I Hate Web Programming!

[crh3675]

I have been in web development for 15 years. I have learned damn near everything for the web including HTML, CSS, XHTML, Javascript, VBScript, Ajax, .NET, XML, XSLT, jQuery, Prototype, ActionScript, Ruby On Rails, MySQL, Postgres, MS Access, SQL Server, Apache, IIS, XSL:FO, Photoshop, Dreamweaver, OOP, SOAP, XMLRPC....and much more

After 15 years of doing web development, I am really not enjoying it anymore. I have worked in about 8 different places and just can't find happiness in my career. Every place wants work performed quickly, and they don't want to pay people what they are worth. Instead of transferring costs to clients and telling them that it actually "takes that long to develop things the right way", the almighty dollar takes over and I spend my time "hacking together" solutions.

So, not only do the positions not pay great anymore, but the companies rush projects so that they can generate revenue. So, I hate web programming! Everyone thinks that they can "get it cheaper". And they sure can:
cheap = crap. I'm not interested in serving crap anymore.

Whoever got the bright idea for "Rapid Application Development Frameworks"? Can we next expect "Rapid Home Builder Frameworks"? Oh, wait, we already have that since modern homes are built with plywood and untreated lumber. Anything that gets created "rapidly", becomes a ticking time bomb after it is completed.

This industry sucks! It used to be enjoyable but now the jobs won't pay, the clients won't pay and everyone wants "instant gratification". That's my rant for the day.

[rhsunderground]

you could always go into business for yourself...

[scragar]

What's the quote, you can have 2 of the 3, fast, cheap or good.

It's just a shame most people want fast and cheap, then complain that it's not good as well, go figure.

[Charles]

Plywood? I don't live too far from Da Capital (DC) and nobody uses plywood any more except for the roof. It's chipboard on the floors and nothing under the vinyl siding these days. Thieves have begun to break into these houses with nothing more than an a box cutter. People now have negative equity in what were half-million dollar pieces of junk that will fall over before the mortgage is paid. All because people thought that they could get ahead by buying a piece of junk and then selling it to some other poor Joe. But I digress.

It's a job, it's not supposed to be fulfilling. If you want a vocation then go become a doctor or a nurse or a porn star. Go do something that benefits other people. Or if you really want fulfillment, there are plenty of people who have been down that path and they are pretty clear about what you have to do to find it. Go sell all your possessions, give the money to the poor and then pursue the spiritual path.

On the other hand, there is the way of Zen. I once saw a Zen master at work though I doubt he even knew what the word "Zen" meant. I was inspecting the roof of a building as it was being installed a few years back and as I looked over the parapet I saw him. He had the humble task of breaking up the chunks of solid tar and feeding them into the melting pot. This was a new construction so there was mud and debris and ugly confusion every where. But working slowly but deliberately he had pulled off the paper wrapping from the cylinders to make a carpet atop the mud. On top he strung out the cylinders and and split each with an ax. The cut face of the solid tar caught the light like a cut gem. In the midst of all that confusion and ugliness and profanity he created peace and beauty while going about that most humble task.

The rest of us make the best of our jobs using the income to seek fulfillment elsewhere. You could donate your web skills to worthy non-profits. You could volunteer your time doing something else. You could bankroll somebody else's good works. Do keep in mind, this is not a good swing in the economy to have an existential crisis and quit your job.

[Mr Initial Man]

What do you mean by "negative equity"?

[Charles]

Now that housing prices are adjusting downward to something more reasonable, quite a few people are finding that they owe more on their mortgage than the value of their home.

http://en.wikipedia.org/wiki/Negative_equity

[drhowarddrfine]

you could always go into business for yourself...
I am in business for myself and it's no better. I'm the definition of the Aquarius astrological sign. A technologist with an artistic bent. I love making code look and act like art and I feel the OPs pain when some people, usually larger businesses, that hand me their crap and want me to work within these restraints.

Fortunately, I'm in a position where I don't have to accept that anymore.

[KDLA]

I am in business for myself and it's no better.

Agreed. The clients you "select" end up being just as misguided as a full-time employer.

[Jeff Mott]

crh3675, some companies have a flagship product—a CMS, a framework, a platform, whatever—that is continuously developed and improved, and a branch is created for each client project. You may enjoy working on a team developing a flagship product, because you're developing for the company rather than any particular client. This kind of team encourages good, clean code since that code will be the base for every subsequent client project.

