Hacker News new | past | comments | ask | show | jobs | submit login
Follow-up on the 'Firefox v3.5 fiasco' (asp.net)
29 points by soundsop on July 11, 2009 | hide | past | favorite | 18 comments



https://bugzilla.mozilla.org/show_bug.cgi?id=501605#c121

They're not removing the stupid tmp-walker, just keeping it from being called on XP (most of the time, maybe?)

They seem to be deluded that it's super important to get more entropy than everyone else (TCP, IE, etc.) on Win2k and WinCE.


Actually they are keeping it as a plan-B for older releases of Windows that can't generate good random numbers.

Keep in mind there is no IE 7 or 8 for Win2K. They should be praised for supporting those users.


If the default entropy source on those platforms was actually inadequate in any way, it would have the everliving fuck exploited out of it, at levels far below their product. Nobody gives a shit about HTTPS seed quality if the entire TCP stack is compromised!

They should be excoriated for delivering a massive (and pointless) performance regression to their users with the slowest computers. Every time they pull shit like this, it just means that fewer users will upgrade, not just now but in the future.


It's actually the opposite; you use TLS exactly because you're not supposed to trust the TCP stack.

But your basic point holds. There's lots of Win2k deployed in sensitive production locations, and if they didn't have secure random numbers, we'd have bigger problems than Firefox.


I thought I used TLS because I didn't want my sensitive data going over the wire in plaintext.


"we'd have bigger problems than Firefox"

It's not like win2K is super-secure...

Seriously, it _really_ scares me when I see medical equipment running any Windows (not only win2k or earlier) being networked in hospitals.


Here's a nickel, kid. Get a real computer.


Here's some info from - https://bugzilla.mozilla.org/show_bug.cgi?id=501605#c98

Here's the results of cold start time; the average of at least 3 tests. Windows XP SP3, Pentium4 2.8G, 1.5G memory

Firefox 3.5 : 34.3 sec Build 1 (file scan completely disabled) : 19.7 sec Build 2 (CSIDL_INTENET_CACHE disabled) : 20.0 sec Build 3 (CSIDL_COMPUTERSNEARME disabled) : 19.7 sec

As you can see, your builds are faster than 3.5, and I guess the difference between builds is the measurement deviation. Both 3.5 and each builds can warm start within 1 or 2 secs. There's no significant impact with antivirus on/off on my side. (Avast! 4.8) Hope this helps.

Even without file scanning it takes around 20 secs !!


...for a 6 year old PC.


On my 5-year old pc, ff 3.5 boots from cold cache in ~4 seconds.

During those 4 seconds, the machine is capable of: Executing more than 8 billion instructions. Reading sequentially trough more than 20 gigabytes of memory. Reading more than 240 megabytes from disk.

If pretty much anything takes more than 10 seconds to load on any hardware even approaching modern, it means it was written badly.


On a machine that's not constantly being defragged, temporary files will be spread out all over the place. No way can you read 240 MB of fragmented files in 4 seconds after a cold boot. Simply not possible on consumer-level magnetic spindle media.


> On a machine that's not constantly being defragged

Windows does "constantly defrag" in the background.


I'm really not sure why I'm being downmodded. "Constantly defragging in the background" is exactly the meaning of the option that you can see listed in XP's TweakUI, labelled "optimize hard disk when idle."


So it's disabled by default?


No, it's enabled; in XP at least, TweakUI is the only place you can go to disable it (besides, perhaps, some Group Policy setting).


Well... Looking into my disk and how badly it fragments, I see Windows is doing a lousy job.


This machine has never been defragged.

Yay for xfs.


"It's all too easy to simply close the eyes and ignore problems reported by perhaps a minority of the users by cooking up excuses for not dealing with them, but that's not the solution: the problems won't go away by ignoring them, the vocal minority might actually be representing a big non-vocal group or worse: a big non-vocal groups of ex-users."

This point about usability generalizes very well and applies to many more software design issues than the issue at hand.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: