How did you guys become competent at PHP

[Jake Arkinstall]

I think one of the best things to learn is not to blindly listen to people without understanding why they say what they say.

You get posts on this forum by people using the most bizarre conventions because someone once told them to, and because they never questioned it they can end up with a mess of code.

That's why I say to try and cut down on tutorials - some of them are pretty poor. A lot of them are written by people looking to score credit and look superior, so sometimes I've even seen articles about the basics of arrays and they throw in quick mentions of polymorphism when using objects as arrays. By doing that they are confusing the reader (who is bound to be a beginner) more than anything. If you feel confused by a tutorial, or advice from another programmer, do yourself a favour and ask about it here, we'll help you clear it up.

Off Topic:

This is one of the things you might want to consider, DeathShadow. Sometimes the guy just wants assistance with PHP - talking about other programming languages (pascal, assembly etc) just makes the original poster lose confidence. I'm not trying to be critical, I'm just echoing some things that people have asked me about. If a 4 year old asks how to make green paint from yellow and blue paint, don't start talking about photons!

Basically, you've shown yourself to be a good programmer, you don't have anything to prove in that respect. Just focus on being a good teacher

Another big thing you need to grasp is that for any one goal, there are a vast amount of ways to accomplish it. Experiment.

[TheRedDevil]

Yeah, but to me that's like leaving out the closing tags or WORSE, HEAD/BODY tags in HTML, I like to see it so that structure is maintained.

I used to think that way as well, a few years ago I changed my mind when a "worm" went havoc on servers (yea, you know those security holes hehe) and one of the things it did in addition to try to infect other servers was add an iframe to the bottom of all web pages it located on the server.

If the ending php tag is not used, then the script will just terminate with an error instead of sending the content of the iframe to the user (usually a virus).

Due to this I now prefer to not use the ending php tag, just as an additional security step as you never know when another guy want to prove something by writing a worm that take advantage in a security hole in the OS.

Because I don't want it to be able to be modified once initialized inside my db.php. I consider that another vulnerability. Much less that if I don't restrict it's scope by passing by reference, it's a global again.

You can archive that by making the adapter wrapper a singleton as well.

Actually, makes it simpler in a way, though really it's a sacrifice to go cross-SQL. mySQL doesn't take the exact same queries as Oracle or MSSQL - hell there are even enough differences for postGre to occasionally need query tweaks. Do you inline those changes with if statements on every query, or do you just include the appropriate values by extending the PDO object with an array containing all your queries appropriate to the task at hand? This part of why I LIKE pdo->prepare a LOT.

We seperate the queries from the application. That way it dont matter if we use MySQL, MSSQL or XML as long as we update the part that provide the application with the data.

It require more work initially, but it gives more flexibility in the end. As you say changing from one database to another does force you to update basically every query if you want to get the best performance out of the engine.

Which is fine for large companies handling it in-house, but not for something joe-six pack is going to install on any of the billion fly-by-night shared hosts. I think the target audience plays a crucial role on that.

That is true, but at the same time can you really prevent joe-six pack from shooting himself in the foot? What I mean is, in the end can you really prevent any kind of screweup done by joe-six pack or his web host?

You might be surprised but there is a lot of crappy hosts out there, especially those offering cheap shared hosting. On many of them, you are able to access any other account on the same server.

They use for example open_basedir restrictions and believe everything is good, but all I need to do to get around that is to write a Perl script.

We might be blessed due to the fact that we normally do not deal wth joe-six pack, but even if we did I do not believe we would ever go to the steps you are taking on the security as in my mind its redundant. Kind of like a company selling a gun to joe-six pack can not provide a security to prevent him from shooting himself with the gun.

The question with security in depth, is always how deep will you go.

Please do not take this the wrong way, if you really want to do the extra work and manage to archive this, then all kudos to you.

I've gotten that for 40 years, getting a little sick of it. As I've said other places "Englisc, modor wyrter! Gedon eow cweþan hit!?!" -- I know my manner of speech is archaic, but DAMN is education really that piss poor now or something? Half the time you make a simile now you get some jackass saying "That's off topic, what's that got to do with it?"... and yes, I said simile, not Smiley.

The lack of general literacy really is one of my pet peeves. Maybe I should just turn my hat around backwards, put my pants around my knees and start typing everything in L33T?

