Discussion Locked

The cause of the degradation of programming

0 helpful
170 replies
Closed to new repliesThis discussion is closed. It stays readable, but it takes no new replies.

In the Discussion about MS targeting apache,

https://www.techrepublic.com.com/5208-6230-0.html?forumID=85&threadID=175795&start=0

A side ceoncern about programming came out.

did ms's basic, visual basic and simple programming cause the degradation of programming from an elite skill to glorified macro writing?

after all, we all know that writing sql queries for mssql, oracle, dbase, mysql, postgresql, interbase, db2......... is now concidered high programming skill, yet in reality, it is little more than a basic part of any office clerks job.
create the queries to get the data you need from the database for the report you are generating, or correspondence you are putting together etc.

is not programming creating the application, such as orcale server, mssql server etc?

how did general clerical tasks become concidered programming?

when are we going to remind employers that macro scripting is for general office staff, programming is creating the application the script runs in?

170 replies

by hal 9000

Where there was a general Dumbing Down of the user base to the lowest level possible. When 95 hit the market there was no longer the need to send highly qualified staff to classes and those that where in the higher paying Data Entry side of the business could be replaced with unskilled personal who cost much less to employ.

They no longer where required to know to actually use a program and only enter data so they became the normal end user and if they could write a Macro in VB they where the "Experts!"

I recently attended a college course where the students where leering Office and in relation to Word and they where not taught how to do a simple mail merge as that was considered as an advanced concept that should not be covered in that particular class as it was only involved in teaching the normal working of Word. I can only suppose that the same applies to other applications in Office as well like designing a template in Access writing any required Macros or whatever.

To me these people are not program users but Data Entry personal which should be treated as the lowest level possible as all they are capable of doing is enter data and not actually use the program/s to even 10% of their potential which was the considered norm when I was working for Big Business.

I now find that I'm constantly being called in to write Macros redesign templates and generally work the programs when changes are required by the so called Specialists Program Operators.

Now this is where it starts getting hard to work as what I find I need now is to lower my skill set to being what was once a entry level computer user instead of the job that I used to be doing which was fixing things when they broke now at the end of every finical year I'm being called in to reset the Accounting Packages, redesign Templates when changes are required and just general basic program user skills a lot of the time.

I've on one occasion even shown a so called Office Manager how to perform a mail merge in Word as a Christmas mail out to all of their suppliers and customers wishing them Season Greetings and she thought it was miraculous that something like this was possible as she had no idea that she could do this from grabbing the data from the already established DB and personalize their form letters. This is not an isolated case either as a lot of form letters that I see now days lack the personal touch that was a basic requirement when we where using Word Perfect 4 or 5 in DOS.

No doubt this can all be traced back to MS marketing where anyone can walk in and use a MS application without training which is what appears to be happening. Now I'm not saying that these people are not required as they are and they can run rings around me when it comes to entering data but they are at a loss when something needs changing and it is this level that is now missing from the business sector so the Data Entry people seem to think that they have far more qualifications than they actually do and the higher level education institutions seem to be catering to this by only teaching the very basics which in all honestly an adverage used could achieve on a home computer in a few days without supervision but then of course they wouldn't have the piece of paper claiming that they have undergone education in using these applications and have reached a level of proficiency that allows them to be rated as passing the subject.

A good example of where having a bit of paper is not actually a good thing and instead of Business employing a second tier of Application users who know how to work the programs they are now calling in what they seem to think of as IT Pros to do these very basic jobs and paying them the big bucks for something that should be a basic prerequisite for using any application.

Well I've got my Fire Proof gloves ready so bring on the masses of people who disagree. :D

Col ]:)

by tagmarkman

This is definate a thorn in my side. It certainly has help degrade our profession.

I'm not convinced that it was the tools that were developed that caused the "dumbing down" of development.

From pioneer to early adapter to mainstream, you are going to see this progression. It bothered me because low skilled people were being called developers even though the pioneers and even the early adapters knew otherwise.

We never pushed for the formalization of program degrees and certification that other engineering fields did. They are more protected than we are because of that.

We never did a good job at segregating our work and advertizing what we did and create ligitamacy. We wanted to be left alone to do the "real" work. In the meantime, the kid down the road was considered a technical expert because she could hook up a modem, the receptionist was a technical whiz because he could turn on the computer and use a spreadsheet, and a person that could mark up some webpage was a developer.

We got lumped into one group be it computers, IT, or software development.

When the "money" started to come in, eveyone tried to pass themselves off as a developer. And they often managed it. When they showed they were less than adaquate, it reflected on us and hurt the industry.

Although, I said that I don't think the software "caused" the problem. I do think it contributed or at least made it easier. Similar to the issue that printers had when everyone "became" a desktop publisher because they had "Word" or "Word Perfect". Or how some people can crack because of the "tools" that were created not because they can really have the skills.

Can we regain our status? Maybe.

by hal 9000

With all the Outsourcing happening we are seeing the lower end of the IT sector loosing out and what is left is again beginning to start being called IT.

It takes a special type to sit a a help desk all day and deal with the general public who think that they know far more than they actually do or are so peeved off by some previous complaint that they just have to take it out on the poor person on the other end of the phone line. I know that I couldn't do it and stay even the little bit sane that I currently am. It was bad enough when I only answered calls that others couldn't handle so I missed most of the problem ones and only got the real ones where I could help or at least try to with some of the problems that arose.

But even still though these people have a definite skill set they really are not IT Professionals but only involved as First String Support. The same goes for the bulk of the HTML coders that we used to see around as well as a lot of the Basic coders.

But I do object to being called in just to change a DB template because there are no qualified people around who understand how the program works to make these minor changes or even write a Macro. The fact that companies are willing to pay me to do this type of work I think tends to point more to the lack of skill within their organizations. With the small business I don't really have a problem although I do try to stay away from their accounting packages but I'm inevitably called in to fix those as well.

But with the larger organizations they seem to lack the skills to do the work themselves and this is a worry as it shows just how dependent they are on outside assistance while being under the mistaken impression that they are self sufficient even for the very basics.

A few years ago I used to sell a lot of the Learn Key products which where heavily used as a training aid for various companies but today even this seems to have fallen off dramatically as these places seem unwilling to even consider wanting to train their staff and while the Learn Key products where good they where not the great products that showed everything possible I just considered them as taking the end users to a stage that was considered as a basic requirement back in the DOS and early Windows days that was up to 3.11.

After 3.11 was replaced by 95 the requirements to learn seemed to drop off dramatically and it seems to have been in a steady decline ever since. The fact that OEM MS Software generally doesn't come with even a manual tends to make a lot of people think that it is easy to use which the basics are but they no longer seem to realize that there is far more that the software is capable of performing and just no longer try.

Maybe I'm wrong but as the MS manuals have shrunk so it appears to me has the end user knowledge of the products.

Col ]:)

by tony hopkinson

we are seeing a progression away from scripting technologies, certainly not from ms anyway.
The any fool can write a program approach has certainly been good business for certain interests.

In the UK though I do believe we are starting to see a business reaction to the cost of this watering down of the discipline.

The slump within IT we've been going through has resulted in a very low level of investment, no attention to refactoring and a hurried series of what I can only call bodges.
Bearing in mind a lot of this code was written by the boom generation with tools that were not designed to code properly but cheaply, there is a shedload of very low quality juvenile code out there.
Now with six years of bodges in top of it what I'm starting to see is a business realisation that quality coding not only gives you a quality product, it gives you a competitive edge and a long term cost reduction.

I think the tools and the boom went together, probably the slump as well.
A whole raft of people were introduced to programming, unfortunately the tools they learnt with weren't the best, their training wasn't the best and on top of that a lot of them entered IT not out of a vocational desire and talent, but because they could and it payed lots.

Based on the evidence I've seen a lot of them should have been prosecuted for fraud.

