You guys making lists like that – would any of you venture a guess on how long it would take to redo 30 fish from scratch, figure out their individual behaviour and incorporate that into the screensaver.

Compared to that, whipping up the lights and finish the crystal plus image importing is a breeze, but how long would you say it should take Jim? A month? Three? A year?

P.S.
A "Check online to see if there is an updated version" button is the one thing missing – and if you phrase it like that, there's little risk of people being angry the program goes online since they specifically asked for it to do so by pressing that button.

I did not say that it would be quick and easy to incorporate those items but Jim's question was do we thing version 3 is ready for public release and my own opinion is no, not until the items I listed have been added.

Obviously it is up to Jim whether he releases now or not or waits until he has added some other stuff.

I just wrote my opinion and thats it.

For what it's worth, I think finishing up logo support, automatic lighting ala 2.6's and some sort of update mechanism would bring it up to a point where it could be released. I'd have to also agree on using a manual "Check for Updates" button approach. I'm rather particular about what programs I allow to access the internet automatically.

I think my solution would address the issue you're having, at least indirectly.

I think the issue is "Can this be released before Christmas?"

Or, money? :)

Michael, - I agree with your sentiments, (Jim has to eat, etc!), but he has asked for our opinions.

Jim is known as a perfectionist, so why break the habit of a life time?

After the initial 'wow' factor has subsided, and the panning is paused or stopped, where is all the magic 3D effect? - Nothing is moving. - The starfish doesn't come on the glass. - The lighting isn't as adjustable as 2.6. - There is no clock or calendar. - The logo isn't working yet. - The fish are not as good as 2.6, and their movement is unconvincing. - And one of the major features of MA3 is its extra width, (enough to span three monitors), yet the bezel size adjustment has had to be put on hold, - Passthroughs, (ie... what all 'other' programs suffer from, but never MA!), we have been asked to stop reporting! - The jumping window is something I have never witnessed on my system, but it seems to be a major problem for those that are affected. - But if we are to accept that the increased width of MA3 is at least partly to accommodate multi-screens, then if possible, I think the FPS of the window should not drop so low when it's not solely on the primary monitor.


I agree that it probably can't all be done, or it may never get released, - so there has to be some compromise. - Where that line is drawn is obviously up to Jim. - But in my opinion, it would be wrong on a number of levels, to send it to market, "as is".

I don't think it is a matter of it taking another month, three months, or a year. - It's what effect marketing an unfinished product may have.

It's stable enough and working well. And I think I already said to get this out before "Windows 7" ships, so deferring major features (fish, background movement, etc.) seems perhaps reasonable. The same for deferring fixes of multiple-monitor annoyances.

However, in my opinion:

The "bugs" (omissions and comissions) in the "About" screen, which is one of the FIRST things that a new user will encounter, will seem amateurish, if they aren't fixed.

The lack of the clock/calendar will seem amateurish, and will disappoint all former users of 2.6 - after they see the "upgrade" has less functionality (if that's not added).

The "pros" will be disappointed that there's no good way to customize the logo, if that remains the case.

And "automatic update checking" is a feature of all good software, today - particularly if you Already Plan Update Features (tm).

Finally, consider how the magazine and online reviewers will react - partticularly after they read this forum. Do you want: "SereneScreen ships MA3 with known major problems?"

Yeah, Jim has certainly done wonderfully, and of course he deserves to eat. So, apparently, SereneScreen has a problem - apparently invested a lot of time and money, new upgrade nicely almost-completed. What to do?

Make a smart business decision, folks.

2 or 3 years. Now, I've had the apparently-mistaken impression that the work to do that, started 4 or 6 years ago. But we've never been told that. So, if it starts tomorrow - yeah, nobody should wait. So, you guys asking questions, perhaps you could tell US "how long would it take", so we have a chance of giving advice that's reasonable.

OK, you asked for my guess: less than a month for those three things, given that much of the code should already be there from 2.6, and some other assumptions. But again, I don't really know all of the situation, so why not try telling US "how long it will take" to do various things, so that we can all try to give better and more realistic advice, rather than shooting in the dark with little or no idea of the effort we're asking for.

I've been doing computer stuff, including software development and IT management, since 1958. So, I didn't exactly get into this yesterday, and I'm giving my best guesses based on the information I have been given.

Come on - asking for our advice is fine. Criticizing our advice because it wasn't "realistic" doesn't help, if you fail to tell us the parameters of reality.

Reality: if you want it bad, you'll get it bad. That goes for products, as well as advice.

Good, fast, cheap - you get to pick TWO of those the first time you pick. The second time, you only get to pick ONE.

What year?

If the issue is SHOULD it be released before Christmas 2008.....

If the issue is really "Can this be released before Christmas 2008?", the technical answer is up to SereneScreen - how fast could they roll it out the door. The more important question, then, would be if Jim would allow his name to be associated with the "product".

