Archives
-
Regions == Evil
I had an email thread at work with a bunch of the guys on regions and this is the concensus we generally have come to (some of the statements are theirs, not mine, just paraphrased here). I was once a convert who liked regions. I enjoyed them. They made me happy. In all the code I would do something like this:
class UserCondition : IActionCondition
{
#region Fields
private int _condition;
#endregion
#region Constructors
public UserCondition(int _condition)
{
this._condition = _condition;
}
#endregion
#region Properties
public int Condition
{
get { return _condition; }
set { _condition = value; }
}
#endregion
#region Public Methods
public bool CanExecute(string action, WorkItem context, object caller, object target)
{
string userName = Thread.CurrentPrincipal.Identity.Name;
return userName.ToLower().Equals("domain\\joeuser");
}
#endregion
}
That felt good and organized and neat (almost in an OCD way).
However I have seen the errors of my ways, as others have before me. Regions are evil. Pure and simple. The absolute incarnate concentrated type of evil that only the Time Bandits would fear and not the watered down, garden variety kind of evil.
They're great for hiding the annoying details of an IConvertible implementation or designer generated code (when it's not already in a partial class). But I often create methods on the fly using ReSharper and it is not going to look for the correct region to place the method. So having everything separated into regions actually slows me down because I have to find where to put the method.
ReSharper is your friend. Ctrl+F12 is all you need to find stuff in a file. Using ReSharper's type member layout to enforce code layout in a file, you can get consistency across all teams so one code file isn't vastly differently organized (say that 3 times fast) than any other project. With the custom pattern on, formatting puts all the members in all the right places and keeps layout and code style somewhat consistent across teams. It makes diff comparisons and merging a more pleasant experience.
Going forward, we're purging them from all projects and forbade use of them in new code. YMMV. -
The CardSpace Value Prop
Last night I took in the Calgary .NET User Group presenation on CardSpace. It was great to see Michele Leroux Bustamante again as she's an awesome presenter. CardSpace is a relatively new technology but basically it makes identity easy for end users. You can find out more about CardSpace in general here.
The thing that I'm not sure about is the value prop for this. Currently CS is really only happening behind the firewall. There's very little penetration in the "real world" so we're not seeing CS logins on Visa sites, PayPal, or even Facebook or Yahoo Groups. That was the one thing I got when I was looked at CardSpace awhile back. I thought it was neat and perhaps solved a few problems (mainly around phishing and issues of users entering ids and passwords in clear text) but there was very little implementation out there.
Discussing it last night Michele brought up an example of how she's using it behind the firewall with a client. Essentially they're looking for a SSO layer that allows them to identify users across multiple disparate data sources, and remove the issue of managing identity on each instance of a data source. If you have SharePoint installed (2007 but 2003 will work to a certain extent) and combine their SSO services with the BDC (Business Data Catalog) you essentially get something like this. However there's the issue of tracking so if you're interested in who actually logged the request this might take some work, whereas CardSpace would help solve this problem.
However behind the firewall I have a problem with CardSpace in general. I already know the user. Sure, sure. In hetrogeneous environments where my users are Mac, Linux and PC I have problems. I also might have problems in environments where I have corporate employees that I can identify (say via ActiveDirectory) but non-empoyees (contractors or external customers) that I can't. Do I force my non-employees to be members of Active Directory? Do I create a cross-trust to other forests or domains to identify them? How do I handle federated identity in the enterprise. Maybe this is the place where CardSpace helps.
Outside the firewall I see there's benefit. There's benefit for managed cards for sure so when Visa, Mastercard, and PayPal come on board (and I'm sure they will) it will make signing into sites certainly easier, and perhaps a little more secure (I'm still debating if there's more security from CS given SSL enabled sites when you're doing banking but there are other advantages). Certainly for managed cards issued by banks and other places, I'm all over that like white on rice. Everyone does it but probably doesn't admit they use fairly weak passwords and probably share the same passwords across multiple institutions. With something like CardSpace in place, it becomes a non-issue for managing paswords (the card is my password, verify me) and really all I have to do is manage my cards, much like how I manage my credit cards in my wallet now. For the geek type we know that an SSL enabled site, a valid URL, etc. all gives us a warm and fuzzy that we can enter our credit card info on ThinkGeek and not expect charges to appear at Phil's House of Bondage. For the non-geeks out there, having them select a card from a friendly UI knowing that it's pretty safe makes me feel better (and cuts down on calls from the Father-in-law about this PayPal site he's never been to).
Of course there's still the roaming issue that needs to be address but that's a different problem. The poor mans solution right now is exporting cards and importing them around (or carrying them around on say a fingerprint enabled USB drive), however it's not a happy-happy-joy-joy scenario for someone like Jason Bourne who just wants to pop into an Internet cafe and log on (okay, bad example as Jason really doesn't want to be identified, but you get the idea).
All in all CardSpace looks fun and secure and will help solve some problems of both external sites and internet identity as well help deal with issues of FedSpace and complex corporate user identification. It's not the silver bullet (has there ever been one for anything?) but it's certainly an enabler. I'm planning on doing some cool stuff with it in the SharePoint space so stay tuned on this towards the end of the year. There's also some neat stuff that I think I'm going to do on a personal level like enabling some of my own sites with it. Even though it's not widespread, it is out there and easy enough for you to just create a personal card to save you the hassle of tracking user ids and passwords all over.
Things I learned last night at the session (maybe not completely related to CardSpace):
-
Michele is Canadian! That just rocks.
-
The iPhone really kicks the llamas butt (thanks JP for the look-see) but not sure if I'm going to ditch my CrackBerry for one just yet.
-
Michele used to work at Canadian Pacific Railway about 4 years before I started in 96 (this was Michele's pre-developer days)
-
CardSpace is simple to implement (web based or services) but does take some code to get tokens and decrypt info. This is all code Michele provides in her demos but will eventually make it's way into the core platform.
-
Garrett Serack wrote the identification code for CardSpace and worked at CP with me for a short time (he's now at Microsoft in the Open Source space)
-
It's a small freakin' world
-
I finally learned how to properly pronounce Michele's full name (and in French too!)
In any case, an interesting technology to track and some cool stuff for developers to try out. Check out CardSpace for yourself and be sure to check out Michele's demos and code as it's one of the few resources out there today for playing around.
-
-
42
The answer to life, the universe, and everything.
Also the number you'll get if you cut me open today and count the rings.
-
ALT.NET Quote #1
Overhead on the altnetconf mailing list:
"Design without patterns is the slowest route to victory. Patterns without design is the noise before defeat."
-
Going Microsoft
We talk a lot about the ALT.NET community and how it's not anti-Microsoft but rather alternatives to it (and better software development in general). Rather than blabbering on about DataSets over Web Services we talk about objects and domains and cool things that go bump in the night. I wonder as I look at my RSS feeds in the last few months how much of the "alternative" crowd are making more common bedfellows in Redmond?
Look at the history here. John Lam, Scott Hanselman, Phil Haack, Don Box, and now Rob Conery (and there are probably dozens of others I missed, sorry). MVPs or community leaders from another world, all bringing their super-brains to Microsoft to fold into the collective. Hey, don't get me wrong. I think it's a good thing. A great thing. It just seems all the cool kids are being gobbled up by the Blue Monster and I have to ask. What are things shaping up to? At this rate of adoption what will Microsoft look like in 5 years?
