flat assembler
Message board for the users of flat assembler.

Index > Heap > About the 64bit 80x86 CPU extension

Author
Thread Post new topic Reply to topic
MCD



Joined: 21 Aug 2004
Posts: 604
Location: Germany
MCD
Most Fasm users prefer the AMD 64bit CPU extension. But to be onest, I don't like neither.

You see, the Intel 80x86 strucure is now about 26 years old, I I think that it really becomes fucking Exclamation time that the whole strucure is updated. I mean, there are much better concepts than the 80x86 one already on the market.
There is too much old stuff breaking down CPU speed or consuming power.

This unfortunately would break existens with nearly all machine code. And most compilers/assemblers (yep, unfortunately, EVEN Flat Assembler) should be rewritten. Be see it this way: Lots of lots of working places would be created this way! You must not think that this wouldn't hurt me in my inner Asm-coding heart too, but I think that it's really time for something new.

For example. the typicall concept of "1 powerfull CPU and 1 big RAM somewhere conected with a bridge" becomes inefficient when both become too powerful. I mean, CPU and RAM should be divided into smaller parts, each one consisting of a combination of a small CPU and a small RAM. The entire computer heart (microprocessor/controller) should be build several such units arranged into an array on a platine, say right angled or hexagonal, or something else.

_________________
MCD - the inevitable return of the Mad Computer Doggy

-||__/
.|+-~
.|| ||
Post 23 Dec 2004, 16:35
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Hear you, hear you. Since we're moving to 64-bit CPUs, we might as well have wiped the slate clean, and get rid of the x86 legacy. And for backwards compatibility, do a software emulator. It's not like the CPU won't be fast enough to emulate things that won't get ported to native 64bit code. But noooo, AMD had to go and ruin that, and lock us with the x86 architecture even longer, forcing intel to follow suit.

Btw, what's your problem with NTFS?
Post 23 Dec 2004, 16:49
View user's profile Send private message Visit poster's website Reply with quote
scientica
Retired moderator


Joined: 16 Jun 2003
Posts: 689
Location: Linköping, Sweden
scientica
>do a software emulator.
please hold 5 seconds while I fire up the emulator to emulate the 32-bit world needed to read this message... (seriouslty, 32-bit emulation isn't that slow, the 32-bit bins for Doom 3 demo runns even if it get's a little slow when many not yet dead victims get on the screen...)

seriously, I think backward compabillity is to strive for, even if it would mean going from 8086 to a fundamentally different mode (via enabling some OM ("Odd Mode") or SBMAGMCFM ("Stop Bitching Medival And Go Mega Cool Future Mode" Razz ) in the flag register).


OT: speaking for my self, you mean against NTFS besides it beeing a non-GPL propertary FS which is getting older and older and probaly wont get younger? Wink

f0dder you're free to choose what ever hardware platform/CPU you like, I think some people still use old VAX machines (and even older machines) - they work and do the job their 'owners' want them to do, and they does it good enought not to have them replaced by newser machines which might not even be able to "speak" with the aperature/machines connected to them. - You don't have to follow the main stream and get you're self an amd64 if you don't want one. Btw, wonder why soo many have gone from IA64 to x86-64/amd64 if it's that evil with the "current" structure...
Post 28 Dec 2004, 17:28
View user's profile Send private message Visit poster's website Reply with quote
MCD



Joined: 21 Aug 2004
Posts: 604
Location: Germany
MCD
Quote:
OT: speaking for my self, you mean against NTFS besides it beeing a non-GPL propertary FS which is getting older and older and probaly wont get younger?


It runs into this direction, but it is mainly because it is NEITHER very portable NOR very fast. There are better FS out there (linux has some newer FS which really impressed me)

Have you ever searched a specific file on a huge new hundreds of Gbytes big HD with NTFS file-system? You can make yourself something to eat before it has found it. In this matter, the old styled FATs where better...

