Sign in
|
Join
|
Help
in
Jeffrey Palermo (.com) (blog)
CodeBetter.Com Emeritus (group)
(Entire Site)
Search
Do you Twitter? Follow us
@CodeBetter
Jeffrey Palermo (.com)
Blog moved to www.jeffreypalermo.com
Published by
Comments
Adam Weigert said:
My guess, and it is just that, would be that they keep a System.WeakReference to the objects.
At least this is the technique I have applied for custom cache controllers. A WeakReference will keep a pointer to the object but won't hold on to it so GC can happen.
I am not too sure if this is the same technique for the ASP.NET cache. So, if your page or object goes out of scope and is GC'd then cache lets it go because it no longer contains a valid WeakReference.
#
April 16, 2004 3:06 PM
Patrick Santry said:
Have you checked out DotNetNuke (
http://www.dotnetnuke.com
)? I run my site on it (
http://www.wwwcoder.com
). DNN provides a lot of what you want, and there is a large community support for it. Just go to the forums on www.asp.net, and you'll see plenty of DNN threads. MS did a major push on at VSLive! this year as well. With DNN you can create portals and child portals, each portal can have distinct permissions and role based security. There are plenty of modules available like shopping carts, forums, blogs, etc. And it's all open source and free.
#
April 16, 2004 4:16 PM
Scott_NO_@_SPAM_Tripleasp.net (Scott Watermasysk) said:
No. Thank YOU!
I got to stay home and write code for the last year. I sacrificed a little of my personal time, but nothing in comparison to you.
Thanks Again,
Scott
#
April 16, 2004 4:18 PM
ScottWater said:
Thank you Jeffery and others who have given up much more!
#
April 16, 2004 5:27 PM
Noah Coad said:
Hey Jeff! Glad to see you up and running with .Text! Great blog, lots of good stuff. Yours will be the first blog I actually read through. :)
Your timing on sending the original e-mail asking how the MVP Global Summit was incredible. I just finished a long day at the summit, when up to my hotel room, checked e-mail, went downstairs, started talking to some of the other MVPs, and found out they were Mark, Rory, Jon Box, etc. Wow. I mentioned you had just e-mailed me and they all wanted to say 'Hi'.
#
April 16, 2004 6:53 PM
Noah Coad said:
I wholeheartedly agree, I can hardly wait for generics! I've seen them in Whidbey and the syntax is even better than I had hoped. One of the C# team members said he thought anonymous delegates were even cooler which was a surprise. They are both kewl, but of course, I am tired of casting out of ArrayList etc.
#
April 16, 2004 6:59 PM
Noah Coad said:
Great looking UI!
#
April 16, 2004 7:02 PM
Paul Wilson said:
This gets much better in ASP.NET v2.0, since you'll be able to specify the a different url for your buttons to post to, and there will even be validation groups.
#
April 17, 2004 1:27 AM
VB said:
I agree it 100%. Nice blog. Keep it up.
#
April 17, 2004 1:53 AM
Jason Lautzenheiser said:
Glad to hear you're getting back and just wanted to thank you for your service!!
#
April 17, 2004 2:34 AM
Noah Coad said:
Hey hey, I agree! Quite interesting... :)
#
April 17, 2004 4:47 AM
Kevin Daly said:
Don't forget that while the default in the PDC bits is inline code, in the release version the default will be CodeBeside (or however they write that), which is similar to CodeBehind but uses partial classes.
It's true that having markup hidden while you edit inline code (a feature borrowed from Web Matrix I think) largely avoids one of the problems of having code and markup in the same file (that is, that it's really, really hard to follow), there are other advantages in separating the two, especially where you have more than one person working on the project. For instance, it enables me to write all the application code that talks to the middle tier and so on, and then hand the .aspx file to our web designer and say "make it pretty" :-)
But then again, for the situation you describe inline code probably does make sense, so I think that it's good that Microsoft have catered for that.
#
April 17, 2004 8:20 AM
Jeffrey Palermo said:
I have recently downloaded the project, but I haven't looked through it yet. Thanks for letting me know that you have had success with it. If they only partially overlap, maybe I can get some good ideas from it. Either I can make mine better or take some ideas and add them to DNN.
