Oldxxx

oldxxx

2023.10.06 02:39 OkDragonfruit3134 oldxxx

2018.10.19 21:52 peter090654 Possible recovery of notebooks when folder is deleted

This post may be specific to 1.5.2 software.
By brushing on the ReMarkable tablet with my wrist when typing on a keyboard placed further away than the tablet, I must have accidentally deleted a top-level folder, and the single notebook within it got automatically deleted. When I tried out the folder delete sequence later it seems there is no warning that all the notebooks within the folder are about to be deleted. A single brush against the three dots on the folder (not that likely but possible), followed by a single brush on the delete key on the top right (much more likely if reaching over the tablet) will do it. Then "undo" appears for a few seconds, but if you miss it because you are slow or not using the tablet, all is lost!
Not quite though. When signing on using the linux ' ssh root@ command it seems that the notebook .lines.backup file is the sole remaining file for each notebook and not deleted as part of the notebooks delete which is part of the folder delete. Your tablet and password can be found by pressing RM on the home screen, following by About / Copyrights and licenses / press General information (top of the left hand of the two columns). Then the password is two paragraphs from the bottom, and the IP address is one of the two displayed (the top is for a cable connection and the bottom is for a wifi connection).
I managed to recover the notebook using this procedure :-

Set up a new notebook with the same number of pages as the notebook accidentally deleted, and make at least two sets of changes with an undo somewhere. Change the name to something not used before e.g. recover1.
Use 'grep -E "visible" *.metadata' to display the list of names (including notebooks, folders and PDF names) and select the GUID of the new notebook file (long string of 0-9, a-f and dashes in the file name before the first dot). Probably the first 3 characters of the GUID are unique, so you can just key those and press the tab key in unix to get unix to fill in the full GUID/file name.
do 'ls -l *.lines*' to find the old .lines.backup files. These will be the files without .lines partners with the same GUID.
Rename .lines and .lines.backup fiiles using 'mv .lines .linesold' and 'mv .lines.backup .lines.backupold'
Do something similar to the .png (to .pngold) and .jpg (to .jpgold) in the .cache and .thumbnails folders.
Then 'cp .lines.backup .lines' and also copy it to .lines.backup.
Reboot the tablet and the new notebook has the deleted notebook contents - hopefully.
Clean up the renames of the files.

There is some risk to this procedure, so be careful. And if you choose to use it you do so at your own risk, not mine.
I also raised a bug report, and suggested they should not include deleting the .lines.backup file in the bug fix!

submitted by peter090654 to RemarkableTablet [link] [comments]

2016.06.09 15:00 sugima Math and Transcendance

Edit : a mistake regarding Chronos.
Edit2 : added some text for the oldXXX values in the formulas.
Edit3 : changed some text about Solomon post-cap.
Edit4 : for non-transcendent players, the old Solomon formula still applies.
Edit5 : for non-transcendent players, the formula for Atman and Kuma doesn't apply, since their HS gain isn't exponential . Progression ancients are still valid
Edit6: Hans139 made some mistakes copying the effect factors of some ancients. The value for Kuma was wrong. Recommended value for Kuma increased by ~3 levels.
Transcendence is out, it is time for more maths.
In this post, I will deal with HS gain and Transcendent Power (TP).
To make calculations easier, I will assume that HS are only gained through TP, and that the HS per primal cap isn't reached yet.
I will also assume that reducing the amount of mobs per zone with Kuma doesn't affect the gold gain, but only the time necessary to reach that zone.
Finally, I will consider that ancients like Argaiv, Siya ... have a linear effect.
See this page for the effects of uncapped ancients.
c1, c2 .... are values that depend on nothing relevant for the calculations.
First, the raw formula of HS gained from TP, if all bosses are primal is :

rawHS = sum(k=1 to (zone-100)/5; 20·(1+TP)^k )
rawHS= 20·((1+TP)^{(zone-100)/5+1} - 2)/TP

At high zones, this can be written as

rawHS = c1·(1+TP)^zone/5

The actual HS gain is HS = rawHS·SolomonBonus·AtmanBonus
To take into account that Kuma makes runs faster, the formula I'll use is