_________________
MCD - the inevitable return of the Mad Computer Doggy

-||__/
.|+-~
.|| ||
Post 28 Dec 2004, 18:03
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Quote:

please hold 5 seconds while I fire up the emulator to emulate the 32-bit world needed to read this message...

Needn't be that slow at all, and loading the VM would mostly be a first-time penalty. Back in the days of NT running on Alpha systems, some x86 applications would run faster in emulation mode on the Alpha than on native CPUs - of course mostly ones that were heavy API (like GDI) intensive, but still - not a bad accomplishment!

Quote:

f0dder you're free to choose what ever hardware platform/CPU you like

If AMD64 hadn't been introduced, there would have been the possibility of a better common desktop processor. AMD has ruined that with their 64bit patch on the aging x86 framework.

Quote:

Btw, wonder why soo many have gone from IA64 to x86-64/amd64 if it's that evil with the "current" structure...

The IA64 (Itanium) was marketed as a server machine, and intel made a number of design mistakes with Itanium, like hardware x86 emulation instead of software emulation - this takes up a lot of die spaces, and makes the CPU even more espensive.

It's because of people like you, that see nothing wrong with AMD's move, that we will be stuck with x86 for even longer. My next machine will probably be an AMD64 because it's the best performance/price ratio, but at the same time it annoys me that I could have gotten better hardware for the same cash. It's not just about the CPU and instruction set, it's the rest of the system hardware as well. Yes, technologies like APIC, PCI express and SATA have made the situation somewhat better, but PC hardware is still pretty cruddy.

Quote:

It runs into this direction, but it is mainly because it is NEITHER very portable NOR very fast. There are better FS out there (linux has some newer FS which really impressed me)

It might not be the best filesystem, but I think it's a bit much calling it "fucking shit". At least it's rather stable (I haven't had any data loss even though I've had crashes and powerouts), and the speed is okay. Perhaps not as fast as, say, ReiserFS, but it runs reasonably. Would be nice seeing a fully working windows port of Reiser or some of the other formats, btw...
Post 28 Dec 2004, 20:37
View user's profile Send private message Visit poster's website Reply with quote
MCD



Joined: 21 Aug 2004
Posts: 604
Location: Germany
MCD
Quote:

It's because of people like you, that see nothing wrong with AMD's move, that we will be stuck with x86 for even longer. My next machine will probably be an AMD64 because it's the best performance/price ratio, but at the same time it annoys me that I could have gotten better hardware for the same cash. It's not just about the CPU and instruction set, it's the rest of the system hardware as well. Yes, technologies like APIC, PCI express and SATA have made the situation somewhat better, but PC hardware is still pretty cruddy.

I fully agree with you, f0dder. But I think that even Intels Itanium is still stuck to close to the x86-architecture. Like already said, this would require some _BIG_ recoding of lots of applications.

Introducing a completely new CPU architecture would also speed up programs. Moor's Law has lost some of it exactitude in recent 2 years, so this could speed everything up a bit.

Another think for speeding up computers would be to have better compilers, especially high-level compilers still having problems with MMX/SSE, what theoretically means wasted power/performence because they are only seldom used!
Post 28 Dec 2004, 20:46
View user's profile Send private message Reply with quote
scientica
Retired moderator


Joined: 16 Jun 2003
Posts: 689
Location: Linköping, Sweden
scientica
f0dder wrote:
Quote:

please hold 5 seconds while I fire up the emulator to emulate the 32-bit world needed to read this message...

Needn't be that slow at all, and loading the VM would mostly be a first-time penalty. Back in the days of NT running on Alpha systems, some x86 applications would run faster in emulation mode on the Alpha than on native CPUs - of course mostly ones that were heavy API (like GDI) intensive, but still - not a bad accomplishment!

