We're building a Kentico 8.2 site using ASPX+portal model. Looking at the rendered HTML on my live site I can see a lot of unnecessary Javascript that Kentico has dumped into my page. What's more this is occurring at the top of my page at the top of the form element.
For example, it's rendering the ASP.NET __doPostBack JS function even though I'm not using any controls that require it. Other scripts are being added as WebResource.axd and ScriptResource.axd includes.
At a glance it would seem these scripts constitute the Microsoft AJAX framework used with UpdatePanel etc. My assumption is that they are there to add portal manager functionality when using the page in the Kentico UI. Presumably they are also used with certain built-in web parts.
However, I am only using custom web parts on my live site so all these scripts are doing nothing and are just slowing down my page and causing poor performance testing results.
I've tried hiding the <ajaxToolkit:ToolkitScriptManager /> and <cms:CMSPortalManager /> controls on my master page when rendering the live site, but this causes templates that have a <cms:CMSWebPartZone /> to break.
Does anyone know how to ensure this bloat is removed when not required? Or at the very least cause these scripts to be rendered at the end of the page so they don't interfere with performance too much?
Unfortunately, building sites within Kentico using ASPX and ASPX+Portal Pages will automatically generate additional markup such as __doPostBack, WebResource.axd and ScriptResource.axd.
I wouldn't recommend removing any of the default code in your Masterpage. This will cause things to break (as you've experienced).
However, having this markup in place shouldn't cause a massive issue in page performance. Understandably, this isn't ideal.
What I do to lessen the hit is the following:
Disable the ViewState wherever possible. For example, either at Page Template or Webpart/User control level.
Move the ViewState to the bottom of the page (in Kentico Settings), so the page is less "top heavy".
Ensure you are caching everything you can. For example, site furniture used by your webparts and templates (images/js etc) at IIS level and at Kentico level using their API.
Reading this article from the Kentico documentation provides some more information in greater depth: Optimizing website performance
If you really want "full control" over the HTML rendered, Kentico does allow you to create templates using MVC. But this won't give you the flexibility to modify Page Templates by moving around web parts within the CMS Administration. I presume you have chosen the Portal Page approach for this very reason.
I hope this helps.
In addition to #sbhomra's great answer I have a few questions, suggestions and comments.
How many seconds or milliseconds are you talking about with performance? If you think you'll gain a few milliseconds back, it's not worth your effort to try to rebuild all the functionality. If you're talking a second or two, there about 15 different things you can change within settings and your code to gain it all back. Think about how much code you're going to write, maintain and upgrade just to gain a second or less back?
The WebResource and SciptResource load resources that are compiled into libraries within the website. So if someone created an external library and that library was loading an image that was compiled into it, you'd get that WebResource.axd reference on your site. You'd have to physically remove those libraries from the Kentico instance.
Although I don't recommend it strictly because you lose so much functionality and have so much extra unnecessary code, MVC will give you the control you're looking for.
Related
There is a pattern in ASP.NET - whenever there is a piece of markup that generates code (like the .aspx/.ascx files in WebForms or .cshtml files in MVC3), these files are dynamically compiled at runtime. aspnet_compiler will produce another assembly for them, which references your code-behind assembly.
This approach seems awkward to me and I don't understand why it hasn't been discontinued already. A much better approach (in my opinion) is like in Winforms or resource files - you have your .whatever file, and then there is .whatever.desginer.cs file. This designer file is created at runtime as you type. When you compile, the compiler doesn't care about your .whatever file, it just takes the .whatever.designer.cs file and produces a single solid assembly.
This provides several benefits:
You can inherit your forms from each other, similar to windows forms;
You can always see the code that is being generated, possibly adjusting your markup to generate better code;
You can easily instantiate strongly typed instances of your forms;
The only benefit I can see from dynamic compilation is that
You can change the markup file anytime and don't need to recompile the app.
To be honest, I've often wanted for one of the first three benefits, but never for the last one. Recompiling the app to see your changes isn't that big of a deal. Especially since you have to do it anyway when working with code-behind, where the most of your time will be. And when you deliver your app to the client, you deliver it as a monolithic, precompiled block. When you update, you update everything, not just individual .aspx files - there's no point to.
So... why is it like this? What am I missing?
It sounds like you are referring to an ASP.Net Website. An ASP.Net Web Application is like an ASP.Net Website, but uses .designer.cs files and produces a single assembly.
See ASP.NET Web Site or ASP.NET Web Application?.
