News for hackers. Stuff that matters.
Re:Software effort estimation is mostly fake research by dkarl on Monday January 18, 2021 @04:26PM (#25826834)
> Even houses always go over time and money and that is fairly straight forward.

This is an incredibly important point. I was "raised" in commercial software at a time when there was a lot of condescension internally towards ourselves as an industry because we couldn't do estimation and planning, and building was the favorite comparison. A $100 million building couldn't run late or over budget. That would be unimaginable, and it doesn't happen because those people are too serious and professional to let it happen, not like us unserious, immature software doofuses. Turns out that was a complete lie, but at one time it warped the profession, because it made people believe the answers already existed, right around the corner, something you would figure out pretty soon just like last year you didn't know regular expressions and now you did.

The problem with that was that when somebody walked in with a button-down shirt, oozing confidence and saying they knew exactly how to consistently deliver software on time and bug-free, those people weren't laughed out of the room. Elaborate heavyweight processes, and their "lightweight" (hah!) cousins, could sell themselves as proven, established solutions. We could believe that bullshit because we had been fed the lie that other industries had conquered these problems.

Re:Software effort estimation is mostly fake research by wheelie_boy on Monday January 18, 2021 @08:31PM (#25828923)
I've seen a lot of industries compared to programming, like construction and manufacturing. For example, people trying to apply lessons from the Toyota Production System to programming.

This doesn't make sense, because we have systematized the construction of executable programs with extremely high reliability and predictability: we call it "Copy File"

It's the design that is the hard part, and the process of coding is the process of iterating on the design. Even turning the design into an executable is extremely reliable, either running via a VM or scripting engine, or compiling. Those are all very repeatable and reliable.

There are other disciplines that focus on creating and improving designs (like writing, painting, architecture, 3D modeling), but it's not like they're reliable or predictable either, unless you're designing very similar things over and over again. That's true of coding as well, but you don't really end up doing that, because once you've done it 3 times you should make a library and then never do that part again, just the unique bits.

In short, yeah, coding is its own thing, and I doubt it'll ever be predictable unless you're wasting a lot of time and effort doing very similar things over and over.

Re:Software effort estimation is mostly fake research by denisw on Tuesday January 19, 2021 @02:57AM (#25830987)
This point was recognized quite early. there is an influential article by Jack W. Reeves from 1992 [1] that looks at the manufacturing analogy and argues that "the code is the design" whereas the "manufacturing" is just the compilation and packaging (that is, super cheap and automated). I believe it‘s this line of thinking that inspired Agile practices, where software developers are seen as engineers rather than code monkeys.

[1]: https://www.developerdotstar.com/mag/articles/reeves_design....

Re:Software effort estimation is mostly fake research by galaxyLogic on Tuesday January 19, 2021 @12:12AM (#25830222)
> It's the design that is the hard part

Design but also requirements. You don't really know what should be built until you have a prototype. And estimating how long it will take to come up with a prototype that does what you think it should do, in other words coming up with requirements, is just the snake eating its tail.

Re:Software effort estimation is mostly fake research by waheoo on Monday January 18, 2021 @11:04PM (#25829872)
I like to compare programming to the marketing industry.

Highly creative work and when deadlines are involved the work becomes spurty (sprints) with corner cutting (scope reduction) to meet unrealistic and over committed deadlines.

Re:Software effort estimation is mostly fake research by konschubert on Tuesday January 19, 2021 @06:06AM (#25832218)
Yes.

Coding is not like assembling the machine. Coding is like inventing the machine.

Re:Software effort estimation is mostly fake research by meheleventyone on Monday January 18, 2021 @04:44PM (#25827029)
It doesn’t even have to be a big construction project to hit budget and schedule issues. Rebuilding our house had us going right back to the drawing board having to get new plans drafted and approved. It worked out much better in the end but the unforeseen cost and time overrun was quite anxiety inducing at the time.

But yeah the idea anyone has magically solved accurately predicting the future is completely bonkers.

Re:Software effort estimation is mostly fake research by gugagore on Monday January 18, 2021 @06:01PM (#25827774)
I like to remember "plans are useless but planning is indispensable". The plan is useless because of the challenge of predicting the future. But the planning is useful because trying to predict the future gets you asking good questions.
Re:Software effort estimation is mostly fake research by saurik on Monday January 18, 2021 @06:27PM (#25828011)
What is then terrifying to me then is when I find myself working with people who manage to, by way of treating "estimates" as "quotes", make it so demoralizing to plan anything--as it will be "used against you" later (which they do even if they were adamant they wouldn't)--that everyone does everything they can to discourage and sabotage any attempt at planning, because planning is valuable :(.
Re:Software effort estimation is mostly fake research by TeMPOraL on Monday January 18, 2021 @06:21PM (#25827965)
Also because planning gets you to make your desires explicit and to put them in a useful framework. Unexpected problems, disruptive as they may be, don't usually derail the whole plan. Some parts may need to be rethought, but others are still valid, and the whole thing keeps you focused.
Re:Software effort estimation is mostly fake research by ethbr0 on Monday January 18, 2021 @09:27PM (#25829291)
There was a submission on this (in the annals of HN) where someone made the point that, more often than not and under normal project scenarios, unexpected problems usually do derail the entire software engineering plan.