Well, it will allways be relativly slow (compared to a "generic" machine operating a the same speed as the emulating machine), as you emulate, but given, a much faster/"better" CPU that emulates an "inferior" machine (say an +3GHz CPU emulating a 26 MHz Z80 CPU, to be extreme). But in the end I guess most users probably wouldn't mind having the emualtion software swapped out most of the uptime, right? (it's only an annouance with the "user notable slow down" - most of us for get the "20 second kernel compile speed up" the freed RAM gives (<-- (probable) unclear reference to a discussion on a "good" value for the kernel swappiness)). And 5 seconds isn't really any time compared to 5 days of running a program, it's just a matter of conveniency/lazyness Smile

f0dder wrote:
It's because of people like you, that see nothing wrong with AMD's move, that we will be stuck with x86 for even longer. My next machine will probably be an AMD64 because it's the best performance/price ratio, but at the same time it annoys me that I could have gotten better hardware for the same cash. It's not just about the CPU and instruction set, it's the rest of the system hardware as well. Yes, technologies like APIC, PCI express and SATA have made the situation somewhat better, but PC hardware is still pretty cruddy.

Sadly you're right about the best solution (general hardware/software) not always (well, rarley historical speaking, even if I personally belive in a "software-darwinism" in the OpenSource movement, guess I'm alone in the world about that --flame me, flame me-- Razz (ie "the best/optimal solution survives") ) beeing the one that is the one used in "reality". (hope you don't missunderstood the "sadly", I ment it as "to bad the best don't win" not "sadly I have to agree with you" Smile)

f0dder wrote:
It might not be the best filesystem, but I think it's a bit much calling it "fucking shit". At least it's rather stable (I haven't had any data loss even though I've had crashes and powerouts), and the speed is okay. Perhaps not as fast as, say, ReiserFS, but it runs reasonably. Would be nice seeing a fully working windows port of Reiser or some of the other formats, btw...

True, perhaps we shouldn't spark flamewars with "Humor for the convinced of side A which sparks hatred with side B".
Well the ReiserFS drivers are open, so I guess it wouldn't be that had for them to implement it-- oh wait bummer, "forgot" about the license... but it might be hope for all you who still use Windows (ahem, still will buy the latest version):
http://www.namesys.com/v4/v4.html wrote:
Yet still we choose to give our software away for free for use with software that is given away for free (e.g. Gnu/Linux). Since we don't have a lot of illusions about our ability to entirely change the world, and it is amusing to sell free software, for those who do not want to disclose their software and do not want to give it away for free, we charge a license fee and let them keep their improvements to our software without sharing them. These fees help substantially in allowing us to survive as an organization.

We don't make nearly as much money as we would from charging everyone for usage rights, but we do make just enough to get by, and that is important.Wink We don't really feel that everyone should follow our example and make their software no charge for most users (it is too hard to survive fiscally doing this), but we do think that everyone should disclose their source code, and no one should design their software to exclude working with other software (e.g. Microsoft's Palladium which makes such a mockery of Athena).


Also, I don't belive yet a new processor would bring "tit ripping"(didn't ol' Hutch-- sayd that long time ago?) speeds in the blink of an eye - new solutions, new problems. Btw, wonder how much the G5 architecture (PPC? Motorola? seriously, what's the propper name for that arch?) could grow if price fell enought?


btw, just out of curiousity, f0dder, have you tried firefox 1.0 and thunderbird 1.0? I find them wonderfull apps, both in linux and on windows machines (I (along with more and more) use firefox in school on their WinXP machines). What are you impressions/disappointments about them?

_________________
... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself.
- Bradley Kuhn
Post 28 Dec 2004, 21:27
View user's profile Send private message Visit poster's website Reply with quote
madmatt



Joined: 07 Oct 2003
Posts: 1045
Location: Michigan, USA
madmatt
Hi Everyone,
scientica, Maybe you can answer some of your own question above to help me. Does FireFox work with winxp sp2? do I have to uninstall Internet Explorer? Have been a bit weary of installing it, not knowing what booby traps microsoft might have set for it.
Post 29 Dec 2004, 00:58
View user's profile Send private message Reply with quote
rea



Joined: 14 Nov 2004
Posts: 92
rea
It can work you dont need uninstall explorer (altough I dont know a way...) you only with the time will not use it more Smile, if MS do some for get down, is a bad way of competition and that should cause legal problems to MS...

get it http://www.mozilla.org/products/firefox/


By the way, if MS do a x thing for stop x people for use the product of his choices, then is passing over your rights Smile.
Post 29 Dec 2004, 01:37
View user's profile Send private message Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Quote:

But I think that even Intels Itanium is still stuck to close to the x86-architecture.