Like I said, this isn't a bad thing by far. I just wonder if Roy, Oren, Scott B., Jeremy, David, and JP (to name a few) are aimed for aquisition? If there's a group working next to the software aquisition guys looking at people and playing "what-if" scenarios with humans. I'm sure Scott would turn over in his grave before that happened but hey, Microsoft isn't a bad place to work. It's actually a positive experience, a great environment, and recently it seems to have been grabbing some of the top talent there is on the street. So where does it go from here?
How "alternative" is it if ALT.NET becomes the norm at the place that spawned the original term? Should be an interesting year. Or two. Or three. Or ten.
-
Taming the ActionCatalog in SCSF
The Smart Client Software Factory provides additional capbility above and beyond what CAB (the Composite Application UI Block) has. A little known gem is the ActionCatalog which can ease your pain in security trimming your application.
For example suppose you have a system where you want to hide a menu item from people that don't have access to it. This is pretty typical and generally ends up in having to scatter your code with if(User.IsInRole("Administrator")) statements, which can get pretty ugly real quick. The ActionCatalog system in SCSF helps you avoid this.
Here's a typical example. I've created a new Business Module and in the ModuleController I'm extending the menu by adding items to it:
public class ModuleController : WorkItemController
{
public override void Run()
{
ExtendMenu();
}
private void ExtendMenu()
{
ToolStripMenuItem conditionalMenu = new ToolStripMenuItem("Conditional Code");
if (canManageUsers())
{
conditionalMenu.DropDownItems.Add(new ToolStripMenuItem("Manage Users"));
}
if (canManageAdministrators())
{
conditionalMenu.DropDownItems.Add(new ToolStripMenuItem("Manage Administrators"));
}
WorkItem.UIExtensionSites[UIExtensionSiteNames.MainMenu].Add(conditionalMenu);
}
private bool canManageAdministrators()
{
string userName = Thread.CurrentPrincipal.Identity.Name;
return
Thread.CurrentPrincipal.Identity.IsAuthenticated &&
userName.ToLower().Equals("domain\\admin");
}
private bool canManageUsers()
{
string userName = Thread.CurrentPrincipal.Identity.Name;
return
Thread.CurrentPrincipal.Identity.IsAuthenticated &&
userName.ToLower().Equals("domain\\joeuser");
}
}
For each menu item I want to add I make a call to a method to check if the user has access or not. In the example above I'm checking two conditions. First the user has to be authenticated to the domain, then for each specific menu item I'm checking to see another condition (in this case comparing the user name, however I could do something like check to see if they're in a domain group or not).
Despite the fact that I could do a *little* bit of refactoring here, it's still ugly. I could for example extract the duplicate code on checking to see if the user is authenticated then do my specific compares. Another thing I could do is call out to a security service (say something that wraps AzMan or maybe the ASP.NET Membership Provider) to get back a conditional true/false on the users access. However with this approach I'm still stuck with these conditional statements and no matter what I do, my code smells.
Enter the ActionCatalog. A set of a few classes inside of SCSF that makes security trimming easy and makes your code more maintainable. To use the ActionCatalog there are a few steps you have to do:
-
Create a class to hold your actions
-
Register the class with a WorkItem
-
Add conditions for allowing actions to be executed
-
Execute the actions
Setting up the Catalog
Let's start with the changes to the ModuleController. You'll add some new methods to setup your actions, conditions, and then execute the actions. In this case the actions are directly manipulating the UI by adding menu items to it, but actions can be anything (invoked or tied to CommandHandlers) so you decide where the most appropriate split is. Here's the modified ModuleController:public class ModuleController : WorkItemController
{
private ToolStripMenuItem _rootMenuItem;
public override void Run()
{
ExtendMenu();
RegisterActionCatalog();
RegisterActionConditions();
ExecuteActions();
}
private void ExtendMenu()
{
_rootMenuItem = new ToolStripMenuItem("Action Catalog");
WorkItem.UIExtensionSites[UIExtensionSiteNames.MainMenu].Add(_rootMenuItem);
}
private void ExecuteActions()
{
ActionCatalogService.Execute(ActionNames.ShowUserManagementMenu, WorkItem, this, _rootMenuItem);
ActionCatalogService.Execute(ActionNames.ShowAdministratorManagementMenu, WorkItem, this, _rootMenuItem);
}
private void RegisterActionConditions()
{
ActionCatalogService.RegisterGeneralCondition(new AuthenticatedUsersCondition());
ActionCatalogService.RegisterSpecificCondition(ActionNames.ShowUserManagementMenu, new UserCondition());
ActionCatalogService.RegisterSpecificCondition(ActionNames.ShowAdministratorManagementMenu, new AdministratorCondition());
}
private void RegisterActionCatalog()
{
WorkItem.Items.AddNew<ModuleActions>();
}
Here we've added an RegisterActionCatalog(), RegisterActionConditions(), and ExecuteActions() method (I could have put these into one method but I felt the act of registering actions, conditions and executing them voilated SRP so they're split out here).
Action Conditions
ActionNames is just a series of constants that I'll use to tag my action methods later using the Action attribute. The conditions are where the security checks are performed. Here's the general condition first which ensures any action is performed by an authenticated user:class AuthenticatedUsersCondition : IActionCondition
{
public bool CanExecute(string action, WorkItem context, object caller, object target)
{
return Thread.CurrentPrincipal.Identity.IsAuthenticated;
}
}
Next are specific conditions for each action. As you saw from the AuthenticatedUsersCondition you do get the action passed into to the CanExecute call so you could either pass this off to a security service or check for each action in a common method. I've just created separate classes to handle specific actions but again, how you organize things is up to you.
class UserCondition : IActionCondition
{
public bool CanExecute(string action, WorkItem context, object caller, object target)
{
string userName = Thread.CurrentPrincipal.Identity.Name;
return userName.ToLower().Equals("domain\\joeuser");
}
}
class AdministratorCondition : IActionCondition
{
public bool CanExecute(string action, WorkItem context, object caller, object target)
{
string userName = Thread.CurrentPrincipal.Identity.Name;
return userName.ToLower().Equals("domain\\admin");
}
}
Both conditions contain the same code as before but are separated now and easier to maintain. Finally we call the Execute method on the actions themselves. Execute will pass in a work item (in this case the root workitem but it could be a child work item if you wanted), the caller and a target. In this case I want to add menu items to the UI so I'm passing in a ToolStripMenuItem object. The ModuleActions class contains our actions with each one tagged with the Action attribute. This keeps our code separate for each action but still lets us access the WorkItem and whatever objects we decide to pass into the actions.
