The Two General's Problem

penny wrote:Um... let's see. The fact that I was still IN the forum, alone before he came?

The fact that I said that posting is a chore now?

Since I shared with everyone that the forum freezes up for me?

Please. When I post he appears instantly lots of the time and grabs the post that I am currently editing.

And yes. The site does alert you when edits are ongoing if you are viewing the post live. The refreshes show up as a WAVE that is unmistakable.

And what reason was it that he could not wait to be sure? Since I had shared with the forum of my editing woes. I cannot fully edit in MS Word. The site does not properly accept my formatting. Quotation marks are incorrect.

Anyway, there had not been a post for over 24hrs and I was the only person in the forum. What was the rush?!

Please. Whatever. I'm done.

.

Penny/Cthia, There is no way to see what another user is doing on the forum. We CAN see if you are active, but it says the same if you are posting, reading... or just wander down to the kitchen for a snack leaving the browser screen open.

If you edit a post... you get what you get. We think your post - now posted - is DONE. Your thoughts are encapsulated, complete. Everyone else can do whatever they want, in total ignorance you've changed your mind, or wanted to expand on a point not fully flushed out. I've been caught in the same situation - a post didn't land right, or I realized that my logic stream is just wrong... and someone beat my edit to the punch.

No need to shake your fists at the nameless sky. The system just doesn't work the way you thought.

edit: (ha ha) you will get a warning if you post to a thread, and some one (as in anyone) has posted another post in the thread. (there is a newer post). but we don't know who it was, especially the originally poster editing.

Theemile wrote:There is no way to see what another user is doing on the forum. We CAN see if you are active, but it says the same if you are posting, reading... or just wander down to the kitchen for a snack leaving the browser screen open.

I keep forgetting we can even see who's online here - so I, personally, am not looking for that info when I'm checking threads.

In part I forget because it's almost never in my sightline because on my laptop "Who is online" is just below the bottom of the screen on the forum topic page (and I rarely have to scroll down because we don't normally have more than a handful of threads with new posts); and when I do scroll down I'm looking over at the list of topics, not the sidebar. Then when I click the 'View first unread post' icon to jump into the thread most likely that section is then off the top of my screen on the thread's page. (Plus I'm focused on just reading the new post)

About the only time I remember it is there is when something makes me think I might not be logged in and need to check

ThinksMarkedly wrote:When I have long replies, I click "preview" so the answer is submitted to the forum once (though not saved) and my browser could reload if needed. I only click "submit" after I've re-read my answer at least once. And I only edit for minor misspellings that escaped even that. Not knowing everyone else's procedures, I assumed they were similar to mine.

I am not sure what you mean by click "preview" so the answer is submitted to the forum once (though not saved) and my browser could reload if needed. So far as I know pressing preview only applies the formatting rules to your edited text, so you get an idea how it will appear when you press submit. It certainly does NOT submit anything and does NOT reload anything; you are still working with just your own copy of text. You can back out with nothing changed.

I will press preview multiple times while editing, particularly to ensure I have the quote attributions correct. But also to check bold, italics and rarely size or color. But I have to wait until submitting, before I can check pointers to outside pictures (IMG) or text (URL).

tlb wrote:I am not sure what you mean by click "preview" so the answer is submitted to the forum once (though not saved) and my browser could reload if needed. So far as I know pressing preview only applies the formatting rules to your edited text, so you get an idea how it will appear when you press submit. It certainly does NOT submit anything and does NOT reload anything; you are still working with just your own copy of text. You can back out with nothing changed.

I was talking about the mechanism: the formatting isn't done by your browser, like you see on other websites like GitHub. If you click Preview, the browser submits the form via HTTP POST, the server formats and sends back, but then discards it. If you never Submit, it's lost forever. However, in case the Submit fails, you can go Back in your browser and press F5, at which point it'll ask if you want to re-send the form.

tlb wrote:I am not sure what you mean by click "preview" so the answer is submitted to the forum once (though not saved) and my browser could reload if needed. So far as I know pressing preview only applies the formatting rules to your edited text, so you get an idea how it will appear when you press submit. It certainly does NOT submit anything and does NOT reload anything; you are still working with just your own copy of text. You can back out with nothing changed.

ThinksMarkedly wrote:I was talking about the mechanism: the formatting isn't done by your browser, like you see on other websites like GitHub. If you click Preview, the browser submits the form via HTTP POST, the server formats and sends back, but then discards it. If you never Submit, it's lost forever. However, in case the Submit fails, you can go Back in your browser and press F5, at which point it'll ask if you want to re-send the form.