You can (well I can) write good code in scripted and or drag and drop environments, the thing is, if you don't know how, you don't have to. Using the old text source compiled applications your scope was much more limited in this regard and the skills required by the environment were much greater. Equally there's no doubt a comptent programmer could write this code much better in the old environments, though of course it would take considerably longer.

by jaqui

isn't that you can write good code, it's that you don't have to.

the majority of drag and drop code is bloat that isn't needed, yet is included in the final executable.

this comes from the libs used, after all if you need a function, use the lib that contains it, then use 35 other libs for one or two functions each.

if they were to build a lib or 5 with the functions they are using, then a lot of the bloat would go away. naturally, this would mean reading the function to see what is requires also, as it may need 3 other functions from same lib. the other 11 are bloat.
( auto configuration / generation of makefiles don't always optimise compiler instructions for final size, and with dynamic loading libs, the extra functions are included even though they are never used. )

by tagmarkman

"I think the tools and the boom went together, probably the slump as well."
Great observation. I suspect this is true.

"Based on the evidence I've seen a lot of them should have been prosecuted for fraud."
I've brought this up in the past as well. If an architect screws up your house, he's responsible. If a car manufacturer produces lemons, they are responsible, if a company's toy, injures a kid, they are responsible. I don't understand (aside from the difficulty with audit), why this has never (to my knowledge) happen in the software industry. Shouldn't Software engineers be held to the higher standards of engineering instead of those of a technical writer? (no disrespect to technical writers intended).

by tagmarkman

I have a lot of mixed feelings of Drag and Drop. I agree it causes a significant amount of code bloat, no doubt about that but I've seen more code reuse with drag and drop than I ever have with straight coding. You can certainly do better code reuse and better coding without it but the nature of most coders simply don't.

However, I suppose that is what this thread is about, did this contribute to a watered down workforce, lower wages, etc...

Probably, but I also believe it has a lot to do with entry level engineers accepting extreamly low wages to get "experience".

by jaqui

but if you are a software engineer, would you really concider macro writing experience? or a job until you can get into software development?

maybe it's a bit of both.

using a drag and drop for designing the interface definately makes sence, but what other libs you need shouldn't just be a reuse code lib, they should be built for the specific application.

the bloat from them is the fault of the user, not the tool.

by tagmarkman

Initially, yes, I would consider macro writing the very first "taste" of programming. Our job is about automation and Macros are certainly automation. However, I certainly wouldn't call it software engineering by any stretch. Macro writing for program experience should be very short lived (diminishishing returns on knowledge through experience on writing macros are quick in Macro writing). I suppose that goes with the job as well... writing macros until you can break into software engineering.

Then again... I have seen people make their careers out of macro writing. They understand their limitation and can be quite valuable to a company as long as they don't think they can suddenly become a software engineer overnight with only macro experience. Saddly, many think they can.

I agree that drag and drop is great for interface design... (I also see limitations in that as well) but as far as RAD is concerned, I'll agree.

For drag and dropping code snippits (asside from skeletons), I'm with you on that, I think that it creates bad programming habbits, however, if coupled with a component, I can see benefits in certain situations.

by hal 9000

Which basically says use this at your own risk. Well from the Big Software Houses anyway and even places like Debian only offer the code as a usable option with no guaranties that it will actually work for your needs.

Now in the case of a place like Debian where there are no companies involved but just a bunch of coders writing code they tend to come up with a better offering in the stability, security and usability stakes. But if the individual coders where held responsible for something that they spent their time on and then gave away for the benefit of others all that holding them responsible for writing bad code would achieve is these people no longer submitting code for inclusion in the Debian range.

However places like MS and their ilk are a totally different story you pay through the nose for their products, generally you also pay for any required support other than the very basic stuff like how do I install this on my computer and then you pay to upgrade every few years to the newest and latest offering just so you can have anything that you put out from a business point of view readable by whoever you need to forward it to. It is places like this who need to be held to much higher standards that they currently are.

Currently MS virtually owns the Desktop in the corporate World and we are stuck with all the shortcomings of a system that was originally built for ease of use at the expense of security.

Even 2003 which was supposedly rewritten from the ground up seems to have its first consideration on ease of use and security a secondary issue as they included IE6 in the offering with already known massive leaks built into that part of the OS so their claim of being a complete rewrite sought of fell a bit flat with me when I heard it. If it truly was a complete rewrite we would have had IE7 made available with 2003 along with a multitude of other little nice and totally unnecessary things that come included in the package.

With the Market Dominance of MS I personally think we have seen a reduction in the really advanced items being integrated into the new versions of the software. Like why would you find it necessary to move from Office XP to Office 2003? There are very few improvements in the individual programs and the addition of One Note hardly makes it a worthwhile upgrade.

Word Perfect has had a tool since version 8 called "Make it Fit" where you can enlarge or reduce an entire publication down to a certain size it was originally intended for those letters that ran a few lines over a page so you could reduce the letter down to full pages and not waste paper for only a few lines of text on the last page. This is yet to be included on the Word Offering offered by MS. Along with the fact that you now have to download plug ins to read some of the older items that have been produced in what is effectively the same program instead of having this option on the install CD to load as required to me seems a backward step. With the almost universal use of MS Office in the latest forms in most of the business around the western world at least I find it had to reconcile the lack of backward compatibility built into the newer programs or even if it is hard which I know it is why not include the plug ins with the install ion CD? Could it be MS attempt to dispose of any of the documents produced on previous versions of one of their or their oppositions products?

Now I think that MS should be held responsible for their OS's and any Service Packs that break things the same applies to patches but this has improved markedly lately and we don't offed see patches recalled because they break more things than they are supposed to fix. But SP2 for XP was a good example of something that did cause problems to quite a lot of people and in all honesty it was more like a rewrite of XP rather than just a security update now there appears to be no visible difference between the Home and Pro versions of XP to the untrained they both look the same on boot up with exactly the same messages being shown on the XP Splash screen.

Col ]:)

by tagmarkman

"Which basically says use this at your own risk."
Of course... and that bothers me.

Commercial products need to stand behind their product and communities should stand behind open-source products. This is where closed souce and profit software should make their stand, "use this product as our risk." It would certainly buy my support.

As far as feature drop and such, this is not limited to Microsoft nor do I think that it's a bad thing. Word Perfect has dropped features with new versions as well. Also, attempting to support older formats can be counter productive... as some point support has to be dropped which in itself is a huge problem "Dark Ages II". A lot of good features are dropped because they were causing problems or were difficult to maintain compared to the number of poeple that were actually using it.

Saddly, there are a number of features in different programs that were dropped even though I appreciated them.

by tagmarkman

I've had a receptionist move through the ranks of call receptionist, to customer support, to training, to technical support (1st,2nd,3rd tier), and eventually a tech manager. She was amazing. No formal education but a massive willingness to learn.

I've even had an office clerk that I notice had a great eye for design. I sent her to an Art Institue and now she's one of the companies best Graphic Artist.

by jerry~beans&bytes

hey, if you have clerical people with the ability to write queries, promote them. if you have users who have access to query functions that don't have that ability, you had better button up your database before something really bad happens. (yes, i used to do *expensive* data recovery work)

by jgarcia102066

Programming (in the computer sense) is writing a set of instructions that a computer can understand with the intention of having the computer generate determined outputs given a set of known inputs.

Given the definition of programming, when a person uses a "Macro" generator (which writes a set of instructions...), are they not programming? At least in some sense, they are.

Which person is more valuable, the person that can record "Macros" or the person that can write the code to produce the exact same results?

In my opinion, the person that realized they needed to solve the problem and found the solution is more valuable regardless of which tool they used.

The real value is not in how the results are achieved but rather in the person's ability to analyze a problem, come up with the best solution possible, and implement that solution.

This implies a process rather than a toolset. Part of the reason for the degradation of this industry is because process has become tainted. The demands for getting applications out NOW have far outweighed the rigorous process required for the work to be of the highest quality. The toolsets we use have contributed to "bad" programming but have also contributed to empowering the end-user. Fast advancements in computer technology (much faster computers) has also contributed because overhead created in some toolsets doesn't significantly impact the computers processing. Finally, influx of incompetent people chasing money associated with this industry at one time or another has also added to the degradation.

Much of the focus has been on the toolsets and,yes, the latest Drag and Drop applications add some overhead, but so does any programming language you use. Let me illustrate:

The code with the least overhead to write is MACHINE cODE. (I can't do this, can you?)
This was difficult to write in so we added a layer of abstraction called Assembly Code.
This too was difficult to write so we added another layer of abstraction called C, C++, Basic, Fortran, etc.
We are being asked to develop more, faster with the same resources so...GUI and event driven tools
We are still being asked to do more with less so...4GL toolsets to use with our GUI and event driven tools

The point is, you will incur overhead when you move up the layer of abstraction to develop your applications. This is not a one way street, however. The higher up in the level of abstraction, the faster you can develop your applications. So as an "Application Developer", not programmer, it is your responsibility to select the appropriate toolset for your environment that will produce the desired results.

