Idea for speed-up

[Jimmer]

I had an idea (posted on another thread somewhere) about how to speed up the game with a small software component being added to the software package: An applet that audibly signals that a turn is ready (or, that a turn has been added).

The idea would be to have a widget that you would tell the location of the game files. For instance, in my games, we use Google Groups. The widget would periodically check Google Groups' "CleverDevils" group (using one of my games as an example) for updates to the location where files are uploaded. Whenever a file was uploaded, the game would check a table to find out if it's my turn yet.

The file naming conventions now are good, but there's no zip file naming convention. For this to work, the game would have to be upgraded to automatically zip up the PBM files and battle files as needed. Then, the player loads them up to the server. At this point, the applet would detect that a file has been uploaded, and proceed to figure out whether it is my turn yet. If so, ring a bell (often, since I might not be in range). If not, then just post a small message of some kind so the user knows a new turn file is on the server, but it's not ready for my turn yet.

The "know it is MY turn" piece can be left out of the first go-around of this. Just signalling when ANY turn is done is a significant step forward.

Marshall? How feasible is this?

[DCWhitworth]

Sounds like quite a tricky thing unless you standardise where the files are held. If you wanted something like this why reinvent the wheel ? Most groups have people post emails saying they're done, why not simply use an local email client and set a suitable audible sound on your PC ?

[Jimmer]

ORIGINAL: DCWhitworth

Sounds like quite a tricky thing unless you standardise where the files are held. If you wanted something like this why reinvent the wheel ? Most groups have people post emails saying they're done, why not simply use an local email client and set a suitable audible sound on your PC ?

Do you leave your email client running all the time? I'm not even near my PC half the night, let alone checking game emails. I have 6 other email accounts I have to check as well.

However, the reason I bring this up is that even with people checking their email twice per evening, the time gap between turns is extremely high. For example:

Let's be generous and assume that all players religiously check their emails twice per night. How often will they be able to take a turn?

Well, on average, a player waits for 6 other players to take their turn. Since everybody is checking twice (at relatively random times), this means each player will take his turn once every three days.

This example is not bad, and works out to be about twice as good as people's real experience (hence the reason I said I was being generous.

Now, where is the slowdown? Is it at the point in time where the user checks his mail? Absolutely not! No, it's BETWEEN those times, while the other players are waiting for him to take his turn. And, when the next guy goes, they're waiting again.

Having watched the times for a while, I've noticed that people tend to check their email once per night, and thus wind up taking their turn every 7 days (assuming no game bugs or other slowdowns).

So, the greatest calendar-hours lost for other players happen I fail to take my turn because I haven't check to see if it is done. This program intercepts that by immediately (or, within a small window) flagging the person that it's their turn again. No time lost = faster games. Possibly 10 times as fast, I would guess, or higher. Less dramatic results would happen if players knew it was "somebody's" turn, but not specifically that it is THEIR turn.

Do we have to standardize file locations? Possibly. Or, perhaps the tool could be made to work with a few locations at first release, with others to be added later.

What does everybody use for turn files? (For example, my two games use Google Groups).

[NeverMan]

I have posted in that other thread while I think this is a waste of time and technically very difficult.

You could get around where the files are posted simply by putting in a URL or ftp site; however, I believe that the file names would be a bit trickier and/or the program would have to keep track of the phase/turn itself, in order to tell who should be next. This would be a fairly difficult task.

As far as using 3rd party tools (which you really shouldn't have to do with a PC game):

2 use Google Groups
1 use Yahoo Groups
1 use TGHQ File Manager

[Jimmer]

Let me explain the time-delay thing further. It's not that clear in the previous post.
 
Let's say it takes a week to get 2 phases done. In that time, 14 people did a phase, or about 2 per night. There are 168 hours in a week. Thus, the 14 phase executions occurred about 11+ hours apart. Each player takes, say, 10 minutes to complete the turn, but that's almost irrelevant in this timing.
 
Now, let's add something that flags the player that it's his turn. Let's further assume that half of the time, he will actually see/hear that notification, and respond by taking a turn. Finally, we'll assume it takes 20 minutes for him to respond. (He'll use the same 10 minutes to take his turn as above, but it will be much more important here.)
 
So, now, for (let's say) 2 hours each night, turns are being taken every half hour when they are noticed. So, the average delay a player adds to the cycle is an hour. That's 11 hours down to 1 hour.
 