One thought that comes to mind is that the primary difference between winforms and webforms has to do with the common development model.
Namely, there is a whole class of developers and designers who work strictly in html/css/javascript. They use a host of tools ranging from notepad on up to competing development products to build these pages.
Further, in the case of WinForms, MS has complete and total control of what can make up a form. Any controls that can be dropped on it have to descend from their specified code.
However, with HTML they don't have this same level of control. Spec changes occur out of sync with VS releases, additional features are added that are browser specific, etc. In short, they can't guarantee that a particular element in the HTML file is a valid element at all. All they can hope for is that whatever is sent was done so on purpose and that the browser knows how to deal with it.
Now they have tried to implement a model that provides visual inheritance. It's called "master pages". However, I believe the only tools that properly work with master pages are VS and Expression. Getting the other vendors to go down this path would be nearly impossible. Also, they've added the concept of "nested master pages" such that you can get multiple levels of inheritance out of them.
The code behind model helps to implement non-visual inheritance allowing people to completely revamp page processing (hence how MVCx works).
Which leaves us with the parts that MS does know about. Obviously they do have a .designer file which backs the .aspx pages. This designer file maintains the server control list that is accessible by the code behind. You could add runat="server" to every regular element in an html page (excluding artifacts like css) but this would increase the amount of processing required for the page.
To sum up, I think they've picked the best model they could for web development given the lack of control they have over how the web works. Other vendors have tried similar paths (e.g. Adobe Contribute, previously by Macromedia). I believe some of those even predate MS's model.
While working on a quite big web application project, I decided that it could gain a little bit of fresh air by marking some of the pages and controls with the CompilationMode="Never" #Page attribute. So far so good, working as expected and then it happened. A corner case scenario that I am going to explain behaved unexpectedly to put it nicely. This scenario is nested master pages.
A quick teaser before continuing. How deep nesting do you think you could go if you mark the top master page as CompilationMode="Always", and all others beneath it with CompilationMode="Never"? No, its not infinite, or some internal number that ASP.NET has. Its 2. Why? - I have no idea, and I was hoping some of you smart guys could enlighten me?
I have attached a project with 5 nested master pages to demonstrate what I am talking about: Nested Master Pages Web Application Test Project.
Another corner case that is working unexpectedly as well - if you have 5 nested master pages, change the second to have CompilationMode="Always" and all others to have CompilationMode="Never". You will notice that the 3rd master pages is being applied twice!
Please help me understand if something I am doing is incorrect, or confirm the issue.
ASP.NET Runtime Version: 2.0, .NET: 3.5
EDIT:
The project attached has all master pages set to CompilationMode="Never". The ASPX page displays as desired. Change the first master (Site.master) to have CompilationMode="Always" to see what I am talking about.
UPDATE (1/21/2010): Good news: after more investigation, it turns out that this issue was fixed in VS2010. The fix was made post Beta 2, so it will be part of the next public build. I don't have an exact date, but it should not be too far out.
Yes, I seem to recall this coming up before, and indeed some scenarios involving nester master pages and CompilationMode="Never" are broken.
Looking at an old mail thread, I think it only happens for certain combinations. It looks like it’s broken for (where NoCompile means compilationMode=never):
NoCompile Page / Compile Master / NoCompile Master
NoCompile Page / NoCompile Master / Compiled Master
At the time, we did not fix this because the fix was non trivial and the scenario is not common.
Note that when it comes to NoCompile pages, most of the benefit is using it for end node aspx pages, and not master pages. Generally, NoCompile pages run a bit slower than compiled pages. Their benefit is that they don't have a first time compile hit, and they use less memory. Also, they can be fully unloaded under memory pressure. That's why they make good sense when you have a super large number of end point pages (Sharepoint uses them). But on master pages (where most apps only have a small number shared by many pages), that benefit would be minimal. And of course you can't have code in NoCompile pages, which is the main reason that few people use them.
So the quick summary is: you're right, it's a bug! :) And the recommended workaround is to avoid CompilationMode=never for master pages.
Can someone please let me know the pros and cons of registering user controls in web.config vs. on top of asp pages?
I am looking for performance issues in particular. Does having all the controls registered in web.config make the pages slower to load (even the pages that do not use these controls)?
Web.config Pros
Only 1 place to add extra bit of code
Web.config Cons
Sometimes, I forget to update web.config on production environment since I never copy the web.config from staging area to live site.