HS = c2·(1+TP)^zone/5·SolomonBonus·AtmanBonus/mobsPerZone

Because sums are easier to maximize than products, the function I'll want to maximize will be :
ln(HS) = zone·ln(1+TP)/5 + ln(SolomonBonus) + ln(AtmanBonus) - ln(mobsPerZone) + ln(c2)
Second part : Optimal Zone

In this part, I will need the growth of the optimal zone with respect to gold and damage boosts.
I assume that the DPS at the optimal zone is proportional to the health at said zone.
This leads to the following formula :

DPS = c3·HPscale^zone
ln(DPS) = ln(c3) + ln(HPscale)·zone (1)

As we already know, DPS can be computed through this formula :

DPS = c4·level·damageBoost·damageFactor^level

I will assume that the level range is small enough to remove level from the product.

ln(DPS) = c5 + ln(damageBoost) + ln(damageFactor)·level (2)

The cost for level is baseCost·(1.07^level - 1)/0.07 ~ c6·1.07^level

ln(gold) = ln(c6) + ln(1.07)·level (3)

Gold gain is tied to zone health :

ln(gold) = c7 + ln(goldBoost) + ln(HPscale)·zone (4)

Mixing (2) and (3), we have :

ln(DPS) = c8 + ln(damageBoost) + ln(damageFactor)·ln(gold)/ln(1.07) (5)

(4) and (5) give :

ln(DPS) = c9 + ln(damageBoost) + ln(damageFactor)·ln(goldBoost)/ln(1.07) + ln(damageFactor)·ln(HPscale)·zone/ln(1.07)

With (1), we finally obtain :

zone·ln(HPscale) = c10 + ln(damageBoost) + ln(damageFactor)·ln(goldBoost)/ln(1.07) + ln(damageFactor)·ln(HPscale)·zone/ln(1.07)

zone·ln(HPscale)·(ln(1.07)-ln(damageFactor))/ln(1.07) = c10 + ln(damageBoost) + ln(damageFactor)·ln(goldBoost)/ln(1.07)

The formula for the optimal zone is then :

zone = (ln(damageBoost)·ln(1.07) + ln(goldBoost)·ln(damageFactor))/( ( ln(1.07)-ln(damageFactor) ) · ln(HPscale) ) + c11

This formula requires the input of HPscale (1.145 for zones 140-500, 1.15 for zones 500-1000 ... )
The value of damageFactor will be detailed below.
Third part (easier than the previous one)

Considering that each ranger cost 10¹⁵ as much as the previous one, and using these formulas for their DPS, we have the following level difference and DPS ratio:
levelDifference = 15*ln(10)/ln(1.07)
DPSratio = 0.007368·10¹⁵ = damageFactor^{levelDifference}
That leads to
damageFactor = 0.007368^{ln(1.07)/(15·ln(10))} ·1.07
ln(damageFactor) = ln(1.07) · (1 + ln(0.007368)/ln(10¹⁵ ) )
Note that damageFactor²⁵ is roughly 4.27, which is close to the x4 multipliers we have every 25 levels.
Fourth part : Damage and Gold Boosts
At this point, we're only missing the damage and gold boosts.

Damage
I assumed that Morgulis, Argaiv and Siya give a linear bonus. We can consider that 2 more ancients give a DPS bonus : Bubos and Chronos (the latter only if we push to the 30 sec timer)
Bubos : it decreases bosses' HP by 50·(1-e^-0.02·Bubos )%, which means boss HP is 0.5·(1+e^-0.02·Bubos )
Or we can consider it as a multiplication of DPS by 2/(1+e^-0.02·Bubos )
Chronos : same idea, more time to kill the boss means more damage on the boss. This time is
30+30·(1-e^{-0.034·Chronos} ) = 30·(2-e^{-0.034·Chronos} )

Gold
Same thing as above, Libertas and Mommon are linear.
If we assume chests are the only source of gold, Mimzee is linear too, and Dora's effect is independant of Mimzee's
Dora : multiplies chest chance by 100 - 99·e^-0.002·Dora
Fortuna : multiplies gold by 1 + 9·(chance of x10) = 10 - 9·e^{-0.0025·Fortuna}
Dogcog : same idea as Bubos, but for gold
Virtual gold gain = 100/(1+99·e^{-0.01·dogcog} )
Fifth part