#
April 17, 2004 10:48 AM
Andrew said:
I found this tool to be very useful, since I can run it on IIS, and access a Sql Server database from any machine (with IE) without having to install any Sql Client tools.
BTW, Cassini is included with the download already.
#
April 17, 2004 9:36 PM
Steve Schofield said:
No Jeffery, the BIG THANKS goes to you and everyone that in Iraq and other places defending our freedom. .Text is a great tool!
#
April 18, 2004 8:45 AM
Jeffrey Palermo said:
I explored .Net Nuke yesterday, and I found it very interesting. My project and .Net Nuke overlap a little. I found .Net Nuke to be geared toward being a portal. My project is geared toward globally controlling the layout of any website/web app. .Net Nuke has great support for the content of every page, whereas my project doesn't focus on each page, but the site as a whole. Now I plan on converting .Net Nuke to be a Plugin for my project. This way, the user can choose "Portal" as one of the plugins to implement for a certain page. Then my user can take advantage of .Net Nuke. Great stuff!!
#
April 18, 2004 4:46 PM
Vijay said:
Maybe you should try to use CollectionGen framework ... A simple XSL based templated collection type generation framework which is completely customizable ! hmm forgot the exact link but Google should get you thro !
#
April 18, 2004 7:01 PM
Stefano Demiliani said:
I think that your application seems god. My little suggestions are:
1) to implement also a layer for Access (.mdb) database (many provider offers only .mdb support). The latest version of DNN has also an Access (.mdb) Layer, but I think it could be improved.
2) Try to have the portal installation more easy as possible, without special permissions for particular folderds etc.
#
April 19, 2004 3:25 AM
J. Michael Palermo IV said:
Great to see you are back home. (And a great blog home too)
#
April 19, 2004 9:15 AM
sangmin said:
Good for you, Jef!
#
April 20, 2004 11:32 AM
Steve Smith said:
Glad you're back ok, Jeff. I just found out Monday I'm heading out for a year myself, leaving for Ft. Leonard Wood on May 22. Send me an email some time - I'd love to hit you up for any tips you can share.
#
April 28, 2004 3:18 AM
Jesse Ezell said:
Sourceforge.
#
April 28, 2004 7:30 AM
Jeffrey Palermo's Weblog said:
Best place to start an open-source .Net project?
#
April 28, 2004 7:50 AM
Jeffrey Palermo's Weblog said:
My project provides master pages features now!
#
April 28, 2004 8:38 AM
Rob Relyea [msft] said:
Happy to read about your interest in Avalon/XAML.
I love to read customer opinion & feedback. Let us know how you like it once you start building things.
Stay safe in Iraq (or maybe you are back?)
-Rob
#
April 29, 2004 7:41 AM
Jeffrey Palermo said:
I am very much interested in XAML, but in my situation, I must concentrate on meeting my customers' needs now, so I am swamped with that work. Whenever it is released, I will definitely develop some applications with it. Thanks for the comment. I like that WinForms interfaces will now be able to have a flow layout like HTML. That is very powerful.
#
April 29, 2004 8:02 AM
Jeffrey Palermo said:
Oh, yes I am back in the states. I've been back for a week now.
#
April 29, 2004 8:03 AM
Mark said:
very cool project you got there.
it took me a while to adjust the IIS settings well enough for it to work...but now it works great.
one question... Is there a way I can remove the information that is displayed at the bottom of the screen? (Trace Information, Control Tree, etc..)
#
May 5, 2004 7:15 PM
Jeffrey Palermo said:
Of course. All it is is trace information. Go to the web.config file and turn trace off.:
<trace enabled="false". . ./> and it'll all go away.
Go to _ezweb/Templates/ folder and create your own customized template.
#
May 6, 2004 12:25 AM
mark said:
thanks a lot. i figured it was something easy like that but I'm still learning the .NET way of doing things.
#
May 6, 2004 3:43 AM
Tye said:
For tons of solutions go to:
http://www.google.com/search?q=scroll+position+post+back+.net
#
May 10, 2004 8:17 AM
JosephCooney said:
I think base pages are one of (possibly THE) best feature about ASP.NET.
#
May 13, 2004 4:15 PM
Peter van Ooijen said:
They are great, but it is a pitty you don't inherit the markup. You only inherit the codebehind, not the aspx itself. You can drop controls on your base form but they will not show up in the inheriting page.