The Action Catalog Itself
public class ModuleActions
{
private WorkItem _workItem;
[ServiceDependency]
public WorkItem WorkItem
{
set { _workItem = value; }
get { return _workItem; }
}
[Action(ActionNames.ShowUserManagementMenu)]
public void ShowUserManagementMenu(object caller, object target)
{
ToolStripMenuItem conditionalMenu = (ToolStripMenuItem) target;
conditionalMenu.DropDownItems.Add(new ToolStripMenuItem("Manage Users"));
}
[Action(ActionNames.ShowAdministratorManagementMenu)]
public void ShowAdministratorManagementMenu(object caller, object target)
{
ToolStripMenuItem conditionalMenu = (ToolStripMenuItem)target;
conditionalMenu.DropDownItems.Add(new ToolStripMenuItem("Manage Administrators"));
}
}
Registering The Action Strategy
Calling ActionCatalogService.Execute isn't enough to invoke the action. In order for your Action to be registered (and called) the ActionStrategy has to be added to the builder chain. The ActionStrategy isn't added by default to an SCSF solution (even though you can resolve the IActionCatalogService since services and strategies are separate). Without the strategy in the builder chain, when it constructs the object it doesn't take into account the Action attribute.So you need to add this to a stock SCSF project to get the action registered:
protected override void AddBuilderStrategies(Builder builder)
{
base.AddBuilderStrategies(builder);
builder.Strategies.AddNew<ActionStrategy>(BuilderStage.Initialization);
}
Once you've done this your action is registered and called when you invoke the catalog.Execute() method.
A few things about actions:
-
You don't have to call catalog.CanExecute for your actions. Just call catalog.Execute(). The Execute method makes a call to CanExecute to check if the action is allowed
-
You have to register an implementation of IActionCondition with the catalog in order to do checks via CanExecute. If you don't register a condition, any action is allowed
Alternative Approach
There are lots of ways to use the ActionCatalogService, this is just one of them. For example in your ModuleController you can set everything up, disable all commands, then execute your ActionCatalog which will enable menu items based on roles and security.The ActionCatalog lets you keep your execution code separate from permissions management and who can access what. This is a simple example but with little effort you can have this call out to say a claims based WCF service, retrieve users and roles from something like an ASP.NET Membership Provider, and make applying feature level security (including UI trimming) to your Smart Client a breeze!
Hope that helps understand the ActionCatalog in SCSF! It's a pretty cool tool and can be leveraged quite easily in your apps.
-
-
Agile in your schools
After the fishbowl session from the Edmonton Code Camp, one of the big aspects that came out of the discussion was around the lack of Agile and good software development practices (aka ALT.NET) in our schools and universities.
Chris Chapman posted a comment on my entry about a series of blogs he did last month on this exact topic. Let me tell you this has got to be one of the most definitive and well-researched pieces on the subject I've ever seen. Chris not only highlights a series of key practices (Agile/lean development, TDD, refactoring, etc.) that each university offers (or doesn't as is the case) but has a scorecard and breakdown of the best-of courses to check out.
If you're about to enter University and feel like you're going to be left out because you're a forward thinker and want to really challenge yourself, check out Chris' findings which may help you see what's out there!
-
Lessons Learned from Edmonton Code Camp 2007
Got back from the Edmonton Code Camp last night around midnight, after having dinner with the group. It was funny because we all sat around the table (James, Justice, Donald, Rockarts, et. al.) and realized that the entire group was also going to be at DevTeach. Not sure if that's a good thing or a bad thing, but I'm sure we'll be sharing a few beers over it again in November. Anyway, a full day of fun and discussion with a few surprises. I thought I would wrap everything up into this entry although some of these could almost deserve their own post.
Education Needs to Get Agile
In a fishbowl session something really came out about learning from a gent who recently (April) graduated from the UofA. He stated that these strange new concepts (MVP/MVC, ReSharper, Domain Driven Design, IoC, etc.) were all alien and he had never heard of any of them during his Comp Sci studies. It was events like the Code Camp that brought them out (or blogs or whatever) and it excited him. There was passion there and I could see a budding ALT.NET developer just wanting to burst out all over the room.
It's true that Java, C#, and Ruby are new when it comes to the academic space. I mentioned that the world was running on Internet time and schools seem to be running on Glacial time, just waiting for something to happen and not moving very quickly when it does. This I think is a huge problem. If people are coming fresh out of University with lofty goals of building huge Enterprise applications using what they've learned from school, it's obvious this is going to be an issue. It says to me that we, the forward-thinking and always-moving-in-some-direction have a responsibility to educate the educators. We need to get out there and let people know that alternatives exist and the universe is not made up of DataSets and XML. This might go a long way to helping foster diversity and knowledge in the community at large, and help make the shift from academic to implementation be a little smoother. People should be coming out of education hitting the ground running, and not being stopped dead in their tracks, told to abandon what they learned and go pick up a few good books from the Martin Fowler signature series. There's also a responsibility of the educators to know that we are out here and enterprise development isn't just regurgitating patterns and practices that were abandoned years ago. There are alternatives and better ways of building the mousetraps, you just have to be open to understand, discuss, and validate them for the appropriate solution at hand.
XNA
I have my dual-head presentation on XNA programming including some remote debugging into the XBox 360 from my laptop. Unfortunately I was in a rush and forgot that I didn't have audio for the 360, so I bought a pair of speakers that morning from the local Staples (I'll use them at work to peeve off the QA people sitting next to me when I blast some Don Ho out on Monday morning). In my fit of excitement I forgot about getting an adapter to plug the speakers into the 360 so no sound during the demos, except when I ran the Windows versions. My laptop wasn't beefy enough to render the Racing Starter Kit (but the sound was there) so it chugged along at 12 frames per second.
I think there was good interest from the community on XNA programming (although a guy asked me if this session was about "games" and proceeded to walk out when I told him yes) so hopefully we'll see some fun stuff coming out of the Edmonton community. I'm really looking forward to XNA Game Studio 2.0 and the networking support but most of all getting it to run on my regular Visual Studio. You don't know how many times I was hitting Ctrl+F12 and Alt+Insert in Express trying to get ReSharper to work.
Domain Driven Design
I think I had fun with my Domain Driven Design session. I struggled with this topic as I couldn't figure out how to squeeze a fairly large topic like DDD into 50 minutes of discussion. I left the session open to talking about issues with development in general, what pain points people were facing, and how DDD might serve to ease the pain (or not in some cases). I covered the basics of DDD which was really only scratching the surface and we spent a little time looking at a fairly rich domain (Ben Scheirman's NHibernate video store series) and dived into writing a builder object and a test using a fluent language to describe the domain better through code.
For future sessions on DDD I'm probably going to spend more time on building fluent interfaces, maybe a DSL using Ruby or something. Writing academic examples of DDD is a little too brief and doesn't really help grok the principles or values of what DDD personifies. I did talk to a few people after the session and a few lightbulbs turned on as a result of it, so it wasn't a complete loss and people did enjoy the session (based on the feedback forms, although maybe they were being nice because it was so late in the day ;)
Community and Adoption
This was probably the coolest sessions of the day and focused on talking about good software design, the community, and how to get the word out (whatever that word is). It's funny having a session at a code camp when there's no code, but this worked (although it wasn't as filled as James' ALT.NET session). One of the reoccurring themes during the session was how does the community get to know about things like ReSharper, DDD, and patterns (to name a few). And more importantly, how does one adopt that in your own organization or community practice when you're the lone wolf?
This is a challenge. Going back to your office and telling everyone "I just saw this great tool/technique at Code Camp and we should all change to using it" isn't going to fly. My message is clear. Practice, practice, practice. And do. Or do not. There is no try (sorry Yoda). Take what you see and practice it against your work. If there's a pain point in how you build or deploy your application then download NAnt and automate the task. There's no need to ask for permission from upper management. How would that conversation go?
You: I would like to use [insert tool here] as it would reduce our deployment time by half.