I don't know Itanium well enough to really comment on that - I did however believe that it had loads of registers, "instruction bundling" and other very non-x86 features? Oh, and that it doesn't do runtime optimizations (out-of-order things etc) and thus depends more on a clever compiler and/or assembly programmer?

Quote:

Like already said, this would require some _BIG_ recoding of lots of applications.

And a good software emulator (ie, JITer) for legacy code. There's already been done impressive things year ago (eg on the Alpha). With a lot of subsequent runs and code analysis and performance data gathering runtime, the code can be refined on each run. Same goes for "virtual machine" languages like .net and java (x86 code can be viewed as just another VM, albeit with a stupid design Smile).

Quote:

Another think for speeding up computers would be to have better compilers, especially high-level compilers still having problems with MMX/SSE, what theoretically means wasted power/performence because they are only seldom used!

Yeah... part of that is because the x86 design isn't really that good for compilers, I guess. The CPU shouldn't include any instructions that can be done faster with a number of simpler instructions. And of course there should be "advanced" instructions like what MMX/SSE/SSE2 offers.

Would also be nice to introduce more and better (more precise) hardware timers, *properly working* DMA (and enough channels that memcpy could be done using DMA), et cetera. x86 hardware tends to suck in these aspects Wink

Quote:

Well, it will allways be relativly slow (compared to a "generic" machine operating a the same speed as the emulating machine)

"Speed" is a relative thing here, and is certainly not measured just in megahertz - you have to take into consideration the cycle-counts of instructions. And if we're comparing vastly different hardware, there's other factors too (bus and memory speeds, etc etc.)

Let's see... the really important and time-intensive applications would of course be ported to the new hardware, to make the most of it. Legacy programs would thus mostly be "somewhat older" things. A "new machine" would probably run legacy code slower than native code, even with a good JITer. But if the JITed code could achieve the performance of, say, a 1.5GHz pentium3, that should be enough to keep legacy code happy, don't you think?

As for the slowdown on VM start... have you tried running a 16bit dos program or a 16bit windows program under NT? Doesn't really take long to start, does it? Both of these launch a form of VM. As for JITting legacy->native code, even the JITter in Alpha-NT used a caching scheme, so it didn't have to re-translate all the time. This is many years ago... can probably be done even better today.

Of course things for the VM would be kept in filesystem cache until the memory was needed for something else ("unused RAM is wasted RAM"). Could also be a good idea to keep the VM loaded even after the legacy code is done (to make further legacy program launches even faster) - perhaps unload it after "some time period" where no legacy code has been run.

Quote:

Sadly you're right about the best solution

...And if AMD hadn't come up with their 64bit extension to the already clumsy x86 patchwork, probably nobody else would - other companies don't have the cash, and intel really did seem to want to abandon this old tech. This *could* have been the thing that would have led us to a world of better utilized processors Smile

I'll try finding a link to a very interesting and informative article I read about the JIT system in NT for the Alpha - it was impressive, especially considering how old the stuff is Smile

As for FireFox, it's okay. I miss some of the functionality that the AvantBrowser (using the IE engine) gives me, but I guess most of this could be added with plugins. I like the standards-compliance (even if this means some websites look wrong - there should be a "IE compatibility mode" Wink), and text-selection works better on CSS-intensive sites than in IE.

However FireFox is slower - this is very visible while scrolling, and it takes perhaps a half second longer to load than IE. This isn't much, but it's noticeable when you're used to launch new browsers often. And this slowdown doesn't go away after the first launch. I blame their use of HTML/XML/Whatever for the UI, and their (over?)use of JavaScript for settings and whatnot.