#
May 14, 2004 2:52 AM
Jeffrey Palermo said:
Yeah, but all this is is class inheritance. I've been able to do some much since I've wrapped my brain around object-oriented concepts. This isn't unique to ASP.NET, it's just a benefit from the underlying object-oriented .Net Framework. From what I read about Master pages, you'll be able to inherit markup by nesting master pages.
Have you checked out my EZWeb Project?
http://workspaces.gotdotnet.com/ezweb
#
May 14, 2004 3:04 AM
Peter van Ooijen said:
Hi.
It is class inheritance. Asp.Net does even more. When it comes to serving the page it will generate a new class which includes members for all controls on the aspx and which includes all <%s > stuff like databindings in plain C#. This generated class inherits from your code behind class.
Before .net I used to do Delphi which was fully OOP from version 1. In there you could inherit visual (win) form classes, including controls. It could be pretty flaky but the idea worked. Sometimes I wished I could do that with an aspx (production) webforms as well. Have to wait till it is accepted to roll out 2005 apps. With masterpages.
In .NET I do use base (form) classes and I like it. Just like you. Havn't looked at you project (yet).
Peter
#
May 14, 2004 6:54 AM
stefan demetz said:
Enterprise grade ASP.NET Development and code reuse
http://dotnetjunkies.com/WebLog/stefandemetz/archive/2004/01/14/5631.aspx
#
May 15, 2004 8:39 AM
Shannon J Hager said:
I have EZWeb in the root of a test web and am running into the problem described here:
http://weblogs.asp.net/dneimke/archive/2004/05/17/133116.aspx
#
May 17, 2004 5:53 AM
Jeffrey Palermo said:
Thanks. I respond here
http://dotnetjunkies.com/WebLog/jpalermo/archive/2004/05/17/13810.aspx
#
May 17, 2004 6:37 AM
Thomas Tomiczek said:
::My colleague insisted that script block
::incur and extra performance hit because the
::ASP.NET engine must check for changes in
::every page hit
Tell your colleague McDonalds ahs a job for him - serviung burgers. He is better qualified for this than programming.
* ASP.NET never checks for changes. ASP.NET subscribes to the events in the file system. When ANY file is changed, ASP.NET is notified and restarts the app.
* Even with codebehind, a change in an ASP.NET page - any change - has to trigger a recompilation. Why? New server side objects may be in there etc.
Common sense, oh so rare today.
Besides, this is pretty basic knowledge, properly documented in the documentation and thoroughly discussed in every message board.
#
June 1, 2004 6:26 PM
Steve Hebert said:
In your post, you stated... "every time I want to make a small change to the UI that involves som C#, I have to fire up VS.NET and then build the whole project"
In your <%@ Page ... %> tag, you can replace Inherits= with src= to get around this. This means you have to distribute your C# code, but there is the workaround if you need it.
If you want a rundown on what's happening on code-behind vs. in-script, check out Chapter 1 on Essential ASP.NET by Fritz Onion. You'll understand exactly why there is no difference and then you can hand it to your buddy.
Then I'd run the following code:
colleague.SpeaksFromSphincter = true;
#
June 2, 2004 2:54 AM
Jeffrey Palermo said:
Explanation of dynamic ASPX compilation and recompilation - level 300
#
June 2, 2004 4:30 AM
Peter van Ooijen said:
Hi,
why not create you own datalist. Inherit from datalist and add your own header property. No big deal :
http://www.dotnetjunkies.com/Tutorial/6614EF8F-2ADB-4AEB-B2B7-D96278B53D74.dcik
Templates are not very nice to work with from code. When some coding on a template is needed I drop an usercontrol on the template and do my coding in there.
#
June 3, 2004 2:44 AM
Jeffrey Palermo said:
That's a good point, and I could easily do that, and I may. Right now I have a UserControl that is only the DataList, so I have added a HeaderText and FooterText property to the UserControl, and I handle it that way. I agree with you, though. What I was perturbed about was that this functionality isn't included in the DataList object.
#
June 3, 2004 6:49 AM
Peter van Ooijen said:
My main problem with the datalist are the templates. Flexible but unfriendly to program. They have been overhauled in Whidbey. Looks like they support events now. But no class members yet (?)