We want to maximize :

ln(HS) = alpha·ln(damageBoost) + beta·ln(goldBoost) + ln(SolomonBoost) + ln(AtmanBonus) - ln(mobsPerZone)

with

alpha = ln(1.07)/(ln(1.07) - ln(damageFactor)) · ln(1+TP)/5 /ln(HPscale)
alpha = -15 · ln(10) / ln(0.007368) · ln(1+TP)/ln(HPscale) /5

and

beta = -(15 · ln(10) + ln(0.007368)) / ln(0.007368) · ln(1+TP)/ln(HPscale) /5

To simplify the formula, we can write :

alpha = 1.4067·ln(1+TP)/ln(HPscale)
beta = 1.2067·ln(1+TP)/ln(HPscale)

From the fourth part, we have :

ln(DamageBoost) = ln(Morgulis) + ln(Siya) + ln(Argaiv) - ln(1+e^-0.02·Bubos ) + ln(2-e^{-0.034·Chronos} ) + c12
ln(GoldBoost) = ln(Libertas) + ln(Mammon) + ln(Mimzee) + ln(100 - 99·e^-0.002·Dora ) - ln(1+99·e^{-0.01·dogcog} ) + ln(10 - 9·e^{-0.0025·Fortuna} ) + c13

Optimize Damage
The cost for Damage Bonus is (approximatively) :

costDamage = Morgulis + Siya²/2 + Argaiv²/2 + 2·2^Bubos + 2·2^Chronos

Gradients are :

G(ln(db)) = (1/Morgulis; 1/Siya ; 1/Argaiv ; 0.02·e^-0.02·Bubos / (1+e^-0.02·Bubos ) ; 0.034 · e^{-0.034·Chronos} / (2-e^{-0.034·Chronos} ) )
G(cost) = (1; Siya; Argaiv; 2·ln(2)·2^Bubos ; 2·ln(2)·2^Chronos)

They must be proportional to get the optimum :

G(cost) = L·G(ln(db))

leads to

Morgulis = L
Siya² = L
Argaiv² = L
100·ln(2)·(2·e^0.02 )^Bubos · (1+e^-0.02·Bubos ) = L
2·ln(2)·(2·e^0.034 )^Chronos · (2-e^{-0.034·Chronos} ) / 0.034 = L

To simplify the equations, i suggest to consider (1+e^-0.02·Bubos ) and (2-e^{-0.034·Chronos} ) as constants, using the old ancient values (this is false, but it's the easiest thing that comes to my mind) .
So we have :

Morgulis = Siya²
Argaiv = Siya
Bubos = (2·ln(siya) - ln(100·ln(2)·(1+e^{-0.02·oldBubos} ) ) )/ln(2·e^0.02 )
Chronos = (2·ln(siya) - ln(2·ln(2)·(2-e^{-0.034·oldChronos} ) / 0.034 ) )/ln(2·e^0.034 )

Optimize Gold

G(ln(gb)) = (1/Lib ; 1/Mammon; 1/Mimzee; 0.002·e^-0.002·Dora /(100/99 - e^-0.002·Dora ); 0.01·e^{-0.01·dogcog} /(1/99 + e^{-0.01·dogcog} ); 0.0025·e^{-0.0025·Fortuna} / (10/9 - e^{-0.0025·Fortuna} ) )
G(cost) = (Lib; Mammon; Mimzee; 2·ln(2)·2^Dora ; 2·ln(2)·2^Dogcog; 2·ln(2)·2^Fortuna)

The optimum is when :

Libertas² = Mammon² = Mimzee² = L
1000·ln(2)·(2·e^0.002 )^Dora · (100/99 - e^-0.002·Dora ) = L
200·ln(2)·(2·e^0.01 )^Dogcog · (1/99 + e^{-0.01·dogcog} ) = L
800·ln(2)·(2·e^0.0025 )^Fortuna · (10/9 - e^{-0.0025·Fortuna} ) = L

Libertas = Mammon = Mimzee
Dora = (2·ln(Lib) - ln(1000·ln(2)· (100/99 - e^{-0.002·oldDora} ) ) ) / ln(2·e^0.002 )
Dogcog = (2·ln(Lib) - ln(200·ln(2)· (1/99 + e^{-0.01·oldDogcog} ) ) ) / ln(2·e^0.01 )
Fortuna = (2·ln(Lib) - ln(800·ln(2)· (10/9 - e^{-0.0025·oldFortuna} ) ) ) / ln(2·e^0.0025 )

