December 2004 - Posts
Today I moved from Visual SourceSafe to subversion, an open source Version Control System for a large number of platforms.
Having used CVS in the past with tortoise I was intrested Subversion especially as tortoiseSVN make the client's control very simple wit it's functionality.
The tools are very powerfull and there are lots of them things like steppign back revisions, creaing and apply patches, browsing modified Files and thier approiate changes, reviewing the repository via IE and Branching too. I am very impresed and I will deffinatly be using this for some time now. And there is even a subversion book.
Although there is a problem (apperently) where Web Projects have to be converted to a Class Library. This doesn't really bother me as work around is very simple and actually sorts out some gripes I have about the Web Projects and thier need to be associated to IIS. Once converted the projects actually load in th VS.net a great deal quicker too.
Thumbs' up to the subversion team :)
Now playing: Pink Floyd - Money - Us And Them
I have been using the WebService.htc to call .Net Web services directly from IE. This utility works perfectly for creating very dynamic (client side) data driven applications (Chubby Clients). Though recently I have noticed that if numerous calls to the Web Service to retrieve and render large lists of Data the response of the client starts to lagg and the memory usage of IE grows very high until the page is reloaded, although the page is not actually any larger. I have discovered that this only occurs when I am creating each item in the list as a htc it self.
I have created an IE Memory leak demo to show the effects of this problem. I gather that the way I am removing the htc's is causing the problem. I am using element.innerHTML = ""; in the example but i have tried the removeChild() method also but i still get the same memory increase.
Has anyone else seen this problem?
N.B. It will only render in IE as MOZ doesn't support htc's
UPDATE:
Having posted this I discovered a post on MSDN Mag reffering this is exact problem!
http://msdn.microsoft.com/msdnmag/issues/04/06/WebQA/
Apperently there is a Fix by altering the registry but only for Internet Explorer 6.0 SP1. by adding this key :-
HKEY_CURRENT_USER\Software\Microsoft\Internet Explorer\Main\Cleanup HTCs
Changing this does work, shame it's not set by default!
According to MSDN the only other solution is to refresh the page!
http://support.microsoft.com/?id=309170
I was just reading a post on Scoble 's blog about a guy on Joel Spolsky's fourm, where he was unable to get a job because he didn't have a blog!
It turns out that the job was working on blog software and I aggree that if the guy had no experiance of writing his own blogs he's possibily not the best man for the job. But are bolgs becoming a new job requirement and should we be including the URL on our CV's to give possible future employees a better understanding of who you are?
It's also made me realise I have to start bloging more as i havn't made the best use of it yet. At first I thought it would be a great place for me to store solutions to problems I spent alot of time on. Thus far I can think of a 4/5 topics that i could have added but was too lazy to add...
So my new years resolution is to make more posts! (though I hope i'll start before January!)
Andy