For the environment where I work VB.NET and ASP.NET work just fine because we have the budget and the fact that our computers run so fast that the benefits we get from the RAD environment far outweigh the performance hit we incur due to the overhead.

As for the Macro writers, that try to pass themselves off as "Application Developers", Good Luck! For simple, tasks, let's just call that empowerment. When you start to add business logic, data manipulation, security, workflow, it's time to call the developer.

Any opinions on this would be appreciated. Also, in this post, if I have offended anyone, it was not intentional.

by apotheon

Programming involves writing programs. Macro writing involves specifying automated sequences of events in an application -- "macros". From certain perspectives, the differences are subtle, but they are significant nonetheless.

The skills involved in writing macros are far more environment-dependent than those involved in writing programs, and they are more limited in scope. The capabilities of your application environment define the outer limits of your capabilities with regard to writing macros. The only limits on programming are, ultimately, the limits of a Turing machine. Such limits are effectively "the sky" within the world of computing. If a computer can do it, a programming language can do it. The same is not true of macros.

To see the damage that equating macro-writing with programming can do, have a look at the PHP language some time. It is Turing complete (though I cringe to say so), so it qualifies as a programming language. Unfortunately, it also perfectly suits the mindset of the macro writer. Thus, its use isn't strictly limited, but its usefulness is. PHP's syntax is anemic, but lots of stuff can be done by researching its bloated library of core functions (around three thousand of them, believe it or not).

I wouldn't use PHP for anything beyond simple database interfaces and server-side embedded scripting for the Web. The problems with the language are just too prominent to use it for solving more complex problems. It creates, in essence, a scripting "application" environment, and its functions become the syntax of what is in effect a "macro" language. It just isn't suited to anything more than some very limited uses, as a result.

I'd still rather use it than Visual Basic, though, any day of the week.

by jgarcia102066

I agree that there is a significant difference between macro "recording" and programming, but they each have their place. Macros are limited in their capability but they do generate the same code a programmer would potentially generate to perform the same function. I agree that for a programmer, macros are not the way to go. Macros can empower a smarter, non-programming end-user.

You make a great example of PHP but this again points to the "process" of application development. Most can learn some syntax and may be able to use many different programming languages as a result. The problem is that many modern day "programmers" (loosely used) don't follow a disciplined approach to their development and, as a result, there is a lot of bad code out there.

Yes, many people, especially in this discussion center would rather eat rusty nails than work in any Microsoft related development platform. I still contend that every tool, PHP and VB included, has its place. If a tool ceases to give you everything that you need, use another tool.

I'd much rather be designing the control systems and writing the applications that control the rides at DisneyWorld, which I would be hard pressed to do in VB. BUT... since I am writing database driven web based applications that need to interface with Microsoft products in a Microsoft dominant shop, VB is a nice RAD platform.

Since you seem to know much about PHP...
I am planning to set up a LAMP environment at home. I have a great interest in entering the open source code arena. What are reasonable alternatives to PHP, if any?

by jaqui

actually, yes it can be done.

we used to play around game programming in rt back in high school days.
writing in machine code and saving directly into a machine executable file.

it takes effort, but it can be done. I wouldn't really want to try to keep up with mdern hardware, as 8 kHz and 3GHz are drastically different clockspeeds, so going into r/t isn't really practicable now.

by tony hopkinson

Well I used to be able to do it. In fact 8 bit on the 6502 was my first experience of programming.
Higher level languages were designed to make writing and maintaining code easier not better. OO was an even deeper abtraction to improve this. Scripting engines are also a deep abtraction. In fact machine code is an abtraction of binary logic.
Syntax and semantics are the constraints and they are environmental. If you ca';t state the problem in the language you can't do it. One of the prinicipal problems we have is people got taught or learnt how to develop without learning how to program.
If the knew how you wouldn't get people doing designs witn 10,000 lines of javascript, or enterprise databases using access etc.

Just because you can drive a nail with by bashing it with your electric drill doesn't mean you should consider hammers outmoded.

by jgarcia102066

I agree in that all of these tools have their place. The point was that using any of these tools provides overhead but it also provides some benefit, faster/easier programming for one.

I have never done machine code programming but...
I have had to build an RPN calculator by designing the binary logic and burning it into a microprocessor, though.

In my college days, I had to write many software applications in C and C++ (drill) to interface with micro-controllers. We used a custom-built compiler, but on occassion, the compiler would remove certain steps that we would have to put back in using Assembly Code (hammer). The compiler was optimizing the code out of business. This would happen when we had to clock a register (set the clock to 1 for a specified time then back to 0). The compiler would see the register being set to a 1 from 0 and back to 0 from 1, so it would remove the line of code that set the value to 1 and would just leave the value at zero (efficiency). The problem was that the controller never received the clock and did NOTHING.

Hammers outmoded? NEVER! As time goes on, however, good luck in finding the people that can use the "Hammers".

Thanks for your feedback.

by debuggist

Wow. I haven't heard that for a long time. That was Apple in the mid-80s, right? I took a class on Assembler for the 8086, and some of the students had worked on the 6502.

Feels like ancient history now...

by tagmarkman

"The code with the least overhead to write is MACHINE cODE. (I can't do this, can you?)"
Uh.. yes... don't use it now but I had to in the past...

"This was difficult to write in so we added a layer of abstraction called Assembly Code."
and we embedded it in C so we could write cool games by writing directly to the metal and take out copy protection :)

"This too was difficult to write so we added another layer of abstraction called C, C++, Basic, Fortran, etc."
Necessary for larger programs...

"We are being asked to develop more, faster with the same resources so...GUI and event driven tools
We are still being asked to do more with less so...4GL toolsets to use with our GUI and event driven tools"
And it was necessary for RAD GUI development.

Library overhead is also somewhat difficult to define because of the implmentation of the overhead. For example, C++ is small and can be embedded with the executable. Java can be small but it requires a RTL to be distributed. A lot of programs in C++ can be larger using repetitive embedded resource than a lot of Java programs using a single RTL.

"Which person is more valuable, the person that can record "Macros" or the person that can write the code to produce the exact same results?"
It depends....

I'll keep chanting my mantra... depends... depends... depends.. depends... :)

I doubt most Macro writers know how to weight those differences. I had one developer who was great in Excel but he use it like a hammer. He wrote great spreadsheets but he also used it as a database and he used pivot tables in it like there was no tomorrow. Sure he got the job done but it was a poor implementation and maintaining it was a nightmare. Luckly, I aquired the funds to build a n appropriate database and interface before it snowballed.

by jaqui

of most current applications the overhead required for any single one is huge. those wonderful things called dlls in windows and shared objects.

unless we optimise by not re-using existing libs where we don't use at least 70% of the code in them, and copy needed sources of those libs into our own to have the smallest libs required for the application.

this does defeat one purpose / benefit of the existing libs, the fact it only has to be loaded once, so 10 apps using it can use same version in ram instad of 10 copies.

the best way to improve the system code is to update the libs to remove bad code.

by tony hopkinson

good code, though doesn't it.

I'd be the first to admit drag and drop is not elegant, it is however extremely pragmatic.
Good code by my definition as far the bulk of my practical experience goes is readable, extensible and maintainable as well as meeting the requirements.

I don't feel that's drag and drop though, it's OO. The more extensible you make a class, the less it does. If it isn't extensible at all then there's no point inheriting from it, therefore it's bad OO, because you would have to rewite those bits of the class you wanted instead of re-using them.

The degradation isn't OO as such. I know a lot of developers who wouldn't know where to start without OO, even more appallingly they don't even know what OO is and why and end up writing bad OO designs with the attendant bloat.

My chosen language is delphi, but I've met very few developers who know how to optimise their design to cut out unnecessary code in the executable. Saying that my knowledge is mainly theoretical as I'm never afforded the time to apply the techniques in the job anyway. In fact If I was required to do it now I'd have to research again because that sort of thing is specific to the compiler not the language.

by tagmarkman

Ah Pascal... there's a language the new guys get lost in... that and Fortran... :) For some reason they ignore the fact that both of those language are alive and kicking... Even Cobol..