Pointy-haired boss: Sorry [insert tool here] isn't on our list of approved tools and your job is to code. Get back to work!Maybe not the best advice (and some upper-management guys may come back to me on this) but just do it. Do something that works and if it works and is better than what you're doing now, isn't it worth it?
All in all, a great day and one that spurred new ideas and things to try that might make your geek life a little better.
-
Edmonton Code Camp - here I am!
Jacked up on Rockstar and hanging out in the Holiday Inn in Edmonton as I hit the Edmonton Code Camp tommorow. Of course in my fit of leaving I forgot the video and still camera, so no pics from me but I'm sure they'll be Flickr coverage of the event from others (sorry Kyle).
Here's to tommorow and I'll see if I can sneak into a few other sessions and blog about them for you while I'm there (the Donald/Justice session on the future of .NET and James' session on ALT.NET should prove interesting).
Ahh.. wireless networking and HBO. What more can a geek ask for?
-
Working Remotely
Kyle just finished up a series of posts on remote pairing and remote working. He's working on our Smart Client project from the Bahamas while we're stuck here in Calgary as the winter months start to show their ugly face. It's a great series and wraps up with some tips (although I have to wonder what he does now as the last few sessions he's been awfully quiet). Now I know the truth.
Top 10 reasons to work remotely:
- You can frag people in Halo 3 during a design session.
- You can frag even more people in Halo 3 during the morning Scrum.
- Did I mention Halo 3?
- Fuzzy slippers! (although I have these and wear them at work anyway, but for those basement geeks here's your big chance to get comfy)
- Halo 3 anyone?
- Matchmaking on Halo 3 while your remote pair compiles
- Waiting for the Matchmaker to pair you up when someone breaks the build and you can't check in anyways
- Uhmm... Halo 3?
- Halo 3 r0ck$!
- And finally you can play Halo 3 without disturbing your co-workers
So as you can see Kyle gets a tremendous amount of work done remote working with us, and you can too! All you need is a pair of fuzzy bunny slippers, a house in the Bahamas and a copy of Halo 3 (oh yeah, and an XBox 360 helps).
-
Kyle, Justice and Me
The race is on. Ever go out and ego-surf? It's the art of finding yourself on the Internet. Either blog posts or websites you author, people talking about you, or the amazing coinky dink that you are the same person as someone else. Just a mind-numbing experiement for those that have no life.
The score so far:
I need to cut down on coffee on Tuesday mornings.
-
Camping out in Edmonton - Code Camp 2007
I'm heading up to Edmonton this weekend for the Edmonton flavor of Code Camp. Seems only fitting as the freaks came down to our Calgary Code Camp, so I'll reciprocate. I have two sessions I'm talking about and will see if I can bring my video camera (where the heck is that charger?) and post some sessions (or at least Justice grooming himself) on my Silverlight space. Here are the sessions.
IEntity<Hello World>(new GettingStartedWith(DomainDrivenDesign))
We'll spend an hour wallowing through aggregate roots, entities, value objects, repositories, and the principles and concepts behind Domain Driven Design. Demos will be on the fly based on what you guys want to see but I'll prep a few things around specifications, identifying aggregate roots, TDDing the domain, and other goodness.So you want to play a game? How about thermo-nuclear war, XNA style!
I won't be teamed up with John The Pimp Bristowe on this, but we'll go double-head on the display as we step through XNA, write a few games and duke it out on the big screen deploying and debugging XBox 360 games from my Windows laptop. It's like rockem-sockem robots, but with less carnage.Watch for a wrapup, pics, and potentially damaging video on the weekend after the dust settles. See you there!
-
Trifecta announcements from the CKS Group
The Commity Kit for SharePoint Group (which I'm a member, but haven't participated recently, bad Bil bad) has three awesome announcements today (along with the software releases that go with them).
FBA has been available on SharePoint 2007 since the release, however it's been a little hard to implement and worst, there are no facilities for managing users or recovering passwords. No longer. With the FBA modules, you can now basically give SharePoint a DotNetNuke facelift and provide that functionality. This includes user login, membership requests and management, password management, and user and role management.
Virtual Earth Maps and the Campus Map App
When I'm at Microsoft, I would sometimes see this uber-cool app which showed the Redmond campus along with facilities, all rendering via Virtual Earth and served up in SharePoint. This app is finally released along with the source code, graciously donated by MSIT Information Services team. Very cool for those that have SharePoint setup and would like to have interactive and live maps added to their sites.
It just had to happen. FBA is nice (now with the new FBA managers) and integrated is cool, but we really want Windows Live ID integration. Here it is in all it's glory.
If there's any doubt you can't do cool things with SharePoint in the public space, take a look at CKS for ideas, inspiration, solutions, samples, and code. And get implementing!
-
Another SharePoint Sucks Post Released into the Wild
I'm in a quippy mood tonight and stumbled across a post titled "Why SharePoint Server is Terrible". Of course you know flame-bait like this just gets my blood boiling, especially since there are points here that are completely inaccurate and don't represent the peoples the author is talking about (like MVPs). A few months ago there were blog posts about how bad a development environment SharePoint is and for the most part I agreed with what was being said. It's not great, but it's better than hand coding the world yourself. Now we get this post that goes deep into why (as the author states it, not me) SharePoint is terrible and a failure.
I'm going to nitpick points here because overall the article is just plain negative on SharePoint. Like I said, in some cases SP blows monkey chunks. Yup, you heard it. A SharePoint MVP saying SharePoint bites. However take that with a grain of salt as it can create major suckage in some areas, it excels in others, so you have to balance what you read (even on this blog) with the context of reality and what business problem you're trying to solve. No technology in the world is going to solve problems, it's how you apply the tool or technology to aid you in a solution.
The author states that SharePoint is a great idea, but failed for three key reasons:
- Far too complex to install, configure, and customize. It is not agile.
- Any particular reason why people associate "agile" with "complex"? I build enterprise systems that cater to thousands of users, written using TDD, built on top of SharePoint infrastructure, and follow agile principles but they're all complex systems IMHO.
- It is being sold as a solution to organize unorganized companies. It is being sold as a system to add process's to organizations that don't have them.
- I'm making an assumption the author is one of these "unorganized companies". I for one (and most of my cohorts in crime) would never recommend a product if it didn't suit the needs of an organization. Installing SharePoint for the sake of installing it (because some CEO read a glossy brochure) is nothing about what SharePoint is or does. I for one don't make a dime on the
millionsbillions that SharePoint has made for Microsoft but would never stand behind pushing the product into a place where it didn't belong. I've also never seen it sold by Microsoft to corporations that didn't need or even want it. Grant you, user adoption can be slow as it's IT that sometimes makes the decision to bring it in but I've never seen an IT organization bring in the product when they weren't ready for it, or didn't have an idea of how it fit into an already existing process. Processes don't fit tools, appropriate tools are adapted to support them. - It's rendered HTML is brutal, along with the CSS files.
- +1 for the rendering of SharePoint pages which can be painful however it has got 10x better with 2007, Master Pages, and better support for ASP.NET. It's still a little brutal and can be cumbersome due to the large number of styles and classes in the CSS. Brutal rendering? Have you seen the HTML output from other portals (Oracle, PeachTree, etc.). I have and SharePoint is a walk in the park compared to them.
