Breaking News
Engineering Pop Culture!

The Case of the Unconventional Coder

NO RATINGS
View Comments: Newest First | Oldest First | Threaded View
Page 1 / 2   >   >>
prabhakar_deosthali
User Rank
CEO
Patches in memory
prabhakar_deosthali   5/6/2014 9:04:19 AM
NO RATINGS
In early eighties , we used manual code by using coding sheets to modify  our programs in binary code using piono key switches on the front panel of the computer. This was because the recompilation was done only once in a while .Any changes of the programs had to be done as patches as the recompilation of the code was not possible every now and then . So we used to keep a record of the patches until the enxt compilation. One of or smart programmers kep all the record of the patches in his brain, never recording it on the paper. This led to many a situations where we used to spend hours in debugging a code until this smart programmer pulling out a patch from his memory and fixing the problem.

Bert22306
User Rank
CEO
Perceptions
Bert22306   5/5/2014 3:57:10 PM
NO RATINGS
Reading the article and the some of the comments, my reaction was, I'm sure glad I don't work in that environment. Took me some time to figure out how to get this across.

Sure, code should be clear in preference to JUST efficient. That's not my problem. My problem is, engineers who feel they are "subordinate" to management, or management that feels it has "ownership" of the work. Both concepts are foreign to me and to where I work.

Where I work, the management of the actual engineering work is done by engineers, typically lead engineers. And "ownership" is definitely with the engineers that did the work. Everyone knows who they are. I wouldn't have it any other way. Surprising that others haven't reacted similarly.

jonnydoin
User Rank
Blogger
Re: Management
jonnydoin   5/3/2014 1:26:21 AM
@Jimmy99Neutron: I agree that expectations should be clear. Actually, the programmer mentioned is a very good Engineer. The case described took place during his integration into the project, and into our Aspect Oriented methodology. It took a little time and effort to convince him that I wanted his brilliant mind on the logic and design, not really on cryptic descriptions. He ended up contributing with many good designs, and also learning a lot with his peers also. Instead of adhering to rigid rules, the person must adopt the style due to the perception of gain. Peer review and code clarity are much more than management concerns. When people realize that they can easily tackle projects with 300k lines of code by practicing a clear style, they embrace it. In a well integrated team, we can write code at ten hands and the code seems to have been written by the same person. That's what peer review does for you: you get all the best contributions to rapidly disseminate through all the programmers. - Jonny

GSKrasle
User Rank
Freelancer
Re: cin >> isReaderCoder;
GSKrasle   5/2/2014 7:22:38 PM
NO RATINGS
Hmmm,

It also matters what you get used to. I can type this useful formula in my sleep:

=IFERROR(MID(L60,FIND("[",L60)+1,FIND("]",L60)-FIND("[",L60)-1),IF(ISERR(FIND("\",L60)),"",MID(L60,FIND("*",SUBSTITUTE(L60,"\","*",LEN(L60)-LEN(SUBSTITUTE(L60,"\",""))))+1,LEN(L60))))

It's probably Bad Style, but I like to think that's because this language never heard of comments or formatting. You can see what it does though?

Seriously, I use stuff like this all the time, but I try not to inflict it on others; I Copy-PasteValues before the data gets to them. 

Jimmy99Neutron
User Rank
Rookie
Management
Jimmy99Neutron   5/2/2014 12:33:33 PM
NO RATINGS
Firstly, as this person was a subordinate, expectations for his work should be clear, a process is to be followed as it benefits the team, the project and its management and the product ( reduced maintenance, robustness, correctness, etc.).  

Knowing this persons history, the guidelines or standards should have been emphasized.  Tolerating deviations incurs risk, re-work, time/schedule implications, quality and money.  As the Leader of the project you "own" the results.  In no way does this condone the Coders preferences and choices.

ghova
User Rank
Manager
Re: Eschew obfuscation
ghova   5/2/2014 11:17:11 AM
NO RATINGS
Re "clarity is the prime virtue."  Amen - writing turgid code that's "efficient" costs more than it saves.  From a cost-benefit standpoint the optimal program is no faster than adequately fast (C and assembly programmers reading this are probably reaching for the smelling salts).  If future additions makes it unacceptably slow, then and only then do you make an analysis of hot spots and work on those that produce the greatest bang for the buck until the program is again adequate.