#
June 4, 2004 3:32 AM
Chris Kinsman said:
Fairly inefficient and unless you change the connection limit it limits you to two connections to the web service at any point in time...
#
June 5, 2004 5:51 PM
Jeffrey Palermo said:
To be honest, I hadn't thought about the connection limit. I'm still searching for the best way to share sessions with ASP and ASP.NET.
#
June 6, 2004 6:46 AM
Jeffrey Palermo said:
Re: Explanation of dynamic ASPX compilation and recompilation - level 300
#
June 7, 2004 4:58 AM
James Curran said:
"and it does not create spaghetti code."
Define what you mean by "spaghetti code".
Unlike "break" & "continue", "goto" giving you no indication where the execution is going to continue. It could be up or down or anywhere.
Doing that also confusing (for the programmer) object lifetimes, and undermines the compilers ability to optimize the code.
#
June 7, 2004 8:22 AM
Mark Bonafe said:
You are most definitely creating spaghetti. I cannot understand why anyone would do this. Simply put the "statement-that-should not happen if condition 2 is not met;" inside the first if condition above the second if condition.
If that is not good enough, then you should consider creating more than just one method so you don't have to "jump" all over the place in your code. You obviously have more than just one purpose for the code. So creating extra methods, one for each purpose, should solve the problem.
#
June 7, 2004 8:29 AM
Jeffrey Palermo said:
Spaghetti code is code that is hard to follow. It does jump around up and down and all over the place. My example of the goto statement is very straight-forward and is used to jump out of a specific IF statement. What method would you prefer to jump out of nested IF statements?
#
June 7, 2004 8:30 AM
Jeffrey Palermo said:
I'm using the goto to simply jump out of the if statement. By my label location, I just down a few lines, that's it. One jump. Straight-forward.
I cannot move the statement-that-should not happen if condition 2 is not met; inside the first IF condition because it has nothing to do with the first IF. It's independent. If both condition1 and condition2 are true, this statement should be skipped. I obviously could avoid the goto with another if checking for them both, but I prefer this simple goto usage. I think this is an appropriate use of goto, and I will defend it. The writers of the C# language thought the goto statement worthy of inclusion, so what argument is there for comdemning the goto statement to a life of nonusage? It goes without saying that just because it's there does not mean it has to be used. Same with xml. It doesn't need to be used with everything, just where appropriate. Same here. Use goto where appropriate.
#
June 7, 2004 8:37 AM
RL said:
Thanks Dude!
#
June 8, 2004 12:35 AM
Mark Bonafe said:
I cannot explain it any better than James. His comments should be all you need to understand why it should be avoided.
#
June 8, 2004 1:36 AM
Jeffrey Palermo said:
In regards to James's comment, I'd be interested to know what compiler optimization is precluded by the usage of goto. Mark, it sounds like you will never use the goto statement for anything. I'll use it where appropriate and where it makes sense.
#
June 8, 2004 2:18 AM
Jeffrey Palermo said:
The method I described above is a good way to make web service calls participate in the same session. This method does not scale for complete Session sharing if every hit of an ASP app is making a web service call. For high traffic like that, I think putting session information in sql server is a better option. And if all you need is the currently logged on user, then adding a cookie with the username is a good option.
#
June 8, 2004 2:26 AM
bale said:
It's not working for me Man... :(
#
June 9, 2004 3:53 AM
Brian Orrell said:
It's just bad. Can you not accept that there is NOT a use that is considered acceptable by your peers?
It's not good practice. Period. And the C# team included the goto statement (my guess) so that VB programs could be migrated more easily.
If you want to get the same label that poor VB programmers have had for a long time (writing crap code) have fun.
Show me one, ONE MS guy who says that he can see a valid use for you using it in production code.
It's not a best practice.
#
June 9, 2004 4:37 AM
Noah Coad (C# MVP) said:
I would tend to agree with Brian. I try to avoid using the statement, if only because it is so controversial. I have tried using goto in programs before, but have found with a little more thought and perhaps an extra line of code, it is worth trying to avoid.
The statement breaks down the natural flow of logic. For the same reason, I try to avoid "break" and "continue" (I never use "continue"). I try to layer my logic to avoid such circumstances.
