Code Frenzy

About a week ago, Chris and I “launched” TravelStash.  By “launched” I mean we have now opened it up for general consumption.  Anyone can join TravelStash and use it just like Rheanna and Sanjay did on their recent South Africa trip, Margaret and Alan have on their trip around the South Island of New Zealand and of course, how Chris and I have consistently used it to keep our friends and family updated on all our travels over the past couple of years.

Basically, there’s a lot of neat stuff already built into TravelStash and we are constantly thinking of new enhancements for it.  The latest addition has been the ability to notify your facebook account (if you have one) when you publish an article or a wishlist item in TravelStash.  That way, if for some strange reason your friends and family haven’t subscribed to your TravelStash RSS feed and aren’t obsessively hitting your TravelStash group or trip pages to see what you’re up to, you can still let them know by posting a small note to your facebook wall and your friend’s news feeds when you publish a new travel post or wishlist item.  We’ll soon be extending this to travel tips too, just to be consistent.

Another feature we’ll be releasing soon is the ability to save an entire trip (photos, map, blog posts and comments) to MS Word or PDF format.  This will enable you to save your holiday stories and memories in a format which you can then print out and make a book out of.  If you save it to MS Word you then have the ability to edit and alter the way it all looks and could even take it down to a print shop to have professionally printed and bound into a beautiful coffee table book.

In any case, at this stage TravelStash is 100% free to use and has no ads.  We think it’s pretty cool (ok, that’s a bit of an understatement… we think it’s freakin’ AWESOME!) and encourage everyone to create an account and use it on their next holiday.

The other day while on a bus in Brisbane, I witnessed a classic error message FAIL:

Clearly in this case letting the end-user (i.e. the general population driving by) know that the modem had failed was not useful at all.  What would an end-user be able to do with that knowledge?  Nothing.  They can’t fix it nor would they care to.  Showing them that message is pointless and makes the manufacturers of the mobile street display look stupid.  SOMEONE ought to know about the failure but in this case that someone is certainly not the end-user.

Another concern with display error messages such as the one I witnessed the other day is that it gives those malicious souls out there who thrive on breaking other people’s software a better idea of what will make the software break.

Personally, I’m a fan of generating more than one type of error message upon software failure and audience.  If the end-user is likely to be the general population (i.e. website users), then you should look at what caused the software to fail.  If you’re allowing users to generate errors by entering invalid data into a form, then you should probably let the user know that they entered some invalid data and that they should try entering something valid next time.  However, if you’ve coded restrictions into the software so that they should never be able to enter invalid data but somehow they do anyway, then you may be a victim of a hacker’s attack.  In this case, you do not want to admit that they have almost just broken your site because if you do you’ll only be encouraging them to try again and you may have just given them the extra information they needed to totally own your site.  Instead, the message you display to the end-user should be a friendly and generic/vague one, something along the lines of ‘Oops!  Your request returned no results.  Please try again later or contact support.’.  That way, the hacker will be none the wiser or at least they won’t think that their last input came so close to pulling your site down.  The same error should be logged somewhere on the system (event log, log file, database error log…) with a much more specific error message.  If it’s a critical error then perhaps it should even be emailed or sms’ed to an administrator automatically.

I guess the moral of the story is that it really doesn’t matter that you’ve handled all the potential errors in your code.  What really matters is how you’ve handled them, in particular, who you’ve alerted and what information you’ve given them.  Telling the world that the last input caused an integer overflow is probably not the best way to keep hackers out and users in!  If you’re a developer, think carefully about every line of code you write and if that line can raise an exception then make sure you know what to do when it happens.

On Tuesday, Chris and I went along to our first .NET User Group (or Queensland MSDN User Group) meeting here in Brisbane. Having attended a few back in Wellington, I thought it would be nice to check out the Brisbane version.