Ok, I understand, yes the browser does not format; I was just not clear on your words. And in that sense I was wrong about not presenting anything.

Anyway, circling back to the question of stopping.
The spider drive is described as grabbing ahold of the hyper wall. That at least raises the possibility that it could stop in some absolute sense - since presumably if you were able to grab points on all sides of the ship and pulled you'd be unable to move in any direction -- just hang suspected in your web of tension.

Though that possibility raises at least a couple of questions - neither of which we have the information to answer.

1) On an engineering level do any of the current spider drive systems have enough ability to alter their grab point to grab points immediately beside, or even angled behind their current position. If not, if they always have to grab points that are angled ahead of their current position then it would seem that the particular implementation of the drive would be unable to grab the hyper wall in such a way to know absolutely that the ship is stopped.

2) What relative movement might exist between the local hyper wall and the local star? (Or, if in hyper, between the next higher hyper wall an the n-space emergence point from there relative to the local star).
Being stopped relative to the local hyper wall doesn't seem especially helpful if the wall is, say, stationary relative to the center of the galaxy (and thus the local star is moving away from any fixed point on it by a couple hundred KPS -- the star's orbital velocity around the center of the Milky Way)

Still, even the possibility of some way to ensure you were stopped dead relative to some spot in space (or hyper) is an interesting one.

Jonathan_S wrote:Anyway, circling back to the question of stopping.
The spider drive is described as grabbing ahold of the hyper wall. That at least raises the possibility that it could stop in some absolute sense - since presumably if you were able to grab points on all sides of the ship and pulled you'd be unable to move in any direction -- just hang suspected in your web of tension.

Though that possibility raises at least a couple of questions - neither of which we have the information to answer.

1) On an engineering level do any of the current spider drive systems have enough ability to alter their grab point to grab points immediately beside, or even angled behind their current position. If not, if they always have to grab points that are angled ahead of their current position then it would seem that the particular implementation of the drive would be unable to grab the hyper wall in such a way to know absolutely that the ship is stopped.

2) What relative movement might exist between the local hyper wall and the local star? (Or, if in hyper, between the next higher hyper wall an the n-space emergence point from there relative to the local star).
Being stopped relative to the local hyper wall doesn't seem especially helpful if the wall is, say, stationary relative to the center of the galaxy (and thus the local star is moving away from any fixed point on it by a couple hundred KPS -- the star's orbital velocity around the center of the Milky Way)

Still, even the possibility of some way to ensure you were stopped dead relative to some spot in space (or hyper) is an interesting one.

But that does not help a ship that is traveling a long distance through hyperspace to meet that ship waiting at one spot, because the traveling ship will still have the same uncertainties accumulate over the course of the flight without stops in normal space to reset the hyper-log. Whether the waiting ship is truly stationary or drifting, errors can make it impossible to meet. Also exactly how do you specify the location of a meeting spot in hyperspace, what coordinates can you use?

The only sure way to meet is for the ships to be together and relatively stationary and after two hyper transitions they are together again in H-space (either they both start in N-space and the fleet transitions followed later by the courier transitioning to give the signal or they both start in H-space and the courier transitions down then back to give the signal). The possibility of anchoring to the wall might only extend the time limit on how long they can be apart.

tlb wrote:

Jonathan_S wrote:Anyway, circling back to the question of stopping.
The spider drive is described as grabbing ahold of the hyper wall. That at least raises the possibility that it could stop in some absolute sense - since presumably if you were able to grab points on all sides of the ship and pulled you'd be unable to move in any direction -- just hang suspected in your web of tension.

Though that possibility raises at least a couple of questions - neither of which we have the information to answer.

1) On an engineering level do any of the current spider drive systems have enough ability to alter their grab point to grab points immediately beside, or even angled behind their current position. If not, if they always have to grab points that are angled ahead of their current position then it would seem that the particular implementation of the drive would be unable to grab the hyper wall in such a way to know absolutely that the ship is stopped.

2) What relative movement might exist between the local hyper wall and the local star? (Or, if in hyper, between the next higher hyper wall an the n-space emergence point from there relative to the local star).
Being stopped relative to the local hyper wall doesn't seem especially helpful if the wall is, say, stationary relative to the center of the galaxy (and thus the local star is moving away from any fixed point on it by a couple hundred KPS -- the star's orbital velocity around the center of the Milky Way)

Still, even the possibility of some way to ensure you were stopped dead relative to some spot in space (or hyper) is an interesting one.