When I get to the Visual Studio team at Microsoft in July, I can ask around. I will also ask someone from the C# team to take a look at this post right now.
Here is an article on how GOTO is harmful.
http://www.acm.org/classics/oct95/
#
June 10, 2004 2:24 PM
Noah Coad (C# MVP) said:
Hehe, very kewl dude! Funny, I just e-mailed you asking if you had read that post. This is just been a great app and made blogging sooo easy!
It will probobly get a GUI if just to see the log output as it runs, but maybe not as I am a command line fan. But it is very easy to adapt as I use custom TraceListeners instead of straight Debug/Trace/Console output.
Want code? Well perhaps after it is polished up just a little more. It is designed to parse my own custom tags so it would require significant non-incode documentation. We'll see... :)
#
June 10, 2004 2:51 PM
Jeffrey Palermo said:
I read the above article, and I agree with it 100%. The author's position is to use extreme caution with goto because it an lead to bad code. I agree with this. This comment thread has been very interesting for me because of the viewpoints expressed. I'm certainly not one to say that I know best - period! I encourage feedback and learning. What I have gathered from the comments is that goto is just bad - period. While I agree with the arguments against common use of goto, I'm struggling to believe that an included feature in C# is really taboo. I made this post because for the first time in my development career I'd thought I'd found a legitimate use for the goto statement. I've gathered that most people avoid goto at all costs. In several years I aspire to be a .Net trainer, so from that perspective, I have to be able to answer the question: Why shouldn't I use goto? I can't just give the answer: It tempts you to overuse. It makes the program confusing if overused. It promotes bad code. It's hard to follow. I can see the students reactions to those answers, so what I'm looking for is a real answer. What is the REAL reason not to use goto? Why is it in C#? Noah Coad has contacted Eric Gunnerson to solicit his opinion on the matter. We'll get to the bottom of this eventually.
#
June 11, 2004 2:29 AM
Andrew Whitten said:
I develop under XP at home at 2003 at work.
Developing against 2003 makes me consider the new security and deployment considerations of IIS 6.
However if I just have a few things to fix in the evening, then using XP is fine.
#
June 15, 2004 7:37 PM
Darrell said:
And FxCop will warn you when you violate the naming guidelines, which is the only way to realistically enforce them.
#
June 17, 2004 7:56 AM
J. Michael Palermo IV said:
IMHO, you should explain the difference between:
All private members...
All private class members...
I would identify the first one as:
All local variables or parameters
I would identify the second one as:
All private type members
my .02
#
June 18, 2004 5:33 AM
. said:
why are you guys wasting your time with such nonsense. Just follow the MSDN Design Guidelines.
#
June 19, 2004 11:50 PM
Noah Coad (C# MVP) said:
And I think profiling (performance testing) should also be avalible. Profiling is in fact of an even more interesting topic to me since the tools are not as readily avalible as unit testing tool are now and I just think it is way cool.
#
June 26, 2004 8:35 PM
Jeffrey Palermo said:
The goto statement revisited - level 200
#
June 30, 2004 6:51 AM
Michael Fitzpatrick said:
I use GOTO all the time in C code. I had to to implement exception handling using error code return values. I will occasionally use GOTO for breaking out of nested loops in C. I use GOTO in VB in place of break and continue, which aren't supported in VB.
In C++ there is SEH, but it is expensive. I have a lot of C code that I ported to C++ that still uses GOTO for exception handling.
I think that if a programmer has done assembly coding that they loose religion about the use of GOTO real fast. The point is that any coding construct can be abused. GOTO has gotten a really bad rap because it was overly abused, and college professors don't allow it at all.
#
June 30, 2004 8:33 AM
Noah Coad (C# MVP) said:
Thanks for the shot out Jeff! This is my last blog comment as an MVP. In a few hours I'll be MS and no longer an MVP. Keep up the great ASP.NET work! Be sure to look me up when you get in the area (won't be long!).
#
July 5, 2004 7:01 PM
muffDaddy said:
I like the tool as well but I seem to have a hard time figuring out the login
#
July 10, 2004 2:24 PM
Noah Coad (MS-VSTS) said:
So is the 21" a CRT? I've used a 21" ViewSonic P-810 for years and love it.
#
July 23, 2004 12:43 PM
Jefrey Palermo said:
Yeah, it's a CRT. Like I said, my company won't slap down $500 just for a display.