The meetings are held in the Brisbane Microsoft offices, in a very very flash building down town. Like usual, the meeting started off with pizza and soft drinks which is always much appreciated by those attending. The crowd started off small but within a few minutes, there were a lot of people there – I’d say at least 40 or so showed up, not bad (of which only about 3 were girls but that’s to be expected!). The couple of .NET User Group meetings I went to in Wellington were probably about half this size but then again I guess Wellington is a much smaller city than Brisbane.

The topic for the evening’s talk was “Silverlight 2.0 and WPF – what’s the same, what’s different?”. The speaker, Joseph Cooney, was very well informed and well spoken. As the title would suggest, he compared Silverlight 2.0 and WPF but not to try and say that one was better than the other. Instead he wanted to try and inform the audience as to why you would want to choose one over the other, what situations suit which more.

What I took away from Joseph Cooney’s talk is that Silverlight is basically a lightweight version of WPF. It’s meant to be a 4~5mb download and you really can’t package up too many libraries in that. In order to keep that size down, Microsoft have removed mundane values such as all but the main HTML colors in the Color namespace. Really, who even knows what CornflowerBlue looks like??

There are some controls which are only available to WPF and like-wise, others which are only available to Silverlight. I guess this means you can’t really call Silverlight a subset of WPF.

I can’t wait to get to use WPF and/or Silverlight in a commercial manner and hope to start on a project at work or at home that’ll let me spend a bit of time investigating these new libraries and what they’re capable of!

And as for the Brisbane .NET User Group, it looks like a great place for networking and meeting other like minded .NET-geeks so I’m sure I’ll be turning up to future meetings.

I’ve just had the toughest time getting SQL Server 2005 WITH Management Studio installed, so I thought I would blog about my thoughts on the matter and my rather round about way of getting it going.

1 – Install SQL Server 2005 Developer Edition. Stick Disk 1 in. Everywhere says ‘click on the link to start the SQL Server Installation Wizard.’ Well guess what, there isn’t a link that says that! There are two links, one to install components and another to install SQL Native Client. Grr.

2 – I click on the components one and it seems to install everything I need to run SQL Server 2005, EXCEPT for Management Studio. Why? I don’t know. I do however wonder why it never asks me for Disk 2… hmm.

3 – Open up Control Panel, Add or Remove Programs. Look for “Microsoft SQL Server 2005”. Click on the “Change” button.

4 – Select the ‘SQL Server 2005 common components > Workstation Components’ option and hit Next. When you are finally given the option to ‘remove’ this component, do it. Just do it.

5 – Once this is all finished (takes forever!), repeat step 3. When you are shown the the same dialog box as in step 4, there won’t be a ‘SQL Server 2005 common components > Workstation Components’ option. Instead there will be some text telling you you should use the ‘To install a new component, click here’ link at the top of the dialog box. Click this link. This will open up a series of dialog boxes very similar to those encountered in step 2 – your basic installation. Select the last option (I think it was ‘Tools’?).

6 – It should tell you it’s going to install a bunch of stuff and then start installing it all… pretty quickly however, it will ask you to insert Disk 2. This is a good sign.

7 – Insert Disk 2.

8 – Wait for the installation of all the different components to complete.

9 – And like magic, you should now have a whole bunch more shortcuts in your start menu under Microsoft SQL Server 2005. Although this seems like all you did was install, uninstall and re-install, believe me, it’s the only way I managed to get it to work.

As I said above, I’m not sure if I’m the only person this has ever happened to but I have seen plenty of other blog articles on how to get Management Studio installed if it doesn’t appear to be there already so I believe I’m not alone. None of the other solutions I saw helped me either so hopefully this will help someone else in the same position as me!

As some of you may by now know, Chris and I have written a website called ‘TravelStash’. It’s essentially a blogging site intended for travelers. It integrates things normally associated with blogging (i.e. writing blog posts, uploading photos, comments) with a map. The structure behind TravelStash is a bit complex but it does make sense. Basically when you join up, you get your own membership. Once you’re a member, you can create a group for yourself and invite others to join your group. Each member can belong to many groups. Someone in the group creates a trip and then all the members of that group are able to log in and post an article about that trip.