- SharePoint OOTB doesn't conform to W3C standards so if that's an issue you'll have to wait or do a lot of customization. Yeah, this sucks and is one of the biggest problems with SharePoint rendering but I don't consider it as a result of HTML vs SomethingElse. Just bad rendering by Microsoft.
- It is one of the most unflexable applications I have ever used.
- Unflexable? Is that even a word. Unflexible? Maybe inflexible? Inflexible is the inability to change. Considering that SharePoint is an extensible platform that I can build solutions on top of to deliver legal document systems, help desks, training portals, or drive the XBox public internet site I don't consider that even remotely close to "inability to change".
Okay, so that was 4 reasons not 3, but who's counting right?
The author goes on to state you have to be a master of many tools. I will admit that SharePoint encompasses a lot of tools and technologies, but I am far from a master in many and some I don't even use. Ever. So do you really have to be a master in everything below to harness SharePoint?
- SharePoint
- Redundant, see redundant. So to be a master in SharePoint I have to be a master in SharePoint? This makes total sense.
- SQL Server
- If you've ever read my blog, tried to access the database directly, or wrote me and asked me what the name of the sproc is to delete a revision of a document from SharePoint you'll know that my motto is "stay out of the database!". So other than setting up (i.e. installing) SQL Server I fail to see what anyone has to master. In any case, there are guys far smarter than I who can make SQL dance and sing. Let them deal with it.
- Internet Information Server
- Again, what must I master here? Install it and turn it on.
- Active Directory
- Why must you be a "master" in AD to make it easy to configure SharePoint? In every setup I've done (both virtual, development, production, and otherwise) you install and generally AD is known (as long as the server is part of the domain, well, duh) and it just works. True, I've had problems with say domain groups with "&" and other characters in the name, but being a "master" in AD might help in knowing SharePoint but then it might help in knowing anything connecting to AD (like say your desktop).
- File Stores
- With each point here, my brain just gets fuzzier and fuzzier. File stores? Master? SharePoint can index a file share. Point it at it and make sure the crawler account has the right priveledges and you're done.
- Indexing
- Indexing in SharePoint can be more of an art than a science, but for all but the largest of installations I've seen or done (meaning 100 million documents spread over 3 data centres), indexing was OOTB and pretty much a brain-dead exercise.
- Software Development
- You must be a master in Software Development in order to use SharePoint. If I said this at a conference, half the room would walk out (the half that didn't already walk out when my demos blew up). I mean seriously, "software development" is such a broad term. Yes, if you want to build web parts and solutions using SharePoint you'll need to know what you're doing, but that's true for most anything. Try doing brain surgery without training. It's a messy subject.
- Search Engines (To help customize the brutal search built into it)
- I think the search engine needs work and customizing it can be difficult, however again it's perception and what is crap to some is fine for others. I once had a client who said "make it more like google" because they didn't like the search. We changed the style sheet to look like a Google search page and apparently that fixed the "problem" they had with it. Go figure.
- Database Design and Development
- I have no idea what database design and development has to do with SharePoint? You don't "design" nor "develop" databases with SharePoint and if you are, then you're doing something wrong.
- XML
- Calling any web service requires some understanding of XML. I don't consider the XML coming out of SharePoint web services or the XML config files to be any more complex than anything else that uses XML.
- .NET 2.0
- Again, if you're doing development you'll need to know .NET to build web parts but for the majority of clients I work with, they just want a tool that works and don't want (or need) to invest the time/money/effort to build custom solutions with code. A lot can be done with templates, packages, features, and what you can leverage OOTB.
- ISA Server
- For front-end web servers knowledge about this is a must, but for the majority of clients SharePoint installs are internal and don't need this. Again, if you need an externally facing system you'll probably have someone that knows ISA.
- Master Pages
- Being a master of master pages does not help you with SharePoint. It merely lets you build dazzling looking websites (if you know what you're doing and have a little design savvy) for use with SharePoint.
I think not.
I just don't get the whole "Document Management Process" excuse. He basically goes on to say Microsoft is so busy with pushing their SharePoint crack they're ignore people who need to understand document management, and people miss the mark. Document Management, or better yet, Information Architecture, is bloody hard. It's not something you can sit down with 2 guys and a small dog and bang together in an hour (unless you really do have 2 guys and a small dog in which case your IA is probably 5 emails a day and 3 PDF files). It takes a lot of work to figure out taxonomies, best practices, placement, structural and organization change, and adoption. All of which completely hinge on your customer and how mature their understanding of their own information is. No tool on Earth is going to help you do this right as it's all contextual to the organization, so blaming SharePoint for this failure is just asinine.
Wiki's Rock. Uhh, yeah. I guess so. Grant you Wikipedia is pretty good and at least you can search for stuff. However wikis suffer from two major problems. First, you have to know what you're looking for before you go looking for it. Second, structurally they suck when trying to organize information in a sensible way. Substituting another tool (MediaWiki) for SharePoint to solve an information problem is missing the root cause of the issues. While it may look better in the short term, years from now when you have thousands (or millions) of Wiki pages and you're trying to discover something useful, you'll kick your IT guy in the head who came up with this flavor of the day.
The author makes a comparison to MediaWiki and BaseCamp. Again, I don't get it. BaseCamp is like an online project management tool (with some document management bolted on the side). True, it's highly customizable from the look and feel but at the core it's a mess and isn't extensible to do anything but what it was designed to do. And MediaWiki, well, it's a wiki. You create content. That's about it. If we want to compare apples to apples, let's talk about SharePoint vs PeachTree vs Oracle vs LiveLink.
What the author mentions isn't anything new. We've known it for years and it's a sore point for us working with SharePoint. However as with any technology, it's got it's growing pains and it's far from being done. We'll continue to learn, adapt, and make things better. Any software the size of SharePoint *is* inherently complex. The KISS principle doesn't apply across the entire system, but it can be applied (and is) in isolated parts of it where appropriate. A comment on the original blog about Photoshop is interesting. I don't go into Photoshop (much) and when I do it's for simple things (resizing an image) and an alien world. I consider myself a fairly bright guy who can pick up most anything, but PS is a crazy beast with all kinds of secrets of the underworld just waiting to be revealed. There are people on this planet that can make it sing and dance and stand on it's head. I am not one of these people. I however can so similar things with SharePoint so it's all complex, from a certain point of view.
As Bill English stated, this is a product with a huge install base and revenues of +billions. Yeah, billions of dollars are going into BillG's pocket (or maybe it's Ray now, I can't keep track of those two crazy kids) as a direct result of this product. That speaks volumes. Millions of documents and thousands of sites spread across multiple data centres housing most everything Microsoft does on a day to day business. That speaks volumes. The drive and demand we put on Microsoft to make the product bigger, better, and faster than it ever was (well, at least easier to work with, faster would be nice too... and works on Vista maybe?) is what we do and what people want, and there's demand for this. If it wasn't we wouldn't be asked to speak and continue to communicate and educate people about SharePoint and we would all be living in a WebSphere world wouldn't we?
I fail to see how this author can write this sort of post without having first hand experience with SharePoint (meaning setting it up in an enterprise environment, not just looking at the pretty pictures on the web or running a VM for a few minutes).
-
DasBlonde, Calgary bound October 29th
One of my favorite speakers, whom I rarely get to see, is coming to Calgary. Courtesy of The Calgary .NET User Group this months speaker is Michele Leroux Bustamante, aka DasBlonde, will be presenting at the Nexen Conference Centre Theatre on October 29th at 5pm (my Birthday! Yay for me). Michele will be presenting about Windows CardSpace.