#
July 24, 2004 2:43 AM
MacSqueeb said:
Thanks for a helpful datagrid post. You cleared up some vagaries surrounding ItemDataBound that I've just lived with until now.
BTW, I just discovered your blog on dotNetJunkies and skimmed back through mid may. This is a *great* ASP.NET blog with a good balance between specificity and general interest. I especially appreciate your 100 - 400 rating system as it helps me identify likely relevance to me at a glance. Keep up the good work.
#
July 27, 2004 8:29 AM
Famil Jones said:
ListItemType Enumeration could also be used to test if the item type is header or footer in ItemDataBound and ItemCreated events (eg. If e.Item.ItemType = ListItemType.Header Then ... End If).
#
July 28, 2004 4:35 AM
Abner Mendoza said:
Instead of using the ItemIndex, it would probably be easier to use ItemType. So in your event handler you can have a statement to test for e.Item.ItemType and then your header would be = ListItemType.Header and your footer would be = ListItemType.Footer.
#
July 28, 2004 4:35 AM
Jeffrey Palermo said:
Those are great comments!! I will, in fact, use the ItemType property. That is a little more readable than testing the ItemIndex. That is great for accessing embedded controls during the data-binding events. I also have to access the button in the Header at other times, and I can't iterate through the Items collection because the Header isn't an item, so I'm accessing my embedded button by calling DataGrid.Controls[0].Controls[0].FindControl("myButton"). The first control in DataGrid is an HtmlTable, and then it has a collection of HtmlTableRow, so I have to get the table, then Controls[0] of the table, which is the first row, and then do a FindControl(). If I find a more intuitive way to do this, I'll post it.
#
July 28, 2004 5:47 AM
Eric Wise said:
When I started using .NET I used datasets for almost all databinding. Once I learned this technique my usage of the dataset is almost nil.
I often wonder if this is the case with other developers out there.
#
July 29, 2004 1:03 AM
Jim Wooley said:
You may want to check into the work of Rocky Lhotka (www.lhotka.net). He as a robust distributed object model that addresses this area and much more.
I've been using variations on his archecture for 4 years now and have been very happy with the results. It takes longer to architect, but is more maintainable in the end.
#
July 29, 2004 3:01 AM
sa said:
as
#
August 2, 2004 1:41 AM
Ellen said:
How do I find out the login information?
#
August 6, 2004 7:50 AM
Art said:
The GotDotNet download has a problem, it downloads as 0 bytes in size...
#
August 7, 2004 4:00 AM
Art said:
Never mind - I found it was Norton Internet Security interacting with GotDotNet. Turned off NIS and it worked fine.
#
August 7, 2004 1:29 PM
Kevin Daly said:
I think MS would argue that they can't guarantee not breaking an unsupported feature, such as Head exposed to the server.
I was doing the same thing with Title for a while but it drove me nuts because any change to the UI would wipe the runat="server" attribute.
#
August 16, 2004 1:12 PM
Anders Norås said:
The head element in the ASP.NET 2.0 page template is by default declared to run at the server. In addition the System.Web.UI.HtmlControls namespace has a new class called HtmlHead defined. When you set the id attribute of the head element to “Head” this becomes a member of your class (which inherits from System.Web.UI.Page).
The reason for you not seeing the code defining this member is that ASP.NET 2.0 uses partial classes to separate user defined code from tool generated code.
As a result of this you get a member naming conflict when defining an additional member called “Head” typed System.Web.UI.HtmlControls.HtmlGenericControl.
To solve your problem using your existing code which I reckon is similar to this:
ASPX snippet:
<head id="Head" runat="server">
Codebehind snippet:
protected HtmlGenericControl Head;
private void Page_Load(object sender, System.EventArgs e)
{
HtmlGenericControl meta=new HtmlGenericControl("meta");
meta.Attributes.Add("name","pragma");
meta.Attributes.Add("content","no-cache");
Head.Controls.Add(meta);
}
Remove the declaration of the “Head” member in the code behind file.