We used TravelStash for our ‘Camp Europe’ holiday and more recently for our ‘Exploring the East’ holiday (even though we haven’t quite finished blogging about that one yet… lazy!). Although we’ve almost finished writing all the code necessary to allow others to join, create their own groups, blog about their trips and upload photos from their holidays, we haven’t quite opened it up to the general public. We’re hoping to do that real soon though so keep posted on that one

In the mean time however, I thought I might add a couple of new features to TravelStash. First is the ability to view an entire group’s photos. Until now, photos were only shown on the article with which they were associated. I figured it might be nice to view all of a group’s photos on one page rather than having to individually click on all the articles to see all their photos. You can see an example of that here. Neat eh?

The other feature I added only last night were ‘Travel Tips’. I had lunch with a good friend the other day who told me that it would be neat if we had travel tips posted on TravelStash. He’s right, Chris and I had thought of doing that anyway so I figured I might as well get onto it and just do it. So a few code changes later and we’ve got ‘Travel Tips’ on TravelStash! As you can see, not many travel tips just yet but I’ll make sure to add some more and hopefully when we open TravelStash up to the general public and people flock to it, they’ll add their amazing travel tips too. I was thinking it might be cool to allow comments on travel tips too in case others have something to say about it. Ahh, so many ideas.

Anyway, that’s about it for new features on TravelStash. Chris and I have heaps of ideas for new features, however, so with a bit of luck there may be some cool new stuff on there soon. And obviously we’re really excited about ‘launching’ it to the general public. I’m sure we’ll both learn lots from that experience.

Part of setting up my new blog was to migrate all my old articles from Textpattern to WordPress. As with all migrations and updates, I was somewhat cautious, thinking it would probably take a lot of time and effort. Ahh well, I’ll never know if I don’t try!

Turns out I was right. WordPress actually provides a script to do the migration for you. This script should take care of importing categories, users, posts, comments, and links (blogroll). All you have to do is fill in a small form with your Textpattern database details – database user, password, name, host and any prefix you might have used on your Textpattern tables – and hit Import! However, when I ran it, it did nothing. Yup, that horribly annoying problem that all developers absolutely loathe. The script ran… and did nothing. It kept telling me it had not imported anything but it didn’t give me any reason as to why that might be the case nor did it give me any errors or warnings. Urgh!

So I searched Google. I found out that a few people have had issues with this script. A couple of other people (such as Alex Brie) have written their own instructions and even scripts (or at least modified the one that comes with WordPress). Most of the advice I read on the matter said you should have both your Textpattern and WordPress tables in the same database for the script to work. Apparently that’s actually not necessary but I tried it anyway. Still nothing imported. I tried using the modified import scripts that I found… still no luck. Argh!

At this point I was starting to get mildly frustrated that something that should be quite straight forward just wasn’t doing anything for me. A lot of the comments on the blog posts I was reading had people quoting errors or saying that they’d finally got it all working. I didn’t get either. It was at this point that I surrendered to my geek-soul and opened up the PHP import script in Dreamweaver. I read through some of the code and eventually found some places where I could stick some debug code in. First thing I tried was to output the values that I’d entered into the import form (my Textpattern database details). BINGO. There was my problem. My database host includes slashes and colons and there was a nice wee line in the script that was scrubbing all those characters out. Grrr…. I changed that line from:

1

add_option('txphost',  sanitize_user($_POST['dbhost'], true));

to:

1

add_option('txphost',  sanitize_user($_POST['dbhost'], false));

I guess whoever wrote the script didn’t realize that some of us have complex database hosts!

Anyway, that seems to have fixed the problem at least in my case. The rest of the script worked great. So now, as you can see for yourself, all my old articles written on mindtrip using Textpattern are now sitting here in Wordpress! Awesome.