Now, adding in the 22 other hours that nobody is checking their mail, and the delay is roughly the same.
 
BUT, 4 player turns were taken rather than 2. The game is running twice as fast, using assumptions heavily weighted to make the old method look its best. Actual speedup would probably be much higher.

[NeverMan]

I understood the speedup concept, I just don't think the program would be as easy as you think it would be to code.

Furthermore, you are assuming that everyone who is not checking their email is still at their computer, so the speedup you are speaking of is mostly theoritical. Personally, IF I am at my computer I check my email every 10-30 minutes; however, it is impossible for me to check my email if I am not at my computer AND therefore also impossible for me to get the widget's reminder that it's my turn.

I think that in concept, it's a very good idea, I just don't think that it's practical for what you would get. JMO.

EDIT: I would like to edit this post by saying that if some genius Java/whatever Coder out there wants to give this a go and it works, then I'm all for that; however, I'm not for Marshall or Matrix taking valuable resources to do this because I don't think the gain > 1.

[Jimmer]

ORIGINAL: NeverMan
Furthermore, you are assuming that everyone who is not checking their email is still at their computer,

Actually, no, I'm assuming the opposite: That the person is NOT at their computer, but that they will pass within hearing distance from it at some point in the evening.

[NeverMan]

ORIGINAL: Jimmer

ORIGINAL: NeverMan
Furthermore, you are assuming that everyone who is not checking their email is still at their computer,

Actually, no, I'm assuming the opposite: That the person is NOT at their computer, but that they will pass within hearing distance from it at some point in the evening.

So then it's safe to assume you leave your computer on with volume turned up?

I don't. My computer is on an entire different floor from where I spend the rest of my time. Also, I use headphones (as not to annoy everyone else in the house).

[Jimmer]

Yes, I leave it on. The volume isn't that loud, but probably loud enough (and I would make it louder if such a widget existed).
 
The times I see it being the most helpful for are the joints between phases. In diplo, Spain goes last, preceded by GB. Then, Spain goes first in Reinforcement, this time FOLLOWED by GB. Many MANY times I have posted my turn, hoping to take my next phase turn within a short time, only to have to wait until the next day (or worse).
 
But, the value of it, for those who would use it, would be that THEY would not be the roadblock. Whether anybody else is is, frankly, not my problem. I have to fix me. This idea would give me the chance to do better at that.

[RayKinStL]

I think the idea is interesting Jimmer, but the programminginvolved would be better spent on a IP/PBEM hybrid system, in my opinion, which could potentially lead to unimaginable time savings when compared to the current PBEM system or even the post-widget system you propose.  If someone is going to take all that time to program such an intricate widget, then use that time to make IP/PBEM play interlaced and streamlined.  Yes?

[DCWhitworth]

I still don't see why this needs a specific application.

Get yourself a client program (Thunderbird, Outlook, whatever), set a rule to sound a nice loud alert when you get a specific email. Job done. Why do you need to write program for this ?

[Jimmer]

You guys are giving me some ideas. Thanks.
 
I'm not done with this idea. But, I've now realized that Matrix does NOT have to be involved at all! All that is really needed is
 
1)  Some code that understands the turn movement orders (for all of the phases)
2)  A way to interact with the storage medium (specifically, to know when another file has been posted)
3)  A way to contact the user (noisemaking, message writing, etc.)
4)  A way to run in the background
 
Optionally, it would help if it had
 
A)  Some code that can zip up a file
B)  Some code that follows the same naming convention for the zipped file as it does with the non-zipped version from the game
 
In an ideal world, it would
 
I)  Be able to actually transmit the file(s) once they exist.
II) Be able to retrieve files
 
The only parts that require Matrix involvement are the lettered ones. The rest can be done outside of the bounds of the game. I have to cogitate this for a bit.

[Marshall Ellis]

ORIGINAL: Jimmer

I had an idea (posted on another thread somewhere) about how to speed up the game with a small software component being added to the software package: An applet that audibly signals that a turn is ready (or, that a turn has been added).

The idea would be to have a widget that you would tell the location of the game files. For instance, in my games, we use Google Groups. The widget would periodically check Google Groups' "CleverDevils" group (using one of my games as an example) for updates to the location where files are uploaded. Whenever a file was uploaded, the game would check a table to find out if it's my turn yet.

