"It should be simple to code"

“It should be simple to code” is fast becoming the most rant-inducing thing (to me) on this forum.

EDIT: Also applies to “It will be too hard to code”

Usually it’s about device firmware.

There are a limited number of people who can judge how easy or hard it is to change the firmware. The list is something like:

  • Developers who have seen the code and made one attempt to change it already.
  • Product/Programme managers closely involved in the launch of the product or the firmware update.

Even product/programme managers can sometimes get it very wrong.

Almost anyone else is guessing about the degree of difficulty involved. (I will happily update the list above if I’ve made obvious omissions).

Maybe extended discussions on the topic are grounds for moderator intervention (on one of 3 recent incidents this did in fact happen).

Anyway … rant over.

74 Likes

So, do you think it is non-coders using this phrase… or over-confident coders (in my experience, most of them)?

7 Likes

Interesting question. I don’t know. To me it makes next to no difference unless they’ve actually seen the code. Even in the case of someone who has seen statements by Ess and Cenk (for instance) they can then go on to make (perhaps unfounded) inferences and over-confident statements.

6 Likes

Thanks for placing this rant in an admirably calm way. It’s been in my head for ages as well.

I guess we really do this with everything though, don’t we? always speculating on business and marketing decisions, sports, etc even though we might not have any expertise.

17 Likes

Miserable git seal of approval.

44 Likes

Being myself on the category of the overconfident coders, my question is, are you working for elektron and got complaints about that, or are you just against innovation and enthusiasm in general? :grin:
I agree with you about estimating the impact on codebases, on the other end you have very negative people that like to say: “no, it’s not possible”, or “not in the current generation” and maybe didn’t even understand what is the subject of the post. What about those, do we need moderation intervention too?

Ps. My stack overflow profile User AndrewBloom - Stack Overflow

1 Like


An aside to this, is that I think the skill(s) of software development have been being devalued over the last decade e.g. “learn to code” programs/bootcamps, learning a language is pointless without the experience to go with it, it just creates a nimeity of inexperienced juniors; the advent of LLM’s means many now can get the “code” for various ideas which again lowers the perception of the worth of hiring a professional.

Then there are the mass layoffs by big tech companies who want to save money (many of whom are rehiring the sames roles at lower pay bands).

I have been a career software developer for about 12 years, and in IT another 10 before that (also programming for most of my life l, started with Basic on a Sinclair Spectrum ZX 48k I got for Xmas one year), I rarely consider a language or framework I haven’t used before to be a barrier, as those are just tools; so those “Learn Python in 12 weeks and get a coding job at the end” programs really irk me.

“Coding” in general feels derogatory, as the writing of the code itself is only a small bit of the work, alot of it is probably staring at the screen/into space and thinking, plus making and drinking coffee.
</opinionated rant>

Apologies for the rant, I’d put it on stack overflow, but it’s a ghost town over there since ChatGPT steamrolled in :upside_down_face:

10 Likes

It doesn’t even matter if a feature is easy to implement. It’s usually just a lack of time and those who give the money decide what’s being implemented in the end.

There are hardly any problems in IT which cannot be solved on a technical level. Sure, in this case, there are hardware restrictions but you can still go a looong way.

7 Likes

Is this where I go to complain about stakeholders asking for a modal to be placed inside a modal? I’m getting mixed signals

6 Likes

I believe it’s a mix of both, and sometimes even experienced coders can fall victim to oversimplification and misled/wrong assumptions of core concepts in a domain they’re unfamiliar. I know I still do even though I’ve been coding for 25 years.

The non-coders kinda assume that code is magic and almost anything can be done if just someone-wanted-to-do-it™ or if a company just wasn’t greedy and didn’t want to invest in better code™, what they lack is the experience to factor how costly it can be to get something done, the limitations hardware has, the current state of the codebase, how well implemented/abstracted some component was to be extended, etc.

On the coders side (which I fall into traps as well) it usually boils down to some wrong assumption being held that oversimplifies the actual implementation, it can be from lack of expertise in that domain, or a failure to model how difficult something can be sometimes. Experience just gave me the intuition of saying “this thought sounds too simple and too good to be true, what am I not seeing?” and analyse deeper, see limitations of my own logic to test what situations could exist that I’m overlooking and might complicate the solution. Another thought I try to hold all along is “a lot of smart people have worked on this, why would they’ve missed this simple solution?” again triggering a “what am I not seeing” overview. It’s good to stay humble, sometimes there’s a lot more constraints to consider that are easily overlooked or only gained with experience.

11 Likes

Coders may come close to estimating the software-level effort, but significantly underestimate the impact on Product (as in PM), where adding one more feature to an already burdened device or UI may significantly detract from, not improve, it.

10 Likes

Yes :). Though its not clear from my subject line, I tried to make it clear in my post that I was including overly-confident statements about ‘easy’ and ‘hard’ .

1 Like

Not at all. I will frequently make feature requests on the forum. I just don’t go on to say “this must be easy” or “I don’t think it will happen because it’s hard”.

EDIT: At least I don’t think I’ve done that :thinking:

2 Likes

I’ve seen this statement as well and all I can think of is “yeah… right…”

  • who and where the person who wrote the original code?
  • can you dig in that easily into what’s already coded?
  • can you find where to insert your “easy code”?
  • can you do that without breaking existing functions?
  • can you test it and make sure it won’t brick devices left and right?
  • and so on and so on.

thinking clear code for a function is easy enough.
integrating new stuff into existing code when you have no idea what it looks like and fearing you possibly destroy hundreds or thousands of units is fucking nerve-wrecking feeling and I wouldn’t want to be the one to do it.

8 Likes

i think its a byproduct of the mild “rockstarification” of coders in the past 15ish years. a big pool of coders think theyre hot shit because social media and AI exist. while no one else thinks its cool at all. but they sure do have an important opinion about things that involve code.

1 Like

I do this quite often, and now I’m a depressed coder. Can’t recommend.

3 Likes

hey chatgpt, write me a code for altering trigs for the digitakt

program alteringTrigs(digitakt);
  begin
    writeLn('Hello, Altering Trig!')
  end.

thanks, wow, that was easy

26 Likes

You could go one step further:

Because it should be simple to code, it should also be cheap/free.

2 Likes

It’s a bit different with some of those at least, because some of the information the decisions are based on is public. With a few rare exceptions (Synthstrom Deluge springs to mind) , firmware code is private.

“It should be easy” is easily the most tiresome line I see around here on a regular basis. If it’s so easy why not save the talk and build us a groovebox? Since these guys know everything surely they’d be instant synth mogul millionaires.

4 Likes