[xvszero]

I think part of it just has to do with the way the business world has expanded in general. I'd imagine 10-15 years ago most companies who wanted websites had some money... it was still too early for everyone and their brother to jump online, it was "novel" marketing.

Now you pretty much need a website to run a business. It's not just the big ones anymore, it's everyone... from small businesses to independent artists to my sister selling sewing services out of her cheap apartment (true story.) Oftentimes the only reason the site even exists is because they know they'd look bad if they didn't have one.

As much as it sucks to look at it this way, for a lot of these type of people it is, financially, a better decision to get a halfassed website that sort of mostly works... than it is to pay a premium to get a high quality website that always works. In general low end products usually have a BIGGER market (yet smaller profit margin) than high end products. Sucks, but a fact of business that may have taken awhile to hit the web development world.

[KDLA]

I think part of it just has to do with the way the business world has expanded in general. I'd imagine 10-15 years ago most companies who wanted websites had some money... it was still too early for everyone and their brother to jump online, it was "novel" marketing.

Now you pretty much need a website to run a business. It's not just the big ones anymore, it's everyone... from small businesses to independent artists to my sister selling sewing services out of her cheap apartment (true story.) Oftentimes the only reason the site even exists is because they know they'd look bad if they didn't have one.

As much as it sucks to look at it this way, for a lot of these type of people it is, financially, a better decision to get a halfassed website that sort of mostly works... than it is to pay a premium to get a high quality website that always works. In general low end products usually have a BIGGER market (yet smaller profit margin) than high end products. Sucks, but a fact of business that may have taken awhile to hit the web development world.

True. We just don't have an online-minded business population yet. Businesses are still struggling with marketing to several media types, and really don't have a firm grasp on integrating all these resources into one general marketing plan.

One of my website clients still cuts out clip art, places it on a piece of paper and resizes using a copier, instead of using a computer program, then pastes it onto a page of text created in Word 97. As to visualizing what their website can do for them (or even thinking it critical because of their diverse customer base), that's pretty much a waste. Their website is a document repository of scanned versions of their cut-n-paste publications; that's all.

[xvszero]

True. We just don't have an online-minded business population yet. Businesses are still struggling with marketing to several media types, and really don't have a firm grasp on integrating all these resources into one general marketing plan.

One of my website clients still cuts out clip art, places it on a piece of paper and resizes using a copier, instead of using a computer program, then pastes it onto a page of text created in Word 97. As to visualizing what their website can do for them (or even thinking it critical because of their diverse customer base), that's pretty much a waste. Their website is a document repository of scanned versions of their cut-n-paste publications; that's all.

Hmm. Reminds me of this place I was working a part-time / temp job at last year. One big part of the job was the guy printed out emails and left the print-outs for me and had me type them into the "comments" section of this software he used to track customers. I was thinking of explaining to him that if he didn't want me in his email (understandable) there were still about 100 different solutions for getting the data from the emails to his software without manually typing all of it (some of it was LONG!)

But since I was otherwise barely employed and I was hourly there and it was only a few hours of work a week, I didn't really want to give him ideas that would give me less work...

[Sunny G]

This came to my mind...

If you meet three 4$$ h0135 in one day, chances are you are the 4$$ h013.
I put that in leet for censorship purposes.

Agreed, the industry has come to the point of instant gratification, and webdevelopment is no longer a true art. Maybe you could do web consultation, or something related? How about being the "cheif web administrator" for a local business? You could help them keep Wallmart and other related places away.

[toicontien]

Web development needs the equivalent of the production line that Henry Ford introduced to heavy industry. Sure mass produced automobiles aren't as nice as the hand built ones, but most people can afford a mass produced car. The object is to create software that does most of the work for the client. Then build off of that. It's not that people want crap, but that they can't afford not-crap.

A well programmed and easy to use content management system is what you need. You've got enough experience, start doing the "good coding" on the side for yourself. Once you've got a really nice CMS made, start selling it. If it's easy to use and inexpensive, you could sell a ton of them. Make it easy to use, easy and quick to set up, market it and watch the cash start flowing in. :)

[crh3675]

A well programmed and easy to use content management system is what you need.


