How to handle technical challenges in work?

jim lahey

I'd be lost without online reference material. I simply don't have the patience to methodically work everything out. It's simplest just to to put your technical issue/requirement straight into google in pidgin English. For example: "Username password authentication WCF" "NHibernate named query field alias Oracle" "Jquery form validation plugin" Aside from the manufacturers websites (MSDN, 3rd party library sites, codeplex etc.) More often than not I find what I need as an article on this site or as an answered, peer-reviewed question on StackOverflow. An important factor is accepting that what you're doing isn't rocket science and isn't so groundbreaking that nobody has done it before. The number of times I find my question has been more or less answered far exceeds the number of times I can't find an answer at all. That's where I can derive another benefit from using existing patterns - because it's an existing pattern, somebody has used it before which greatly increases my chances of finding the help I require and I don't needlessly reinvent the wheel, something I used to excel in.

Eytukan

Go to StackOverflow, Click on Ask a question. And type your question, but don't submit. You'll see your question has been asked a 1000 times, and it's been answered a 4000 times. And the next question you are going to ask in succession will also be there in the same thread. wait a minute, oops StackOverFlow :rolleyes: Just kidding, CP first always, but lately I referred a bunch of Q&As from StackOverflow. It was quite useful. And even Google lists them on Top. Don't know why CP forum messages are pushed down. And about typing the search key, Never do the girls way:

I Want to know how can I add multicast delegate in my C# sharp project

Go by the man's way:

Multicast delegate + C#

You get more returns.

Starting to think people post kid pics in their profiles because that was the last time they were cute - Jeremy.

SoMad

There are already a lot of good answers to your question, but you should also check out this article written by Christian Graus: How to Use Google and Other Tips for Finding Programming Help[^] Soren Madsen

Anna Jayne Metcalfe

In short: intuition, experimentation, research (Google/books/asking around - whichever is likely to be quicker for the subject), followed by proof of concept prototyping. Add in a dose of dogged bloody mindedness and you're done. A couple of recent cases in point:

I had to update a dev build of our Visual Lint[^] plug-in to support the Atmel Studio 6[^] IDE a couple of weeks ago, and although the Atmel IDE is based on Visual Studio, there are parts of the interface which are different, and of course it uses a custom project file format (which we have to parse to pull out preprocessor and include folder settings). None of that was documented, and it turned out to be far too esoteric to be being talked about on the web (in other words, my google-fu failed completely). I was able to get the plug-in bits going just by reference to our code, so that was nothing clever. The project file format however presented quite a challenge as the device settings and includes just aren't documented, so I had to reverse engineer them from the Atmel header files, and test my assumptions by building unit tests (I wrote those first using TDD) and analysing test projects to check that the results matched my assumptions.
Visual Studio 2012 adds theming to the IDE, and as our code is generic (we integrate into both Visual Studio and Eclipse) and written in C++ using WTL, we can't use anything but plain Win32 to re-theme controls. Win32 UxTheme (i.e. XP style) controls do not support custom colour schemes (ever seen an XP style button in any colour other than OS default?), so we had to implement custom versions of every control we use in the main IDE, and to do that figure out how to draw UxTheme control parts with custom colours. The UxTheme API is poorly documented at the best of times, but *nobody* has described how to do this - I had to work it out from first principles (it involves a memory DC and colour transformations) and prove the concept. To cut a long story short, the proof worked and then the google-fu kicked in - find examples of custom controls of each type we need (button, combobox, list control, header control etc.) to use as references while designing WTL templates t

NormDroid

Even better way RTFM, which very few people seem to do nowadays.

Software Kinetics - Dependable Software news

Eytukan

Yeah RTFMSDN in most cases here.

Starting to think people post kid pics in their profiles because that was the last time they were cute - Jeremy.

Anna Jayne Metcalfe

I have to admit I didn't know about Christian's article. A big 5!

Anna :rose: Tech Blog | Visual Lint "Why would anyone prefer to wield a weapon that takes both hands at once, when they could use a lighter (and obviously superior) weapon that allows you to wield multiple ones at a time, and thus supports multi-paradigm carnage?"