CardSpace is a client technology that is part of the .NET Framework 3.0 that allows users to create, manage and share their digital identities in a secure and reliable manner. CardSpace makes it possible to create personal identities that replace the common user name and password for an application - with the strength of certificates and without the complexity. CardSpace also supports installing identities issued by third parties for authentication. This session will provide an overview of the identity metasystem in which CardSpace plays a role, and describe how it helps to prevent identity theft and increase trust in online transactions. You'll learn how to create and manage information cards, how they are used to generate tokens with the help of a Security Token Service (STS), and the role of the STS when CardSpace is incorporated in the authentication story for ASP.NET Web applications and WCF services and clients. You'll also learn how to trigger CardSpace from ASP.NET or WCF applications and services.
Should be fun so come on out and welcome Michele to Calgary. You can register for the event (free) here.
-
Deep Thought of the Day
Some people are like slinkies,
They don't really have a purpose,
But they still bring a smile to your face
when you push them down the stairs.(found on Facebook, where else)
-
RYO AltNetConf
There's a tremendous amount of goodness (the "new" goodness?) that's circulating around the 'sphere. Martin Fowler chimed in with his take on it and I'm glad we're all generally singing from the same song sheet.
Jeremy MillerJeffrey Palermo brought up mention of how the original Code Camp spread like wildfire as the format and idea was easy to implement. As time goes on, I think this is true for the AltNetConf idea. Jeff summed the idea of the AltNetConf best with this quote:AltNetConf's are open spaces conferences where DotNetters get together to discuss how to build better .Net software.
Short and sweet. Just the right amount of description.
Given this the idea of new conferences springing up and spreading the new goodness is a great idea. What does it take to start up your own AltNetConf? The passion and desire to do so. So why not? There's nothing stopping you.
On the heels of the first one in Austin there are a few good ideas that you could use when you're building your own AltNetConf:
- Keep the size manageable. I think the 100 person limit was great for the Austin one. This also helps you locate a place for it.
- Self-organizing agenda. Rather than pre-canned agenda of topics, the first day/night of the conference is the time to collaborate and drill out what people are passionate about. What bugs people, what do they want to talk about. This is an agenda driven by both speaker and speakee (as I would consider everyone a speaker for each session, with someone keeping the conversation on topic rather than coffee-talk, much like a Scrum Master does during the daily standups)
- Nothing but .NET. This isn't Alt.JAVA so the conversations follow building on Microsoft platforms using the most appropriate tool, technology, and technique that makes sense for the problem at hand.
- Don't turn it into a vendor fest. While it may be Microsoft related, I think the last thing an AltNetConf needs is "Brought to you by [insert .NET vendor product here]". True, it should be free and things cost these days, but there are too many ideas that spiral out of control and become product showcases rather than guys and girls talking about software development.
- Follow the OpenSpace approach to organization and flow. Just resonates on the ideas above.
I'm at a disadvantage as I didn't directly attend the conference in Austin so I'm looking for those that were there to maybe bring out a AltNetConf retrospec. What worked well? What didn't work. What can we do better?
So spread the news, pick a location, and start doing it. For me, I'm looking to see if we can get an AltNetConfCalgary or AltNetConfEdmonton (or AltNetConfAlberta for that matter) going so ping me if you're interested. Let's keep the momentum going!
Hopefully lessons learned and ideas here would be applied to future conferences like this (which we all hope to see soon everywhere as we don't all need to coalesce to one single place once a year).
-
Blogging Time Machine
Okay, I'll be the first to admit that I ego-surf. Type your name into Google and see what comes up. With Google's blog search, I like to see who's referencing things I've mentioned and whatnot.
However today I saw this link on some Windows Live space site (not even sure what that is, some bastard child from MySpaces?) posted October 11. My name came up so I started skimming through the entry. I know it was about DotNetNuke and SharePoint so I assumed the writer was referencing my blog post from January 2006. After reading through the Windows Live space I realized it was a copy of my own blog entry I was reading (I *thought* it looked familiar as I was skimming). I got down to the end and it said "Published Tuesday, January 31, 2006 by Bil Simser".
Right. I get it now. It's some aggregator that copies content. I've seen them before. However what befuddles my meager brain is why in the name of all that is holy is my post from January 2006 showing up in October 2007? Slow mail delivery or something. The guys crawler *just* got around to finding my blog entry?
I just don't get it.
-
In the Queue...
Stuff that is swirling in my head and being worked on coming shortly to this blog:
-
Updates on SharePoint Forums and KB Web Projects for Office 2007 support
-
Continuous Integration Build Indicators (via X10 and some fancy integration with Cruise Control.NET)
-
What does Alt.NET mean to you and growing the Alt.NET community
-
Greenfielding Agile in the Enterprise
-
Making sense of the *DDs
-
Sessions for Edmonton Code Camp (Oct. 20)
-
Getting ready for XNA 2.0 and driving out games with TDD
-
Fixing a dead TFS server with Subversion
Feel free to toss your own ideas on the pile that you would like to see...
-
-
Join me for my first geekSpeak show in November
I'll be presenting a webcast in November (November 14th at noon PST to be exact) via the MSDN geekSpeak group. Here's the blurb about what geekSpeak is from their site:
geekSpeak is a new kind of webcast series, hosted by Jacob Cynamon and Glen Gordon (from the MSDN Events team). Dispensing with slide decks and scripted demos, geekSpeak webcasts bring you industry experts in a sort of "talk-radio" format. These experts share their knowledge and experience around a particular developer technology. You'll hear about industry trends, new technology, real-world experiences and more. During the webcasts you will be able to have your questions answered realtime, hear lively discussion and debate, and add your comments to the fray. Who knows, you might even see a whiteboard sketch or an off-the-cuff demo. It's another way for you - the developer - to engage with Microsoft in an interesting and effective way!
So tune in and hear me blather on for an hour or so about Tricks of the WPF Programming Gurus. We'll play around with WPF, look at what we can (and can't) do with it, build some cool apps, talk about Mort (kidding!), and generally nerd out.
I'm looking for a lot of the content to be driven by the listeners as that seems to be the geekSpeak way and hey, when in Rome...
You can register for the event here.
-
Calgary Agile Project Leadership Network (APLN) 2007/2008 Kick-Off
Calgary APLN is a local chapter of the Agile Project Leadership Network (APLN). The APLN is a non-profit organization that looks to enable and cultivate great project leaders. I've worked with Janice before and Mike is a well known person in the Agile community and an awesome presenter so check this event out.
Description: The Calgary chapter of the Agile Project Leadership Network (APLN) invites you to the APLN 2007/2008 Season Kick-Off Meeting.
Guest Speakers: Janice Aston and Mike Griffiths
Date: Wednesday, October 17th, 2007
Time: 12:00pm - 1:00pm
Location: Fifth Avenue Place Conference Room, Suite 202, 420 - 2 St. S.W.
