Posted on 09/20/2010 8:52:32 AM PDT by RogerFGay
Pessimist; I think you’re spinning. OK, if you don’t like Java, I’ll agree to accept that you don’t like Java.
[You’re experience isn’t with J# is it?]
Woops, that’s one I completely forgot about, along with C#. I started to try them, but got out quickly.
I wanted a language that would still be around in 20 years so I wouldn’t have to continuously rewrite code from scratch, so I think I guessed right going to Java (everyone said I was a fool for not going Microsoft). Now I just do in house stuff for my real estate, so code longevity is more important than bleeding edge or keeping up with the microsofties. Also free development platforms like netbeans save a ton of money.
All applications on Android devices are Java applications.
Amen!
If you liked that story, sometime I’ll tell you about implementing version control software at the same company. They were using multiple machines with multiple copies of the code as backups and development sandboxes and using the unix date and time stamp for version control. And recompiling code in production.
I’m terribly sorry, but I couldn’t quite figure out just what it was you were proposing. Is it some sort of library that can be shared across enterprises? If so, that is going to be a non-starter for many, many applications due to the proprietary nature of much commercial software. Are Apple and Microsoft going to share their programming secrets with one another? Wells Fargo and Citi Bank? You get the idea.
It’s a bit like another much more prosaic issue which is ride sharing. Anybody can sit on a crowded expressway in rush hour and see car after car with exactly one occupant each and think to himself or herself - Gee what if there were 3 or 4 occupants to a car - wouldn’t that be better? And also to think that they had the “answer” while everyone who came before was “stupid” for not figuring this out. But still, people still drive their own cars and probably will continue doing so.
I disagree. I work in an industry where unneeded and unused functionality is anathema - and will prevent certification of the system containing the software. Adaptive reuse - using the design as a template for the construction of a tailored solution - is a better alternative in my industry.
Technology cannot eliminate the difference in requirements for different applications.
Real programmers work at assembly level. Most kids don’t even know what a register is these days.
I spent a decade administering a version control database depot for an IT dept of a billion $ software company. Despite numerous efforts by lower level staffers including myself, no code reuse project ever got off the ground. Redundant code is being written to this day, while the company upper management whines about the costs of the IT department, forces layoffs and exports jobs to Bangalore. Why?
Rush, artificial deadlines that are never met anyway, emphasis on projects that benefit directly and immediately customers outside of IT, the yes-man corporate culture. For the same reasons the code is written undocumented using unstandardized variable names. I tried to promote code reviews? Are you kidding?
At the same time, believe it or not, the engineering department which produces the company’s products adheres to strict standards of coding, code reviewing, release management. Go figure.
The “advertising” might work. The hard part is of course that the developer needs to know there’s an in house solution to Problem X before he realizes he has Problem X. Because developers tend to be natural problem solvers once they know they have Problem X their natural tendency is to tackle it, not to step back and wonder if somebody in the building has already solved it. And of course the other part of the problem is people tend to have tunnel vision, until they have to tackle Problem X most people don’t give a crap if anybody has a solution. So it all combines to put you in a position where you need to tell everybody you have that solution, knowing they don’t actually care, and hoping they remember when it matters.
Amen +1
I totally agree with you on the basics. Compare though with high level languages and frameworks. Every programmer knows that you refer to the API to get the most knowledge about modern programming - to use the language or framework that they’re using. Projects are defined all the time to include such tools. Engineers are interviewed to determine whether they can use them, or learn them. When a system comes along like HLL, that is designed to support sharing and reuse, then a similar situation exists for specialized application components.
In the circles I travel in APIs are almost treated like languages, people know C, and C++, and WinAPI, and .Net. So if you can teach your developers to think of the in house stuff as an API then it could work, of course then you have to really package it that way. At my first company, where we did the hardcore code reviews, we did that, we approached these common tools as our own version of the standard libraries that came with VC (that’s how long ago this was). We still ran into problem of people not knowing all the stuff that was there, but it did help. Since then nobody I’ve been with has been that ambitious.
[Real programmers work at assembly level. Most kids don’t even know what a register is these days.]
Oh yeah? Well real programmers use bit toggles on the front of the machine to even boot er up!
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.