Lost User

All I can say is be very careful. Recently, while working with a large log file on a linux server, I googled 'man tail' seeking info about switches to the tail command. The results were not pretty.

biop codeproject

Yes. Will learn this. Sounds useful.

biop codeproject

Forum is good to ask questions, but once you post, you just do not when you will get an answer back. Google allow you to actively search for answer. At least you can tell your boss about that. :-D

BobJanova

Is this junior/senior as in time? Because, no disrespect intended, this doesn't sound like a very senior question! When you're given a requirement you should be able to mentally break it down into the rough steps you need to do in order to complete it. A lot of those will be simple stuff that you can Google if you're not familiar with the language/framework, e.g. file or database access, updating UI controls, simple aggregations or filtering of data. Possibly, buried right in the depths of the requirement, might be something where you actually have to think – but in my experience at least 95% of code is simple and has been done before. The skill is in translating a requirement into steps that you can look up if you don't already know how to do them (though, really, anyone in a senior position should know how to do most of the common ones already). When I'm searching I use the language or framework and the thing that I'm trying to work out how to do. For example I've been having some trouble caused by an OS upgrade lately so my search history is full of 'java character encoding' and 'ant character encoding' and the like. If you're partially familiar with the framework you might know where to look for inspiration; for example if there were a C# module involved I'd know to search for 'C# StreamReader' or 'C# TextReader' as a place to look to similar issues.

RaviBee

I focus my technical searches at StackOverflow and CodeProject - this works when I'm trying to answer questions like "How do I do XYZ using foo?" For example, figuring out how to implement sorting in a Silverlight DataGrid is easily accomplished by Googling SO/CP for "silverlight datagrid sort" which should lead you to a CollectionViewSource. Then, it's a simple matter of hacking together a small POC to determine how to use this class - MSDN comes in handy here, because you already know what class you should be using and are just searching for its properties and methods. Unfortunately, most of the time I'm trying to determine answers to more nebulous questions, like "What's the best/recommended way to design a large-scale JS app?" In this case, I'm not looking for a cookie cutter solution (there isn't any), but rather opinions from acknowledged authorities and people in the know. I find these in blogs, white papers and (sometimes), discussions on SO, and discuss my findings with others before arriving at a conclusion. Hope this helps. /ravi

My new year resolution: 2048 x 1536 Home | Articles | My .NET bits | Freeware ravib(at)ravib(dot)com

charlieg

google-fu ... I like that.

Charlie Gilley You're going to tell me what I want to know, or I'm going to beat you to death in your own house. "Where liberty dwells, there is my country." B. Franklin, 1783 “They who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” BF, 1759

biop codeproject

OK. I see your point. Let me go over the technical aspect. There is a sale system. There are about several hundreds of transactions in a day. The data is stored in a database. For each transaction, there is transaction header (usual stuff like customer id, transaction id, number of SKU involved, etc). There is also transaction detail (like item id, quantity of this item customer wants, etc). As you may see it, header to detail is a 1 to many relationship. One item we have to produce is an executable. This program will be run at end of each business day to output all transaction data to one XML file. This XML file is 3 level deep. It is similar to this:

It looks simple at first. Without much experience in processing this kind of XML, I think we could just let the program to write the XML out using something like XMLWriter(). This basically output a database table to XML format. Well in the database, the header and detail of each transaction are stored in two different tables! Somehow the person suppose to help me out use another way. He write a stored procedure to output 3 tables. In the program, he use DataAdapter to hold 3 tables (da.TableMappings.Add()). He then set mapping type of certain DataColumn (to be MappingType.Attribute or Mapping.Hidden). Afterward, use XMLDocument.LoadXML() to write out the XML file. Well there you have it.

BobJanova

There can be more than one answer to a problem. Like you I'd have written the XML file manually, from two queries (select from transaction, and select from transaction_detail where transaction_id in [list of IDs from the first query], grouped by transaction_id on the code side).