The file naming conventions now are good, but there's no zip file naming convention. For this to work, the game would have to be upgraded to automatically zip up the PBM files and battle files as needed. Then, the player loads them up to the server. At this point, the applet would detect that a file has been uploaded, and proceed to figure out whether it is my turn yet. If so, ring a bell (often, since I might not be in range). If not, then just post a small message of some kind so the user knows a new turn file is on the server, but it's not ready for my turn yet.

The "know it is MY turn" piece can be left out of the first go-around of this. Just signalling when ANY turn is done is a significant step forward.

Marshall? How feasible is this?

Does not sound too hard BUT I need a little more explanation of what you are trying to acomplish? Are you wanting to be able to tell when it is your turn as the turn files come in?

[Jimmer]

Yes. Somehow to contact the <fill in the blank> (mail server, FTP server, Google Group, whatever), and then flag the user (ring a bell, ring 10 bells, flash neon tie-dyed shirts on the monitor, whatever).
&nbsp;
But, like Neverman said, it shouldn't take priority over the other stuff. This is an idea. IF a "proof of concept" can be done that would show it's feasible, without a lot of time invested in it, then MAYBE we would move to the next step.
&nbsp;
Interestingly, in my last post I noted that it might be able to be completely independent from the game. But, if it WERE in the game, two major things would be required:
&nbsp;
1)&nbsp; A communication tool for "talking" to whatever server the files live on (email server, group server, ftp server, local system, whatever). That might mean putting an smtp program in the game, which could make it unwieldy. Plus, it would only support those things for which you would have written communication tools for.
&nbsp;
BUT, Matrix may already have done at least part of this. Check it out.
&nbsp;
2)&nbsp; The file names would have to be standardized, and the files would have to be pre-zipped. Otherwise, we would be dependent upon the user properly zipping them AND applying the right name to the file. I would rather have a program doing that, but, for proof of concept, maybe not?

[delatbabel]

I think that a lot of this would be viable with a web component somewhere out on the web (holding the turn files and distributing them as necessary) and a Java/Delphi PC program on the client end to collect the files.&nbsp; Eventually the web component could do some in-game activity such as managing die rolls and chit picks, etc, however that would require in-game modifications while most of the above would require only out of game code.

I'd be happy to write the web based stuff but someone else would have to write the PC stuff as I'm not a windows programmer.

[bresh]

I do think, most of the delay in a game will be timezones.
And battles, would be nice if more battles could be submitted at the same time. (I know current code does not allow this, so would need a whole deal of recoding).
&nbsp;
2-4 Battles in 1 land combat phase witch requires fileexchanged, and one might go for more than 1 battleday. This can take weeks, even months.&nbsp;During witch&nbsp;the interest could go downhill then for those other players who are&nbsp;not involved.
&nbsp;
Regards
Bresh

[NeverMan]

I can't wait to see this program since everyone thinks this is going to be "easy" and how "easy" it will be to make this thing not work once it's programmed.

I do; however, think this would be an AWESOME idea with IP play, since so many complain about "sitting around for an hour" (which is a total exaggeration).

[Jimmer]

ORIGINAL: delatbabel

I think that a lot of this would be viable with a web component somewhere out on the web (holding the turn files and distributing them as necessary) and a Java/Delphi PC program on the client end to collect the files.  Eventually the web component could do some in-game activity such as managing die rolls and chit picks, etc, however that would require in-game modifications while most of the above would require only out of game code.

I'd be happy to write the web based stuff but someone else would have to write the PC stuff as I'm not a windows programmer.

Correct. But, I'm also thinking of ways that Matrix can sell more copies of the game. If something can be (easily) made part of the game, or as an add-on for licensed users, that's a selling point (provided it saves any real time).

[Jimmer]

ORIGINAL: NeverMan

I can't wait to see this program since everyone thinks this is going to be "easy" and how "easy" it will be to make this thing not work once it's programmed.

I do; however, think this would be an AWESOME idea with IP play, since so many complain about "sitting around for an hour" (which is a total exaggeration).

Considering this would automatically be a part of an IP enhancement (any version of play over IP would require a way to tell you when it was your turn), I would think you would be jumping all over this one, since you want IP play so badly. The only difference would be that this one has to interface with more than one potential server type (email, ftp, etc).

[Marshall Ellis]

Careful here guys because once you get sucked in the dev trap you'll never get out LOL!
&nbsp;
&nbsp;