================================================
--- [This E-mail was scanned for viruses by Declude Virus (http://www.declude.com)]
--- This E-mail came from the Declude.JunkMail mailing list. To unsubscribe, just send an E-mail to [EMAIL PROTECTED], and type "unsubscribe Declude.JunkMail". The archives can be found at http://www.mail-archive.com.