I have always felt this but in the recent times with fworks like RoR, Cake PHP and many, many others I figure, why bother?

None of them "solve" any problems, they just inhibit the programmer by requiring them to learn a "new way" of thinking or a new set of rules and requirements to follow. And, most of them use non-typical programming methods. Maybe I will try to build something, something that actually solves problems without inhibiting creative programming.

[Jeff Mott]

RoR, Cake PHP and many, many others I figure, why bother? ... they just inhibit the programmer by requiring them to learn a "new way" of thinking or a new set of rules and requirements to follow

The same can be said, and was said, of object orientation. OO isn't a direct solution either. It's a new way of thinking, with a new set of rules and requirements. And yet OO is one of the greatest innovations in software development.

You're absolutely right that frameworks don't solve an immediate, specific problem. They aren't libraries to be included. They're a new way of thinking, with a new set of rules and requirements, exactly as you said. But don't make the mistake of believing that a new way of thinking about software is a trivial thing. Frameworks may be, and is already shaping up to be, the next great innovative stride.

It's unfortunate you feel inhibited by them. They may indeed be your answer to match quality with speed of development. Today, few developers working in teams would even think of not using one. All that is debated is which to use.

[drhowarddrfine]

@Craig
Just wanted to say I absolutely agree with everything you said.

[Webnerd]

It's unfortunate you feel inhibited by them.

We have the debates all the time at work and RoR seems like the choice but I find that only certain components speed up process, the rest complicate it. Not to mention I despise the Ruby language. I appreciated the object connectivity that ActiveRecord provides but hate the obscurity to the database. I also don't enjoy having to install buggy plugins for things as simple as "enum" type columns.

The other thing I really can't bear with in Rails is the entire MVC (should be MCV as that is how it's processed) structure because it requires me to modify 3-4 files to accomplish anything. Newbies may appreciate Rails but as an experience programmer, I find it more time intensive when I need to create anything rapidly.

The only one language or way of thinking that I have fully adopted and use with great pleasure is jQuery as it eliminates redundancy, has not been error prone, allows me to solve problems and doesn't lock me in to it's paradigm. That's probably a whole other topic though.

[drhowarddrfine]

From all the stuff I read every day, Ruby seems to be falling out of favor.

[rirara]

Webnerd I totally agree with you.
I spent too much time with Java, PHP and so on.
Currently I welcome every new change with fear. There are a lot of frameworks but if you need something a little less common but logical and useful you must learn to hack.
And I live in cheap outsourcing destination - of course I'll be pissed.

And the forums are full of offers like - EASY job - get 5-10$ for some kind of coding. Are they joking or what?

[tivrfoa]

what do you consider a fair wage for a good programmer?

have your own business is good (more freedom), but you have to manage many sites to earn well. just one site it's capable too, but it's more difficult because it's a very competitive market.

I think that the best way to have a good remunaration is to work for a good company. How much do you think yahoo's, google's, m$'s pay for their developers?

I think it is possible to make a fast development. the key: reusability. but to achieve this you need experience and be organized. many people (most teachers) compare soft developing with construction (eg: house), but soft is much more easier (if you don't agree, you don't have logical reasoning :D)!!! you can not use the same brick, wood, cement again.

there are many programmers out there (most are not so good, better for us :)), so you need be excellent to become remarkable.

[tivrfoa]

The other thing I really can't bear with in Rails is the entire MVC (should be MCV as that is how it's processed) structure because it requires me to modify 3-4 files to accomplish anything. Newbies may appreciate Rails but as an experience programmer, I find it more time intensive when I need to create anything rapidly.
I agree with you that should be MCV, but not the rest! MVC provides a more understandable and MAINTAINABLE code. take care, or newbies will leave you behind.

[rirara]

You haven't seen enough sh*ts in the code. Of course you can reuse s**** many times ...

[gundelf]

Isn't this at least partly another side effect of the economic times? When investors were willing to throw money into web technology it was easier to do the job right. It's not just the web sector that is suffering this way. It's harder to find money for other software development projects as well.

As for development tools, have you had a look at the Seaside framework? If you're looking to enjoy the development process itself it might be what you're looking for. I've been using it for a couple of years on my web projects and I think it's pretty amazing.