Also, it seems that FFox does't like heavily nested tags without close-tags (like the <FONT> madness output by the first version of my gfx->html convertor Wink). But an okay product, which I'm currently using since I for the first time in my XXX years of internet explorer finally got hit by an exploit.

I'm using PMMail2000 for mail, and I kinda like it. Didn't check thunderbird since 0.9 or whatever, and I didn't like it. Was too slow and bloated for my like... oh well.

I hope they make mozilla leaner and meaner, and fix the scrolling slowness. It might eventually become better than IE Smile

MadMatt, FireFox works fine on XP SP2 and can co-exist peacefully with IE. It's up to you which one you want to use as your default browser. And microsoft doesn't set booby-traps for other applications... they've actually gone (too much) out of their way to ensure compatibility with people who write utterly retarded code... just read some of the microsoft employee blogs Wink (google for "old new thing")

[edit]
Think I found the article about the FX!32 technology... http://www.realworldtech.com/page.cfm?ArticleID=RWT122803224105
Post 29 Dec 2004, 01:49
View user's profile Send private message Visit poster's website Reply with quote
scientica
Retired moderator


Joined: 16 Jun 2003
Posts: 689
Location: Linköping, Sweden
scientica
madmatt wrote:
Hi Everyone,
scientica, Maybe you can answer some of your own question above to help me. Does FireFox work with winxp sp2? do I have to uninstall Internet Explorer? Have been a bit weary of installing it, not knowing what booby traps microsoft might have set for it.

yup, it works finr in school (where they've got WinXP SP2 - not putting my personal oppion about their choice...)
You don't need to uninstall Ie (in fact, relly you can't, unless you want to cripple your system - wonder if there really is eny difference between "iexplore" and "expler".
At worst you'll need to uncheck a check box when (for some reason IE lauches, eg, for windows update), the "bugg me to death with asking if you want to use IE as default browser" - in firefox at fist start you'll get a similar prompt, just answer yes (and let firefox check if it's the default browser). Note, I'd recomend installing the user agent switcher extension/plugin - some sites are terrible broken and wont work unless beeing told a lie that the browser is IE.


Quote:
Let's see... the really important and time-intensive applications would of course be ported to the new hardware, to make the most of it. Legacy programs would thus mostly be "somewhat older" things. A "new machine" would probably run legacy code slower than native code, even with a good JITer. But if the JITed code could achieve the performance of, say, a 1.5GHz pentium3, that should be enough to keep legacy code happy, don't you think?

Hopefully, but then again, emulation isn't for die hard gaming fans -- so as you said ""Speed" is a relative thing here", so I guess that would be ok for most "common" apps during a transition time (ie wiating for the next release optimized/compiled for the new platform).

Quote:
As for the slowdown on VM start... have you tried running a 16bit dos program or a 16bit windows program under NT? Doesn't really take long to start, does it? Both of these launch a form of VM. As for JITting legacy->native code, even the JITter in Alpha-NT used a caching scheme, so it didn't have to re-translate all the time. This is many years ago... can probably be done even better today.

iirc that VM wasn't by far 100% -- many games wouldn't run (ie do "illegal operations"), or at least not that I recall (speaking of old DOS games). But maybe that was a "design choice" in favour of speed/code complexity?


Quote:
I'll try finding a link to a very interesting and informative article I read about the JIT system in NT for the Alpha - it was impressive, especially considering how old the stuff is Smile

Sounds interresting, hope you find it. Smile

Quote:
As for FireFox, it's okay. I miss some of the functionality that the AvantBrowser (using the IE engine) gives me, but I guess most of this could be added with plugins. I like the standards-compliance (even if this means some websites look wrong - there should be a "IE compatibility mode" Wink), and text-selection works better on CSS-intensive sites than in IE

It might be sites thinking that non-IE can only handle the most basic HTML, I've read an article about such sites on www.spreadfirefox.com some time ago (near the 1.0 release), try installing the User Agent extension and select MSIE and try the sites again.
Btw, could you specify some of the features you're missing? (I might know what plugins/extensions you need/could try).