Over the past few months, my article on GPS Tracking with Google Maps & AJAX has received many hits and attracted a multitude of comments requesting that the code for it be made available. I apologize for the delay but I’ve been away on holiday and far away from my trusted laptop. Now that things are somewhat back to normal and I can spend hour upon hour on my laptop again, here is the code that so many of you have asked for. I hope it makes sense and that it helps you all with what you’re trying to achieve. Obviously I haven’t posted the entire HTML and PHP coz it’s not all relevant to the GPS plotting but I hope what I have posted gives ya’ll a general idea on how to solve this particular problem. I’m sure there are many different ways to do this but this was just one way that I tried.Please feel free to ask questions and I’ll try to answer as best as I can!

...

Within the head section of the page that contains the gps tracking map:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45

<script language="JavaScript">
 
function setTimer() {
   window.setTimeout("show_data();",5000);
}
 
function show_data(){
   //Append the id (just a simple count) to the requestURL
   var requestURL = "http://www.sitename.com/getGpsPoints.php";
   var count = document.getElementById("counter").value;
   count = parseInt(count, 10);
   var queries = "?id=" + count;
   var url = requestURL + queries;
 
   //Increment the hidden counter variable
   document.getElementById("counter").value = count+1; 
 
   var request = GXmlHttp.create();
   request.open("GET", url, true);
 
   request.onreadystatechange = function() {
      if (request.readyState == 4) {
         var xmlDoc = request.responseXML;
         var markers = xmlDoc.documentElement.getElementsByTagName("marker");
 
         for (var i = 0; i < markers.length; i++) {
            var point = new GPoint(parseFloat(markers[i].getAttribute("lat")), parseFloat(markers[i].getAttribute("lng")));
 
            // Draw the MapMarker
            var mapMarker = new GMarker(point);
            map.addOverlay(mapMarker);
         }
 
         // Recenter the map
         map.centerAtLatLng(point); 
      }
   }
 
   request.send(null);
 
   //Reset the timer so that the page keeps refreshing itself
   setTimer();
}
 
</script>

Within the body section of the page that contains the gps tracking map, right at the bottom just before the /body tag:

1
2
3
4
5
6
7
8

<script language="JavaScript">
   var query = window.location.pathname;
   if (query == "/mapping/gps-tracking-with-google-maps--ajax") {
      var map = new GMap(document.getElementById("map"));
      mapSetup();
      setTimer();
   }
</script>

Separate PHP script that queries your database to retrieve the GPS points to plot:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

<?php
    header ("content-type: text/xml");
 
    $link = mysql_connect ("localhost", "username", "password");
    if (!$link) {
        die('Could not connect: ' . mysql_error());
    }
 
    mysql_select_db ("dbname");
 
    //If no search string is passed, then we can't search
    if(empty($_GET["id"])) {
        echo " ";
    } else {
        //Remove whitespace from beginning x%x% end of passed search.
        $search = trim($_GET["id"]);
 
        //Query the DB and store the result in a variable
        $query = mysql_query("SELECT * FROM gps WHERE id=".$search);
 
        //If no rows are found...
        if(mysql_num_rows($query) == 0) {
            echo " ";
        } else {
            //Stick the returned rows into a handy array for easy use
            $row = mysql_fetch_array($query) or die(mysql_error()); 
 
            //Write out XML using values returned by the query
            echo "<?xml version=\"1.0\" encoding=\"UTF-8\"?><markers><marker lat=\"".$row["lat"]."\" lng=\"".$row["lng"]."\" /></markers>";
        }
    }
 
    mysql_close($link);
?>

Since Chris and I have become addicted to Facebook, we decided it would be cool to have a plugin for our respective blogging engines (Chris uses WordPress, I use Textpattern) that would update our Facebook accounts’ mini-feeds when we posted something on our blogs. Thus we both set off on our epic missions to create Facebook applications and WordPress / Textpattern plugins that would support this behaviour.

Here I’ll relate my experience with my Textpattern plugin. I’ll leave Chris to talk about his experience with WordPress.

So yeah, basically, my experience was a nightmare.