The gist being: estimates are given with zero as a lower bound, and some upper bound. In reality, there is no upper bound. Software engineering can and does involve unexpected problems that would take infinite time to solve.

Ergo, project timelines are more accurately averaged between {min_time} and {infinite}, and, given that, remaining risk is the only true metric that should be reported. I.e. how far are we along in terms of project completed, such that we can say only x% of the project remains, and therefore that 1-x% did not explode.

Re:Software effort estimation is mostly fake research by imtringued on Tuesday January 19, 2021 @09:01AM (#25833443)
The fundamental problem with estimates is that the recipient believes that the estimate will always turn out to be true, which is obviously not correct.

So, why can't we express the upper bound in combination with uncertainty?

Example: "There is a 50% chance that the feature will be delivered within 1 month."

As time progresses confidence starts increasing and we get closer and closer to how much time it actually takes (be that 5 days, 5 months or 5 years). On the day the feature is completed, confidence is 100% because the real delivery date has become known.

Re:Software effort estimation is mostly fake research by galaxyLogic on Tuesday January 19, 2021 @12:16AM (#25830253)
Programming is isomorphic to theorem proving (according to Curry-Howard).

Therefore trying to estimate a software project is a bit like trying to estimate how long it will take to prove a new theorem in mathematics, assuming you would know which exact theorem you want to prove.

Re:Software effort estimation is mostly fake research by garethrowlands on Tuesday January 19, 2021 @09:44AM (#25833825)
Er, isn't it just getting your program to typecheck that's equivalent to proving the theorem (by construction)?

I tend to find this isn't the main consumer of my software budget :-)

Re:Software effort estimation is mostly fake research by galaxyLogic on Tuesday January 19, 2021 @05:46PM (#25839962)
In a way I think yes but just getting the type-check right would be like proving ANY mathematical theorem. It is very easy to prove mathematical theorems if you don't care what those theorems are.

The challenge in mathematics, and programming, I believe, is what ("lemmas") to prove so that you can then prove more things that are useful for engineering, science, etc. . What functions to write in your program so that you can get its "main function" to do something useful.

Add to that the fact that much of the "type-checking" with current practical programming languages still has to be done informally, to describe what we expect of the result and then prove (to yourself and the client) that the program does that.

A lot of the thought-process that goes on in programmers' heads is informal reasoning proving things to yourself, "If I make this change here then that thing there must have a correct value, and then ..."

Re:Software effort estimation is mostly fake research by convolvatron on Monday January 18, 2021 @06:50PM (#25828187)
its more than that. there is a loose correspondence between plan and work. so changes in plans can often be projected onto changes in work. the same goes for expectations and dependencies.

the other thing about the plan is that it attempts to capture what completion might look like. where i come from, the things software developers do right before releasing code look very different than the things they do when they are starting a new project. the plan can allow you to try to do some orchestration, ensure minimum quality bars are met, etc.

having a fuzzy map of the future that you're constantly having to correct sucks. having no map at all leads to the kind of carpe diem development organizations that never seem to finish anything.

Re:Software effort estimation is mostly fake research by meheleventyone on Tuesday January 19, 2021 @02:10AM (#25830798)
More than that we need to be flexible to adjust a plan. You’d be amazed at how many people who quote that aphorism aren’t actually flexible. At least with a building they are also motivated by not leaving a physical monument to their failure.
Re:Software effort estimation is mostly fake research by smichel17 on Monday January 18, 2021 @08:16PM (#25828831)
I'll raise (shrink) you an estimated-at-1-week landscaping project for my house that ended up taking just over 2. Wasn't particularly anxiety inducing, though, as we expected some overrun and both parties (re)negotiated in good faith.
Re:Software effort estimation is mostly fake research by reallydontask on Tuesday January 19, 2021 @03:31AM (#25831184)
Oh man, this hits home so much.

I used to lead a software department in a small company that sold software to, mainly, the construction industry, so the go to comparison was building houses, which according to my boss, were always on time and on budget, unlike software.

The funniest thing about it was that our software was used by huge projects: Hinkley Points C, HS2, etc ... which obviously have an great track record of being on time and on budget ಠ_ಠ

Re:Software effort estimation is mostly fake research by eru on Tuesday January 19, 2021 @02:02AM (#25830767)
From what I've heard, in construction most of the delays happen with the foundations.

Everything that's below the ground is hard to predict, because you seldom know exactly what's in the ground.

Once your construction has progressed above ground, it's somewhat more predictable.

(But delays are still possible.)

Re:Software effort estimation is mostly fake research by tilolebo on Tuesday January 19, 2021 @03:26AM (#25831151)
There's also the problem of of a company working on multiple construction sites in parallel.

In software engineering, this also happens of course, especially if you develop in house. And you also need to factor in the maintenance, as it's usually performed by the same team.

Re:Software effort estimation is mostly fake research by amadvance on Tuesday January 19, 2021 @03:30AM (#25831175)
I think that adding a feature to an existing software product is more like adding a new floor to an existing house.
Re:Software effort estimation is mostly fake research by mathw on Tuesday January 19, 2021 @04:56AM (#25831723)
Sometimes it's like adding a new floor to an existing house in between two of the existing floors without any of the occupants noticing.