Definition of good code... That's the main difference of RAD development and more elegant solutions. Quick development has it's place. But it's up to both development and management to agree what needs to be done for the goals of the company and client.

Personally, my business doesn't optimize anything until we stamp the product as code completed to specification and we met the minimum of test bar. After that, we'll shoot for optimization. We usually have time to do that for about 25% of our new products and about 75% of our updated products.

by tony hopkinson

doesn't work or isn't finished is a newbie mistake.
I should point out that I don't do RAD, even thought the tools were designed for it. RAD all too often concentrates on the front end while sacrificing any semblance of quailty behind the scenes.
I use Drag and Drop to quicky do front ends but I do it in conjunction with the back end code, that way they don't get out of step.
My development method is effectively functional iteration. I add a a ui and the backend for new functions, and any refactoring of existing functionality in each pass.
It's the way I learnt and all the gatekeeper crap left me unmoved.
RAD = QAD in practice as far as I've seen.

by eth0

Generel office, clerks, secretaries, accountants, execs. Do you really expect these guys to script. Granted, it is not programming, but it still takes that intuitiveness.

It is junior level programming to say the least. Remeber your first C tutorial? Awfully similar, and the learning curve from beginner to real efficiency can be just as steep.

I do agree that VISUAL STUDIO made it easier for everyone. There were electronics kids at my school (ITT) showing me stuff I didn't even know you could do with VB. ;)

by tony hopkinson

Programming is no harder or easier than any other 'science'. It is a science though.
I started my career as a general office clerk by the way. I became a software engineer because I was that clever, perhaps not. The probelm isn't that some can't program it's that some programmers can't program.
The electronics students I've met showed me things that they've done in VB that I'd never do in any language, I'm sure they'd have similar feelings about one of my designs for an 11 kv step down transformer.

by jaqui

in the original discussion, where he mentions using 64kb eproms, that's a perfect way of explaining the differentce.

the script kids / drag and drop / visual developers can't work with that cap.
a software engineer can.

most of us that have been around for a while, didn't start as software engineers or it peope, we started in other industries, or capacities and moved into it. expanding our knowledge base and skills as we went. the boom hit with win95, this started the business school ( ITT etc. ) focus on gui only programming techniques, and cost the industry the skills for good code in newer people.

takes us old timers years to break those bad habits in them, if we get the opportunity.
:)

by jaqui

if you cannot script a macro to use the application and do the job properly, then you do not work for me.

if you cannot built an lfs system with the book, a bootable cd, and the time, then you ain't working as a programmer for me.

no real intuitiveness required to write a macro to insert data or extract data.
you know what flieds you are working with, you know what you need to do.
it's pure logical construction.

application development benefits from intuitiveness, but it doesn't require it.
a methodical, logical approach will get the job done properly.

I'm lucky, my software engineering is for myself.
and I have very strong opinions on what's important if my name is attached. quality of product being high on the list.

security being number one.
commenting / quality documentation being number two.
code quality number three.**
platform requirement: linux
cross platform support: *bsd, unix, irix, linux.
( open source platforms* )

* when m$ starts putting out fully posix compliant products I'll concider them, when apple stops corrupting beos so the open source software won't run without recoding I'll start concidering adding support for macos
** though it definately ties into security very closely.

by tony hopkinson

and culpability.
Well bisiness's have in the past forced me to produce what I consider very poor code. If they employ me and choose to ignore my knowledge then that's there fault.
If they employ/use someone for a task they are not capable of it should be both parties fault. However, when both parties are ignornt of the skill requirement assigning 'blame' would be pointless and difficult.

Personally I put much of the blame down to greed.
I've applied for jobs in the past with a stated requirement of my level of experience to be turned down in favour of a cheap recent graduate.

I've never yet met a recent graduate who can write a decent application, they aren't taught how to do it.
You only learn that by doing it wrong and that's where the boom cost us.

The bust howver means that, that code hasn't been fixed, some of the examples I've seen are so bad they have me rolling around on the floor.
When I get up I use them as useful examples for mentoring junior devlopers. My juvenile mistakes were in technologies that the poor fellas haven't heard of and think are irrelevant.

by tagmarkman

I've been at companies like that. The problem was that I simply couldn't sleep at night unless I was comfortable with what I produced. So I would spend countless (unpaid) hours getting it at least to the point to where I wasn't embarrassed. I eventually left the first company that was like this, but luckly at other companies I was promoted to the point to where I could put more pressure on change.

Companies should be held responsible for the product that they sell. Employees should be responsible for their credentials. In my company, I can tolerate a whole lot, I've keep on employees that I shouldn't have do to their work but they were good team players. However, I do not tolerate deceit and will be the quickest way to get fired - good developer or not.

I agree you must learn by doing. I personally never met a recent graduate who can write a decent app (front to back) though, I have seen some that are quite good and show tremendous potential. I put more weight into the persons personality, thirst for knowledge, and willingness to work hard more than what they have on their degree.

Still, I do put wieght on degrees as well (ex: I toss out the first 2 years of college and then do a 2 to 1 ratio: 4 years of college equates to 4 years experience, 6 years of college is 8 years experience, 10 years of college = 16 years experience.) I know they are not exact and in many cases I have found the opposite to be true but in general that seems to be true for a least 60% of the time. However, give me a dev with a 4 year degree AND 2 years experience and chances are, I'd choose them over a fresh master's student.

I've seen some pretty bad code out there. I mean REALLY bad code. You could randomly slap keys and produce better code. And I've seen some code that was so awesome and amazing and did such a good job only to be "obsoleted" because of a compiler or architecture change.

I know what you mean about fellas that think old technology is irrelevant. They don't understant you need to know where you are comming from and "WHY" the industry evolved the way it did. Then again I suppose the is the difference between the degrees, bachelors: I know the basic, please give me a job... Master: I know how to implement everything at least in theory and to some degree practics... PhD: I know where we came from, and why, and I have a good idea where we are going and why.... However, they all still need "practical" experience to equate to a anything in industry.

by jaqui

people learn from thier mistakes is another way to look at it.

this is why when hiring someone it's hand them the 300 page printout and livecd and park them in front of a workstation with no partition structure on hard drive. if they can set thier workstation up on the first day they definately understand 3 things.
1) how to follow direction
2) linux
3) how to use compilers

they may not be great programmers, but if they can't get this done in a day then they won't really be a benefit to an open source os only company. learning more programming skills is an expected part of working, you never know everything. ( don't you hate getting a "know it all" applicant? specially just ot of school? )

my own formal training was only a few months long.
( back when 286 was new ) but it gave me the foundation I needed to be able to learn as the tech changed.

by hal 9000

At every opportunity when I go to the big places for any reason and back in the old days it used to be for new product launches I would constantly be missing in the museum playing with the old stuff.

The new stuff didn't interest me in the slightest as I could play with it every day to my hearts content but that old stuff in their museum's was something that I would very likely never get another chance to play with so I made the most of it.

I particularly liked a developer that I taught several years ago now come up to me at a presentation he was giving for one of the big boys in town and claim that Boolean is dead! I just asked what language was used to store the data on any digital storage surface like a HDD or CD/DVD, he went into great detail about what language was being used to actually write the code but had simply over looked the machine language that it was converted into for storage. :D

He was one of my biggest failures although since then he has began to understand that at the very least he doesn't any longer know it all and is starting to get better. :)

Col ]:)

by tony hopkinson

graduates, ones who understood the points I tried to get across first time.
For the ones who wonder what they could possibly learn off someone without a degree and don't pay attention I provide a shoulder to cry on, when the screw up big time after not following one of my suggestions. Any that screw up the same way twice I figure bought their bit of paaper instead of earning it and treat accordingly.
My estimation is any number of years at college = 0 years experience.
Two critical skills missed every time as far as I've seen at least in the UK, how to read someone elses code, and how to read your own six months later.
Readability is step one, without it, kiss everything else goodbye

by tagmarkman

I've noticed something... Most of use ARE old timers... I didn't expect that when I first started posting.

Most of us seem to have been around to see the progression. We understand true code (you know... binary).