-Carl

No commercial links permitted

[Webnerd]

take care, or newbies will leave you behind.

Ahhhhh, but that is where the naivety lays. These newcomers will "lock-in" to Ruby and have their own "tunnel" to go through. Ruby will fade as a language, it has no stature and certainly cannot compete with languages like PHP (albeit can be sloppy) and the immense libraries of code and support. Hey, I could be wrong but after 16 years of programming, I knew that Cold Fusion had a limited life span when it came out; hence, I never learned it's over-glorified XML style.

Programming web apps and sites need to work from a "concept" not a framework.

This thread got me motivated so I started my own "less-than-perfect" but not over-engineered concept of programming that anyone can download/use here:

http://code.google.com/p/piwi-tools/

It's not perfect as I did not spend a lot of time on it. But it meets my needs and works as intended. It is for developers who like to program. Enjoy it or trash it, it's my concept that I wish to share and I hope that someone else can benefit from it.

[Mig]

I came to the exact same conclusion about 6 months ago. When your employer constantly wants you to rush things and hack together solutions, it becomes very frustrating and you lose all job satisfaction. I wanted to be making things I was proud of, not crap that 'will do the job' (just about).

I left my job for this reason. I took a break over the summer of 3 months, and have now been in a new job for 6 weeks. After clearly identifying the reason for my unhappiness (as you appear to have done), I decided I would set about finding my 'ideal' job (or close to), and be stubborn in my search.

My new pay package is a vast improvement, and I'm in charge of the company's web dev, so I have more control. I made it clear in my interviews that the most important thing I was looking for was job satisfaction, and my manager has already noticed that I am keen to produce work to my own standards, that I will be proud of.

I'm aware that not everyone can simply leave their job and search for a better one for 3 months, but I am lucky in the sense that I am young and have no financial commitments. I'm glad I did this early (I've only been in the industry for 4 years), and I would urge anyone who is feeling the same way to do something about it - however you can. Don't just sit around at your old job thinking about how much it sucks that you don't get to create work that you're proud of; do something about it ASAP. I think a couple of my old colleagues are actually kind of envious and wish they had the guts to leave and try finding a more fulfilling job.

[buntine]

The other thing I really can't bear with in Rails is the entire MVC (should be MCV as that is how it's processed) structure because it requires me to modify 3-4 files to accomplish anything. Newbies may appreciate Rails but as an experience programmer, I find it more time intensive when I need to create anything rapidly.
Considering you seem familiar with the concepts behind MVC, it seems strange that editing multiple files is a problem for you. The separation of business logic from presentation is key here. And I can only imagine trying to effectively write unit tests for an old-style mishmash web application...

If multi-files is a killer for you, why not look into some other, lighter frameworks such as Sinatra?

[Webnerd]

Yeah, I'll lear yet another framework that will eventually fail or require some bizarre modification. Our clients "tell us what to do" regardless of the framework we build on so in the best of interest of our projects, scalability and flexibility, I'll stick with PHP

[Jeff Mott]

...regardless of the framework we build on so ... I'll stick with PHP
That doesn't make sense to me. "Sticking with PHP" has nothing to do with using or not using a framework. There are frameworks in PHP too. Plenty of them.

The other thing I really can't bear ... is the entire MVC structure because it requires me to modify 3-4 files to accomplish anything.
I have to admit, this is one of the weakest reasons I've heard. Do you believe the same of HTML, CSS and JS? Should you go back to pre-1997 HTML, little or no CSS, and only inline JS, all because you don't want to modify three files to accomplish anything? Because that's what your reasoning implies. I hope you realize now how terrible that reasoning is.

Programming web apps and sites need to work from a "concept" not a framework.
If it's only you on a project, then sure, you can probably get away with that. But if you're working on a team with perhaps a dozen developers, then the code needs to have structure. There needs to be a well defined place for everything. Otherwise each developer will be writing code in different places and in different ways, and the code will slowly turn into an unmanageable, incompatible, mess.

[buntine]

Yeah, I'll lear yet another framework that will eventually fail or require some bizarre modification. Our clients "tell us what to do" regardless of the framework we build on so in the best of interest of our projects, scalability and flexibility, I'll stick with PHP
I can't help but think that if you really find yourself limited by "obscurity to the database" (by which I am assuming you are talking about the Object-Relational mapper) then it's most likely because you haven't taken the chance to truly understand why they have been developed in the first place.

