Mac Oil Price Widget

By · Published · apple, dashboard widgets, dashcode, mac, osx

Because there doesn’t seem to be a good, simple way to track oil prices on the Mac dashboard anymore since the previous widget I used quit working, I whipped up a quick little widget that allows me to monitor the price of Crude Oil on the New York Mercantile Exchange.

You can download it over on its own page.


PHP Filtering: Validation, Sanitizing and Flags

By · Published · php

PHP's filter functions are really, really great. I've started using them almost any time I need to get input from a user and, personally, I don't think you should use the old $GET, $POST unless you know what you are doing and there is some specific thing you're trying to accomplish that you can't with filter. Filter forces you to think carefully about what inputs your script takes and what format it takes them in.

But there are also some behaviors of filter that can bite you in the rear if you aren't really careful. One of these is knowing which flags you need to pass and what the difference between validation and sanitizing, when is the right time to  use each, and what flags to use. I ran into a good example of this today where I messed it up.

I had configured filterinputarray to pull in a variable as FILTERVALIDATEFLOAT, probably because I wasn't thinking like a user and instead was thinking like a developer. I'm the type of person that, when a form wants to know my phone number, I only enter 10 digits without parentheses or dashes. But users are different. They like friendly things. In this case, the user was entering "16,473.54" and the like into that box.

Now, I can look at that and say, "yeah, that's a float" (actually, it's currency). It should be considered a valid value. But FILTERVALIDATEFLOAT will throw this out because it has a comma in it, unless you pass FILTERFLAGALLOW_THOUSAND. Then, and only then, does it return the above as a valid value (in this case "16473.54").

But I looked at the code again. In this case, the value doesn't need to be there except in a specific case, which I handled in error checking in the code, so I switched it to a Sanitize value instead. It's probably a good idea to only use  FILTERVALIDATE* functions when your user has to give you a valid value and your script would fail if that wasn't the case. If a validation returns false, you should fail the process and return a (nice) error message to the user. Sanitize functions allow you to accept a little wider range of data and still return a valid value from it. The docs have a great example of this involving email addresses.

So if you're writing PHP these days, definitely use filter. Just be careful and mind the flags and the difference between validation and sanitizing.



dystill 0.2 released

By · Published · dystill, e-mail, python

Version (do those really matter anymore? :P) 0.2 of dystill has been released.

This version brings a significant change to dystill. Namely, it breaks the unofficial association between dystill and Postfix that has existed since I first wrote it last year. I did this for a couple of reasons:

  • To hopefully increase adoption. Dystill now (really!) stands independent of any MTA. Use it with whatever you want (sendmail, Qmail, etc). You actually always could, but you'd have to ape some Postfix tables. You don't have to do that anymore.

  • To make it easier to write web-based front-ends to dystill's MySQL database, enabling users to add rules.

This was done by adding an "email" column to the filters table, updating that field with the recipient address, and dropping the old userid field. Also, a "maildirpath" config variable was added to the config, specifying where the maildirs live.

There was also a minor bugfix I came across the other day where certain uncommon (but legal) characters could result in unreadable maildirs.


Do Version Numbers Matter?

By · Published · linux, ramblings

The recent announcement by Linus Torvalds that the next release of Linux will be 3.0 has provoked rather furious discussion around the Internets about whether or not the incrementing of the version number is warranted. Linus himself has said that "absolutely nothing" has changed. "It will get released close enough to the 20-year mark, which is excuse enough for me, although honestly, the real reason is just that I can no longer comfortably count as high as 40."

This got me to thinking about the nature of version numbers. Once upon a time (when versions were driven more by engineers and convention, and less by marketing), a version number meant something. Major, minor, revision. A major new release that modified significant portions of the code from the previous release incremented a major version number. Version numbers less than 0 were beta releases.

Linux has been at 2.x since 1996, and at 2.6.x since 2003. Mac OS has been at 10.x since 2001 (even though the current version of OS X is significantly different from the original release in 2001).