Libertas/Siya ratio
It's easy, because we want to maximize alpha·ln(siya) + beta·ln(lib)
G(effect) = (alpha / siya ; beta / lib)
G(cost) = (siya; lib)
=> lib/siya = sqrt(beta/alpha) = 0.926 roughly

Solomon, Atman and Kuma
Their optimal levels depend on TP.
We want to maximize

f = alpha·ln(Siya) + ln(Solo) + ln(1 - 0.75·e^{-0.013·Atman} ) - ln(2 + 8·e^-0.01·Kuma )
cost = Siya²/2 + Solo^2.5 /2.5 + 2·2^Atman + 2·2^Kuma

Gradients are :

G(f) = (alpha / siya ; 1 / solo; 0.013·e^{-0.013·Atman} / (4/3 - e^{-0.013·Atman} ); 0.01·e^-0.01·Kuma / (0.25 + e^-0.01·Kuma )
G(cost) = (siya; solo^1.5 ; 2·ln(2)·2^Atman ; 2·ln(2)·2^Kuma )

We want

G(cost) = L·G(f)
siya² / alpha = L
solo^2.5 = L
2/0.013 ·ln(2)·(2·e^0.013 )^Atman · (4/3 - e^{-0.013·Atman} ) = L
200·ln(2)·(2·e^0.01 )^Kuma · (0.25 + e^-0.01·Kuma ) = L

which leads to

solomon = siya^0.8 / alpha^0.4
Atman = (2·ln(siya) - ln(alpha) - ln(2/0.013 ·ln(2) · (4/3 - e^{-0.013·oldAtman} )) / ln(2·e^0.013 )
Kuma = (2·ln(siya) - ln(alpha) - ln(200 ·ln(2) · (0.25 + e^{-0.01·oldKuma} )) / ln(2·e^0.01 )

TL;DR :
Here is a table with optimal values :

Ancient	Optimal level
Argaiv	Siya
Atman	2.832·ln(siya) - 1.416·ln(alpha) -1.416·ln(4/3 - e^{-0.013·oldAtman} ) - 6.613
Bubos	2.8·ln(siya) - 1.4·ln(1+e^{-0.02·oldBubos} ) - 5.94
Chronos	2.75·ln(siya) - 1.375·ln(2-e^{-0.034·oldChronos} ) - 5.1
Dogcog	2.844·ln(siya) - 1.422·ln(1/99 + e^{-0.01·oldDogcog} ) - 7.232
Dora	2.877·ln(siya) - 1.4385·ln(100/99 - e^{-0.002·oldDora} ) - 9.63
Fortuna	2.875·ln(siya) - 1.4375·ln(10/9 - e^{-0.0025·oldFortuna} ) - 9.3
Kuma	2.844·ln(siya) - 1.422·ln(alpha) - 1.422·ln(0.25 + e^{-0.01·oldKuma} ) - 7.014
Libertas	0.926·Siya
Mammon	0.926·Siya
Mimzee	0.926·Siya
Morgulis	Siya²
Nogardnit	Libertas^0.8
Siya	Everything is based on it
Solomon	Siya^0.8 / alpha^0.4

with

alpha = 1.4067·ln(1+TP)/ln(HPscale)

oldXXX is the current value for the ancient (Atman ...)
HPscale is the increase of monster health between 2 zones at zone of ascension (1.145 for 140-500, 1.15 for 500-1000, 1.155 for 1000-1500 ... )
Nog is based on Libertas, but uses Siya's level and the correct goldRatio (see below) in the calculators.

PS : I proved here that you can use these formulas even without Morgulis when you have some levels in Chor.
In this case, the amount of HS you need to keep if the cost of the optimal value of Morgulis.
PS2: Kragnir did some post-cap math here. Siya to Solomon formula stays the same.
PS3: After Wep 8k, there are only x4 multipliers every 25 levels, so damageFactor = 4^1/25 . That leads to alpha = 1.1085·ln(1+TP)/ln(HPscale), and gold Ancients = 0.905·Siya

submitted by sugima to ClickerHeroes [link] [comments]

http://swiebodzin.info