Not a bad rant, you actually made my day

On that note though, you should remember that the beauty with forums is that people from different nationalities comes together to discuss a topic they like. This means that not everyone has English as their first language.

My comment was not regarding your way of writing, and I am sorry if it sounded like that. What I meant was that you from time to time can jump around between topics, making it harder to follow your point.

[SituationSoap]

Off Topic:

Is there a way to ignore users on this forum?

[Jake Arkinstall]

Off Topic:

Yep.

But I wouldn't recommend it.

[Salathe]

Off Topic:

> Is there a way to ignore users on this forum?

Hopefully not me!

P.S. How would one go about requesting that [QUOTE] tags within [OT] tags get styled more sympathetically?

[SituationSoap]

Definitely not you, Salathe. I've got a lot of respect for the work that you do.

Edit: I also just found out that you can embed the quote tag inside the OT tag. It looks a bit funky, but works well enough.

[Ramiro S]

This is the How, for me:

1. Found an interest in learning: I wanted to develop a gaming site. That was my motivation.
2. Found a resource (Book, Internet, Lessons, etc) that worked with me: in my case it was a book.
3. After I did some hobby stuff I started working as a junior developer... work, jobs gave me lots of experience.

Another advice: do not let others tell you about structures, things like: you need a degree to to be good or you need to learn design to understand programming or you need to learn other languages to perfect ..... bla bla bla... follow your own thing and try to get the job done. There is no a single way to accomplish this.

Decide on what you want to do and what you like. We are all good in some things and bad in some others.

[Jake Arkinstall]

things like: you need a degree to to be good or you need to learn design to understand programming

Ironically, some of the worst programming I've come across has been from people with degrees, and the worst programmers are usually designers at heart. It's also usually the case that the worst designers are programmers...

[deathshadow60]

Ironically, some of the worst programming I've come across has been from people with degrees, and the worst programmers are usually designers at heart. It's also usually the case that the worst designers are programmers...

I'll second that. As a rule of thumb computer science degrees are worth less than a sheet of bog roll. In an industry where three years is obsolete and five years is the scrap heap, what good is a four year program using ten year old textbooks?

Some of the worst design practices and coding I've seen came not from people with degrees though - you want really bad look at the garbage put out by educators.

Though when both combine to do something; Hell, just look at your average college website for examples of educational ineptitude.

Or programs like the one a friend of mine took at the local state college that teaches nothing but Dreamweaver and Flash Pro, and doesn't even tell the students that there's this thing called HTML underneath the WYSIWYG.

[oikram]

I think one of the best things to learn is not to blindly listen to people without understanding why they say

This is, a mon avis, a great advice and me, as a newbie, I always found this part the hard to deal with.

We have different backgrounds, and different professional structures, on doing our programming. When I ask some questions about (to me, expecting to be) simple tasks, sometimes, I got answers... quite advanced, that require a lot, on our structural development environment (resources equipment, background knowledge, etc....).

So, many times, I end up desperate that I'm unable to accomplish a given advice. And of course we can't. We need (literally) years to properly understand the answer.

1) Sometimes, good/best programming answers are not the best on a newbies point of view.

2) However, this doesn't mean that as newbies, we should forget about them. No we don't. Many times, I got my answers understood some months later.

3) When we doing this alone, knowing what is good and not good programming practice, seems hard to find. My advice to solve it is to, cross data. And compare. And study.

Example of 3)
I see many database connection tutorials here and there, showing different ways for establishing a connection to it. Many use mysql_connect, and if we do not stop there, we look and we see that mysqli is better then the other and if we keep looking we can see that PDO is actually even better... and so on... How do we know if it's better or not? We listen the arguments, the details, and generally, found that something that some articles ignore, and others, don't.

[starbar]

Anyone ever considered that some of the advice written is not to help others but to confuse others?

Also that it creates books, and is another article when someone is stumped for a bit of filler content for their blog or other type of site.

To become competent i would say that its wise to believe half of what you read and hear.

[TomB]

This is, a mon avis, a great advice and me, as a newbie, I always found this part the hard to deal with.

We have different backgrounds, and different professional structures, on doing our programming. When I ask some questions about (to me, expecting to be) simple tasks, sometimes, I got answers... quite advanced, that require a lot, on our structural development environment (resources equipment, background knowledge, etc....).

So, many times, I end up desperate that I'm unable to accomplish a given advice. And of course we can't. We need (literally) years to properly understand the answer.

