As a business owner in development projects, there are days where I just wonder if I’m doing the right thing and being a good businessperson.  I recently had a project which just was one bad thing after another happening, and while it wasn’t really anything I caused, the customer was very upset.  It led me to think about why things go wrong in web projects and what can I do to prevent them?

The simple fact is that Web Projects are very much like kids.  They each need to be raised, educated, and disciplined in pretty much the same way.  However, like kids, not all Web Projects are ever exactly the same since there are always different players involved, new technologies to explore, and plenty of kicking, screaming and crying when things don’t go exactly the way they are supposed to.

I get where my clients are coming from, and you always want to focus on the positive and not the negative for a new project, which is why in my contracts I put a “bugs” clause which basically says that the web is a huge place filled with plenty of facets and the slightest blemish in one of those facets can cause the entire project to be delayed, not function properly, etc.  However, I think it’s fair to say that even when I’ve been forced to sign a “sh*t happens” document when going to the doctors, it doesn’t make it any better if things go wrong.

So what can we do?  Do we play Scotty from Star Trek and just over-exaggerate the timelines so when things do go smoothly we’re heroes?  But that can limit the clients who need things done quickly and also force them into unneeded delays for their business as well.  Do we tell them about all the things that can go wrong and hope they sit back and say, “Wow thanks for letting me know all hell could break loose so I can prepare for it?”….not likely.  How about just giving up web development and just taking up selling pet rocks at a carnival…those people have it made!

But what happens when the google eyes fall off that night from the rock?  Or the little feet bend and snap and now looks like something out of a horror movie?  Do you get less upset because its a pet rock?  Or do you still get upset just a bit because money was spent and things didn’t go exactly as expected.

I guess the revelation here is that things go wrong in any business no matter what that business is.  I’m sure some engineers can weigh in here about how their project got completely screwed up, or even a restaurant owner who unfortunately had a shell from an egg make it into a food critic’s dish.

I think the key here is how you handle those issues as both a consumer and a vendor. If you are a vendor…do what you can to make things right by understanding the customer’s frustration.  If you are a customer…if you see someone trying to make an effort to right a wrong (whether its themselves or just in general) work with them to get to the other side.  I’ve had great relationships made by hanging in there with a vendor and seeing just how good they do in a crisis situation to make everything turn out right.

So….just like kids….there will be good days and bad days….but it will never change your love or understanding towards them.  I think we as a society get too wrapped up in the “me” and really need to step back sometimes and not allows ourselves to take out our frustrations for other aspects of our life on these situations when they present themselves.

Hopefully this will help you as a customer reading this to know we will always work with you as a client to make things right if something should go wrong, or if you are a business owner…know that you are not alone and things can and do go wrong at times.  Try your best to rectify the situation and hopefully your customer will see your efforts.

Doug

Why PayPal Needs a Quality Assurance Wakeup Call

I spent literally 4 whole days going back and forth from blog to blog, developer forum to developer forum looking for the answer to this seemingly simple scenario.  Create a recurring payment and charge a single payment for items in the same transaction.  According to PayPal, this is a piece of cake.  Not so young Skywalker.

I started by going to PayPal’s website and discovered their old Express Checkout site had changed quite a bit and some of the familiar code snippets they had available before to perform tasks such as this were gone.  I wasn’t that concerned and proceeded to their Express Checkout area which, again according to PayPal, was an easy and great way to perform the tasks I wanted to complete.

PayPal’s Code Wizard…Merlin’s Red-Headed Step Child

Ok so here’s the point of the blog post.  This code wizard of PayPal.  Now most wizards are pretty friendly.  They are supposed to walk you through the basic features of what you’re trying to accomplish and give you a base to start from so you can then maybe go into the advanced features and adjust your code accordingly.   Theirs is completely a mind-screw and you’ll just love what I found out.

Help for You Fellow Coders in ASP.net

If you found this article because you got the “Token is Invalid”, or “Timeout Error”, or “Amount is Undefined”, dude…I feel your frustration.  You also probably spotted a few articles where these other devs said, all you have to do is call the SetExpressCheckout, change your version to 42 or higher, etc. What they failed to mention was in the PayPal’s own generated code wizard, the version number is not a NVP (Name-Value-Pair) you put into your code…it’s in the .CS code itself way down at the bottom of the friggin’ class.