I Gave a speech at a development conference about 5 years ago, and if you could have saw the look of confusion when I was talking about pokes and peeks on the Commodore 64 you would have been rolling on the floor. And the looks of horror on their faces when I was talking about Matrix/Grid computing (programming using jumpers = not the new fancy stuff) was priceless.

by jaqui

have gotten a video of thier faces.
cause I would have loved to see it.

my favorite was the old apple II plus, pokeing memory address c080 or peeking it would lock the entire system.
it was a memory hole.

we used to make sure people remembered to save often by inserting a peek or poke if they left the station un-attended. one person was hit 32 times in one week before he started saving it and taking disk with him. ~evil grin~
but he did get the message.

by tagmarkman

Now that was just cruel :)

And yes... I would have loved to got video of their faces but would they sign the releases ;) hmmmm...

by tony hopkinson

poking was what we did with our pens.
In the paper tape copies of obnoxious students work.

I remember a really fun conversation with one new fella storing tonnes to three decimal places as floating points because they was more accurate than storing kilos as integers. Took me ages to get it into his head. When I started explaining BCD he went off in a huff.

by hal 9000

:D

Actually that brings back memories of a meeting that I went to for a Fuel Injection thing by Bosch for the Ford Motor Corp while I had escaped up north and was working in obscurity where no one knew me. God it felt good but I digress.

At this meeting the guy running it was someone that I had taught many years previously and at one time he threw up a CRO trace of a Injector opening that was measured from + 12 V to -12 V which looked like a 24 V swing, so he was quite rightly asked how a 12 volt system could generate 24 VDC by one of the confused guys there.

This was something that he had not thought of so he stammered around and then did exactly what I had told him at the door never ever to do he asked me. Unfortunately I was given a lift down to this meeting by the owner of the place and the head mechanic and I was sitting between then at the time they where both looking at me and wanting to know why I had been asked this question by the presenter. While trying to hide I muttered something about Back EMF and said I'll talk to you latter. He then went on to finish that session of the meeting and I followed him into the toilets where I tried to flush him away, but unfortunately someone followed me in and saved him namely my immediate boss who wanted to know why he had asked me the question. Being somewhat damp and disoriented at the time he just answered that I was the one who taught him everything that he knew which was clearly wrong as I had never seen one of these things previously but I suppose I knew the basics.

Well that meeting was the end of a perfectly great job where I could just put up my hand and ask for help when I ran into trouble and have someone assist me. From that time on I was considered the resident expert on all things electrical which was ridiculous particularly as they didn't have the proper circuit diagrams or any form of schematics just all the wires going into a box and I was expected to know what was happening inside that Bloody Box. :(

When ever I asked a question after that I was laughed at and constantly told that I was only testing them when in actual fact I was doing something for the first time and wondered if there was an easier way of doing things. Of course I then got every first time job to hit the place and was constantly being asked if I could repair burnt out circuit boards with in house markings on the IC's.

The only good thing to come out of that was that I did get real schematics for the wiring and control boxes as well as the sensors so I could better fault find and repair those boards with the charred IC's that I was constantly being presented with. I think that the last 3 months that I worked there where mainly spent with the owner who was an auto electrician explaining just how the things worked and why the things where breaking. When I quit I left the circuit diagrams there but I'm sure that the owner had no idea of how to read them and I never mentioned the differences between Electron Flow and conventional flow as it would just have confused him too much. :)

Col ]:)

by tagmarkman

"When I quit I left the circuit diagrams there but I'm sure that the owner had no idea of how to read them and I never mentioned the differences between Electron Flow and conventional flow as it would just have confused him too much."
:( such a sad story...
Sounds a lot like what happends to most software engineers I know... If someone think you know anything about a computer, they ask you how to fix this or that in the system... I wonder if they ask autobody people to fix engines as well :)

by hal 9000

But there are now so few of us still around to do the right thing even if we are given the chance.

Personally I walked away from Punching Code when I started seeing Basic used for everything under the sun yes it's easy but it isn't elegant or small and has a very high overhead but it made a lot of people think that they where programmers or developers. Luckily I missed the worst of the HTML rubbish that was pushed down our throats as I switched to mainly hardware and now I only script things for one off applications for clients who are willing to pay for very small foot prints and low overheads in their existing systems.

I've built a few Web Pages with a compiler though but I would never ever consider calling myself a Web Page Designer the fact that I can use these tools is only secondary to what I can do and even when I do push out something I never consider pushing out Bad Code with Bad Coding Practices.

The closest thing that I've come to some really sloppy coding was when I tried to install a version of Linux onto a SCSI only machine apparently the install app could only recognize IDE drives until it was installed so I pushed out some really lousy code for that but it was only for one of my own test machines and then only for testing. If someone was to actually see what I did I would disown it and claim that I never did it and in all honesty if it was for a production application it would never have been done that way but for a single install for myself I knew what corners I could cut and still get away with it without causing any unnecessary problems but it really was sloppy.

What scares the hell out of me is that there are some people who would think that that piece of code was great.

Col ]:)

by jaqui

has thrown a hack together to accomplish a one-time task.
this is a normal part of life.

if the current trend of larger and larger applications through bad coding practices continues, then we may see a ressurgence in mainframe sales, to have the capacity to run the apps.

personally, I hope the ai development groups are all software engineers, not script kids.
cause there is no way a script kid will get a functional ai that won't go insane and start killing people.

by hal 9000

And we all know what happened to them don't we? :D

Actually I visit QU a few times a year now days and generally have a look in at the AI section when I'm there and at least the people there are not the Script Kiddies that I've seen all too often but then again being Under Grads they don't know all that much that isn't already in the text books either.

But they are doing some interesting work and their main aim seems to be keep the foot print to a minimum and the overheads as low as possible for the thing to develop and be able to use the available resources without limiting its development. While they may not be the most advanced they are doing some interesting work but even then their work is confined to mainframes for the processing power required.

As I know the guy who runs the place I've been asked the odd question when they have run into some minor coding problems and while I may have solved a specific problem I don't know if the solution that I offered actually worked when integrated into the complete system. But they continue to ask so at the very least they are learning something even if it doesn't suit their current project.

Without being able to see the full code and flow charts it's very hard to know what is happening but from the very little that I've seen it certainly looks interesting.

Anyway a Hack is one thing but the mess that I put together for that job was really terrible. I don't think it would even qualify as a Hack just a quick and dirty work around for the current problem. It was so dirty in fact that you couldn't see it as it was covered by at least 10 feet of mud. :D

Col ]:)

by tagmarkman

"any software engineer has thrown a hack together to accomplish a one-time task"
So true, even if it was just to learn a specific item not to be used in production. Ever one hacks soemthing together from time to time. A "real" software engineer knows the difference WHEN to hack and when not to.

I have faith that AI development groups will be "software engineers" they have to be in order to get past a standard Hanoi tower puzzle, however, the Drag and Drop AI features that will evolve from that will probably wipe out have our population. ;)

by tagmarkman

That's pretty good with the "set up your own workstation" bit. I bet that works absolutely great at a smaller company. I may actually "borrow" that technique from you. At the bigger one it's not that easy, too many controls and oddities.

Even though I would never expect a sales person to perform that task, I'll use them as an example if I did... The salesforce at my company is completely run on Blade servers. They have no physical access to the machine, just the input devices, the keyboards don't even have USB extentions. This is done to help protect the company information theft and to reduce maintainance cost. It's worked so well compared to standard desktops we are considering it for most of the departments (IT, Dev, GA departments are excluded). This is just and example... other controls are in place that would make it extreamly difficult for a person to get their box up and running in my company without sometype of IT governances.

Also... once in a blue moon, I'll come across a brilliant developer or software architect that either refuses to set up their own box or is so far removed from it that they would weeks get it up and running the way the need it (usually they have a strong background in larger companies where that was always done for them).

by tagmarkman

Boolean is dead :P that's a good one! Just give them a copy of "CODE". I suppose that it's just a coinedent that decimal numberics on computers tend to be exponential binary equiviants ]:).

I'm a nut with the old computer stuff too. I tend find myself collecting computers. I have about 40 unique computers systems (game consoles included) in my hobby room. I even have 9 track reel to reels, you'd get such a kick out of it.

by tagmarkman

Very true... This used to be a real sticking point for me. It doesn't bother me nearly as much as it should anymore because the priorities of my worries are elsewhere.