1) Sometimes, good/best programming answers are not the best on a newbies point of view.

2) However, this doesn't mean that as newbies, we should forget about them. No we don't. Many times, I got my answers understood some months later.

3) When we doing this alone, knowing what is good and not good programming practice, seems hard to find. My advice to solve it is to, cross data. And compare. And study.

Example of 3)
I see many database connection tutorials here and there, showing different ways for establishing a connection to it. Many use mysql_connect, and if we do not stop there, we look and we see that mysqli is better then the other and if we keep looking we can see that PDO is actually even better... and so on... How do we know if it's better or not? We listen the arguments, the details, and generally, found that something that some articles ignore, and others, don't.

What you always have to do is look at an article date. (Now this is really an argumentum ad novitatem but it does somewhat apply here.) In software engineering, things move fairly rapidly. If you're looking at an article from 5 years ago, it probably uses mysqli, from 10 years ago? mysql. From 6 months ago? PDO. This isn't a coincidence. Things change.

In our industry things move fast. There are always improvements. Both in the underlying technology and the principles behind the way we do things. The key is to embrace change. If you get set in old ways you will get left behind. I like to continually critique my own work... I'm constantly improving the way I do things, even when I reach a point where I don't believe I can improve I do. Usually significantly.

[Jake Arkinstall]

Unfortunately, the qualifications required to write web articles are simply nadda, zilch, nought - nothing.

I STILL come across articles which explain things using register_globals and magic_quotes. I even emailed one author explaining this and he called my style 'out of date' because of "the amazing invention of register_globals which many developers are yet to realise the full potential of". When I replied with a dozen or so links to articles backing my points up, he replied 'just because they can write articles doesn't mean they know what they're talking about'.

Ironic.

[Stomme poes]

^ he was also right (with the quote "just because they can write articles doesn't mean they know what they're talking about.").

I can find a dozen articles claiming the "62% font size means 1em = 16px" myth.

This is absolutely a problem newbies to anything web dev related run up against, as a few have mentioned above. This is why it's good to have a decent community, with people recognised as experts in their field posting information so that newbs and not-so-newbs can find the right way. Discussion with peers after getting more advanced is also essential.

[deathshadow60]

When I replied with a dozen or so links to articles backing my points up

Did you include PHP.NET where they have the big red box at the top saying "This feature has been DEPRECATED as of PHP 5.3.0. Relying on this feature is highly discouraged." ??? I love it when people argue with the page defining what things are

I can find a dozen articles claiming the "62% font size means 1em = 16px" myth.

10 SP, Ten. 16px is the 100% default size at 96dpi - and technically it's 62.5% aka 10/16ths -- but yes, the articles claiming a fixed relationship between px and default font are pure bull. Doesn't help when you have people who should know better saying complete nonsense like "This is why browsers use the 96 dpi rule when deciding how all of the absolute units relate to the CSS pixel."

BULL - Camel mannered tunic wearing molly-coddle. :P

Given who said that, is it any wonder things are such a mess? Apparenly never used IE or Opera on systems of different DPI, or changed the default font size in his browser to match that behavior... Basically changing the system metric - something I've been doing since the release of Windows 3.0... Though I seem to be saying that a lot lately about 'revolutionary new features' in software.

Really that's one of my big pet peeves, being I've been a 8514 (Win 3) large fonts (Win9x)/120dpi(winNT)/medium 125%(Win7) user for two decades...

[Jake Arkinstall]

Did you include PHP.NET where they have the big red box at the top saying "This feature has been DEPRECATED as of PHP 5.3.0. Relying on this feature is highly discouraged." ??? I love it when people argue with the page defining what things are

I sure did

He claimed that PHP5.3 was "purely experimental" and lacked "key" features of PHP such as register globals ( ), not because it was bad practice but because it was harder to program into the new core, or something like that.

Fortunately he made a few dangerous errors; no escaping, no magic_quotes with his script requiring magic_quotes, having the same password on his demo admin account as his blog, bit of SQL injecting found me the password and I emailed his password to him. I think he got the hint, his domain is now 'parked', 3 months on.