Meanwhile, Google Chrome has blasted through major 11 "versions" in three years. Mozilla is planning to release versions 5, 6, and 7 of Firefox this year. You can't tell me that they are going to change major parts of Firefox three times this year. In this case, version numbers are purely being driven by marketing. They need to "catch up" to Chrome and Internet Explorer.

But we live in a different world now. One where, arguably, version numbers are becoming less and less important. The growth of "app stores," I think, is desensitizing your average user to a version number. While apps in the app store still have versions, I couldn't tell you what "version" any of the apps on my iPhone are (other than the OS), and I bet you can't either. Any of the apps I've installed from the Mac App Store I could not tell you the version of them. I just know that, when I see the number on the icon, I know I need to do updates. The updates happen, and I get a new version with whatever new features are there (or, in the case of the Twitter app, whatever features have been removed).

Then there are web apps which are versionless. What version of Gmail do you use? You don't. You use Gmail. Sure, there's probably a revision number or something in the background, but the user has no clue what version they're using. And they don't need to, because there's no action they need to take.

So version are numbered in a wide variety of ways depending on the product and overall seem to be becoming less important as the growth of broadband, "app stores," web apps, and automatic updates make thinking about version numbers less important. So why does it matter if Linus ups Linux to 3.0? Ultimately, it's just a number.


MySQL mathematical operations and NULL values

By · Published · mysql

So I came across an interesting quirk in MySQL the other day. Let's say you have a table schema and some values that look like this:

+-------------------+------------------+------+-----+---------+-------+
| Field             | Type             | Null | Key | Default | Extra |
+-------------------+------------------+------+-----+---------+-------+
| page_id           | varchar(30)      | YES  |     | NULL    |       |
| clicks            | int(10) unsigned | YES  |     | NULL    |       |
+-------------------+------------------+------+-----+---------+-------+

+---------+--------+
| page_id | clicks |
+---------+--------+
|       1 | NULL   |
+---------+--------+

And then let's say you pass the following SQL statement to MySQL:

update page_click_count set clicks = clicks + 1 where page_id=1;

If you come from a loosely-typed language such as PHP, you would probably expect clicks for page_id 1 to now be 1. But that's not the case in MySQL. After the query is run, the table will still look like this:

+---------+--------+
| page_id | clicks |
+---------+--------+
|       1 | NULL   |
+---------+--------+

Not only does the query fail, but it fails with no warnings given. It appears that mathematical operations on null values silently fail.

There are a couple of ways around this. The first and most obvious is to set NOT NULL and a default value on the column. In the example above, this would work. The NULL value in that field becomes a 0 and you can to normal mathematical operations on it. But what happens if, for whatever reason, you can't do that? We actually have this situation in a few places at dealnews, where NULL represents a distinct value of that field that is different from 0. In this case, you can use COALESCE() to fill in the appropriate value for the field.

update page_click_count set clicks = coalesce(clicks, 0) + 1 where page_id=1;

Edit: Brian Moon informs me that this is actually part of the SQL specification. So hooray for specifications. Still, it's kind of arcane; in working with MySQL (and PHP) for a decade now, this is the first time I've ever actually encountered this. Hopefully this helps someone who was as confused as I was.


Interview Questions for Programmers

By · Published · php, ramblings, business

Over the years, I've seen a number of blog posts relating to common questions that should be asked of programmers. Obviously, this is going to depend on exactly what position you are hiring for, but there are some good "gateway" questions that can be used to determine whether or not an applicant you are interviewing can ... well ... even program at all. If they even have the mindset that makes a good developer.

A common one I've seen tossed around is Fizz Buzz. The challenge goes something like this:

Write a program that prints the numbers from 1 to 100. But for multiples of three print “Fizz” instead of the number and for the multiples of five print “Buzz”. For numbers which are multiples of both three and five print “FizzBuzz”.

