Frans Bouma's blog - All Comments

re: The Firefox 3.5 fiasco

FransBouma — Thu, 09 Jul 2009 20:29:12 GMT

@mac user: it's a normal font? I wasn't aware that this page uses a stylesheet with a font that wasn't available anywhere else. (I based the style on an existing style of this blogengine here at asp.net). I'll check if I can change it.

FF 3.5 seems to suffer from the slowdown delay only on windows indeed. Which is makes it even more weird why on windows NSS uses these odd file-system based randomization. Anyway, I installed v3.0 over 3.5 and everything is back to normal

re: The Firefox 3.5 fiasco

A Mac user — Thu, 09 Jul 2009 19:46:01 GMT

FireFox 3.5 seems just fine on the systems I use(WinXP and MacOSX).

Btw, why do you have that awful Windows font on your page? Did you know that for those(like me right now) that don't have that Cambria or what-ever it is called, the page is unreadably compressed.

re: The Firefox 3.5 fiasco

Bryan — Thu, 09 Jul 2009 15:30:04 GMT

Wow. Just, wow. And you know what? I just started Firefox and watched it in the Resource Monitor. Sure enough, it appeared to touch every single damn file in the fonts folder.

This is a completely amateur mistake. How can we trust a security system that is built by somebody making mistakes like THIS?

My God...

re: The Firefox 3.5 fiasco

Jeff — Thu, 09 Jul 2009 14:11:56 GMT

No such problems on the Mac version. There was some weirdness caused by the Google toolbar, but uninstalling and reinstalling that cleared it up.

Weird though that they'd end up taking a step back like that.

re: The Firefox 3.5 fiasco

Mark — Thu, 09 Jul 2009 13:18:48 GMT

I hate to sound like an MS fanboy, but if you haven't used it yet, IE8 is a great offering. It's much faster than FF3 (not sure how it compares to FF3.5), has some nice additions compared to IE7, and the developer tools are great! They're *almost* as good as FireBug, but are a lot easier to use.

re: Linq: Beware of the 'Access to modified closure' demon

FransBouma — Thu, 02 Jul 2009 06:29:51 GMT

@Mike: It doesn't reset the reference, as Where returns a new IQueryable<T> :) so after an iteration of that loop, 'customers' _was_ an IQueryable<Customer> and still _is_ an IQueryable<Customer> however, the contents of that object has changed: it now contains a call to Where() with as input the filterlambda and the original customers IQueryable. It's a fluent interface :)

re: Linq: Beware of the 'Access to modified closure' demon

Mike — Thu, 02 Jul 2009 00:18:42 GMT

@Frans

Ok, maybe I'm missing something else about the code then. I know that the query execution is deferred until the 'customers' is enumerated, but doesn't:

foreach(var search in searchCriteria)

{

customers = customers.Where(p => p.Country.Contains(search));

}

just keep resetting the customers reference? It doesn't look like it is appending to a list of customers there.

re: Linq: Beware of the 'Access to modified closure' demon

FransBouma — Wed, 01 Jul 2009 17:17:18 GMT

@Mike: that's because of the 'deferred execution' of a linq query: it's executed when it's enumerated. So when the query returns an enumerable, it's executed when it's enumerated, otherwise it's executed directly. Yes, this sucks, but that's how MS designed it, unfortunately.

re: Linq: Beware of the 'Access to modified closure' demon

Mike — Wed, 01 Jul 2009 14:12:25 GMT

I don't understand this statement:

"As there are no more values, the loop ends and the query is executed."

This must be something I don't understand about Linq...why isn't the query executed during every loop interation? Why is it that the loop ends and then the query is executed? If after the Linq statement, you did:

customers.Count, wouldn't that only work if the query was executed inside the loop?

re: Linq: Beware of the 'Access to modified closure' demon

FransBouma — Fri, 26 Jun 2009 09:34:11 GMT

@Think: that's what I mentioned in the first line of the post.