[Salathe]

I love it when people argue with the page defining what things are

I don't meant to be the pedant who swoops in to pick on very minor, off-topic, niggles (or, maybe I do! Perhaps even in this case, the issue is with a word choice rather than the point you're making) but the warning boxes stating that the feature is deprecated are not defining anything. The manual, in an ideal world, gives information on the current and past states of PHP, and is always playing "catch up"; information in there can be outdated, missing or just plain wrong (of course, the majority isn't). If one really wants to point out a definitive source of whether a PHP feature is deprecated or whatever, make that the source code itself (e.g. to show that the register_globals setting is deprecated in 5.3, point here). That said, a big, red box in the manual is, as it should be, considered authoritative and, as previously noted, is generally correct. At least enough to be a useful citation.

[teckyhead]

I started messing with PHP when it was PHP/FI and Rasmus was developing it. Initially it just seemed pretty cool for form processing. In 1996 I had to up-scale our 24 line dialup BBS that was running on a couple of OS/2 Warp boxes to a full blown ISP service with Cisco routers, Ascend MAX access server etc. By that time PHP was doing quite a bit more than just form processing so it was a really good fit for a backend admin system that had to be able to manipulate MySQL data and unix shell accounts. It was also fairly natural for me as I'd been programming in C since the late 80's.

So - in my case, becoming proficient with PHP involved creating a number of business systems. Since the start of the 2000's I've been doing all sorts of other things as well. I do a lot of modifications on existing sites as well as building some new stuff from time to time.

I kind of like what an earlier poster said - think of something that is beyond your skill level and then don't give up .

[Alex]

But seriously, to become competent at PHP, practice practice practice.
Challenge yourself, and if someone tells you you're doing something wrong, or that there's a better way of doing something, be glad for their help.

That's exactly right. That pretty much sums it up.

Ironically, some of the worst programming I've come across has been from people with degrees, and the worst programmers are usually designers at heart. It's also usually the case that the worst designers are programmers...

I'll second that.

I'll third that. Looking back on my former employment... It's the developers I worked with who held computer science degrees that had the worst practices, most limited knowledgebase and slowest learning curve. The core of the development department lied with me and one other self-taught developer.

EDIT: @Jake --- I'm developer at heart (As evidenced by my recent commitment to outsource design and focus on development) but most consider me to be a fairly decent designer, too. While it's often true that designers make horrible developers and developers make horrible designers - I don't think it's a given.

[baileylo]

I'll second that. As a rule of thumb computer science degrees are worth less than a sheet of bog roll. In an industry where three years is obsolete and five years is the scrap heap, what good is a four year program using ten year old textbooks?

The point of comp sci major is not to teach you how to program. If you want to program in php buy a 40 dollar book and go to your hearts content. Comp Sci degrees teach you fundamentals(or are ment to). There are plenty of people who scrape by with cs degrees.

I think the simplistic way to form a computer science degree is a book will tell you how, a degree will tell you why. This also applies with ITT and other "career" colleges. They'll teach you everything you need to get in the market now. But you can learn cs with a 30 year old book because it's not like bubble search has changed, or Big O has changed in the past 20 years.

[xtiansimon]

When you go to school they teach you how to read things like this from #man trap:

BASH BUILTIN COMMANDS:
bind [-m keymap] [-lpsvPSV]
bind [-m keymap] [-q function] [-u function] [-r keyseq]
bind [-m keymap] -f filename
bind [-m keymap] -x keyseq:shell-command
bind [-m keymap] keyseq:function-name

Someone always suggests the best answer to a problem is read the man page. I know there's a logic to this. I just don't get it intuitively. I know there's got to be some info document explaining how to read this logic, and its never in a published how-to book. That's why I like yous guys. golly gee, you're swell. you tells me what I need to know in real words.

[gemini181]

Build something that is beyond your skill level. Don't give up on it, persevere till you figure out how it is done and go forth and build it.

Through this process you will realize how much you are capable off, and what you can accomplish. To reiterate the fact, do not build something that you are familiar with. You need to throw yourself in the deep end.

Yes, great advice!

I'll tell you when I get there.

True, I'm only part way there.
Try everything

Books which used too look way too hard, suddenly start too make sense after spending countless hours modifying open-source code.
Code which used to look way too hard, tends to get easier when taking a second (or third) look, a few months later.

[Mittineague]

Things getting easier sure is true. Of course I keep raising the bar as I go so I still have plenty of difficulty at times.

I've noticed by looking at my older code - what I thought was great at the time - often looks like total kludge now so I guess I must be getting somewhat better.