AZdecrypt 1.19

[Jarlve]

Hey Marclean,

Happy new year!

No, there is not. Will put it on my to do list for the next version.

[Marclean]

Will put it on my to do list for the next version.

This is really cool Mr Jarvle!!
I know I can use the matrix for this, but as you know, it's laborious. I also suggest functions similar to this with the lines, as well as rotations. In my "studies", I have observed that, in the movement of some columns, in my "theory vigenere" the presence of words like: kill, die, fun, etc. I know these appear, even with random keys, but, I have some particular keys to be tested.
Always grateful
Marcelo

[Jarlve]

AZdecrypt source code release: https://drive.google.com/open?id=0B5r0r ... mljRWJvYkU

A product of a self-taught and mostly isolated Quick Basic hobby programmer (since the early 90's) with a chaotic mind and a "Die Ludolfs" way of storing and retrieving things. It probably looks like a terrible redundant mess to any seasoned programmer. Take whatever you need. I hope you find something useful. The "sub azdecrypt_perfmode_5g" is a good place to start if you want to learn about my solver. There may also be allot of unused work in progress stuff inside the code.

Included is the FreeBASIC 1.05 64-bit package and my FbEdit with compiler flags. And if you set the FbEdit paths correctly you should be able to compile it on a Windows machine.

Updated source link.

[Largo]

Hi Jarlve,

many, many thanks for these great news! I still don't have as much time as I would like, but this release is a really valuable thing! I'm already curious about the source code.
In the little sparse time I have, I'm working on an idea for a new tool. But it will be a while before its release. However, your decision motivated me to make this tool open source as well. So we all work together more effectively.


Translated with www.DeepL.com/Translator

[doranchak]

WOW! Jarlve, thanks so much for releasing this. I am sure there is a lot for us to learn from the source code, especially all the clever ways that you have optimized your hillclimber for speed.

Many, many thanks for your valuable contribution!

[beijinghouse]

Thank you Jarlve! This is wonderful!

I'll be extending the code and sharing it with the community soon.

[Jarlve]

Thanks everyone!

I am sure there is a lot for us to learn from the source code, especially all the clever ways that you have optimized your hillclimber for speed.

That was the main objective I had in mind.

I'll be extending the code and sharing it with the community soon.

Feel free to do so, or anyone else for that matter. However, it is probably best to start a new project from scratch that is programmed more professionally. And take from my code what you need.

[Largo]

Hey Jarlve!

I found a little bug, maybe you'd like to take a look at it. Here are the steps to reproduction:

- Loading a (solvable!) cipher
- Copy a transposition matrix from the clipboard to the output window. Ideally, this should make the cipher unsolvable. (I'll post one for your convenience)
- Select in the menu "Functions -> Transposition -> Use Transposition Matrix -> Transpose".

The text in the input window is now transposed correctly. If you now click on "Solve", the cipher will still be solved. Apparently the previously loaded cipher is still stored somewhere. If you copy the transposed cipher and insert it again, "Solve" does not lead to any result anymore (as desired). Something is obviously not updated correctly.

And another small request, even if it may sound ridiculous: If you select the content of a window with ctrl-a, a confirmation sound is played every time. I know this is not dramatic, but in the long run it would be nicer if it were without sound (Windows standard).

Maybe you will find some time and desire to have a look at these things. Many thanks in advance.