This was my first attempt ever at writing a Textpattern plugin so I am by no means an expert at this. It took me a while to figure out how to compile my php code so that I could deploy it to my Textpattern instance on my web server. Then it took me another while to figure out how to hook into Textpattern ‘events’. I must say, the documentation for this is really poor and I found myself trawling through the Textpattern source to figure out what functions did what. Not pretty. The best (and only) sites I found that helped me somewhat were this article on threshold state and this Textpattern source repository.

When I finally figured out how to call the ‘register_callback’ functions and that I had to use ‘article’ as the event I wanted to register my own events against, I was able to get stuck into it.

My next challenge was how to add a checkbox to the ‘write’ tab of the admin screens. I wanted a checkbox that when checked, would essentially call my code to notify my facebook account of my post. If it wasn’t checked, it wouldn’t notify facebook. However, adding a checkbox and positioning it where ever I want in the flow of elements on the page was NOT an easy task. Below is a sample of the code I had to use to get this to work:

1
2
3
4
5
6
7
8
9
10

function mdt_event_insert_checkbox($buffer) {
   $fb_notify = graf('<small><label for="notify_facebook">'.
      checkbox("notify_facebook", '1', '1', '', 'notify_facebook').
      'Notify Facebook?</label>&nbsp;');
 
   $find = '<input type="submit"';
   $replace = $fb_notify.$find;
 
   return str_replace($find, $replace, $buffer);
}

NASTY! I had to find the ‘Save’ button, then replace the HTML with my new HTML that includes the checkbox. OMG.

Anyway, I finally got that working (see screenshot below).

Next I needed to grab the title of the new post, the friendly-url to the article on my site and the excerpt of the new article. This again was no easy task. The following code seems to do the trick:

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21

function current_url() {
   return 'http://' . $_SERVER['HTTP_HOST'] . $_SERVER['REQUEST_URI'];
}
 
function get_latest_post() {
   include_once txpath.'/publish/taghandlers.php';
 
   // get the latest article posted
   $rs = safe_row("ID, Title, Excerpt", 'textpattern', "1=1 ORDER BY LastMod DESC LIMIT 0,1");
 
   $title = ' has blogged: ' . urlencode(stripslashes(urldecode('<a href="' . permlinkurl_id($rs['ID']) . '">'))) . substr($rs['Title'], 0, 45) . '</a>';
   $msg = substr($rs['Excerpt'], 0, 180);
   $url = current_url();
 
   if (strtolower($step) == "publish") {
   $url = $url . "?event=article";
 
   } elseif (strtolower($step) == "edit") {
      $url = $url . "?event=articlex%x%step=editx%x%ID=" . $rs['ID'];
   }
}

So now the tricky part – how to connect to Facebook, log in as myself and call the Facebook API function that updates my ‘mini-feed’ with some content. I first attempted this by embedding the Facebook API into the Textpattern plugin so that the integration would be completely seamless. This, however, was difficult on many levels. First of all, I couldn’t get the Facebook API to compile into my Textpattern plugin. The php compiler would tell me that certain server variables weren’t defined, etc etc. I did iron out all those error messages and warnings eventually though and finally got the code to compile. However, I later realized that this was never going to be a feasible solution because in order for this to work, my Facebook developers’ API key and secret key would need to be hard-coded into the Textpattern plugin and if I were ever to make this Textpattern plugin public (which I intend to do), I would be sharing my API and secret keys, something I don’t necessarily want to do. Doh! So I had to change my thinking somewhat.

The solution to this was to basically write the Facebook part of the plugin as a totally separate web application and have the Textpattern plugin call this web app, passing through to it the title, url and excerpt in the form of a query string. This method lets the Facebook web application handle all the Facebook functionality (logging in and authenticating, remembering you once you’re logged in, calling the appropriate API functions, etc etc) and the Textpattern plugin basically doesn’t need to know anything about Facebook.

What a mission.

However, it all seems to be working now and I’ve used it once already. Those of you who are on Facebook and have added me as a friend will be able to see little posts on my mini-feed (on my profile page) saying “Annie has blogged…” with a link through to this article. Very cool When I am comfortable that things are working as intended and I tidy things up enough to make my Textpattern plugin publicly available, I’ll put it up on my here somewhere for people to grab.