If you haven't already, I recommend you read The Pragmatic Programmer (no, it's not a biased Ruby book and was published long before Rails existed).

I'm not necessarily telling you to switch over to the flavor-of-the-month framework -- but keeping such a closed, harsh view of frameworks in general may really inhibit your productivity.

Cheers.

[Webnerd]

I am glad there are different perspectives on this topic. Everyone's real world experience is different and I hope one day I can actually use a framework that solves a problem but for the time being, I'm going to focus on enjoying my work the best I can - for as long as I can - without them.

I wouldn't say my perspective of frameworks is harsh or closed as I have had a lot of experience with many of them. My opinions of frameworks are based on real experience and I have done a few analysis' of how much more "productive" they really made our projects.

In the places I have worked that have a vast range of experience, frameworks have proven to be more trouble to integrate and work with and they certainly did not increase productivity. I'm sure others will of course disagree or explain "why" that is but I really am not interested in finding out why as I did not do the hiring.


And just to clarify my response

The other thing I really can't bear ... is the entire MVC structure because it requires me to modify 3-4 files to accomplish anything.


Comes from the fact that I have to frequently view and edit files on using Unix vim and having to navigate a file structure up and down to locate models, views and controllers, becomes rather inefficient.

[Jeff Mott]

[can't bear MVC] Comes from the fact that I have to frequently view and edit files on using Unix vim and having to navigate a file structure up and down to locate models, views and controllers, becomes rather inefficient.
Gotcha. Navigating through directories is hard, therefore MVC is bad. Makes perfect sense.

I can only hope that one day you will look back and feel appalled at the reasoning you used.

But getting back to today. Obviously MVC isn't your real problem. It's developing from the Unix command line that slows you down. And that is what you need to change. You could work from Windows or Mac, whatever your preference, and FTP your changes to the Unix box. You could commit to a code repo and deploy to the Unix box. There are plenty of options. You don't need to keep doing what you've been doing.

[buntine]

But getting back to today. Obviously MVC isn't your real problem. It's developing from the Unix command line that slows you down. And that is what you need to change. You could work from Windows or Mac, whatever your preference, and FTP your changes to the Unix box. You could commit to a code repo and deploy to the Unix box. There are plenty of options. You don't need to keep doing what you've been doing.
Whoa, whoa! You and I are going to have to disagree here, Jeff. Nothing is faster and more proficient than working directly from a terminal!

WebNerd, I use Vim for ALL of my development and text editing. I have no idea what you are talking about. Navigating directory structures with BASH and Vim is super fast and efficient.

[TheTeenScripter]

Whoa, whoa! You and I are going to have to disagree here, Jeff. Nothing is faster and more proficient than working directly from a terminal!

WebNerd, I use Vim for ALL of my development and text editing. I have no idea what you are talking about. Navigating directory structures with BASH and Vim is super fast and efficient.

Everyone has their own web designing preference, and if their preference is not possible on a unix then uploading the data from a machine that they could use proficiently to a unix could be more efficient. For example, if it takes 2 hours to work on the unix machine and 2 seconds to update it, compared to working on the code for 1 hour on a windows machine and 5 minutes to upload it, you still get it done alot faster.

[scragar]

Everyone has their own web designing preference, and if their preference is not possible on a unix then uploading the data from a machine that they could use proficiently to a unix could be more efficient. For example, if it takes 2 hours to work on the unix machine and 2 seconds to update it, compared to working on the code for 1 hour on a windows machine and 5 minutes to upload it, you still get it done alot faster.

As long as you use well done sym links for getting around and know your way around vi/vim it takes much less time on a command line that it does in a gui(since I can delete 5 lines, the paste them out later with a grand total of four key presses, no need to hold control, no using the mouse, it's just faster, by far), and since time loading etc is minimised you save a lot more time there. No matter how annoying the directory structure is it is still faster, you just have annoyance navigating the paths.

TIPS btw, if you alias different things to "cd /whatever" then you can quickly type something in to change directory, without having to remember where you are, or type the full thing every time. For example I have something like:alias Wi='cd /home/web/includes';# quick hop to my standard includes dir
alias W='cd /var/www';# to web root
alias DBu='mysql -h localhost';# mysql as user
alias DBr='mysql -h localhost -u root -p';# mysql as root
That way I can quickly get around and manage anything I need.

