Getting Real<br>Elijah Potter

Sign In<br>Getting Real

While I am on PTO (as I am right now), my mind rarely strays far from Harper.<br>Perhaps you could call this “poor work-life bal­ance”, but I think oth­er­wise.<br>I be­lieve that liv­ing a ful­fill­ing life is hard, so hav­ing one takes ef­fort.<br>For me, part of hav­ing a ful­fill­ing life is to strive for great­ness.<br>Harper has the po­ten­tial to be great, and I am partly re­spon­si­ble for get­ting it there.<br>There­fore, part of liv­ing a ful­fill­ing life (for me) is to imag­ine ways to po­si­tion Harper so that it can best achieve its po­ten­tial.

That’s all a com­plex way to ex­plain why I am read­ing Getting Real by DHH and Jason Fried while on PTO.<br>I apol­o­gize for all the cor­po­rate lan­guage.

A few years ago, I read Rework, an­other book by the same au­thors.<br>In short, Rework is a col­lec­tion of ideas which, taken to­gether, form a set of ideals which must be em­bod­ied to build great busi­nesses .

In a sim­i­lar vein, Getting Real is a set of ideas which, taken to­gether, form a set of ideals which must be em­bod­ied to build great web ap­pli­ca­tions .

The two books are struc­tured sim­i­larly: As a col­lec­tion of (exceptionally) short chap­ters, each con­tain­ing a sin­gle idea or piece of ad­vice. They are both ex­cep­tion­ally fast reads. I do not think that read­ing both is nec­es­sary, but I do rec­om­mend that every en­gi­neer read at least one.

If you do not have the time (or the at­ten­tion span) to read ei­ther book, there is one idea which, I be­lieve, ac­counts for 90% of the ben­e­fit you stand to de­rive from them.<br>That is the idea of mass .

Mass

What is mass?

In physics, mass is the qual­ity of an ob­ject to re­sist ac­cel­er­a­tion.<br>The more mass an ob­ject has, the more dif­fi­cult it is to ac­cel­er­ate.<br>Con­versely, the less mass an ob­ject has, the eas­ier it is to move.

Why is this rel­e­vant to our jobs as soft­ware en­gi­neers?<br>If you al­ready know the an­swer to this ques­tion, you’re se­nior enough to skip this es­say.

Our job is to pro­duce prod­ucts which ad­dress con­sumer de­mand.<br>If we do not pro­duce prod­ucts which ad­dress a con­sumers needs, we are bad at our jobs.

The trou­ble with this is twofold.

We of­ten do not know what a con­sumer’s needs. Consumers of­ten do not know what they need. We must use ex­per­i­men­ta­tion and it­er­a­tion to find that in­for­ma­tion.

Once we have iden­ti­fied con­sumer needs, we can­not rest. Their needs change over time, so we must change with them.

How does this all re­late to mass?

In or­der to iden­tify a con­sumer needs and track it over time, we need to be able to change and it­er­ate quickly.<br>Soft­ware and sys­tems with a high mass are able to do this by de­f­i­n­i­tion , so to be able to do our jobs well, we need to also ad­dress this sec­ond-or­der prob­lem of min­i­miz­ing our mass.

Reducing Mass

So, I’ve ex­plained how re­duc­ing mass re­sults in sys­tems that are eas­ier to change, and sys­tems that are eas­ier to change are bet­ter po­si­tioned to ad­dress con­sumer’s needs.<br>How then do we re­duce mass?

Here are some off the top of my head, in no par­tic­u­lar or­der:

Create less soft­ware with fewer fea­tures.

Remove any­thing weigh­ing you down.

Don’t worry about mak­ing the wrong de­ci­sions and fo­cus on mak­ing more of them.

Choose to do one thing at a time.

Embrace con­straints.

There are sev­eral more listed in the books, with far more de­tail than I can af­ford here.<br>As I said be­fore, you should read them.

Published July 24, 2026 at 8:48 PM<br>Proofread by Harper.<br>RedditBlueskyLinkedInXFacebook<br>Comments

Comment<br>Name<br>Email<br>Message<br>SubmitOther Stuff<br>Do Not Type Your Notes<br>It didn't work for me, and if you reading this, it probably won't work for you either.

Hacker News sans AI<br>I like HackerNews, but I don't love that so much of it has turned into discussion of a single topic: AI. This is a version of HackerNews, filtered to remove any article focusing on __AI__. Refreshes about every ten minutes.

Markov Chains Are the Original Language Models<br>Back in my day, we used math for autocomplete.