Lab computers displaying the same research log on each hacks

[DillonSimeone]

Describe the bug

When you're hacking lab computers to read the research logs, you'll get the same log on each hacks.

Steps To Reproduce

1. Break into a lab.
2. Find a working computer in lab.
3. Hack the research log over and over, noting that you're getting the same log each time.

Expected behavior

Different logs are shown at random.

I know about the other logs because that was how it used to be a long time ago, with your odd of failure going up the more you hacked to get at the deeper logs.

Screenshots

Versions and configuration

• OS: Windows
  • OS Version: 10.0 1903
• Game Version: 0.E-2131-g3ac02c5 [64-bit]
• Graphics Version: Tiles
• Game Language: System language []
• Mods loaded: [ Dark Days Ahead [dda], Disable NPC Needs [no_npc_food], Aftershock [aftershock], C.R.I.T Expansion Mod [crt_expansion], Magiclysm [magiclysm], XEAS [XEAS], DinoMod [DinoMod], Crazy Cataclysm [crazy_cataclysm], No Rail Stations [No_Rail_Stations], sees-player icon, retrodays [sees_player_retro], Stats Through Kills [stats_through_kills], SpeedyDex [speedydex], Arcana and Magic Items [Arcana], Graphical Overmap [Graphical_Overmap], Advanced Gear [adv_gear], Blaze Industries [blazeindustries] ]

Additional context

[Qrox]

For some reason the random engine used in snippet_library::random_from_category returns the same value for different seed values.

Edit: seems that uniform_int_distribution produces significant bias when the random engine is cata_default_random_engine aka minstd_rand0 and the seed is small (see this stackoverflow question). We should either revert snippet_library::random_from_category to use std::mt19937 again or use std::mt19937 for cata_default_random_engine instead. @jbytheway What's your opinion?