How I Solved the Issue

Go into the PayPalFunctions.cs the wizard generates for you and scroll down to the  buildCredentialsNVPString().  Change the following code from:

codec["VERSION"] = “2.6″;

to:

codec["VERSION"] = “54.0″;

You can see they were off by just a few versions.  Now while I understand that the wizard is not made for recurring payments, it would have been nice for them to mention that in maybe the final screen.  I would also imagine that this change of version will work for the Express Checkout regardless of whether you’re doing recurring payments or not, but maybe not.

Other Items I Had to Solve

Other things they don’t really point out to you are the things you should include in the NVP for the SetExpressCheckout:

NVPCodec encoder = new NVPCodec();
encoder["METHOD"] = “SetExpressCheckout”;
encoder["AMT"] = “45.00″; //amt;
encoder["RETURNURL"] = returnURL;
encoder["CANCELURL"] = cancelURL;
encoder["PAYMENTACTION"] = “Sale”;
encoder["L_DESC0"] = “12 Month Subscription to Your Store”; // THIS MUST MATCH THE BILLING AGREEMENTDESCRIPTION
encoder["L_BILLINGTYPE0"] = “RecurringPayments”;
encoder["L_BILLINGAGREEMENTDESCRIPTION0"] = “12 Month Subscription to Your Store”;  // THIS MUST MATCH YOUR L_DESCn FIELD…IN MY EXAMPLE THE L_DESC0
encoder["L_NAME1"] = “Item Number 1″;
encoder["L_NUMBER1"] = “XY1238″;
encoder["L_DESC1"] = “Some Description”;
encoder["L_AMT1"] = “15.00″;
encoder["L_QTY1"] = “3″;

You’ll notice there’s a few things they exclude in their version that is code generated and such.  Now in a separate function I called DoRecurring, I have the following NVP:

NVPCodec encoder = new NVPCodec();
encoder["METHOD"] = “CreateRecurringPaymentsProfile”;
encoder["TOKEN"] = token;
encoder["DESC"] = “12 Month Subscription to Your Store”; //YEP ADD THIS AGAIN
encoder["PROFILESTARTDATE"] = “2009-03-16T13:00:00Z”; //MAKE SURE THIS IS GMT FORMAT
encoder["BILLINGPERIOD"] = “Month”;  //Month, Year, etc.
encoder["BILLINGFREQUENCY"] = “1″;  // This is the “cycle” so each month the customer will be billed once
encoder["AMT"] = “25.00″;
encoder["L_BILLINGTYPE0"] = “RecurringPayments”; //Has to be this for recurring to work
encoder["L_BILLINGAGREEMENTDESCRIPTION0"] = “12 Month Subscription to Your Store”;  //YEP HAS TO MATCH

So there ya go.  Now if you’re still a bit confused at the way PayPal’s Code Wizard said things like insert this on your billing and confirmation pages, etc.  Don’t let that bother you.  Here’s a simple version to do what I was trying to accomplish.  Hopefully this helps make it a bit clearer for you:

• Take the image they put in a form and make it an Image Button.
• On click of this image, in your code behind, insert the SetExpressCheckout code to setup your transaction and send the user to paypal to either enter their credit card or log into their PP account.
• After they do this they will be returned to a return URL you specified in the Wizard.  Create something called checkout.aspx and in the .cs file in the page load save the token and payerid into session variables.
• Put another button on this page and say something like “Finalize Payment”.  (Above this you’d put something like a list of the items or subscriptions they’ll be purchasing….just so they’re really sure about purchasing.
• On the click of that button put in the call to the CreateRecurringProfile function by copying the call as the FinalizeCheckout does.

Hopefully that helps you out a bit.  If you have any questions please email me at info[-at-]justicesolutionsllc.com or simply comment below.  I’ll really try to respond and help as much as I can because this was really frustrating.  I hope a PayPal developer reads this and rather than criticizing the hell out of my code which may not be perfect, really takes this back to PayPal and does a real QA on their code wizard and their manuals.  Because there are critical pieces and better explanations needed.

I sat back and said a prayer for a moment since I believed that getting angry was not going to solve my problem.  And just like that I had a thought….approach this problem as if I were my 6 year old son using the form.  Within minutes I had the problem figured out, put in a quick javascript validator to make sure the values would be entered correctly and just like that…the problem was solved.

So I decided to blog about this since many times not only developers, but business owners will have their websites so complex that it will cause more problems than successes simply because they are not thinking like a user who’s prime source of income would be a nice chocolate chip cookie.