Come and experience Agile planning in action at the Calgary APLN season kick-off meeting. If you are interested in how to run effective agile projects here is your opportunity to help choose the presentation topics for this season's talks and workshops.Agenda
-
Welcome and overview of Calgary APLN group
-
Report on new initiatives from the APLN
-
Brainstorming of topics for 2007/2008 season
-
Affinity grouping and ranking of topics
-
Top 5 list identified
About the Speakers
Janice Aston has over 16 years of project management experience with an emphasis on delivering business value. She is passionate about building high performing teams focused on continuous improvement. Janice has a proven track record delivering on project commitments with a heart for leadership and people. She has recently founded Agile Perspective Inc. specializing in creative collaboration.
Mike Griffiths is an independent project manager and trainer with over 20 years of IT experience. He is active in both the agile and traditional project management communities and serves on the board of the APLN & Agile Alliance, and teaches courses for the PMI. Mike founded the Calgary chapter of the APLN in 2006 and maintains the Agile Leadership site www.LeadingAnswers.com .
Please visit www.CalgaryAPLN.org for more details and to sign up for this event. -
-
Justice Gray IS Kyle Baley! (and Tyler Durden too)
For those of you who have never seen Fight Club, ignore this post. If you haven't seen it, go rent it then come back and read this blog entry. It's okay. I'll wait.
Back yet?
No?
It's okay, I'm still waiting.
Back?
Good.
There is a conspiracy on the Internet tonight and it's name is Justice Gray.
Perhaps you've seen the metrosexual hunk of a software consultant present at an Edmonton Code Camp. Or have you? Or were you really looking at this man:
You see dear friends, it is my experience and understanding that we've all been shim-shamed. Hoodwinked. In reality, Justice Gray is Kyle Baley!
First let's take the letters for Justice Gray and Kyle Baley and mix them up a bit. What do we get?
"A celibate sly jerky guy"
Yup.
You read it here. Justice Gray is not the metrosexual guy you think he is, he's really a eunuch. Definitely a man. But is he real or not?
Kyle I can vouch for. I've worked with him and we've stood in the same room (but not the bathroom). He's not just a disembodied voice on the phone from the Bahamas (although recently he has been) nor is he some crazed lunatic writing blog entries furiously in the dead of the night (although he does sometimes). No, he's quite real.
Justice on the other hand falls in "another" category. You see I don't recall ever meeting him myself. Ever. He's bailed on my Edmonton User Group presentations and was absent from our spectacular-spectacular dual-screen head-to-head XNA presentation at Code Camp. Again, where's the proof that he's real?
And more importantly, and I can categorically state this to be true, "I have never seen Justice and Kyle together in the same room at the same time!".
Just like Batman and Clark Kent.
Go figure.
But wait dear reader, the conspiracy does not end there!
I believe that Justice Gray is really the Tyler Durden of Kyle Baley. How so? Because any time Kyle wants to meet Justice, Justice just happens to not be able to show up. Edmonton User Group. Calgary Code Camp. Alt.net conference.
The proof? I have it right here...
Take the letters from Kyle Baley and Justice Gray and they form the name “Tyler Durden”.
Well, almost.
Okay. So what if we’re missing a couple of ‘D’s and a few other letters but there’s a conspiracy here I tell you!
ORIGINAL: JUSTICE GRAY AND KYLE BALEY
RESULT: TYLER DU??EN
Missing: RD
LEFT OVER: JSICEGAYAKYBA
When you take the remaining letters and put them together they form the phrase “Say Big Cake Jay”.
So in summary:
- Kyle Baley's real name is Jay
- Justice Gray's real name is Jay
- Kyle Baley is Justice Gray. They are the same person!
- At some point in Jay's tortured life he was asked by his mother to "Say Big Cake". I believe he could not pronounce this phrase and rather it came out as "Play Fig Make" which emotionally scarred him for life and has branded him to live behind this facade I'm revealing today.
- Justice Gray is just a figment of everyone's imagination. You have never seen him present nor does he really exist. The man you were looking at was Jay.
Don't believe me? Invite them both to dinner and see who shows up.
Well that just says it all doesn’t it? Jay, wherever you are man, I love you.
-
Alt.NET, stop talking just do it!
Hopefully the last of my Alt.NET soapbox posts for the day. There was a post by Colin Ramsay that while was quite negative about the whole Alt.NET thing (it was called Abandon Alt.NET) but it contained a single nugget that I thought was just right for the moment:
If they really wanted to change things then they should be writing about their techniques in detail, coming up with introductory guides to DDD, TDD, mocking, creating screencasts, or giving talks at mainstream conferences, or producing tools to make the level of entry to these technologies lower than it is.
I argue we've been doing this. Just visit the blogs at CodeBetter, Weblogs, and ThoughtWorks (these are just three aggregates that collect up a bunch of musings from Alt.NET people, there are others as well as one-offs). There's noise to the signal, so you have to sift through it but the good stuff is there if you look hard enough.
I totally agree with Justice (and others) in what he said on the mailing list:
Looking at this from a perspective of the conference participants being the developers and the general .NET community being the "client" in this case, how much value is the "client" going to derive from either:
a) what our mission statement is
b) what we choose to name this group?
in comparison to actual involvement with devs, recaps of sessions, evangelism efforts?So just do it. Enough with the name bashing, mission identity, who is and who isn't, and all that fluff. No fluff. Just code. Just go out and write. Blog. Present. Mentor. Learn. And if you're already doing that, you're ahead of the game.
-
!Alt.NET
I live in a Alt.* world. Have been all my life. I prefer the alternate movies over mainstream. I would rather sit and watch an art film from 1930s German cinema than the latest slap-fest from Ben Stiller. I prefer alternate music over mainstream. Give me Loreena McKennit or Mike Oldfield over Britney Spears and Justin Timberlake anyday. So it's only natural I'm sucked toward the Alt.NET way of software development.
Back before I was into this software thing I was an artist. I jumped from graphic design to commercial advertising. During my 7 year itch I spent a good part of it in comics, and more precisely the alternate comics. I never tried out for Marvel or DC (although a Marvel guy who shall remain nameless liked my stuff and invited me down to New York to talk to them) so the alternate scene for me was Dark Horse, Vertigo, and Image. These were the little guys. The guys who preferred glossy paper over stock comic newsprint. The guys who were true to reality and weren't afraid to show murder, death, kill in the pages. I did a comic once about drug dealers in Bogota, Columbia (and the band of 5 guys [think A-Team but cooler] who would bring them down). The writer put something simple on the page like "the drug industry in Columbia was everywhere". If I was at Marvel or DC, one might draw the factories and lots of trucks, people packaging up the drugs, and shipping them off to the Americas. However this was Alt.Comic land and we told it like it was. I thought showing drug addicts (including one guy shooting up in an alley on one panel) mixed in with the tourists was the way to go. It was deemed a little racy and I was asked to tone it down, but it wasn't censorship and in the end I got to express what I really intended to do. I felt like I had made a difference and wasn't going to let the mainstream way of doing things cloud my judgement.
Alright, back to software development. I still don't know if I'll call myself an Alt.NETter simply because I'm not sure it's clear what that means. Any Alt.* movement in the world has it's basis in reality. Alternate art, movies, and music were created as a way to exercise expression of freedom, not just to be different. What is it that we look at in the "mainstream" way of software development that bothers us (enough to create an Alt.NET way). This really doesn't have anything to do with Microsoft does it? However many people have tagged MS as being the "evil empire" and using Microsoft tools is the wrong way, Alt.NET is the right way. Even the name seems to resonate against .NET and way Microsoft does things.