DLLS was a great idea with a bad implementation. I always like the idea of sharing resources, but I had so many headaches about it not freeing those resources when it was finished (luckly fusion help clear up most of that). Still, you see issues with versioning, and rebooting, and resource bloat (using only 2% of the dll), and so many other issues.

by jaqui

but then why aren't people demanding that the dll's get fixed?
( with linux I have the sources so I can edit any shared lib and submit the changes as a proposed patch. developers say thanks test it and if they like it they apply the patch )

though, a lot of linux programmers are even more guilty of bad code than it sounds. they code to use gcc "extentions"** wich is not something I myself agree with. I beleive that coding to ansi or iso standards is the correct way.
( I have to code every line myself to do this though, as very little existing code is ansi / iso standard )
working on getting my entire installation of gcc redone to be ansi standard only., but it means having a double copy of it, one left alone for building the kernel. ( absolutely requires gcc extentions )

** gcc extentions are better called bugs

by tagmarkman

They are demanding it. That was the main reason behind fusion. But there is still so much to do it's not even funny. Dealing with legacy issues and third party development (their dlls and dependancy on popular dlls) is just one of the dozens of issues holding back advancement. Their is nothing so annoying to me than to see my box taxed at 2+ gigs of filled RAM after I have closed out of all open programs (ignoring services, dlls, and such). Reboot, and it's down to 250Megs. (same programs running). In fact, on the windows web servers I had to maintain, I used java servlets to get around the DLLs hell... but I must say, it's better than it was several years ago.

I agree with you about the gcc extensions... I try to stay ANSI/ISO myself, but sadly, we don't have the luxury of always doing that... so we tend to mix and match (yes... two copies :) ).

by jaqui

perl, python, any server side scripting language.

lamp isn't really a specific set of tools as a concept.
what is sent to the browser is only html for display.
( core of php design concept )

by jgarcia102066

Since my first exposure to web development was in integrating with IIS (ASP), I haven't been exposed to any of the open source scripting languages.

This is cool, thanks!

by jaqui

a discussion here on php formatting for html, linked to a download here.
in it I posted an entire set of php scripts to create all html entities needed in a page.

the same idea can be used in perl ( existing ) and python. ( not sure if it exists yet )

apache even has support for asp.
mod_asp so you can use existing code in apache, on linux, as the module is not available for windows.

by jaqui

the idea of proving practical skills isn't my property.

it does present interesting problems with the setup you describe, but think of the fun you can have tormenting people that really shouldn't be in it.

by jaqui

or s&%t?
;)

a quick and dirty piece of code is the definition of a hack, how can it be not even a hack if it's that dirty, means it's an extremely impressive hack.

after all, the dirtier the more impressive.

by jaqui

origin of "hacker"?

after all, before anyone used programming skill to break a systems security, techs were hackers.
we were the only ones with the knowledge and skills needed to accomplish most tasks.

throw together a hack to get something towards releasable code, then spend the time cleaning it up into good code.

by gabriela2

I don't think that a office guy can't do that and these are db queries too. Simple select statements, eventually using a tool which let you click on the columns you want for the query, but for complex stuff... I don't think so.

by jaqui

using the db server is using it, if you are not capable of using it then you are not qualified to work.
creating the database, which is when the blob ( data type in defining fields ) comes in is the db admin,, not the same thing as generating a query script against existing db at all.

by gabriela2

What is the matter with you guys?I just wanted to say that you need some knowlege to extract data like BLOB OR BFILE from a db like Oracle. Actually, it is not possible to see a blob or a bfile by using a select statement in sql*plus at all, so you need a tool( say Oracle Forms) isn't it? And Oracle Forms is not a easy tool to learn.
And when you have 3 ReallyBigTables to query you must know how to write that query otherwise you'll slep 3 hours on your desk, waiting the result, isn't it? Optimizer stuff...
It is funy. Imagine a user(MS Office specialist with an excellent typing skill) writing such a query. Hmm, I don't think so.

by tony hopkinson

It's the consequences of
Select * From ReallyBigTable where Id in (select RBT_ID from EvenBiggerTable)
Usually on the live system at peak time.
Or my personal favourite missing where clause in a simple join.

by www.pinoyfun.com

I was in a net cafe a long time ago watching these babies play one of the net games. One baby came in and spent his precious 20 mins entering things into the command line and bingo! When he entered the game he got sly looks from the other babies who were there before. Now if that's not coding what is?

Something came in my mind. My these babies are already coding and who knows what holds in the future. Computers come naturally now to the young kids that pure coders or as we say programmers will wither away. They are now replaced by developers - those with the skill to dig the loads of code snippets scattered in the internet and everywhere be it macros or whatever you have plus knowledge in scripting, a language or two, analysis and design and the determining factor - the business knowledge of the requirement at hand, and presto he's got a glue to hold these snippets together.

Programming is not getting degraded. It's just in a perpetual state of evolution and the delineation between the pure coders and developers is just getting broader.

I'm wishfully thinking my four-year-old will code in PERL in the next three years. whew..

by jaqui

is the core of this discussion.

tying together snippets of code does not create a decent code quality. doesn't exclude it, but doesn't mean it's there either.
actuall skill of individual coders that wrote the snippets and he script to tie them together controls code quality.

I'm not saying the macros are bad in and of themselves, I'm saying app specific macros, like an oracle application set, are not application development, they are scripts running in an application.

by tagmarkman

We are talking about something a little bit different. We know that kids can do cool things and they have great potential. I'm always impressed with children that have the desire to learn technology (you'd be supprise how many don't).

I also think we have different definitions of what a pure coder, developer, script kiddy, and software developer are.... Heck, we can barely define what a hacker, cracker, etc are... Let alone the differences between Usability Engineer, Interaction Designer, Graphic Artist, Information Architect, or Web Editor.

This is on of the reasons our profession has been watered down... everything is lumped together. A good developer of device drivers is probably a pretty poor developer of web services. The rare person that can develop a good Windows Service shouldn't be in the same category as someone that can only drag and drop a label on a web page using (gasp) Frontpage.

by hal 9000

I didn't realize that they where still using Frontpage. :D

Actually at one MS meeting I was listing to several Web Designers talk about just how great Frontpage was and how they could do this and that without a problem. After about 30 minutes of this and the subject I realized that I was wasting my time in that session so I went one of the other ones running. It was surprisingly easy to just pickup what they had been discussing while those around me where asking very basic questions I was wondering why it was that I had taken time off to attend. :(

Col ]:)

by www.pinoyfun.com

Why do you need to break any habit like that?

It's a natural progression that macros will evolve as computing power increases. Never mind that macro does more and initializes and stores things more than what we really need. Today's computing power makes these garbage real puny. What matters are solutions even if it comes from somebody who doesn't even know that i=i+1 is not really equal in the real sense. He's got to have the business sense in this case.

by jaqui

if you are given instructions to fit the end result in a specific amount of space, and you can't follow them you have a habit that needs to be broken, ignoring instructions from your boss.

the options, do it to spec, or find another job usually get that message across.

I avoid having to deal with it, as my post about first days task for developers details.

by tony hopkinson

the argument our computers are so resource heavy we can afford to write inefficient code.
Course when certain types insist on writing so much inefficient code that it uses up those resources, you are left with a hardware upgrade or employing someone who's capable of rewriting the interpreter that runs the macros.
The rise of scripting might have been a short term boost to business, but it's a big time long term negative.
You can write i= i + 1 in script that is n't the problem, the problem is i is rarely an integer.

by stan20

If clerks can write queries, etc thats great for me! I won't have to waste time on the easy tasks that other people can do, and I can spend more time on serious development. Whenever possible I teach my customers how to do whatever tasks they are capable of.

Programming and development are maturing, and developing a hierarchy, the same as other professions. The CEO and a group leader are both managment, but one has more skill and is higher paid than the other. It should be the same in IT.

New tools don't degrade a profession, they make it possible to be more productive (and therefore worth more).

by tony hopkinson

it's the misuse of them.
Scripting for instance is far easier than coding to get started in. VB is easier than C

Programming is breaking a large task into small steps. The size of the step depends on the syntax of the chosen language. The finer the granularity of the language the more steps you need admittedly, but you can make those steps much more precise.

