davideisinger.com

My personal website
Log | Files | Refs | README

taylor-town-5siv9a.txt (7922B)


      1 [1]taylor.town
      2 [2]about [3]now [4]spam [5]rss [6]hire
      3 
      4 cloaca maxima
      5 
      6 When to Build Millennia Sewers
      7 
      8 In the mid-1800s, [7]every building in central Chicago was raised 10ft (30m).
      9 Yes, they literally used [8]jackscrews to lift entire city blocks up
     10 one-by-one.
     11 
     12 Chicago had to [9]hotfix production because they built the city on the
     13 shoreline of Lake Michigan, where filth accumulated without natural drainage.
     14 They lifted the entire city after it was built so they could add sewers and
     15 prevent flooding.
     16 
     17 For comparison, Rome’s [10]Cloaca Maxima (“Greatest Sewer”) is still in-use
     18 after 2,400 years.
     19 
     20 So why didn’t Chicago just build it right the first time?
     21 
     22   • [11]Irreversible Decisions
     23   • [12]Unintended vs. Unforeseen
     24   • [13]Always Scale Down
     25   • [14]Labor & Materials
     26   • [15]Awful Architecture
     27 
     28 Irreversible Decisions
     29 
     30     Some decisions are consequential and irreversible or nearly irreversible –
     31     one-way doors – and these decisions must be made methodically, carefully,
     32     slowly, with great deliberation and consultation. If you walk through and
     33     don’t like what you see on the other side, you can’t get back to where you
     34     were before. We can call these Type 1 decisions. But most decisions aren’t
     35     like that – they are changeable, reversible – they’re two-way doors. If
     36     you’ve made a suboptimal Type 2 decision, you don’t have to live with the
     37     consequences for that long. You can reopen the door and go back through.
     38     Type 2 decisions can and should be made quickly by high judgment
     39     individuals or small groups.
     40 
     41     As organizations get larger, there seems to be a tendency to use the
     42     heavy-weight Type 1 decision-making process on most decisions, including
     43     many Type 2 decisions. The end result of this is slowness, unthoughtful
     44     risk aversion, failure to experiment sufficiently, and consequently
     45     diminished invention. We’ll have to figure out how to fight that tendency.
     46 
     47     – [16]Jeff Bezos
     48 
     49 The Cloaca Maxima didn’t magically start out as the Greatest Sewer. It began as
     50 an open-air canal, then was modified and renovated and connected to the
     51 aqueducts.
     52 
     53 The Romans probably made mistakes, but they didn’t make any wrong irreversible
     54 decisions. To build something that lasts, make sure the architecture is correct
     55 where it counts.
     56 
     57 The Chicago sewage disaster was technically reversible, but extremely expensive
     58 and painful.
     59 
     60 Put “wiggle-room” in your architecture. Plan for repairs. Add backdoors,
     61 engine-hoods, seams, and spaces. Emergency plans are generally cheap to include
     62 in early phases of design.
     63 
     64 Unintented vs. Unforeseen
     65 
     66 [17]Exxon executives knew that CO₂ emissions would harm Earth.
     67 
     68 Exxon willfully ignored its own research. Climate change was unintended but not
     69 unforeseen.
     70 
     71 Prophets are silenced when apocalypses seem bad for business.
     72 
     73 But remember – all apocalypses are opportunities for entrepeneurship. Exxon
     74 could’ve made billions by diversifying themselves with renewable energy. They
     75 acted against their own self-interest by ignoring their facts.
     76 
     77 To prevent long-term disaster, solve the hard problem of aligning incentives.
     78 Build systems so that all constituents predict and prevent impending doom.
     79 
     80 Transparency thwarts [18]own goals. It’s difficult to do stupid things when you
     81 do stupid things publicly.
     82 
     83 Always Scale Down
     84 
     85     There’s really two ways to design things. You can either sort of start with
     86     small things and scale them up or you could start with big things and scale
     87     them down…
     88 
     89     So suppose you want to build a system for like 10,000 people to use
     90     simultaneously. One way of doing it would be to start with the system,
     91     design it for 10 people and test it like that and scale it up 10,000. The
     92     other way would be to design it for like 100,000,000 people – I mean do the
     93     design for that – and then scale it down to tens of thousands. You might