I’ve actually had my son test quite a few of my websites and even though I may have to read the form to him to fill out, he has discovered some issues with my sites and I’ve coded appropriately to deal with them.  So the next time you’re ready to launch a new form, or a new functionality of your site, think to yourself….if I were 6 how would I look at this.  Or better yet, if you can snag a youngster from your family or your close friends, use them to test it themselves and pay them with a nice big cookie for their good work.

Happy Coding!

Doug.

So you quickly run down to the greeting card aisle, quickly survey the demilitarized zone of Valentine’s Day cards, find one that looks nice and has words that seem appropriate and zip over to the checkout counter.  Ok, we’re nearly complete…but wait….we need to write something inside of the card and sign it….oh no!

So we look frantically through the glove compartment box, flip the sun visors, look in the change drawer, the middle console, under the seat, all in the hopes of finding that single writing implement that is most likely missing a cap, has a glob of ink dangling from the tip and has been severely chewed on the end so much that you could dislodge a penny from the ashtray by using it as a chisel.

This is usually the moment where for some strange reason you’ve luckily managed to hit all of the green lights for a change and are actually hoping for a red light for that mere 30 seconds you need to write those unending words of love and compassion that will sweep your honey right off of her feet. 

Ok so you get a red light finally and then warped pen in hand you stroke those loving words on the open side of the card right on the steering wheel.  However you then push too hard on the Love…Me….which of course beeps the horn at the car in front of you and awards you the friendly one finger wave from the driver ahead of you.

So you get back to the house in time to go out to dinner, you calmly stroll on into the house, flowers, chocolates, and your novelette of a Valentine’s Day card in hand, and proudly give the items of your affection to your sweetheart.

She gasps for a moment, just so happy she has married a man who still likes to make Valentine’s Day special…grabs the flowers, the chocolates and the card sets them down to give a very welcomed and very earned Valentine’s Day smooch.

Finally without another word, she sniffs the elegant scent of the roses, mouth waters over the chocolate treats she’ll be enjoying later tonight, and then opens the card…which magically says on the front….

“This Valentine’s Day I wanted to say Something Special…”

and on the inside…

“You’ll Do.”

Funny enough…I don’t think she ever read the other side of the card I wrote on.  Happy Valentine’s Day Everyone!

Ok, it’s Friday and it’s my birthday…so woo hoo for me.  However, I wanted to quickly make a blog about how many times I’ve heard people say, “Well at least I have a website” or “It’s not a great site, but at least I have a website.”  These are the same companies that can’t understand why they are getting low conversions and low sales.

The simple fact of the matter is, image is everything.  Whether your site is just a few pages, or a full fledged application portal, it must look and behave as though it cost a million bucks.  Otherwise, people will quickly dismiss your site feeling as though you probably don’t have the ability to either service their needs, or deliver the products or services you are selling.

Take for example these two BBQ style restaurants: 

www.thehornytoad.com

www.santanflat.com

as you can see they are similar in the style of restaurant and even most of the items on the menu are pretty similar.  However, which one are you more likely to take your family to?  I’m assuming you said San Tan Flat….and I’m assuming that most of you reading this blog have never been to either establishment before.  But faced with the tantalizing decision as to which BBQ restaurant you’ll be taking your family to on Friday night, chances are you’d first try San Tan Flat before you’d try the Horny Toad. 

This is also the case for many companies who even may outrank some of their competitor’s sites in the search engines.  Why?  Because the site says it all.  One is a site that clearly has taken the time to give itself a very professional appearance, and the other, well just wants to have  website that they can put on their take out menus and coasters.

Keeping your site fresh, up to date, and professional looking is just as important as the quality of products and services your business is trying to promote or sell on the web.  You’d also be amazed that many times a site “facelift” isn’t as expensive as you would think.  So when you’re looking to either create a new website, or have its look changed…please keep this example in mind.  It could mean the difference between a successful business and a failing one.

Have a great weekend everyone.

Doug.

The tool is called an Object Generator or an O/R Mapper.  Some also like to call them “CRUD” Tools.  CRUD is an acronym we geeky coders like to use that stands for:  Create, Read, Update, Delete.  In almost every basic web application out there today, we are asked to interface with some type of database or datasource.  And in almost every instance of that interfacing we (the coders) are doing some type of call to the database.  It could be reading one record, getting several records, updating a record, or even deleting one. 