[Jeff Mott]

Whoa, whoa! You and I are going to have to disagree here, Jeff. Nothing is faster and more proficient than working directly from a terminal!
I work from text editors in Windows. I press Ctrl+S and the document is automatically and immediately uploaded to the appropriate place on the linux dev box. It doesn't get much faster or simpler than that.

[buntine]

I work from text editors in Windows. I press Ctrl+S and the document is automatically and immediately uploaded to the appropriate place on the linux dev box. It doesn't get much faster or simpler than that.
That is pretty standard from any platform. Infact, simply uploading via an ancient protocol like FTP on every save is potentially sloppy and inhibiting. I can think of a few ways to match or exceed that:

1) Work from the Linux dev box via SSH.
2) Work from remote terminal with the exact same functionality (FTP files on save).
3) Work locally (on any platform) and use deployment tools to automatically sync the server with your working copy when you commit your work to whatever version control software you use.

[Jeff Mott]

simply uploading via an ancient protocol like FTP on every save is potentially sloppy and inhibiting

Please be specific. Sloppy how? Inhibiting how?

[buntine]

1) Totally unencrypted and insecure
2) Totally blind and without integrity. Files are not verified on the receiving end (checksums, etc). If data is lost during transit, you will not know.
3) Initiates multiple connections when transferring, which would suck severely if you save anywhere as often as I do!

[Jeff Mott]

1) Totally unencrypted and insecure
See SFTP.

2) Totally blind and without integrity. Files are not verified on the receiving end (checksums, etc). If data is lost during transit, you will not know.
See SFTP.

3) Initiates multiple connections when transferring, which would suck severely if you save anywhere as often as I do!
I save dozens of times each day. So does everyone else on our teams. But that's still nowhere near enough traffic to bog down a server.

[buntine]

I know very well what SFTP is. I have written large wrappers for SFTP.

I am, however, skeptical that YOU know what it is... SFTP is not at all FTP. It is implemented using SSH.

[Jeff Mott]

SFTP is not at all FTPNot under the hood, no. I tend to lump them in the same category because so do applications. In the Dreamweaver site manager, you would enter an SFTP address in the FTP host field. Also in FileZilla, you can enter an SFTP address just the same as FTP. So the way we as end users use FTP and SFTP is the same.

[Webnerd]

Wow, topic changed direction a bit. I use a Mac for development by default mapped to our dev servers. I only have to use Vim when some idiot screws up the production sites using FTP instead of "svn up". My problems are more environmental than anything. (I also need more experience using Vim.)

[chestertb]

I was discussing just this issue with a colleague at my local dog park yesterday.

The problem is that code cutting is just plain boring. I mean, there's a limit to the number of times you can get satisfaction from writing an SQL query that works, and an even bigger limit to the number of times creating workable cross platform javascript is challenging.

We compared writing for the web to the sort of work we were doing 10 years ago, and we really miss the superb IDEs we had then.

I used to use a thing called Omnis 7. It was an intuitive relational RAD tool that allowed me to, for example, create a graphic multi-route ferry reservation system in a long weekend. To do the same in php, html, javascript and mySQL would take months.

It had features that allowed the creation of page elements, like drop lists, in a single function. It had stored array fields. It had an intelligent code editor that allowed you to create a line of code in just a few key strokes. And because it wrote the code for you, there weren't any syntax errors, or hours trawling through lines of code looking for a missed comma or semi-colon.

It also had a wysiwig design module that was quick, flexible and powerful. Point, click, drag, done.

My dog-owning friend at the park agreed. Omnis 7 was just one of a whole raft of application development tools that took the drudgery out of programming and allowed us to concentrate on delivering applications that worked. I really miss it, and I only stopped using it because its web support wasn't great. (Though with benefit of hindsight, maybe I should take another look.)

It seems that application development has taken a few huge steps backwards to accomodate the web, and I'm not seeing any client/server web development tools that even come close.

[NogDog]

Sounds like what Delphi for PHP (http://www.codegear.com/products/delphi/php) is trying to do, though I've not had an opportunity to play with it.