Phew

Ever wondered why your application is chewing up so much of the processor? Or where all that memory is going?

Introducing dotTrace – a profiler for .NET applications, both desktop and web apps. It lets you run up a web app and take snapshots of it’s profile whenever you want. One view shows you a summary of the bits of code that the threads are busy executing complete with precise execution times, while another view shows you where the memory is being allocated, specifically which types of objects are using up the memory.

Something I particularly liked about dotTrace is its ability to show ‘hot spots’ – essentially a summary of functions that are using up the highest percentage of the thread’s time and effort. It shows how many calls to these functions are being made and where from. This quite nicely highlights some functions that are perhaps being overused in your applications (functions such as Concat and ToLower), allowing you to fine-tune your (already perfect, of course) code for performance.

Anyway, I’ve downloaded it and had a bit of a play – so far so good. Unfortunately, you either have to buy it (and it ain’t cheap) or deal with a 10 day trial license key. Still, I guess there’s no such thing as a free lunch!!

About a year and a half ago (October 2005), I tentatively applied for a Software Engineer in Test position at Google. When, to my surprise, a recruiter actually emailed me moments later and scheduled my first interview, I panicked and dug out all my old Computer Science notes and textbooks from my 4 years at uni. I had read that Google’s interviews were tough, so I thought I had better get studying. Over the course of a few months, I stressed out heaps, reviewed all my old comp sci notes (really, who remembers how to implement B-Trees, A* Search algorithms and MergeSort after leaving university?) and had 5 hour-long hard-core technical phone interviews with software engineers and testers at Mountain View.

Anyway, although the recruiter I was dealing with told me I had done very well during my interviews and that everyone was very impressed, I was eventually given the bad news – there wasn’t really anything for me at that time. I was pretty gutted but glad that I had gone through it all for the experience – no interview from then on would ever be as difficult as the ones I’d just undergone.

How wrong was I? About a year later (December 2006), one of the recruiters I’d dealt with during my interviews emailed me and asked if there was any chance I could come in for an interview for the same position at their Zurich office while on my holiday in Europe. Apparently they were expanding their test team and thought of me :). Well, since I was going to be in the area anyway, I said sure. However, since I was going to be on a much anticipated holiday, study and stress were not an option. I just accepted the fact that this meant I was going to totally suck at any tricky technical interview questions – but I thought, what the hell, I might as well.

All I can say is: yikes. As you can probably guess, I didn’t get the job, AGAIN. But this time I felt I actually did quite well in the interviews, despite the lack of study. Google flew me (and Chris) down to Zurich from Copenhagen and put us up in a really flash hotel for a night. The following day I had a full day of interviews on-site at the Zurich office. Basically this translates to 5 hour-long interviews with 5 different software engineers and testers. The questions they asked were very technical and covered a lot of different technologies. Towards the end of the day I was totally knackered and found it really hard to think straight, whether in code or in plain old English. 5 hours of hard-core technical questions when you have only a few seconds to think up a good answer and write code on a white board is TOUGH, no matter how much of a geek you are. NOW I’m pretty sure I’ve experienced the toughest interviews I will ever have to.

Anyway, I guess the moral of all this is that if you are given an opportunity to try out for something you really want but are pretty sure you’re not going to get – give it a go anyway. Sure, it’s gonna be hard and you’re gonna stress about it. Sure you’re gonna feel pretty crap about it if it doesn’t work out. But you will gain SOMETHING from it – even if it’s just the experience. In my case, apart from being forced to review a whole bunch of pretty cool geeky stuff that I’d totally forgotten about, I got a cool trip to Zurich, paid for by Google, I got to meet a bunch of really fun and talented software engineers (who by the way took Chris and I out for a great Swiss-fondue dinner after my day of interviews in Zurich) and basically I feel pretty good about the fact that I even got the chance to interview with Google. And as I mentioned above, I am now confident enough to believe that I can sit through any technical interview known to man (or woman)