Re the manager who wanted three clever lines replaced with one that would be familiar to other readers: they promoted the right guy to manager.

However, the managers referenced, who insist on rewrites to improve clarity, are thin on the ground, at least in my experience (1959 - present).  The ones I've encountered would not take kindly to the programmers taking the time to (in their view) "beautify" code.  And it wouldn't win them any friends among THEIR managers.

BrainiacVI
User Rank
CEO
He followed a Magazine Article on Management
BrainiacVI   5/1/2014 6:01:42 PM
NO RATINGS
I had a (terrible) manager once who insisted one of the programmers on my team recode the three move statements he used to make a string and use sprintf instead.

So we went from three small, fast MOV statements (when compiled) to a slow, interpreted function call that would have added K's of code if it wasn't already used elsewhere.

All for the edict of "One line of code is easier to maintain than three."

perl_geek
User Rank
Freelancer
Eschew obfuscation
perl_geek   5/1/2014 5:32:24 PM
There's a place for code like that; the "Obfuscated C" competition, where programmers can get the perversity out of their systems before producing real work.

People acquire all sorts of ideas about what constitutes good code, from all sorts of haphazard sources, but managers should make it clear upfront what constitutes merit. (It took me a while to figure it out, and I ask forgiveness of my victims.)

A program is not just a set of instructions to the compiler, it's an open letter to the next unfortunate who has to maintain it. As long as it does the job in a reasonable time, clarity is the prime virtue. It should be obvious what it's doing, and why.  This can take a surprising amount of work, distilling the essence of the problem's structure and using the most powerful tools available.

One of the unfortunate consequences of CS education is that people feel compelled to recreate compilers and usurp the functions provided by the operating system, presumably to justify the pain of studying them. KISS, dammit!

 

 

betajet
User Rank
CEO
Re: C? No!
betajet   5/1/2014 1:43:25 PM
NO RATINGS
anon wrote: This man is an APL programmer at heart.

I'd say more likely a LISP programmer at heart.  The LISP conditional expression is COND, which is a superset of C's conditional expression.

If this were APL, there'd be a bunch of outer products and none of us would be able to understand it :-)

Sheepdoll
User Rank
Blogger
Re: C? No!
Sheepdoll   5/1/2014 1:42:14 PM
NO RATINGS
% For a moment I thought perhaps this programmer was a postscript programmer.  We use lots of % as these are comments.

% Normally I code my lashups in postscript. This last week I have been coding in C.  Must remember to put the ifs and elses at the top of the bracketed {} blocks not the bottoms.  << and >> are used to denote dictionary structures.

% Perhaps this is why I prefer AVR ASM for embedded.  It feels really strange to use C on an 8 bit micro with only 2K of SRAM.  printf and fprintf are not the best solution in such a limited environment.

Page 1 / 2   >   >>
More Blogs from Engineering Pop Culture!
This collection of places from technology history, museums, and modern marvels is a roadmap for an engineering adventure that will take you around the world.
A future engineering student gives his advice on making the most of the time-honored tradition of the college visits road trip.
Manufacturing engineer Jeremy Cook discusses a few machine failure problems that seemed complicated at the onset, but were quite simple to solve in the end, and the lessons he learned.
Jeremy Cook's Cigar Box Creepster opens up when triggered with a knock or other sound, then slams back down as if to express “his” annoyance.
How to deal with a meatball of a manager who issues an edict that makes no earthly sense? Ignore it.
Top Comments of the Week
August Cartoon Caption Winner!
August Cartoon Caption Winner!
"All the King's horses and all the KIng's men gave up on Humpty, so they handed the problem off to Engineering."
5 comments
Like Us on Facebook

Datasheets.com Parts Search

185 million searchable parts
(please enter a part number or hit search to begin)
EE Times on Twitter
EE Times Twitter Feed
Radio
LATEST ARCHIVED BROADCAST
David Patterson, known for his pioneering research that led to RAID, clusters and more, is part of a team at UC Berkeley that recently made its RISC-V processor architecture an open source hardware offering. We talk with Patterson and one of his colleagues behind the effort about the opportunities they see, what new kinds of designs they hope to enable and what it means for today’s commercial processor giants such as Intel, ARM and Imagination Technologies.
Flash Poll