On every page pros
When you deploy/publish pages and/or copy pages from dev/test/staging to live server, you don't need to worry about updating web.config
One very page cons
Its on every page, so if something changes, it's a pain to go through every page and fix it
You have to add it manually to each page
I'd recommend doing it in the web.config. It's less work for you since you don't need to remember to add it to every page. Although, if you have a ton and lots are only used on a single page, then you could do a bit of both to keep your web.config a bit less cluttered. I usually put my controls in web.config, if I use them in more than one page. If it's just used on a single page, then I usually just declare it in my asp.net page.
Old question, but finally an answer on performance! There is no difference:
http://weblogs.asp.net/scottgu/archive/2006/11/26/tip-trick-how-to-register-user-controls-and-custom-controls-in-web-config.aspx#1056687
The good news is that there isn't any performance difference between
registering them in a web.config file vs. at the top of a page.
The runtime performance is exactly the same (they get compiled down to
the same instructions in both scenarios).
Compilation should also compile at the same performance.
Not a heck of a lot cons to putting it into the web.config. It just can't be done with a drag and drop, so many just don't do it. Keeps the page less cluttered and makes change easier in the long term.
Scenario:
I have a pretty standard master page for all my pages. It includes the usual login forms and other dynamic lists to be extracted on each page. Webdesigners can already modify the central content place holder of each page. But still, the design and layout for the master page is still in my project and any modification to the design must be made in Visual Studio and the project re-compiled and re-deployed.
What is the best way to provide near-full access to designing the master page through a CMS? Some of the problems I can identify is the inclusion of any dynamic lists or specific controls such as a login form.
Thanks.
Unless you want to host your content within a portal I don't know of a perfect answer to this.
If the bits they design just amount to look and feel for the page then this can be controlled by css and you could allow them to create themes using different css files.
This is indeed an interesting question, and there is no perfect solution. I worked for an ecommerce shop with this issue, and frankly, I just asked the designers in many cases to provide me there html and css, then I would grab the html pieces and css and add them to my project. Yes this was tedious....
Then we we built a cms where the designers could copy and paste their html into html editors, and we would store those pieces of html in a database. My web app would grab those from the database at run time. This solve some issues, but not all, since it did not give them complete control of the design of the web page.
The bottom line is you need to standard as to how the designer will submit their work to you. If you have that, and you can count on the html and css, then you can star to think of possibly building a CMS around that. In this days of RAD, I have found it easier to just work with the html and css delivered to me and simply copy and paste the pieces into my master page and other pages as needed.
While this is not a CMS answer, you do have the ability to allow designers to open the master pages in Expression Web. I will not say it is the greatest tool in the world, but I have had designers work up the master page designs in Expression with good results.
There is a pain point, however. If the entire project is opened, the designer will see the code behind files as separate items, not like the treeview view seen in Visual Studio.
I imagine you could have the master page checked out for use with Expression through a CMS, but there is no built in way to do this, nor do I know of a third party tool to do this. Hopefully Expression Web 3 will make things easier.
If you have a CMS, you may be better to give it full control over page content. If there are things the CMS cannot do, you could look to write extensions or plugin modules for the CMS that your designers can then drop onto the page in the CMS's page editor.
If your CMS doesn't support plugin modules, you may be trying to force both the CMS and master pages to do things they were not intended to do.
If the above doesn't work in your situation, here's another thought: place inline frames on your master page that host pages that are edited in the CMS.
Hope that helps.
Would it be possible for you to put placeholders in the Master Page in place of the areas that designers should be allowed to edit? Since Master Pages are only editable in Visual Studio, it may be your only feasible option at this point in time. One problem with this approach is that the content put in the placeholders is unlikely to be valid, since you would probably have tags left open in one placeholder and closed in another.
<html>
<head>
<title></title>
<asp:PlaceHolder ID="headerContent" runat="server"></asp:PlaceHolder>
</head>
<body>
<asp:PlaceHolder ID="beforeContent" runat="server"></asp:PlaceHolder>
<asp:PlaceHolder ID="centralContent" runat="server"></asp:PlaceHolder>
<asp:PlaceHolder ID="afterContent" runat="server"></asp:PlaceHolder>
</body>
</html>
I know it's ugly, but it might give you the control you need (as long as you don't mind the XHTML validator warning you all the time). What you put in the placeholders could be your user controls or literal content or whatever, but you'd have to load it dynamically.
Thoughts?
EDIT: This won't work. The PlaceHolder is going to render <div> tags that would mess things up. Maybe you could extend PlaceHolder and override how it renders its HTML.