Sorry. I guess some of my own frustration shone through in my post. I didn't need to be so negative. :silent:

Speaking of impression of MA3 – I've come across people on the Dream Aquarium forum who were convinced Jim had all the new features of MA3 all finished and just trickled them out in small updates for some reason. :loco:

...and there are people who think that everybody who is participating in the Beta test, was formerly a participant in the Alpha testing.

Not so. I don't have that background. I struggled and searched to find out what "timing clamp" and "sleep mode" were about, etc. etc. And I certainly don't have the Alpha background that might give me better clues as to where the development stands, what's "totally not started yet", what the target rollout dates were, how fast things have(n't) moved, etc. etc.

Nor what the target features were/are, what the "essential feature set" was/is, etc. What the PRIORITIES were/are, Or if those (rather important) things were even defined.

Sorry, frustration runs "both ways". :D

Yes, 7i is stable on 7 machines (all XP Pro, with ATI and nVidia cards, one with dual monitors).

As a software developer myself, I think this would be poor interface design; it doesn't really solve the problem either - it simply lessens the likelihood of it occurring. Personally, I don't ever use windowed mode myself, so I will pose the following suggestion (again), and ask those of you who like windowed mode if this would work for you:

1) Eliminate the *automatic* (or timed) hiding/display of the title bar, and replace it with a check-box on the options page. ("Display title bar in windowed mode: [x]") Default: checked.
2) Allow the window to be moved by clicking anywhere in the window and dragging it. (I've written code to do this before - it's not that hard.)
3) Add a pop-up menu when the user right-clicks on the window, with options like: Close, Minimize, Maximize, Menu.

I understand that current behavior of the program is that right-click simulates "Escape" and closes the program immediately. This is non-standard behavior for a Windows program; however, since the pop-up menu should appear under the mouse, and the first option should be "Close", the impact on the [minority?] of users who close the program this way, is that the will have to right-click, then click on Close.

The advantage of this approach is that it's relatively clean (from a Windows standard interface perspective), it eliminates the jumping issue completely, it's easier for new users to understand, and it's fairly easy to implement, requiring a minimum of coding.

Thoughts?

~Ralph S.

I unfortunately have to agree with cjmaddy's assertions- I've used Marine Aquarium since 2.0 first came out, and although 3.0 is better than most of what's out there, in it's current form it doesn't really stack up as a major new release. This is in no way a slight on Jim, (or anyone else who is contributing, for that matter). I can't begin to imagine the amount of work he's put into creating this software, and I certainly do appreciate every last bit of it :-). But, the question was asked, and I don't think Jim would really benefit if everyone just said "It's great! Release as is!"

Lots of good suggestions to mull over.

MA 1-2-2.6 was always for me an asociation with no interferences.
MA3 will be the first released version with some interferences and therefore less realistic as needed for the most of MA users. I'll have no "power" to explain to a child why the fish now and for some times from now are swimming one throgh the other... and think that MA 2.6 will remain for me on the top place.

pvernam, - Yes, it may be stable enough, (...though that depends on your definition of 'stable') - But is it FINISHED enough?

rps, - A perfect solution! :) ... Even though I don't have any problem with window 'jumping'! ;)

RPS - yes, I agree with your solution. Not sure how long it will take to research how to drag the window without a titlebar, but if you say you've done it, then at least I know that it's not a wild goose chase.

You have to write event handlers for the following events*:
mouseDown, mouseUp, and mouseMove

You need 3 global variables:

mouseIsDown: boolean;
mouseDownX,mouseDownY: integer;

And then the code for the event handlers looks like this:

event handler: MouseDown(X, Y: Integer);
{
mouseIsDown = true;
mouseDownX = X;
mouseDownY = Y;

}

event handler: MouseUp(X, Y: Integer);
{
mouseIsDown = false;
mouseDownX = 0;
mouseDownY = 0;
}

event handler: MouseMove(x,y)
var
dx, dy: integer;
{
if mouseIsDown {
dx = X - mouseDownX;
dy = Y - mouseDownY;

window.top = window.top + dy;
window.left = window.left + dx;
}
}


Hope this helps!

~Ralph S.


* If you're not using an Object-oriented programming language/framework that provides these event handlers for you, then you'll have to hook WndProc and look for the messages WM_MOUSEDOWN, WM_MOUSEUP, and WM_MOUSEMOVE

Alternatively, when you run your drawing loop, you can check on the mouse position and button status using the functions GetCursorPos and GetAsyncKeyState (with a key value of VK_LBUTTON). However, if you take this approach, you'll have to make sure that the cursor position is within the bounds of the window, and that your application has the focus during when you check for the mouse's new position.

Let me know if you have any questions - I'm a long time (dual monitor) MA fan, and I'm happy to help!