I aren't saying people who use scripts necessarily can't program, I'm saying you don't use a pickaxe to make a hole in a wall tile and expect the best result.
And that's what people have been doing.

by jerry~beans&bytes

in the old days (yes, i'm that old), you could hardly get into an application without knowing a little something about computers. now, thanks to GUI, and the corresponding emphasis on intuitive interfacing, lots of people are using computers without the kind of background that we old dogs had to have. so a lot of people are able to pass themselves off as something they're not. so of course they are writing bad code.

add to this the massive improvements in hardware (that mask much of the cost of bad code), mix in buying decisions made by people who don't know anything about computing, and you get where we are today.

all we can do is keep doing it right, and hope that, eventually, hte cream (and not the scum) will rise to the top.

by stan20

The result depends on the skill of the person using the tools. And of course anyone with any understanding of the field (whatever the field) will choose tools appropriate for the task at hand.

I frequently write programs that include parts done in several languages because using the best tool increases productivity. VB is great for user interfaces while C++ is much better at system
tasks for example. I even use scripts when its the best way to do the job. If I'm doing a real-time operating system for an embedded processor where performance and code size are the most important criteria, I'll use assembly language.

Any tool can be misused. The results seperate the people who have the skills to use the tools correctly from the rest of the crowd.

by jaqui

you mean the applications that are hosting the macros?

becuase macros have been around since the first multi use application. a very usefull tool. after all the conditional defines in an application are macros.
the include directive is a macro.
a makefile is a series of macros

it's not the tools, it's the mistaken abuse, and promotion for skills not actually there that is the problem.

by stan20

An assembler is a tool that makes programming easier (and faster) than working in machine language. Which I have done back before assembliers existed for micro processors. New symbolic languages (Fortran, Basic, C, VB, C++, Java, etc) are simply tools that make (good) programmers more productive. Wizards, script languages, etc. are also just tools to make simple tasks easier and faster to implement.

And by the way, macros have been around since about the day after the first assembler was written. Macros, library routines, dll's...the principle is the same. Reusing code makes programming easier and programmers more productive.

A skilled programmer knows how to use the available tools to get the desired results. Amateurs don't.

by tony hopkinson

Never built the calculator but my teacher did. He taught me alot of the basics, floating point representation, boolean algebra, address decoding and so forth.
I ran in to RPN again years later when learning compiler theory with parsing an infix equation as the example.
Haven't done anything with micro-controllers, have done a lot of work with plcs and intelligent devices though.

by jaqui

I always do one, on paper yet.
at times, I have flowcharted individual functions.
( for the data flow between functions. )

by jaqui

an attitude than an ability issue though.

a good attitude is the one that listens and learns.

the problem is, most recent graduates have the attitude that they don't have anything to learn now.

only met a few formally trained people that don't have that attitude, unfortunately.

by jaqui

this was school stuff, so it was a way to develop good habits for when finished.

wouldn't you rather have a new employee that has a foundation of good work habits? I would.

by stan20

Of course, to do anything interesting you have to understand how computers work (and I have yet to meet a kid who does).

And I still love programming in assembly language because it allows me full control over the hardware, and doesn't prevent me from doing things that compilers won't allow.

But, I charge my customers confiscatory rates and use whatever tools are appropriate to get the desired results in the minimum amount of time.

If some kid with little education and less understanding can do the job, its probably a job that I'm not interested in. What makes it fun is doing things that I haven't done before (or better yet, things that no one has done before) instead of things that I did 30 years ago.

Wizards, GUI, intuitive coding tools mean that anybody can do the easy things.

If you want something that takes real programming you have to go with a pro. Usually, companies that want my services have tried a number of far cheaper "programmers" and have nothing to show for it. And finally realized that its cheaper to pay $250 an hour and get the job done by a pro.

by jaqui

so you have someone who has learned orcle forms, that does not make then a software engineer, it makes then an end user that has mastered a difficult piece of software.

it's the fault of the software engineers for oracle that it's such a difficult tool, as with all the sql engines I've used a simple select will allow access to any field's contents.

by hal 9000

Not so long ago things like this where expected of the end users as a minimum requirement to get a job. Now however being able to fill data into some fields is all that is required. Then when a new queries are required it then falls to someone else who is called an IT Pro to come in and write the Macro for that required job.

To me there is a whole part of the industry that has simply disappeared they have either walked out in disgust or been promoted to a higher lever where they are not capable of working in but are called something that they simply are not.

The question here is not can we write a Macro to do something for the end user but should we be required to be doing this in the first place. It wasn't so long ago that a question like this would never have been asked of any of us unless something had gone very wrong. The general level of experience of the End User would have done this and we would have been left alone to produce better code.

Col ]:)

by tagmarkman

That is the real issue... not if the receptionist can do a query.

Personally, I have to limit what people can do in my business. I don't want the receptionist to be able to build a query and look into my database. I want to control that view.

Taking time away from a seasoned developer to write macros is a managers issue of using resource improperly. My junior guys, test, or support guys (they love to get away from the phone) tend to write macros and such for other departments, it lets them grow and it spares my more senior developers from being misused.

I've worked for companies that assign work to who ever is avaliable at a given time without putting any thought into "if" that person should do the job to begin with.... junor devs would get in over their heads, and seniors were getting ticked off because they had to write some crappy one off macro. I vowed to avoid doing that to my subordinates.

by hal 9000

"A skilled programmer knows how to use the available tools to get the desired results. Amateurs don't."

Today I see far too many amateur's getting jobs and calling themselves Programmers/Developers who are misusing the tools available. Sure eventually the job gets done but at a cost of lost clock cycles and overheads that are just plain unnecessary. What is the frighting thing that is happening now is that these people are turning in Script Bloated to Hell and are considered as good because they have brought in a project either on time or under time but the code is so bad that the application runs slowly. Because it works they just are not bothered by the footprint or overheads that they have created and it is these people who the penny pincher's see as useful to the business not the ones who create it properly in the first place with smaller footprints and lower overheads because it takes a bit longer to get it working properly.

While in the short term it is better to get things out fast in the long term this is a disaster waiting to happen when the apps get so big and resource hungry that there is no system powerful enough to actually run the things when they are eventually all installed on a unit and are expected to run. We see this now and it is laughingly called "Undocumented Features!"

The current situation very much reminds me of an old SiFi film that I saw many years ago where there was a computer system controlling a enclosed environment that catered for all the people inside the enclosure. Everyone studied different parts of the system and became even more specialized in individual components so they started instigating changes to the various subsections that they studied but no one any longer under stood how the whole system worked and why it was not appropriate to change something in one part of the system and break things in several of the other parts of the system. Naturally in the end because no one really understood the entire system they broke it and they all died because the system could no longer keep running with so many of its resources cut off or disabled. If I remember correctly it was brought about by no one wanting to be involved in the sewerage treatment processes and that all the changes made broke the waste treatment and with that broken there was no possibility of recycling anything.

I see exactly the same thing happening today with computers so many understand how one thing works like just one application so they make changes to make it better but break other things in the process. In this case there is no physical damage just a general slowing down of the systems while the badly written code consumes more and more system resources for no real benefit. At some point in time things will have to change and what worries me is that when that time comes there will be no one capable of knowing where to begin to bring things back to what is possible or acceptable.

Col ]:)

by tagmarkman

Drag and Drop front ends with back end code is still what I consider RAD. In fact, I can develop program much quicker the way you describe than editing properties.

I think drag and drop UIs are fine for internal or quick development. I feel they are great for many projects.

However, for truely professional products, I rarely see good components (although there are a few out there especially when it comes to generating report data). I say this because my people in the usability lab almost always pulls their hair out when drag and drop UI programs are handed to them. They almost always tend to be "program interfaces" and not "user interfaces".

by tagmarkman

"but think of the fun you can have tormenting people that really shouldn't be in it."
And so much fun it would be... Luckly, I get all of my tormenting out during the interview ]:). After I hire them, it's a little hard to kick them out so I better make their lives at that point as smooth as possible.

by tagmarkman

I've been fortunate with the people that I've hired. All of them have a desire to learn, and all of them understand that we know very little compared to all of the information out there so we better listen to one another... even if we don't always agree.

If you get stuck on a development issue for 30 minutes, you get help from someone. Too many times have I seen a good developer get suck on a problem for hours only for a junior dev to notice the problem in a few minutes.