Quote:
However FireFox is slower - this is very visible while scrolling, and it takes perhaps a half second longer to load than IE. This isn't much, but it's noticeable when you're used to launch new browsers often. And this slowdown doesn't go away after the first launch. I blame their use of HTML/XML/Whatever for the UI, and their (over?)use of JavaScript for settings and whatnot.

Odd, most of the imes I've heard firefox is faster than IE, but well, no biggie. Also, the UI is made with XUL (XML User Interface Language, http://www.xulplanet.com/ ) - I've heard "guess who" have "invented" (just as the vitrual desktop manager and the double clicking *cough* *cough*) XAML (or what ever they call it, from the description it sounds really similar to XUL) -- hope there will be a reall W3C standard for it soon.

Quote:
Also, it seems that FFox does't like heavily nested tags without close-tags (like the <FONT> madness output by the first version of my gfx->html convertor Wink). But an okay product, which I'm currently using since I for the first time in my XXX years of internet explorer finally got hit by an exploit.

just a side note, <font> is deprecated as of HTML 4.0 (or was it even in 3.2?)
But then, if some one writes a site with (m)any uncolsed tags it's either a non-finnished, broken or bad written app - and as of XHTML sanity in the web has been give new hope Wink

Even if I'm not 100% happy with firefox my self I'd say it's the best "general purpose browser" I've ever used this far. elinks/lynx might be the best text mode, but it's dependning on the purpose.

I see you found the article, I'll read it after dinner Smile

_________________
... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself.
- Bradley Kuhn
Post 29 Dec 2004, 17:13
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Quote:

iirc that VM wasn't by far 100% -- many games wouldn't run (ie do "illegal operations"), or at least not that I recall (speaking of old DOS games). But maybe that was a "design choice" in favour of speed/code complexity?

Indeed it isn't 100%... VESA linear framebuffer modes aren't supported, only some devices are emulated, etc... but it's good enough to get most stuff running, for badly behaved software you'll need a better emulator (virtualpc, vmware, dosbox). IMHO the OS doesn't need to (nor should it) support severely broken software.

Quote:

Hopefully, but then again, emulation isn't for die hard gaming fans -- so as you said ""Speed" is a relative thing here", so I guess that would be ok for most "common" apps during a transition time (ie wiating for the next release optimized/compiled for the new platform).

Well, die hard gaming fans would have to wait for native versions of their games. The most popular games would probably be ported relatively quickly. For the remainder of the games, hopefully emulation would be fast enough. Considering that a lot of games are GPU rather than CPU bound, this might very well be the case... but of course then there's badly designed games like DOOM3 that are very CPU bound because of bad design decisions Smile

Quote:

t might be sites thinking that non-IE can only handle the most basic HTML

Nah, more like sites not using proper HTML/CSS and fonts or stuff. Like www.lunarstorm.dk . Works fine in FireFox, but looks somewhat different.

Quote:

Btw, could you specify some of the features you're missing? (I might know what plugins/extensions you need/could try).

After grabbing a bunch of plugins, I think the two things I miss most are... the close-button should send to tray rather than exit the browser, and it should remember (and re-launch) open tabs on exit. Both of those should be easy to code, I guess, but I don't have time nor interesting in learning how to do plugins, at least not right now. Other than that, dunno... would be nice if it handled downloads and "are you sure you want to run" differently, more like IE - have default choice "no" (so an accidental enter/spacebar can't wreak havoc), and NOT having the "yes" button disabled for 2 seconds.

Quote:

Odd, most of the imes I've heard firefox is faster than IE, but well, no biggie.

It's noticeably slower in starting... or well, the very first launch of IE after a system boot is somewhat slow because of the way temporary internet files are stored, and it seems firefox does this a bit better.

Quote:

just a side note, <font> is deprecated as of HTML 4.0 (or was it even in 3.2?)

Dunno, but it's still widely used (especially for stuff like PHP-generated pages) - even if CSS makes general management a lot easier. It's no biggie that it chokes on the deeply nested font tags, after all they're not closed - just a bit weird.