To me, Alt.NET means doing things differently than some whitepaper or robotic manager tells you how to do it because that's how it's been done for years. Alt.NET is any deviation from the "norm" when that norm doesn't make sense anymore. Maybe it made sense to shuttle fully blown DataSets across the wire at the time, because the developer who wrote it didn't know any better. However in time, as any domain evolves, you understand more and more about the problem and come to a realization you only need these two pieces of information, not the whole bucket. And a simple DTO or ResultObject will do. So you change. You refactor to a better place. And you become an Alt.NETter.
It's not about doing things differently for difference sake. You see a flaw in something and want to correct it (hopefully for the better). Perhaps BizTalk was chosen as a tool when something much smaller and easier to manage would have worked (even a RYO approach). Dozens of transactions a day instead of thousands and no monitoring required. If there's pain and suffering in using a tool or technology, don't use it. When you go to the doctor and say "Doc, it hurts when I do this" and he replies with "Don't do that" that's what we're talking about here. If it pains you to go in and maintain something because of the way it was built, then there's a first order problem here in how something was built (but not necessarily the tools used to do the job). That's my indicator that something isn't right and there must be a better way.
As software artists we all make decisions. We have to. Sometimes we make the right ones, sometimes not so right. However it is our responsibility if we choose to write good software, to make the right decisions at the right time. Picking a tool because it's cool doesn't make it right. Tomorrow that tool might be the worst piece of crap on the planet because it wasn't built right in the first place. Software is an art and a science. There's principles we apply but we have to apply them with some knowledge and foresight to them. Even applying the principles from the Agile Manifesto require the right context. Individuals and interactions over process and tools. We stay true to these principles but that doesn't mean we abandon the others. I use Scrum everyday and pick the right tools for the right job where possible. It's a balance and not something easy to maintain. If all you do is stick your head down and code without looking around to see what's going on around you, you're missing the point. Like Scott Hanselman said, you're a 501 developer and don't really care about what you're doing. You might as well be replaced with a well written script. For the rest of us, we have a passion about this industry and want to better it. This means going out and telling everyone about new tools and techniques, demonstrating good ways to use them, explaining what new concepts like DDD and BDD mean, and most of all being pragmatic about it and accepting criticism where we can improve ourselves and the things we do.
I suppose you can call it alternative software development. I think it's software development with an intelligent and pragmatic approach. Choose the right tool for the right job at the right time and be open and adaptable to change.
-
Quotes from the ALT.NET conference
Unfortunately I couldn't make it out with my Agile folks to the ALT.NET conference but from the blogs, various emails and IM's and the photos it sure looked like a blast. 97 geeks (Wendy, Justice and myself couldn't make it but there were probably others) got together and partied only like geeks can do. While I wasn't there, here are some quotes that came out of the conference. Some to think deeply about, others to just... well, you decide. Remember to use this knowledge for good and not evil.
"Alt.net is in the eye of the beholder"
"Oh I spelled beer wrong" -Dave
"Savvy?" -Scott Hanselman
"Scott, it's Morts like you..." -Scott Guthrie to Scott Bellware
"Programmers Gone Wild"
"There's the butterflies: then there's the HORNETS" -B. Pettichord
"I think 'grokkable' is more soluble then solubility" -Roy Osherove
"MVC is that thing that wraps URLs"
You can view (and contribute!) the altnetconf Flickr pool here. There's also a Yahoo group setup here if you want to carry on with the discussions since Alt.net isn't only about being at a conference.
-
Scrum for Team System Tips
As I'm staring at my blank Team System setup waiting for the system to work again, I thought I would share a few Scrum for Team System tips with you. SFTS has done a pretty good job for us (much better than the stock templates Team System comes with) but it does have it's issues and problems (like the fact that PBIs are expressed in days and SBIs are expressed in hours, totally messes up with the Scrum concept and makes PMs try to calculate "hours per point"). I'm really digging Mingle though and will be blogging more on that as we're piloting it for a project right now and I'm considering setting up a public one for all of my projects (it's just way simpler than Rally, VersionOne or any other Agile story management tool on the planet, hands down).
So here's a few tips that I’ve picked up using Scrum for Team System that might be helpful to know (should you or any of your Agile force be put in this position):- When a sprint ends, all outstanding PBIs should be moved to the next sprint and the sprint marked as done.
- The amount of work done in the sprint (SBIs completed) gives you the capacity (velocity) for the team for the next sprint.
- Only allocate one sprints-worth of PBIs to a sprint when planning and try to estimate better based on previous sprint data.
- When the customer decides the product is good enough for shipping have a release sprint where the goal is to "mop up" bugs and polish the product ready for shipping. This would include new PBIs like:
- Fix outstanding bugs
- Create documentation
- Package/create MSIs
- Fix outstanding bugs
- Never attach new SBIs to previously closed PBIs. If the customer changes his mind about the way something is implemented, it is recorded as a new PBI because the requirement has changed.
Definitions:
PBI: Product Backlog Item, can be functional requirements, non-functional requirements, and issues. Comparable to a User Story, but might be higher level than that (like a Theme, depends on how you do Scrum)
SBI: Sprint Backlog Item, tasks to turn Product Backlog Items into working product functionality and support a Product Backlog requirement for the current Sprint.
- When a sprint ends, all outstanding PBIs should be moved to the next sprint and the sprint marked as done.
-
Previously on Fear and Loathing...
Bil was struggling with the question. What would you rather have? No source control or no tests.
Let's go back a bit to Friday morning when Bil came into the office to find himself alone (it was 5:30AM and what ungodly soul would be at work at that time?) and unable to connect to the Team Foundation Server. It seems the drive that contained the database files filled up. Oh it gets better. No only did the data drive fill up (and BTW, the data files AND log files were on the same drive, not the way I would have set things up) but (drumroll please...) no backups. Somewhere along the way the SQLAGENT was disabled and frankly, when it's disabled nothing really works, including backups. As the data drive filled up, the transaction logs filled up and eventually became corrupted, unreadable, and unrecoverable.
The short of it that I'm left with a Team Foundation Server and it's databases with no log files. Not the end of the world (at least I don't think so) and there are techniques for re-mounting a file without the associated .ldf file. Grant you, if there were any transactions in play they would be lost, but this was the morning and the drive filled up sometime throughout the night (probably during a build).
At this point we've tried a variety of things to restore the databases and strangely enough we got them all back online. All but one. TfsVersionControl is the table that (yup) holds all the source code for Team System projects. A single table that just refuses to restore. The single remount trick (which worked for all of the other databases) doesn't work for this one (of course) so we're turning to PSS to help us fix this. There are a couple of "hacks" where you rebuild the database and swap out the data, but again for some reason it won't work. The best and closest we got was getting TfsVersionControl back online, but checking a solution out (any solution) ends with an error about "downloadUrl" being null, and the checkout stops.
So tune in again tommorow as the geek suffers. We'll see how things go, but as a last ditch effort we do have the latest code on the build server (a separate box) which has the latest checkout so we technically *could* rebuild the projects, we would just lose source control history which in the grand scheme of things, isn't the worst thing in the world.
-
What would you rather have...
A system with no source control, or a system with no tests?
Personally I would rather have a system with no tests. At least with tests I can write them. With no source control, well, welcome to my Monday morning.
It's going to be a long week.