But that does not help a ship that is traveling a long distance through hyperspace to meet that ship waiting at one spot, because the traveling ship will still have the same uncertainties accumulate over the course of the flight without stops in normal space to reset the hyper-log. Whether the waiting ship is truly stationary or drifting, errors can make it impossible to meet. Also exactly how do you specify the location of a meeting spot in hyperspace, what coordinates can you use?

The only sure way to meet is for the ships to be together and relatively stationary and after two hyper transitions they are together again in H-space (either they both start in N-space and the fleet transitions followed later by the courier transitioning to give the signal or they both start in H-space and the courier transitions down then back to give the signal). The possibility of anchoring to the wall might only extend the time limit on how long they can be apart.

True enough.
If you don't end up where you thought you were then holding stationary doesn't help. And if the ship(s) trying to meet you don't end up where they thought they were then it also doesn't help.

The only scenario I can see it helping is it potentially lets you extend a wait in hyper indefinitely. So if you got close enough to where you intended you could then lock yourself in for arbitrarily long until the other ship(s) show up (if they get close enough you can see them). Whereas without the ability to make some kind of absolute stop you'd most likely be coasting in some unknown direction at some (likely quite low; but also) unknown velocity away from your intended coordinates; so the longer the wait the harder it'll be to link up.



Though now that I say that I realize that Spider ships might actually be worse at trying to rendezvous in hyper than a normal ship. Grav sensors see vastly further in hyper -- up to around 20 LM +/- 3 LM (call it an average or 360 million km) -- than non-grav sensors which can't see beyond maybe 5 million km. So you'd want to use grav sensors to detect ships approaching the rendezvous area -- except spider drives don't show up on grav sensors.

So a pair of spider ships attempting to rendezvous have to hit their coordinates 72x more accurately to link up than impeller powered ships would. (And even then, at least one of the spiders would have be broadcasting active sensors or a high-power beacon or something so it could be detected despite its optical and radar stealth)

Jonathan_S wrote:True enough.
If you don't end up where you thought you were then holding stationary doesn't help. And if the ship(s) trying to meet you don't end up where they thought they were then it also doesn't help.

The only scenario I can see it helping is it potentially lets you extend a wait in hyper indefinitely. So if you got close enough to where you intended you could then lock yourself in for arbitrarily long until the other ship(s) show up (if they get close enough you can see them). Whereas without the ability to make some kind of absolute stop you'd most likely be coasting in some unknown direction at some (likely quite low; but also) unknown velocity away from your intended coordinates; so the longer the wait the harder it'll be to link up.



Though now that I say that I realize that Spider ships might actually be worse at trying to rendezvous in hyper than a normal ship. Grav sensors see vastly further in hyper -- up to around 20 LM +/- 3 LM (call it an average or 360 million km) -- than non-grav sensors which can't see beyond maybe 5 million km. So you'd want to use grav sensors to detect ships approaching the rendezvous area -- except spider drives don't show up on grav sensors.

So a pair of spider ships attempting to rendezvous have to hit their coordinates 72x more accurately to link up than impeller powered ships would. (And even then, at least one of the spiders would have be broadcasting active sensors or a high-power beacon or something so it could be detected despite its optical and radar stealth)

Another item not mentioned yet is we are assuming that the "fabric" of each layer of hyper is perfectly fixed in association to each other. If each layer has a slight drift or the relationship between layers varies subtly from point to point, over time (and travel) additional errors will add up - and cannot be accounted for.

This has not been mentioned, and doesn't seem true - but....

Theemile wrote:Another item not mentioned yet is we are assuming that the "fabric" of each layer of hyper is perfectly fixed in association to each other. If each layer has a slight drift or the relationship between layers varies subtly from point to point, over time (and travel) additional errors will add up - and cannot be accounted for.

This has not been mentioned, and doesn't seem true - but....

I am not sure if that matters to a ship that is traveling in a single band (except for going to or coming from normal-space). But if a rogue wave is encountered that required a shift to another band for some time, on return (without a position check in N-space) the assumed position might be wildly off.

Another case we that could be significant would be a ship with streak drive trying to catch a ship with a head start, but a normal hyper-generator. Assuming the heading was known, when the chase ship drops down to intercept, the slower ship might be nowhere to be seen.