Oh well, at least with the plugins firefox is pretty usable, and my remaining gripes can probably be fixed with extensions. It's sucky that it can't use MS-java as I refuse to install suns crappy java on my box (unless I'm going to do cellphone coding Razz), but fortunately I only need it for my webbank. I hope they'll improve loading speed and scrolling speed soon... and that they fix all the "unknown" exploits before firefox is so widely used that it becomes an attractive attack vector Smile
Post 29 Dec 2004, 19:56
View user's profile Send private message Visit poster's website Reply with quote
scientica
Retired moderator


Joined: 16 Jun 2003
Posts: 689
Location: Linköping, Sweden
scientica
f0dder wrote:
Quote:

t might be sites thinking that non-IE can only handle the most basic HTML

Nah, more like sites not using proper HTML/CSS and fonts or stuff. Like www.lunarstorm.dk . Works fine in FireFox, but looks somewhat different.

Surprised lunarstorm Surprised Razz (dunno how many either have been using it in school - only that we how haven't are a minority...)

f0dder wrote:
Quote:

Btw, could you specify some of the features you're missing? (I might know what plugins/extensions you need/could try).

After grabbing a bunch of plugins, I think the two things I miss most are... the close-button should send to tray rather than exit the browser, and it should remember (and re-launch) open tabs on exit. Both of those should be easy to code, I guess, but I don't have time nor interesting in learning how to do plugins, at least not right now. Other than that, dunno... would be nice if it handled downloads and "are you sure you want to run" differently, more like IE - have default choice "no" (so an accidental enter/spacebar can't wreak havoc), and NOT having the "yes" button disabled for 2 seconds.

I know there was an old plug in for having tabs auto restored - but it was lost/died prior to 0.8 Sad the closest now is bookmarking all tabs (using the checkbox) and then the open in tabs menu item in the saved bookmark folder. Well, actually I *love* that delay - you can't install two plugins at once, you ahve to wait for the previous 2 sec delay befoer the next one pops up (good thing to prevent "install floods").

f0dder wrote:
Quote:

Odd, most of the imes I've heard firefox is faster than IE, but well, no biggie.

It's noticeably slower in starting... or well, the very first launch of IE after a system boot is somewhat slow because of the way temporary internet files are stored, and it seems firefox does this a bit better.

don't forget that parts of IE is loaded during system boot, and if you got a jpeg image as background I'd guess most of IE is loaded (the "hid the loading time trick' - that was a feature that was enabled in mozilla some time ago, but I know the firefox branch removed that dead pice of code (which only loaded all of the app at first start and then kept it loaded))

f0dder wrote:
Quote:
just a side note, <font> is deprecated as of HTML 4.0 (or was it even in 3.2?)

Dunno, but it's still widely used (especially for stuff like PHP-generated pages) - even if CSS makes general management a lot easier. It's no biggie that it chokes on the deeply nested font tags, after all they're not closed - just a bit weird.

Yeah, sadly php it self (by default) used <font> for some error messages! (which is horrible, I think they've fixed it in php5 - haven't tried it yet, gotta remember to unmask it some day and test it.). I'm a member of those who say "malformed code shall be punnished by errors" - as you've probably guessed, I like xhtml, maybe I'm a bit sado-masochistic Wink Very Happy

f0dder wrote:
Oh well, at least with the plugins firefox is pretty usable, and my remaining gripes can probably be fixed with extensions. It's sucky that it can't use MS-java as I refuse to install suns crappy java on my box (unless I'm going to do cellphone coding Razz), but fortunately I only need it for my webbank. I hope they'll improve loading speed and scrolling speed soon... and that they fix all the "unknown" exploits before firefox is so widely used that it becomes an attractive attack vector Smile

Well, I think suns java is better (even if I try to avoid java, quite a space waste for now), but that's just me, maybe MS own java is better on windows (ie gives "greater access to windows internals") . btw, as for scroling, not sure what you mean, but check the settings and look at the scrolling settings (some might be under the advanced pane).

