Wed Oct 17 - Avoiding Code Smells by Walter Bright

[Laeeth Isharc]

On Thursday, 1 November 2018 at 22:37:59 UTC, unprotected-entity 
wrote:
> On Thursday, 1 November 2018 at 03:10:22 UTC, H. S. Teoh wrote:
>>
>> Actually, code within a module *should* be tightly coupled and 
>> cohesive -- that's the whole reason to put that code inside a 
>> single module in the first place.  If two pieces of code 
>> inside a module are only weakly coupled or completely 
>> decoupled, that's a sign that they should not be in the same 
>> module at all.  Or at the very least, they should belong in 
>> separate submodules that are isolated from each other.
>
> How does one determine, whether a 10,000 line module, is 
> tightly coupled and cohesive?


You take a look through it and make a judgement.

> Only the author can make that statement - which they naturally 
> will, even if it's not true.

?

> An outsider, seeking to verify that statement, has a hell of a 
> job on their hands...(I for one, think code smell immediately).

There is a basic question in life.  Do you believe in discernment 
(if possible informed by data) or are you someone who makes 
decisions on the basis of evidence and believes that anything 
else is completely arbitrary and nothing more than a matter of 
opinion.

My impression is that the values of the D community tend more in 
the direction of recognising the importance of discernment. If 
someone is somebody who believes more in 'evidence', policy and 
rules then it probably isn't going to be satisfying expecting 
one's values to be shared on a wide scale here.

People also differ in their working memory and the degree to 
which they naturally think associatively.  Chopping up everything 
into small pieces favours those who have a smaller working memory 
and who think more analytically but it's a disadvantage for those 
who have a large working memory and think associatively.  For a 
private project that's something to be resolved between the 
relevant people, but I don't think it's reasonable to say that 
large files per se are wrong, just because they aren't your cup 
of tea personally.

I think that lots of things seem clear in theory but the 
difference between theory and practice is often quite large.

In practice Phobos is very readable.

And on the other hand, I have seen an experienced and capable C# 
programmer struggle to understand an intranet site written in the 
approved way by an experienced person who was well-trained at 
Microsoft.  I couldn't understand it either so I concatenated all 
the little itty bitty files, pulled out the data structures and 
then it was easy.

> I don't use a particular language. I'm more interested in 
> design and architecture.

Can one really speak of that kind of design in the abstract ?  
Language features shape your choices and lead to large shifts in 
the optimum.  If you don't have design by introspection as a 
possibility you are going to pick something else.

> In the age of 'lock-down-everything', increased modularity is 
> becoming more important. A monolithic module approach, is 
> already outdated, and risky, in terms of developing secure, 
> maintainable software....

If you can't understand the program does that make you more or 
less secure?  Security requires also to understand the behaviour 
of the system as a coherent whole.  I think D programs are pretty 
easy from that perspective.

> Is that really such a bad idea? Are there no programmers out 
> there in the D world that might think this could be a good, 
> additional tool, to give programmers, so they can better 
> architect their solution?

Burden of proof is on you to write a DIP and make an argument for 
it.  I am not sure you would find it easier to get a change into 
C++.  Look at how difficult it is for Walter sometimes ; and he 
has just a little earned credibility.  Same thing for Guido - he 
had such little fun with a recent PEP he decided to retire from 
BDFL of python.

> The amount of push back in the D community on this idea, is 
> really odd to me. I'm still trying to understand why that is.

Persuading people isn't easy even if it's a good idea.  Look at 
the pushback from C++ over static if.  They crippled it when they 
finally did relent.  It's a bit entitled to think that if you 
can't persuade people without having written a DIP that it's them 
not you!

> Are D programmers just hackers, insterested in getting their 
> code to work, no matter what? Are their not enough Java/C# 
> programmers coming to D - and bringing their design skills with 
> them?

Would you mind explaining why you think that people from mass 
communities have design skills by virtue of having come from a 
mass community?  Walter and Andrei are just a little bit known 
for their design capabilities so the bar is quite high.  I think 
it's possible you might have things topsy turvy.

Making D code work is rarely a problem.

Every nation has its own strengths and weaknesses and the same is 
true of language communities.

Having worked with D professionally since 2015 and with a decent 
size codebase in relation to what I have heard from others, my 
experience of working with D programmers is that I have never had 
to say please don't make this dirty hack.  Whereas you know you 
could copy memory just this once, please don't bother making this 
bit fast, I really think we can just make a start on a prototype 
before doing a full design, do we really need to refactor this 
part of the code now - those things whilst not common are more 
frequent sources of disagreement with D programmers.

So like I say I think you have it upside down.  The D programmers 
I work with have a strong focus on design and doing things in the 
right way.  Take a look at Atila Neves work on excel-d or dpp.  
It's a model of clean code.  There are some outstanding Java and 
C# programmers but they are much sparser than in the D community.

I spoke to the then CTO of the company where I am now involved.  
I was telling him about my project and he said ooh you want to 
spend a couple of weeks mapping out your class hierarchy.  I said 
I don't have any classes.  Well I did have two, but I took out 
one and John Colvin got rid of the other.  The expression on his 
face was priceless.

You can write D like Java or C# if you like.  Depends what you 
are doing - you certainly don't need to, and many people who have 
been using D for a while often find writing code using structs 
much better for their goals and given their values.

When one encounters a new idea that's unfamiliar sometimes it's 
easy to think that because it's unfamiliar it must be unsound.  
That can be a mistake.  It might be better to suspend judgement 
for a while and keep an open mind.