We always respect the person and that allows us to slam the hell out of the code that is produced without people getting defensive about their code. I try to move projects around to different developers often. It broadends the developers understanding of the whole system, it discourages an individual from becoming too personal with the code, and it forces everyone to write clean readable code (or else their peers will freak out on them.)

by tagmarkman

"wouldn't you rather have a new employee that has a foundation of good work habits?"
Nah... we always need a scape goat ;)

by tagmarkman

"I always do one, on paper yet.
at times, I have flowcharted individual functions.
( for the data flow between functions. )"

Use cases provide a senario and broad understanding of the task...

However, instead of a Flowcharts I recommend activity diagrams. Flowcarts are binary which can make big cumbersome diagrams, the also are pretty antiquated just take a look at some of the symbols that are never used. However, activity diagrams are much more versitile.

by tagmarkman

"An assembler is a tool that makes programming easier (and faster) than working in machine language. Which I have done back before assembliers existed for micro processors. New symbolic languages (Fortran, Basic, C, VB, C++, Java, etc) are simply tools that make (good) programmers more productive. Wizards, script languages, etc. are also just tools to make simple tasks easier and faster to implement."

I like the definitions.

by tony hopkinson

drag and drop certainly has a bearing on usability.
I did five years as a data entry clerk though, so I don't neglect that side of the job.
Also I'm allergic to things like db aware components and I write my own components in preference to using an unsuitable one.

by tony hopkinson

just waiting for the people who control the purse strings to figure it out.
My last job involved Fortran, C++, VB, PHP and Perl & MySQL in one project. I was enviromentally constrained to use Fortran & C++ and forced by a non-technical person to use VB6. The latter in my opinion is never the right choice, Delphi is always better.

by stan20

The last place I worked as an employee said "we are a microsoft shop" and that was that. People with the least knowledge made the basic technical decisions.

Finally I decided that I was never going to work for a chimp in a suit again and started my own company. That was the best move I ever made! Now I work on what I want, having fun doing it, making 5 times as much money, and best of all, I don't have to put up with managment types.

by tony hopkinson

Must confess I rarely bother. I hated doing them. For event driven aplications they aren't even suitable. I'd rather spend time making sure my code is readable, than trying to figure out how to lay all the boxes on a piece of A4 so the lines don't cross.

by hal 9000

What I meant to say was flow charts for the entire project or something similar as just being handed several reams of fanfold paper with code on it is a bit of a job to wade through to find out what the wanted result is.

I know that these are horrible things but no matter what they look like or how they are drawn to me they are Flow Charts and what I mean here is a laid out foundation of how they expect to develop the complete program not just a section of it.

Personally I don't like drawing the things either so I just do a rough sketch free hand of what I want to achieve and then set out to figure out how to get to the end point reliably with as few things in between as possible. The original bit of paper ends up looking like a dogs breakfast and generally ends up in the shrewder as I work along and redraw the thing at the early concept stage. Once I start punching code it tends to be forgotten or at best is only there as a rough guide as to what I want to achieve but I only use these things on a Big Project not the little ones where I've already got the basic solution worked out in my head before I start.

Just how would you guys like to look at a part way developed project when your only option is the code and a development plan. If you don't know what they are aiming for it gets very hard to understand just what it is that they are attempting to do with the code. :p

Col ]:)

by apotheon

As Jaqui mentioned, the LAMP platform isn't just PHP. That P can also stand for Perl or Python.

The LAMP platform refers to use of Linux as an OS platform, Apache as the webserver, MySQL as a database back end, and Perl, PHP, or Python as an embedded scripting language. Perl and Python can both be used as embedded scripting languages with Apache modules for that purpose, though for maximum portability across different webhosts you're probably going to find that PHP is the only universally available option in that regard. You'd be hard-pressed to find a credible webhost that doesn't support PHP in all webhosting plans (static-page plans notwithstanding), but it wouldn't be difficult to find such hosting that only supports Perl through CGI and doesn't support Python at all.

Interestingly, Perl, PHP, and Python are all capable of running as either embedded or CGI scripting languages on an Apache webserver. If you have ultimate control over your webserver environment, I'd recommend going with Perl or Python over PHP, because both are much richer languages and don't suffer the same kind of security SNAFUs and bloat that PHP deals with. PHP can make for easier rapid development when doing things you're already used to doing, though, because whatever you're trying to do is probably accomplished by a function. Just keep in mind that rapid development doesn't always translate to quality development. You might have to write your own functions in Perl when the same task might be accomplished by a core function of PHP, but you'll also have more control over how the task is accomplished and, if necessary, it will be easier to translate to another scripting language.

If you're looking for alternatives to LAMP for much the same benefits, you should look into alternatives for each individual letter in the acronym, and consider the benefits and detriments of these alternatives.

L: Linux
Instead of Linux, you might look into FreeBSD or OpenBSD. FreeBSD is a very credible server OS. Its performance capabilities are about on a level with those of Linux, and it is nearly as cutting-edge. New versions of various server applications might lag a few months behind implementation on Linux, but for a stable server environment that may not be a concern. OpenBSD lags quite a bit, but is widely hailed as the most secure OS in the world. There has only ever been exactly one remote exploit vulnerability discovered for OpenBSD in default configuration, compared with the dozens for Linux and the hundreds for any version of Windows. You might also consider what flavor of Linux best suits your needs, which is a very involved discussion beyond the scope of this post.

A: Apache
There are other webservers available than Apache. Something that is increasingly being used in some instances is a combination of Apache and TUX servers. TUX is used to serve static pages, and Apache to serve dynamic pages. This, obviously, is only of use in instances where you're running a site that makes significant use of static (X)HTML pages, however. The major benefit here is in page loading speed for static pages, and in removing all static site load from the Apache server so that all its resources can be directed toward dealing with dynamic pages. For my purposes, it's generally a better idea to just run straight Apache sites and be done with it. There are other dynamic-site capable webservers out there, but generally Apache is the best combination of power, stability, speed, and flexibility that you're going to come across.

M: MySQL
In general, your best options for SQL database management systems are MySQL and PostreSQL. I have yet to come across any particular need for MySQL over PostreSQL, but it tends to be more widely used in large part because there's a greater user base. PostgreSQL is generally accepted to be the more scalable of the two, though you'll find some who claim MySQL is just as scalable (or at least should be as soon as better database clustering support has been incorporated in MySQL). There are bigger (and far more expensive) DBMSes available, such as Oracle, and smaller DBMSes like mSQL, but generally in any implementation where mSQL and its smaller siblings would handle the load, loads will be small enough on MySQL and PostreSQL implementations that you wouldn't notice any significant performance increase. In fact, if you're doing something small enough where MySQL starts to look too big, you should probably just be using flat files anyway.

P: Perl, PHP, Python
As already mentioned, each of these is an alternative to the others. There are other options as well. JSP is a big one, but generally suffers performance problems in comparison with LAMP under most circumstances since the server-side code must run on top of a virtual machine, and doesn't grant any particular benefits in exchange for that. JSP is basically the non-Microsoft answer to ASP. ASP can run on Apache, but that raises the question "Why?" Aside from simply migrating away from the vertically integrated Microsoft solution a piece at a time, I have a hard time coming up with any reason to do so. If you already have an entire site written in ASP, ASP on Apache might be a good direction to go, as an intermediary step toward migrating off ASP entirely, but it's not a final step by any means. You might want to look into other solutions, however, such as Ruby on Rails. Ruby is a "purely" object oriented interpreted language that is absolutely fantastic for web development on the server side, and Rails is a web development framework for Ruby. Look into it.

For use on a home-run website, PHP will probably suit your needs just fine. Ruby on Rails might be a good alternative for you, however, if you're willing to learn a language (though Ruby is an incredibly easy language to learn, as I'm discovering). Python is also easy to learn and quick in development, though I personally find it very difficult to look at. Perl is great, quite possibly the most flexible non-Lisp language out there. There's probably more Perl code already written for web applications than any other scripting language, as well, though you should be careful to check out the code to make sure it's not all BAD code (as you should with prewritten code from dubious sources in ANY language). Perl thus makes it easy not only to write your own code, but to avoid reinventing the wheel when code already exists for what you want to do.

I think that pretty much sums it up for now.

by jgarcia102066

I really appreciate you taking the time to detail this post as you have. This is great information and will help me, tremendously.