Code: Select all

1 2 3 256 4 5 6 257 7 8 9 258 10 11 12 259 13
14 15 260 16 17 18 261 19 20 21 262 22 23 24 263 25 26
27 264 28 29 30 265 31 32 33 266 34 35 36 267 37 38 39
268 40 41 42 269 43 44 45 270 46 47 48 271 49 50 51 272
52 53 54 273 55 56 57 274 58 59 60 275 61 62 63 276 64
65 66 277 67 68 69 278 70 71 72 279 73 74 75 280 76 77
78 281 79 80 81 282 82 83 84 283 85 86 87 284 88 89 90
285 91 92 93 286 94 95 96 287 97 98 99 288 100 101 102 289
103 104 105 290 106 107 108 291 109 110 111 292 112 113 114 293 115
116 117 294 118 119 120 295 121 122 123 296 124 125 126 297 127 128
129 298 130 131 132 299 133 134 135 300 136 137 138 301 139 140 141
302 142 143 144 303 145 146 147 304 148 149 150 305 151 152 153 306
154 155 156 307 157 158 159 308 160 161 162 309 163 164 165 310 166
167 168 311 169 170 171 312 172 173 174 313 175 176 177 314 178 179
180 315 181 182 183 316 184 185 186 317 187 188 189 318 190 191 192
319 193 194 195 320 196 197 198 321 199 200 201 322 202 203 204 323
205 206 207 324 208 209 210 325 211 212 213 326 214 215 216 327 217
218 219 328 220 221 222 329 223 224 225 330 226 227 228 331 229 230
231 332 232 233 234 333 235 236 237 334 238 239 240 335 241 242 243
336 244 245 246 337 247 248 249 338 250 251 252 339 253 254 255 340


Translated with http://www.DeepL.com/Translator

[Jarlve]

Hey Largo.

- That is a relatively new undocumented "feature". If there is a valid transposition matrix in the output window then some of the solvers will automatically use it. I better disable that since it can obviously lead to issues.

- The simple GUI library I use does not support Ctrl + A. Someone at the FreeBASIC forum hacked something together for me. I also commented about the added sound but he said it was not possible to remove it.

Thanks for reporting!

[Jarlve]

I have made some improvements to the "Substiution + by rows" solver for the upcoming AZdecrypt 1.14 release. This is a solver that does not cross row boundaries (original idea by smokie treats). To test the improvements I let it solve a 408 formatted in rows of 3 characters. At that point the solver can only use 3-grams and since it cannot cross row boundaries it must try to solve it with only about 1/3 of the 3-gram information available.

1.13 best output is 26.9% accurate (can't make out a solution):

Code: Select all

ILINONUGLKSBWARWG
APARYEDEISITSOMER
HUEDUCKOMPLOREFTH
OUNINGISBFULFBEMA
KDTHOUENVADSPARDE
TAMYFISCHEMRDTCOU
BOLCEAESDMYGPRONG
TENINGTRMASHUDBBK
WOSMACHAMODTTHNIN
GIFBABWEVOURAUSKD
OWASPACTALTHEDBES
CIFBYPENVERNTRURF
ITHDBULLTHOPASSWY
NCOUKTIDTHEOFHOUI
CUAKFINGPAVAPPLSI
DWONEFUREOFCYLLCH
OKHOWANILLOFFINGP
ARRMAMYDNDWATIFIG
LSPTBUWAYEEMYDOMA
PORYESAYPEFILNTLY
SODGRICRFFEVOTOWM
YRENNERTISBRRDLYW
OTUPLMYDUCANGIRAE
PARNIACEMASHHWUCK


1.14 best output is 55.3% accurate (rough partial solution achieved):

Code: Select all

ILIDIDAGLSCHREORG
ERACAUSEITITSOMUC
HNUTASSUMOREDUNTH
ONDINGICHWALTHEME
STTHINALBESTRECOU
TAMANISSHEMOOTDON
HERSUEECOMAGODONG
TADINGTOMETHATHHS
WISMESHEMOOTTHLIN
GINHAVREBENCEATSS
IWECRESTARTHETHET
SINHYOULBACDTONDW
ITHOHARLTHERESTRA
LSONSTIOTHEEWHINI
DAESWINGREBARORCI
TROLETACEUNDALLSH
ESHOWEDILLITWINGR
ECOMEMYSNOWATIWIG
LCOTHAWAYAUMYTOME
RICAUSEYOUWILNTRY
TOSGOIDOWNABUTORM
YCANNECTICHODSLAW
ITNORMYONSALGIDEE
REOLIESEMETHHRASS