Interesting question,
been dabbling in that area myself a while ago.
How knowledgeable are these web designers when not in the realm of not-inside-Photoshop-or-flash?
If using a DIY-CMS, perhaps you can template the most susceptible objects, e. g. making a generic (as in whatever you feel like, not whatever they feel like ;-)) list and a way of entering design, if applicable.
As long as you have a thorough framework set up, that deals with the attributes available to the designers through the CMS, there shouldn't be any need for recompiling... but of course, I can easily see a developer (read : me) stumbling
into the gap of nitty-pitty-perfection....
I'm afraid the easiest, and only manageable, path is to standardize how the designers express their needs&wants to you...it just won't code itself...
Could you provide some examples?
Well, with a Web Application Project pages are not compiled until accessed (link is 2005 but it still applies). This means that the actual .aspx (and .ascx etc) page is deployed in its original state. A designer can update the format of the page on the server and the updates will be compiled the next time someone requests that content.
It would be relatively trivial to allow designers to download the current pages and upload replacements through your website's UI. However, it isn't very secure (and probably should never be done). It would be better to allow designers access to the virtual directory over the web so they can connect to it using a tool such as Expression Web. This way the designer can open the current website, edit pages, and push the results directly into production (scary tho that thought may be).
As I'm getting downvoted for having a correct answer, let me point out something.
Website projects compile codebehind and pages on demand. If you need to update code regularly, its an okay solution.
Web application projects can be configured to be updatable. All codebehind and classes are compiled into an assembly, and all aspx, ascx, etc pages are deployed and compiled on demand. This means that a designer can connect to the website, update the layout and static content, and see the changes on the next request.
This is my preferred method of deployment. I have a few web application projects out there in the wild, with updatable aspx files deployed alongside my dll. The idea being that users of the website can alter the UI without having to submit updates to me so I can recompile it for them.
a .master is just a text file. They can edit it however they like. There's certainly nothing they'd do to it that would require you to recompile the code just to view it. That's the big win with Master Pages in the first place: designers and other non-programmers can edit them manually without breaking anything.
Give them access to the file under source control and let them go nuts.
I wonder what the best practice for this scenario is:
I have a Sharepoint Site (MOSS2007) with an ASPX Page on it. However, I cannot use any inline source and stuff like Event handlers do not work, because Sharepoint does not allow Server Side Script on ASPX Pages per default.
Two solutions:
Change the PageParserPath in web.config as per this site
<PageParserPaths>
<PageParserPath VirtualPath="/pages/test.aspx"
CompilationMode="Always" AllowServerSideScript="true" />
</PageParserPaths>
Create all the controls and Wire them up to Events in the .CS File, thus completely eliminating some of the benefits of ASP.net
I wonder, what the best practice would be? Number one looks like it's the correct choice, but changing the web.config is something I want to use sparingly whenever possible.
So in that case I would wrap it up in a feature and deploy it via a solution. This way I think you will avoid the issue you are seeing. This is especially useful if you plan to use this functionality within other sites too.
You can also embed web parts directly in the page, much like you do a WebControl, thereby avoiding any gallery clutter.
What does the ASPX page do? What functionality does it add? How are you adding the page into the site? By the looks of it this is just a "Web Part Page" in a document library.
I would have to do a little research to be 100%, but my understanding is that inline code is ok, providing it's in a page that remains ghosted, and thereby trusted. Can you add your functionality into the site via a feature?
I would avoide option 1, seems like bad advice to me. Allowing server side code in your page is a security risk as it then becomes possible for someone to inject malicious code. Sure you can secure the page, but we are talking remote execution with likely some pretty serious permissions.
Thanks so far. I've successfully tried Andrew Connel's solution:
http://www.andrewconnell.com/blog/articles/UsingCodeBehindFilesInSharePointSites.aspx
Wrapping it into a solution is part of that, but the main problem was how to get the code into that, and it's more leaning towards Option 2 without having to create the controls in code.
What I was missing:
In the .cs File, it is required to manually add the "protected Button Trigger;" stuff, because there is no automatically generated .designer.cs file when using a class library.
Well, it's a page that hosts user controls. It's a custom .aspx Page that will be created on the site, specially because I do not want to create WebParts.
It's essentially an application running within Sharepoint, utilizing Lists and other functions, but all the functionality is only useful within the application, so flooding the web part gallery with countless web parts that only work in one place is something i'd like to avoid.