Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Software Architecture Guide

martinfowler.com

Post reply on HN
Print view View original

51–60 of 303 posts

Re: Software Architecture Guide

#51

Post by tootie »

blub wrote: post #7

Whenever I see yet another article by Martin Fowler, Kent Beck, Robert Martin & co I ask myself - where is the evidence for what they are preaching? What are the graphs based on? What is the evidence behind the proposed rules? These authors are clearly accomplished blog and book writers, but anyone whose software-related accomplishments are not open source or at least well known should have their statements scrutiniz…

It's good point, but most of what I read on his blog feels like a distillation of best practices that I've seen mature in the industry. It all rings true and matches my experience of what works best. But can I prove it? No, not really.

Re: Software Architecture Guide

#52

Post by legulere »

scarface74 wrote: post #15
chrisseaton wrote: post #8

Earlier quoted context omitted.

For some of these speakers, I’ve literally never seen any evidence that they’ve written a single line of code. Why should we listen to Fowler? What large successful projects has he created using his ideas are there that I can look at? Where’s the data behind his slides? It’s all just talk talk talk. I’d listen to someone like Chris Lattner talk about how to design and write software because I can see whatever approac…

Chris Lattner knows how to write system software and compilers which are completely separate beasts from the typical bespoke enterprise or software as a service apps that most developers write every day.

System Software isn’t all that different from enterprise software. The business part is simply of technical nature. The biggest difference usually is that the software developers don’t have a deep understanding of the business rules.

Re: Software Architecture Guide

#53

Post by scarface74 »

vfc1 wrote: post #50
scarface74 wrote: post #15

Earlier quoted context omitted.

Chris Lattner knows how to write system software and compilers which are completely separate beasts from the typical bespoke enterprise or software as a service apps that most developers write every day.

Exactly, many times these famous authors advocate for practices that make sense for building compilers and frameworks or libraries, but that make little sense for building applications. Building applications is very different, because an application is the very end of the software tree dependency. Some practices make sense for both types of software, but many others don't and some of these authors don't seem to be ab…

And sometimes it makes sense to take shortcuts that don’t make sense from a long term maintenance standpoint just to survive long enough to get product market fit. Get funding and survive.

Twitter is a perfect example of this. Everyone agrees that Twitter made some bad architectural decisions that caused the frequent appearance of the “fail whale”. But eventually, they got enough funding via combination of the VC funding and later the public market to rearchitect their system.

Re: Software Architecture Guide

#54

Post by geofft »

streetcat1 wrote: post #49
geofft wrote: post #40

Earlier quoted context omitted.

That's, again, not what I'm asking. I'm not asking for the theoretical underpinnings of why this is a good way to represent 200-500 classes. I'm asking why, if I and two coworkers need to code up 200-500 classes by next month to meet a promise to a customer, spending time on UML will make the customer happier. I have a programming language that itself has strong theoretical underpinnings. I also have a pretty good hu…

First, I must say that the inventors of UML saw it as the last layer. The grand vision was a complete code generation from UML diagrams. And this was the overall grand vision that drove OO in general. I think that this is was happening now with the "low code" startups. The whole idea is to separate the global decisions (which are hard to change) - e.g. architecture, what classes, what each class do, from the local on…

Again, I am not asking how to use UML, or why the inventors of UML designed it in a certain way, or why it sounds like it could be a good idea, or whether there are other systems like UML that are good ideas (I can just use those systems directly then!).

I am asking what evidence there is that UML helps me deliver better products more efficiently. Are there successful software projects that do this complete code generation thing (and how do they compare to directly writing real code in more efficient programming languages with rich typesystems)? Are there development efforts where giving up and deciding that certain architectural elements are hard to change has been the right approach (in my experience it generally never is)? What are their names? Do they have experience reports? Are there numbers?

I am looking for, at the least, a specific name like "The Chrysler Comprehensive Compensation System took these approaches to development." (But hopefully not a massive and expensive failure like C3.)

Re: Software Architecture Guide

#55

Post by mobjack »

millerm wrote: post #9
blub wrote: post #7

Whenever I see yet another article by Martin Fowler, Kent Beck, Robert Martin & co I ask myself - where is the evidence for what they are preaching? What are the graphs based on? What is the evidence behind the proposed rules? These authors are clearly accomplished blog and book writers, but anyone whose software-related accomplishments are not open source or at least well known should have their statements scrutiniz…

I have had this thought for many years. Where is all the perfectly designed, bug free, maintenance-bliss, fully documented, fully tested, future-proofed code located so we can all marvel at its glory?

You are comparing the perfect against the good while Fowler is likely comparing the good against the bad.

Taking an extra week or two to clean up bad design decisions early in a project can pay dividends in the future.

Spending months extra trying to do the perfect architecture is harder to justify.

Re: Software Architecture Guide

#56

Post by scarface74 »

legulere wrote: post #52
scarface74 wrote: post #15

Earlier quoted context omitted.

Chris Lattner knows how to write system software and compilers which are completely separate beasts from the typical bespoke enterprise or software as a service apps that most developers write every day.

System Software isn’t all that different from enterprise software. The business part is simply of technical nature. The biggest difference usually is that the software developers don’t have a deep understanding of the business rules.