Writing these calls can be very mundane.  Ones I usually save for my new coders to do <insert evil laugh here>.  However, these CRUD or Object Generator tools can help things along tremendously in development time, which then allows us or your web developers, time to focus on things that are usually overlooked or zipped through at the end of a project such as field validation, styles, SEO, etc.

For ASP.net we personally like to use O/R Mapper (www.ornetmapper.com).  The thing I personally like about this open source tool is the simple fact that the developer continues to really improve this tool with newer and more awesome releases than its prior.  As of this updated blog, the latest version is version 6, and it’s come a long way and really could be the easiest most comprehensive way to develop the business logic for your next .NET project.

A Screen Shot of O/R Mapper (the ASP.NET Object Mapper Tool)

And using the code is even easier than using the tool itself in .NET.  So here’s a quick example.  Let’s say you have a table called Users and in your datagrid you have your typical select button which triggers your onSelect event in your code behind.  So now you want to populate a few text boxes in your form for editing.  Here it goes….in C#.net

          Users u = new Users();
          UsersAssembler a = new UsersAssembler();
          UsersCriteria c = new UsersCriteria();
          c.Set_Username(CriteriaBase.AVAILABLE_STRING_OPERATORS.EQUALS, “doug”);
          u.ReadSingle(c);
         
Now that you have the information loaded into the “u” object…you just assign your textboxes accordingly….

          tbUsername.Text = u.Username;
          tbPassword.Text = u.Password;
          tbFirst.Text = u.FirstName;
          tbLast.Text = u.LastName;
      
          etc.

Did you blink and missed it?  Yes it’s that quick.  You can also retrieve lists of results and do things like auto-bind it to your gridview, etc.  Really really cool.

For PHP, we personally like to use POG.  (www.phpobjectgenerator.com) This nifty little online tool will allow you to not only create the code classes you’ll need to interface easily with your web application, but also build your tables for you.

      

A Screen Shot of POG (the PHP Object Generator Tool)

Once all of your code and tables are built, its a snap to interface with them.  Let’s say you have a table which has your user’s vital information in it and you want to retrieve a list of users that have signed up in the past week.

       <?PHP
                  
                   // Create a new instance of our POG class for Users
                   $u = new users();

                   // Call the GetList Function to bring us back a list of users meeting our criteria
                   $userList = $u->GetList(array(array(“signupdate”,”>=”,”2008-02-16 00:00:00″)));

                   // Now loop through the results and display them for the end user to view
                   foreach($userList as $user)
                   {
                         echo ‘<p>User Name: ‘ . $user->username . ‘ <br />Name: ‘ . $user->first . ‘ ‘ . $user->last . ‘<br />Signup: ‘ . $user->signupdate . ‘</p>’;
                   }
          ?>

Pretty simple huh?  And you can send as many criteria you want in additional arrays all in that same GetList() function.  So the speed at which you get results and have the ability to update them are fast and easy.

So take some time to check out POG or send this blog link to your favorite PHP coder.  They very well may thank you for it, however knowing most coders they’ll probably say they already knew about it, thanks for the link, and then when people aren’t looking will start to use it all the time.  Ha!!  Happy coding!

Doug.

Ok, this week on Thursday Tips Day, I have a rather unique situation I was working on for a client.  The issue:  To get Tomcat 6 working on a Windows 2003 server running IIS and have ColdFusion 8 running on the Tomcat install.

This is not as easy as it would seem.  According to the Adobe ColdFusion 8 site (www.adobe.com/coldfusion) Tomcat is not directly supported as is was in the past.  The only mention of Tomcat in any of the installation documentation is only in a JBOSS type instance.  So here’s the first question…why would someone want to do this?

The client’s decision is actually a very good one.  Those of us familiar with ColdFusion over the years would agree that JRun is not the best Java server around.  In fact it sometimes causes more harm than good.  And in some unique situations it really ends up crashing the server and causing all kinds of Coldfusion errors when it gets too bogged down.  Now in CF8′s Enterprise Edition, the server monitoring tools are very good in letting you know what is causing the overloads.  Sometimes it’s your own code doing it.  Sometimes it’s not.  However for those of us not willing to spend the $4,000+ to purchase a copy of Enterprise Edition, we’re back to the same page as always.

Anyway, by choosing a Java server such as Tomcat, which has plenty of public support and has been running Java apps for quite some time very reliably on both Linux and Windows platforms (and even a Macintosh here and there I’ve seen), the client will be able to run their JSP apps, plus leverage Tomcat’s excellence in running CF8′s Java based programming technology.  Therefore a very lean and mean system and as the client puts it, “a smaller footprint” on their server.