Another option is to add your metadata using the Page.Header.Metadata collection like this:
Page.Header.Metadata.Add("name", "pragma");
Page.Header.Metadata.Add(“content”,”no-cache”):
#
August 17, 2004 3:32 AM
Jeffrey Palermo said:
Yeah, that's exactly what I did to fix it. I decided to make this post to make this issue public. The more I dive into Whidbey, the more issues I encounter. Yes, there are always workarounds, but I wish v2.0 wasn't touted as 100% compatible.
#
August 17, 2004 5:34 AM
Jeffrey Palermo said:
Whidbey *BUILD* web performance very slow - level 100
#
August 17, 2004 10:17 AM
vank said:
We've been investigating this and the prognosis looks good. It looks like we're doing about twice as much work as we should be doing.
Do you mean that it waits 10 seconds before announcing that it's begun, or it waits ten seconds until there's any disk or CPU activity? There are two different, real problems. First, the Output window seems to only be updated during idle time, so even though we post the starting announcement right away, since we get busy building there isn't any idle time for a while. This is fixable. Secondly, we wait for any pending Build Manager AppDomain resets before starting; this could potentially be waiting for ten seconds and we plan to eliminate this.
We don't have our perf goals set for this one test case yet, but we've got sample timings of 48 seconds while the command line build on the same machine takes 14 seconds, and our objective is to approach the command line time. We just don't know how close we can approach it yet.
#
August 17, 2004 10:21 AM
Jeffrey Palermo said:
CodeBehind attribute mysteriously reappears in VS 2005 - level 100
#
August 17, 2004 11:18 AM
Romualdas said:
I thought it this way as well, when looking first time, but there is an option for that so You can change it.
If You do that, it will create separate file (code beside now?) with partial classes.
#
August 17, 2004 7:15 PM
J. Michael Palermo IV said:
Jeffrey,
I just read all of your blogs leading up to (and including) this one. Thank you! The power of blogging is having information such as this "war story" available for all of us to either learn from or help you with. I am also in the middle of porting code over, and we share much misery :)
(not to mention a last name)
Michael Palermo
#
August 18, 2004 3:48 AM
Jeffrey Palermo said:
SQL Server Reporting services integration with ASP.NET - level 100
#
August 18, 2004 8:52 AM
Jeffrey Palermo said:
EZWeb v1.1 released on GotDotNet! - level 100
#
August 18, 2004 9:42 AM
Jeffrey Palermo said:
Sure, no problem. I'm glad my blog is being helpful. I'm also trying to combat the slew of other posts that only seem to announce and reannounce current events.
#
August 18, 2004 10:01 AM
Anders Norås said:
There is no such thing as one solution to all problems. In some scenarios it would make the most sense to set top-level page properties such as the title, target and similar in code. I believe that this is an exception and not the rule. If it is more of a rule, such as with generated titles in ASP 1.x, where master pages aren’t supported. I would consider injecting or altering the HTML code for the TITLE element in an HttpModule. This would separate the presentation rules from the content.
This would of course be more difficult to develop than a extension to the System.Web.UI.Page class, but you would benefit from this in the long run since it would make the application easier to maintain.
#
August 19, 2004 9:46 AM
Lakshmi Narayanan.R said:
wow thanx la..very simply...but the complete..
#
August 19, 2004 9:14 PM
Jeffrey Palermo said:
Successfully using Server 2k3 on my development laptop - level 100
#
August 20, 2004 10:18 AM
mXskweeb said:
Here here. I just tried SSRS this week thanks to your previous post. I didn't look too closely at the setup for the datasource, but I think it had a handful of providers to choose from. So, theoretically, you could run reports against any supported data provider. That's from memory though, and I only looked at it briefly on Wednesday, so I could be totally wrong.
#
August 20, 2004 5:20 PM
Darrell said:
I would have to say it depends on the laptop. I abandoned the very same thing last year as I couldn't get any of the XP multimedia drivers to be stable on my Toshiba Satellite Pro 6100.
#
August 21, 2004 1:32 AM
Chad Smith said:
Thanks for the heads-up and time taken to blog this. I believe this will be of use to me in the near future.
#
August 21, 2004 3:31 PM
K.Senthil Kumar (Sensoft2000) said:
Wow. Very nice article..
How microsoft allow this?.
#
August 22, 2004 2:18 PM
Veerakumar said:
Hi,
It is really super. Thank you for ur information.