I’ve had to maintain a bespoke compiler/IDE/VM for Windows Mobile/CE before and have spent two decades writing business software and architecting around the entire stack from networks, to databases, to web servers, load balancers, monitoring tools etc. I’ve also written SOWs, Done project plans, led teams (hated every minute of it). I’ve seen both sides.

The types of optimizations and thought processes around enterprise software is completely different than the optimizations around compilers.

The “entire stack” with enterprise software is much different than the “entire stack” with the compiler. I haven’t had a need to look at the assembly created from my compiler as a business software developer since around 2002.

For context, my first hobby programming was in the 80s optimizing 65C02 programs by using zero page memory access to eke a little more speed out of the 1Mhz processor and trying to use branch statements instead of JMP statements because they were faster.

Re: Software Architecture Guide

#57

Post by streetcat1 »

traderjane wrote: post #43
streetcat1 wrote: post #32

Earlier quoted context omitted.

Sure. So UML is based on two main research areas - UML class diagram is based on rational algebra theory. A correct UML class diagram should follow the same normalization roles as defined by Codd seminal paper - "A Relational Model of Data for Large Shared Data Banks". You can ask, how does Prof Codd knows that this is is the correct way to describe and implement data. Well this follow directly from set theory. Again…

geoftt isn’t asking whether metaphorical TLA+ is academically fruitful. He’s asking if Amazons et al. have _continued_ to use it, and whether there’s some volume of evidence attesting to this.

Sure. So if amazon did not continue to use it. Does TLA+ suddenly become not useful?

My point here is that those tools are theoretical CS tools based on the underlying nature of computers - which are a discrete digital state machine.

Why do you even use high level languages, why not just program in assembler?

Re: Software Architecture Guide

#58

Post by rhizome31 »

chrisseaton wrote: post #47
rhizome31 wrote: post #44

Earlier quoted context omitted.

I might be wrong, but did Fowler pretend to follow the scientific method? I work on developing business web applications and Fowler's work has been very helpful to me, especially when I was a beginner. I still sometimes refer to PoEAA when picking up a framework that implements a pattern described there. However I'd be surprised if Fowler had anything useful to teach to someone working in your field.

> did Fowler pretend to follow the scientific method His company calls him a 'scientist' - a 'chief scientist' no less - https://www.thoughtworks.com/profiles/martin-fowler and presumably bills accordingly.

Oh I see, that looks like marketing BS indeed. Not cool that he agreed with this job title. But to be fair, these days you also see a lot of 'data scientists' whose 'scientific' work consists in putting some statistics together and plot them with Python.

Anyway what I was saying was based on reading his books, which again I think have real-world value for some areas of software development. As I said in another comment to me his work consists in collecting and documenting existing practices. I don't think he pretends to teach magic recipes. Each technique is usually presented with potential drawbacks and context where it might apply. That's why I was telling you to use your judgement because these techniques are not supposed to be one-size-fit-all silver bullets. The point is more to identify and describe recurring patterns so that we can communicate more clearly, hence my light-hearted joke about how it helps writing your commit messages.

I agree that it would be great to have strong scientific studies showing that such practice works better than such other practice and I bet Fowler would agree too. In fact I watched a talk recently where he advertised a book called "Accelerate" (not by himself) where apparently the authors attempted to evaluate the impact of practices associated with DevOps based on statistical data.

The thing is it's much easier to demonstrate or measure the performance of an algorithm than the impact of a code change on the whole life-cycle of a software project. So we resort to experience and intuition not because we don't value the scientific method, but because the scientific method is too difficult to apply in this context.

Re: Software Architecture Guide

#59

Post by EdwardDiego »

> All I see is yet another reasonably sounding yet unsupported piece of folklore.

Another artefact of the fact that software developers are still craftspeople - as in "someone skilled at making things by hand".

Software development output is something that no-one has yet successfully quantified. There's no easy metric to assess a non-productive software developer from a productive one. We can all tell the difference, but good luck measuring it. In my country, which (thankfully) lacks America's "at-will" employment contracts, it does mean that it's quite hard to fire an underperforming dev, because how the hell do you prove they're underperforming? What metric do you use?

Bug rate is a bad one, because very smart people working in a new problem domain may register an significant uptick in bugs - doesn't mean that they're incompetent, it's just the result of doing new stuff in an unfamiliar domain in a highly complex environment.

Lines of code? The less said about that metric and how gameable it is, the better.

Us still being craftspeople is also why we can command such high wages relative to the rest of the economy - what we do is not yet automated, so we can still draw a premium.

Re: Software Architecture Guide

#60

Post by traderjane »

streetcat1 wrote: post #57
traderjane wrote: post #43

Earlier quoted context omitted.

geoftt isn’t asking whether metaphorical TLA+ is academically fruitful. He’s asking if Amazons et al. have _continued_ to use it, and whether there’s some volume of evidence attesting to this.

Sure. So if amazon did not continue to use it. Does TLA+ suddenly become not useful? My point here is that those tools are theoretical CS tools based on the underlying nature of computers - which are a discrete digital state machine. Why do you even use high level languages, why not just program in assembler?

If the Amazons of the world tried and dropped TLA+, that is signal to people who were on edge of technological adoption. If we received data on its efficacy during its trials, we'd be even more informed, as people are interested in more than the internal consistency of a system.

Right now some people have a theory that Uncle Bob is noise in the sea of architectural opinion, and we can't move forward because we don't have enough data. The call for more data is welcome by me.

Post reply on HN