The Installation Sequence

Ok, so to start, go ahead and download the latest version of Tomcat (http://tomcat.apache.org/download-60.cgi) and install it on your Windows2003 Server.  You can do the direct services install or the longer more in depth binary install, but both work just fine.  After the installation is complete.  You should easily be able to browse to http://localhost:8080 to get the nice little Tomcat welcome page.

Next download a copy of CF8 from Adobe’s website (http://www.adobe.com/products/coldfusion/) and install that on your server next by using the “Deploy as J2EE, and as a .WAR” installation.  When the installation is complete you should have a .WAR file most likely if you used the defaults in your C:ColdFusion8 root directory.

Stop the Tomcat Server via your services or direct command line. 

Now copy the .WAR file into your {TomcatRoot}webapps directory. 

Start the Tomcat Server and within a few minutes, Tomcat should have expanded all of your directories into the webapps folder.  So at the end you should have a /cfusion directory sequence in your webapps directory.

Stop the Tomcat Server again. 

Ok, now for a fun part.  Open up notepad and create a .bat file called setenv.bat and place it in the {TomcatRoot}bin folder.  Place the following code in that .bat file making sure to change the bolded areas to your exact folder structure:

rem Set a few variables.
set JAVA_HOME=C:Program FilesJavajre1.5.0_14
set CF_HOME=C:Program FilesApache Software FoundationTomcat 6.0webappscfusion
set CF_WEB_INF=%CF_HOME%/WEB-INF

rem Concatenate binary file directories into a single variable.
set CF_SHARED_LIB=%CF_WEB_INF%/cfusion/lib
rem The following variable must be on a single line.
set CF_SHARED_LIBS=%CF_SHARED_LIB%;%CF_SHARED_LIB%/_nti40/bin;
%CF_WEB_INF%/cfusion/jintegra/bin;
%CF_WEB_INF%/WEB-INF/cfusion/jintegra/bin/international

rem Add libraries for binary files to the Windows system path.
set PATH=%PATH%;%CF_SHARED_LIBS%

rem Set JVM options to enable sandbox security (all on one line).
set CF_SECURITY_JVM_OPTIONS=-Djava.security.manager
-Djava.security.policy=”%CF_WEB_INF%/cfusion/lib/coldfusion.policy”
-Djava.security.auth.policy=”%CF_WEB_INF%/cfusion/lib/neo_jaas.policy”

rem Set JVM options for CORBA. Uncomment this line if vbjorb.jar is not in
rem your JRE’s lib/ext directory.
rem set CF_CORBA_JVM_OPTIONS=-Xbootclasspath/a:”%CF_WEB_INF%/lib/vbjorb.jar”

rem Consolidate JVM options.
rem * Uncomment this line if you’ve configured CORBA
rem set CF_JVM_OPTIONS=%CF_SECURITY_JVM_OPTIONS% %CF_CORBA_JVM_OPTIONS%
rem * Uncomment this line if you haven’t configured CORBA
set CF_JVM_OPTIONS=%CF_SECURITY_JVM_OPTIONS%

rem Populate JAVA_OPTS, which will be used by the catalina.bat file
rem when starting the JVM.
set JAVA_OPTS=%CF_JVM_OPTIONS%

Ok, go ahead and start Tomcat back up, and browse to localhost:8080/cfusion/cfide/administrator and with any luck, you should have the nice welcome screen of the ColdFusion 8 admin staring you in the face.  Go ahead and configure your instance with datasources, etc. and place any .cfm files you want to run in the /cfusion/cfide/ directory.

That’s it.  I’m now in the process of getting the .cfm files to render in other directories and folder structures in the rest of the Tomcat so you can browse to other directories with .cfm files and get them to render.  Until next week!

You have a contract which outlines what a site will need to do….allow a company to login, create a job for a customer, do the job, bill the job.  Sounds simple right?

However throw in the variables such as, they don’t have any standard billing methods (not ones that could be followed by an application at least), multiple scenarios of “exceptions”, and a staff that doesn’t seem to talk to one another.

Now from a legal perspective, if the contract is simply written outlining the issues above, we’re sunk.  The client can take as much time as they want to test, to work out their own billing practices, etc.  However if you add a paragraph which specifically outlines how the system will be tested and what to do if “exceptions” to those rules come up, you’ll be in much better shape.

So here’s my paragraph (feel free to use in yours, and send me an email letting me know you found it helpful):