Now, to anyone who even has a basic understanding of programming, this is super simple to solve using a modulus operator. But apparently many people applying for even entry-level development jobs cannot solve this problem. According to the article linked above, even one "senior developer" took 15 minutes to solve this problem.

Earlier today, a friend posted something on Facebook that inspired what I think it another good, intermediate to difficult level programming question that also looks for pattern recognition. The relevant part of the post began by stating: "This year July has 5 Fridays 5 Saturdays and 5 Sundays." There is the question! It would go something like this:

The month of July 2011 has 5 Fridays, 5 Saturdays and 5 Sundays. Calculate the next 50 times there will be a month that has 5 Fridays, 5 Saturdays and 5 Sundays.

Woah, so how to go about solving this problem? Well, look at a picture of July 2011. Notice something interesting about this month in relation to the question? This month has 31 days (the most any month can have), begins on a Friday and ends on a Sunday. And that's the solution! It's any month with 31 days that begins on a Friday!

With this in mind, it's pretty easy to come up with a PHP solution:

<?php
$count = 0;
$num_found = array();

while(count($num_found) < 50) {
    $count++;
    $ts = strtotime("$count months");

    if(date("t", $ts) == 31 && date("N", strtotime(date("Y-m-01", $ts))) == 5) {
        $num_found[] = date("F Y", $ts);
    }
}

print_r($num_found);
?>

Note that I make use of PHP's strtotime function, because it is the Swiss Army Knife of date manipulation in PHP. This would need to be adapted for use in another language.

So now tell me: what are some other questions you've been asked or asked in an interview?


Xcode 4

By · Published · apple, microsoft, objective-c, ramblings, windows, xcode, mac, osx

So today, out of nowhere, Xcode 4 finally landed as an official release. After seemingly forever in beta, and me quipping more than once about it's similarity to Duke Nukem Forever, Apple finally pulled the trigger and released it. But something changed.

Xcode now has a price. And that has left me, as both a Mac user and a Mac developer, with a lot of questions.

It's either $4.99 if you're not a registered, paid Apple developer, or free if you are a registered, paid Apple developer (with all its $99 per year price tag glory). Supposedly there's some crazy accounting reason that they have to charge for it. This, of course, leaves open the possibility that Xcode will soon be free again once OS X 10.7 arrives. But, it also leaves open the possibility that Xcode will no longer be distributed with OS X and will always have a price tag. It may not even stay $4.99. It may be $49.99 or $499.99.

There are additional questions, too. Does this mean that Apple is still distributing Xcode as a bundle with GNU GCC? Because there are things (such as MacPorts) that rely on the underlying foundation provided by the developer bundle that don't actually use Xcode. Before, those were completely free. Now, they cost $4.99 unless they have split the underlying compiler from the IDE. And if they are still distributing it with GCC, that leads to all kinds of crazy interesting licensing questions.

But I think the worst part is that there is now a barrier to entry, however low, to being a developer on a platform that is already a minority in market share. I can't understand how Apple potentially believes that it is good and right to trade short term profits for long term growth in the number of potential developers. For the future of the Mac platform, I sure hope this isn't their line of reasoning.

So, let me tell you a little story.

My first dabbling in programming came courtesy of QuickBASIC back in the MS-DOS and Windows 3.1 days. This was the late 80s or early 90s, so I would have been 10 or 11 at the time. I stumbled across the Qbasic environment included with MS-DOS by accident and found Nibbles. And, after playing it, I discovered that I could change things by making changes to the strange text presented on the screen. I could change colors and speeds. But it would be a couple of years before I really understood what I was doing.