     94     not get the same architecture. You might get a completely different
     95     architecture. In fact, you would get a different architecture.
     96 
     97     And I think it’s a really bad idea to start at a design for 10 or 100
     98     things and scale it up. It’s better to start with an architecture that you
     99     know will work for a few trillion things and scale it down. It will
    100     actually be less efficient when you’ve got your 10,000 things than when you
    101     scaled up, but you’ll know that you’ll be able to scale it up later. So
    102     it’s good.
    103 
    104     So rather than ask, “how do we get to five nines?”, let’s make it more
    105     interesting! Let’s start at 9,999 nines reliability and scale it down.
    106 
    107     – Joe Armstrong from [19]Systems that run forever and self-heal and scale
    108 
    109 If you can afford it, throw a few extra zeroes on your designs.
    110 
    111 Labor & Materials
    112 
    113 Carefully compare lifetime, labor, and materials.
    114 
    115           lifetime   repair   labor materials
    116 asphalt  20 years   moderate  $     $
    117 concrete 30 years   difficult $     $$
    118 stone    100+ years easy      $$$$  $$$
    119 
    120 Pay particular attention to labor – 9 women can’t make a baby in 1 month.
    121 
    122 Exercise for the reader: Which is cheaper, a Nespresso machine or a [20]
    123 percolator?
    124 
    125 Awful Architecture
    126 
    127 Sometimes there are no tradeoffs.
    128 
    129 Some decisions are awful in every dimension.
    130 
    131 [21]Dvorak keyboards reduce finger fatigue using the same materials as QWERTY
    132 keyboards.
    133 
    134 [22]Juicero famously launched a high-tech product that was inferior to
    135 traditional juicers [23]in every comparable way:
    136 
    137     After taking apart the device, venture capitalist Ben Einstein considered
    138     the press to be “an incredibly complicated piece of engineering”, but that
    139     the complexity was unnecessary and likely arose from a lack of cost
    140     constraints during the design process. A simpler and cheaper
    141     implementation, suggested Einstein, would likely have produced much the
    142     same quality of juice at a price several hundred dollars cheaper.
    143 
    144 If you want to create lasting sewers, study sewer architecture and its impacts.
    145 What do good sewers have in common? What do bad sewers look like? What
    146 tradeoffs exist with sewage systems? Are there any promising-yet-untested sewer
    147 designs? Why do sewers go into disrepair? What societal factors prevent sewers
    148 from being made in the first place? Who truly controls the sewers?
    149 
    150 Great architects think ahead, but don’t let ambitions run amok. They anticipate
    151 irreversible changes and second-order effects. They consider all the costs –
    152 labor and materials and maintenance and environmental impact. They always stay
    153 ahead-of-schedule and within their budget. And despite the overwhelming
    154 constraints, great architects build millennia sewers whenever and wherever they
    155 can.
    156 
    157 
    158 References:
    159 
    160 [1] https://taylor.town/
    161 [2] https://taylor.town/about
    162 [3] https://taylor.town/now
    163 [4] https://newsletter.taylor.town/
    164 [5] https://taylor.town/feed.xml
    165 [6] https://taylor.town/hire-me
    166 [7] https://en.wikipedia.org/wiki/Raising_of_Chicago
    167 [8] https://en.wikipedia.org/wiki/Jack_%28device%29#House_jack
    168 [9] https://www.reddit.com/r/ProgrammerHumor/comments/blhec6/fixing_bugs_in_production/
    169 [10] https://en.wikipedia.org/wiki/Cloaca_Maxima
    170 [11] https://taylor.town/millennium-sewer#irreversible
    171 [12] https://taylor.town/millennium-sewer#unintended
    172 [13] https://taylor.town/millennium-sewer#scale-down
    173 [14] https://taylor.town/millennium-sewer#labor-materials
    174 [15] https://taylor.town/millennium-sewer#awful
    175 [16] https://www.sec.gov/Archives/edgar/data/1018724/000119312516530910/d168744dex991.htm
    176 [17] https://en.wikipedia.org/wiki/ExxonMobil_climate_change_denial
    177 [18] https://en.wikipedia.org/wiki/Own_goal
    178 [19] https://www.youtube-nocookie.com/embed/cNICGEwmXLU?start=433
    179 [20] https://en.wikipedia.org/wiki/Coffee_percolator
    180 [21] https://en.wikipedia.org/wiki/Dvorak_keyboard_layout
    181 [22] https://en.wikipedia.org/wiki/Juicero
    182 [23] https://en.wikipedia.org/wiki/Juicero#Criticism