Regards,
msveera
#
August 22, 2004 5:29 PM
Noah Coad (MS) said:
That is great info Jeff! Did not know about being able to step outside the box like that. I've always used my own "Config" class with "Save" and "Load" methods that serialize/deserializes the class' properties. My big deal with the Configuration namespace's ability is the lack of even notification. Frequently I'll have settings that several modules of code need to track. I use the property/event design pattern so they can subscribe to change notifications. But! This is to change in v2, now one can subscribe to config setting changes (though its not easy).
Cheers!
#
August 23, 2004 12:29 PM
Scott Galloway said:
Cooperate and use common sense I'd say. If you app server has hung that's a classic case of needing to cooperate to find a cause. Same goes for your DB connection pool running out of resources, the underlying cause could be either code, DB, Drivers or any combination thereof. Trying to work independently on such issues is just asking for trouble.
#
August 25, 2004 2:06 AM
Eric Wise said:
Been there done that.
The flagship moment was when an exe was executing on the task schedule on the main server and was running as administrator. When they changed the password for the administrator account it broke it. Course they didn't notice for weeks, by then they'd forgotten about the password change. Took me a whlie to figure out what happened but I ended up doing most of the troubleshooting since it was "my program" and the network administrator was exceedingly unhelpful.
Similar experience with DBAs, they always try to blame the code first, even if said code was previously stable. I also have a major peeve with dba's that don't let programmers have admin rights on test servers. I worked a contract once where anything I needed changed in the test database I had to work through a busy dba. Sometimes my development would come to a grinding halt b/c I couldn't move forward without something, and the guy wouldn't have time until "tomorrow".
In the end, I think that whoever is in charge needs to be very clear that a problem is a problem, and discourage finger pointing. I've experienced alot of "us vs them" for programmers and dbas, and it's extremely wasteful of time and resources.
#
August 25, 2004 2:55 AM
Jeff said:
I found this (devAdvantage) which I like much better than FXCop. You can easily adjust the rules to do what you want so it isn't as annoying. the only downside is that it does not handle spelling yet. the really cool thing though is that it can do the renaming for you. definitely check it out.
http://www.anticipatingminds.com/Content/Products/devAdvantage/devAdvantage.aspx
#
August 25, 2004 5:18 AM
Jonny said:
int c = 0;
if(conidtion1){
statement;
stetement;
if(condition2){
c = 1;
statement;
}
}
if(c==1) statement; // again wont happen if condition2 is not met
Surely using a flag such as above would be far cleaner then using the goto statement. Really it would be more sensible to use seperate functions here but you get the idea. There is always an alternative to using goto and generally it is a good idea to use it unless you are being lazy.
#
August 28, 2004 4:21 AM
Darrell said:
Sometimes a blog post will spur action when an email will not. And it is still better than stopping listening without saying anything.
Plus, as the old saying goes, bad press is better than no press. :)
#
September 1, 2004 8:41 AM
-e said:
So this is what it feels like to be a Democrat? Dissent isn't good, eh? :)
Anyway, I posted because I want the show to improve, and as Darrell points out, sometimes the blog post course of action gets more response. The show has had some really good subject matter, but as I pointed out, the stuff "around" the show isn't as good as it once was. You are more than entitled to believe things are fine with the show as it is, and that Carl can tweak away to his hearts content, until it becomes nothing more than Carl Rocks (which, if you recall the Coding Slave show, is what Carl wants the show to become - and if that's fine with you, great, but at that point I will vote with my feet).
If I'm allowed to post book reviews (which seems to be heavily encouraged, and it should be), then why can't I review a radio show. After all, I'm not any more qualified to write a book than I am to host an internet radio show... :)
#
September 1, 2004 8:49 AM
Jeffrey Palermo said:
"why can't I review a radio show?"
You can. And you did. I never suggested that you should not have posted what you did. I just commented on it. And you've commented back. And we're participating in "freedom of the press". I'm not sure what the reference to being a Democrat is (and dissenting). I'm hardly an authority from whom to dissent.
#
September 1, 2004 9:19 AM
Sahil Mailk said:
Well, I have to say I liked the show better when Mark Dunn was the co-host. But I wouldn't turn it down with Rory either.
To be frank - yes I do see at times Carl having a big head and my propensity to listen to the guest rather than him, because he does do shows in wh