Released now.

Any update on when we can expect Mercury 4.7.  It's been in Beta for a while now.  Any progress?

Thanks

Yes there has been a number of betas, and last two weeks with complete installer.

We put in production yesterday what I believe is Davids' final build of 4.71. It was released just a few days ago

About nine month ago it was announced that the beta test is going to be opened to the public:

[quote user="David Harris"]After consultation with my closed test teams, we've reached the conclusion that the community and the quality of the programs would be best served if we opened up the beta testing process: making betas available publicly would allow wider testing, as well as giving a clear, visible indication that the programs are alive and in active development.[/quote]

Has this decision been withdrawn for some reason or does 'final build' mean a public beta?

Best regards,

Nico

David wanted to release one of the betas as public, just before he left for a vacation (don't quite remember when that was, a few months ago I guess) but, that build had some issues that would have caused more trouble than the benefit a public beta would. With the build I reported as I believed would be the final 4.71, there are some minor questions and touches left to deal with. As I'm writing this, I know David is contemplating on some of these questions before deciding on either a public beta or a formal release.

Since I started testing the native service mode in production - one server has had that for over a month now, I've not experienced any hickups at all. But, I run a very clean server and a clean native Mercury setup, as stability is the most important factor. I know that Rolf, who has a few daemons in play, has had one or two service restarts in that time, which is one thing that we keep an eye on. For all the other testers, that run Mercury as a normal Windows App with GUI, none has reported issues with the latest build - so - with a little yellow flag about native service and daemons, in my view, Mercury 4.71 is ready.

Will you give some hints about native service mode and Vista?  As I understand it, in Vista, services run on desktop 0 and cannot interact directly with the user's desktop.  If that's true, I'm imagining and hoping for some front-end, probably using IPC to communicate with Mercury/32, to give the user access to, at least, the various configuration dialogs and, at best, a full-blown GUI for monitoring the activity of each module.  Please comment on how it will work.  Thanks. ... Vince

One of the most attractive uses with Mercury is the visible action of mail flowing through each modules' console window.

It is true that running Mercury as a native service under Vista/Srvr2008 has issues in running with interactive gui, due to the new Windows security model. David has though created the base foundation to communicate with a Mercury installation in near real time through a new module, MercuryR. But this module is to my knowledge not anyway near completion, so we face to run under a native service, for all cases, with the extreme power that the disconnection of a gui gives, and for managed mode - effectively when an admin actually logs on to a server, and starts the desktop to do gui interaction, the best recommendation, is to stop the service, and start Mercury by double clicking the icon. When management is done, you shut down Mercury gui, start the service and log off.

If you have lots of day to day tasks with manual intervention, you will get by easier until MercuryR  reaches us all, by using MercuryB for list management, and Rolf Lindby's HttpServer.

Mercury 4.71 has been running rock steady here in production since april 20th.

Currently Symantec is actually preventing a release, as the release archive .hlp files are trapped falsely by their heuristic engine - and Symantec responsiveness seemes to give plenty to wish for.

Thanks for the clarification, Peter.

Not that I want to delay 4.71 further but I'm having a issue with my ClamAV mailinglist and Mercury 4.62 that is, if a bounce comes in to VERP (I'm using MercuryC and sometimes my ISP's SMTP is listed at dumb dnsbl's like SORBS) as long as Mercury is still busy processing outgoing mail to the list, Mercury crashes in the same moment the bounce arrives. I've seen this repeatedly meanwhile so it would interest me if the issue was already addressed (if it's already known)?

Thanks and best regards,

Nico

Pls update me with your issue when 4.71 is out.

[quote user="Peter Strömblad"]

Pls update me with your issue when 4.71 is out.

[/quote]

Looks good. The 4.71rc survived two mailinglist runs with the usual bounces without a crash :)

Thanks and best regards,

Nico

Glad to hear it works ok.

To those wondering; there remains one issue with address completion in 4.72. The address completion is currently being tested.

I'm hoping the next release isn't all that far away now.

4.72 looks like it will never happen then??? does this mean 4.71 now has a bug that will never be fixed?

David has found the reason for the address completion problem, and it seems there is a good chance for release of Mercury 4.72 quite soon.

/Rolf 

[:D]

[quote user="Rolf Lindby"]

David has found the reason for the address completion problem, and it seems there is a good chance for release of Mercury 4.72 quite soon.

/Rolf 

Quick update: A fixed version is now being tested!

/Rolf 

[Y][quote user="Rolf Lindby"]
Quick update: A fixed version is now being tested!
/Rolf 
[/quote]

When [:D].... or never [:'(]?

The new address completion feature appears to have partly broken functionality in SpamHalter for messages arriving via MercuryD. Lukas is having a look at it, though, and I hope there will be a fix shortly.

David has written an important open letter to the user community, and in case you didn't already read it you can find it here:

http://www.pmail.com/openletter.htm

/Rolf 

any idea if finally the new version 4.7 will be delivered???

The SpamHalter/MercuryD issue is fixed so it should be very soon.

/Rolf