When Windows 95 came out (and along with it, Visual Basic 4), I talked my parents into getting me a copy. I don't remember how much it cost but it was probably a lot because it was one of the few Christmas presents I got that year. But boy did I run with it. I've periodically felt guilty over that expense because I didn't actually make anything really useful with it, but it was instrumental in furthering my education. Now I could do things on my computer far beyond what poor ol' Qbasic was capable of. So I wrote lots of silly little programs. I put together a "family newsletter" one year that was installed and ran as a piece of software. I was pretty proud of that. I even wrote some software for my high school as part of a software development and AP Computer Science courses.

Eventually, I would move on to other things. Other versions of Visual Basic, Java, C, a brief foray into LISP and Forth-based languages for programming MUDs, and eventually web programming. First in Perl, then in PHP. I even landed my first paying programming job while still in high school, writing applications for a local transit contractor. At first, these were Visual Basic applications. But by the time I left (August of 2000) everything was going to the web and so were we.

But I can trace everything - my entire career, and my consuming passion for software engineering - back to Qbasic and Nibbles. A silly little game about a block snake, and a free development environment included with the operating system. Had I not stumbled on Qbasic and Nibbles, there's a chance I would never have been a developer.

This is not about $4.99. I spend more on coffee in a week than that. My worry is about that 11 year old kid out there somewhere who may never get the opportunity to stumble across Xcode or the sample applications in /Developer and realize the raw power they possess. This is an area where Apple, a company with billions in cash on hand, should be happy to show a loss. It would be to the benefit of their platform, both now and in the future.

One of the great benefits of the Mac platform has been it's low barriers of entry to developers. Sure, one could argue that the hardware is more expensive (and I could counter-argue that, for the quality of the equipment you are getting a bargain), but the development tools have always been freely available online and included with the machine. You could dabble in programming to your heart's content. Sure, if you want to put something in the app store(s), you had to pay for admission, but there was nothing stopping you from getting all the way to that point, or even distributing your creations on your own.

But this new trend of charging for the development tools - even if it is a paltry sum - sends, to me, a worrying signal about the course Apple intends to tread. They've now moved the gate from the last step to the first step. It's a course that Microsoft, as above, once tread.

Microsoft? They now give away a version of Visual Studio for free.


BASH Quickie: Backing Up MySQL Databases

By · Published · linux, mysql, bash

In some ways, after years of doing programming and scripting, I'm now sort of rediscovering the power of the shell.

Tonight, I was working on my server and remembered that I needed to start backing up my MySQL databases (which you do also ... right?). So instead of writing a script to do that, with a little research, I was able to come up with a way to:

  1. Dump each database to a separate SQL file, with a timestamp.

  2. bzip the file.

  3. Keep 5 days worth of backups for each database, rotating the oldest backup off.

Here's what I came up with:

cd /backup/mysql; 
for i in $(mysql -BNe 'show databases' -u root -p<password>); do 
    mysqldump -u root -p<password> $i | bzip2 > $i-`date +"%Y%m%d"`.sql.bz2; 
    rm -rf $i-`date -d "-5 day" +"%Y%m%d"`.sql.bz2; 
done > /dev/null 2>&1

Shoved that in my crontab. Works great. Linux rocks.


Automatically Setting Adium IM Status with AppleScript

By · Published · apple, applescript, mac, osx

I have more than 20 various IM accounts set up in Adium on my Macintosh. But during the day, the only one I really want to be active is the one I use for work. The remainder, I want to leave logged in, but showing as away with a warning not to bother me unless it is important. But half the time I forget to set all those accounts as away, or I forget to set the work one as available, or some other issue that would arise out of a manual process interferes and too often it doesn't get done.

Enter AppleScript. I whipped up a surprisingly easy AppleScript to do just this:

tell application "Adium"
    go away with message "Working. Please do not disturb unless necessary."
    go available first account
end tell

Because the work account is the first one, this makes it easy. It just sets all accounts as away and then sets the work one available. I shove this in my MarcoPolo ruleset to fire when I arrive at the office.

The script to reverse the change when I leave is even easier. This is fired when I leave the office:

tell application "Adium" to go available