_________________
... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself.
- Bradley Kuhn
Post 29 Dec 2004, 22:06
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Quote:

don't forget that parts of IE is loaded during system boot, and if you got a jpeg image as background I'd guess most of IE is loaded

I dunno how much of IE is really loaded - I don't have active desktop enabled (.bmp wallpaper), and I don't have webview loaded... so the HTML/WEB stuff of IE shouldn't be loaded at boottime... haven't checked, though. And it doesn't really matter that much... 2nd time launch of IE is instantaneous, while Firefox still takes some time - this means more processing is done at startup (windows is pretty efficient at filesystem caching).

Quote:

Well, I think suns java is better (even if I try to avoid java, quite a space waste for now), but that's just me, maybe MS own java is better on windows (ie gives "greater access to windows internals").

SUNs java is more standards-compliant (doh Wink) and supports more of java (again doh, since MS froze the VM because of suns bitching). I'm looking at it purely from a speed, size and efficiency point of view. Suns java VM is slow and bloated, Microsofts runs a lot better.

Quote:

btw, as for scroling, not sure what you mean, but check the settings and look at the scrolling settings (some might be under the advanced pane).

"Smoothness" of scrolling - and I'm not referring to IE "smooth scrolling" setting, but rather the CPU power spent on the scroll operation. This is very smooth in IE (feels like they're just blitting a bitmap), while it's substantially slower in Firefox; I can clearly feel the difference on a 2.53GHz P4 with a r9600xt graphics card: this means sloppy code (they might be re-rendering the HTML while scrolling?). It's most easily seen if you press the middle mouse button to enter that "funky scrolling mode"...
Post 30 Dec 2004, 02:59
View user's profile Send private message Visit poster's website Reply with quote
scientica
Retired moderator


Joined: 16 Jun 2003
Posts: 689
Location: Linköping, Sweden
scientica
f0dder wrote:
Quote:

don't forget that parts of IE is loaded during system boot, and if you got a jpeg image as background I'd guess most of IE is loaded

I dunno how much of IE is really loaded - I don't have active desktop enabled (.bmp wallpaper), and I don't have webview loaded... so the HTML/WEB stuff of IE shouldn't be loaded at boottime... haven't checked, though. And it doesn't really matter that much... 2nd time launch of IE is instantaneous, while Firefox still takes some time - this means more processing is done at startup (windows is pretty efficient at filesystem caching).

I think that means IE isn't fully unloaded (just killing the windows, kleeping it in memory - probably not swapped out...

f0dder wrote:
Quote:

Well, I think suns java is better (even if I try to avoid java, quite a space waste for now), but that's just me, maybe MS own java is better on windows (ie gives "greater access to windows internals").

SUNs java is more standards-compliant (doh Wink) and supports more of java (again doh, since MS froze the VM because of suns bitching). I'm looking at it purely from a speed, size and efficiency point of view. Suns java VM is slow and bloated, Microsofts runs a lot better.

ok, so a feature for speed trade off, well, i guess it depends on the needs of the applets one uses. btw, "because of suns bitching" - pardon me, but didn't Microsoft violate an agreement (and got sued and convicted for it so MS made their C# to "kill" java?) - and thus, it's more approperiate with "because of Microsofts bitching"...

f0dder wrote:
Quote:
btw, as for scroling, not sure what you mean, but check the settings and look at the scrolling settings (some might be under the advanced pane).

"Smoothness" of scrolling - and I'm not referring to IE "smooth scrolling" setting, but rather the CPU power spent on the scroll operation. This is very smooth in IE (feels like they're just blitting a bitmap), while it's substantially slower in Firefox; I can clearly feel the difference on a 2.53GHz P4 with a r9600xt graphics card: this means sloppy code (they might be re-rendering the HTML while scrolling?). It's most easily seen if you press the middle mouse button to enter that "funky scrolling mode"...

well, I don't know the internals of the scrolling so I donät know how it's done. But for me the middle button is used for a gpm feature (the "select and midle click pastes", you either love or hate it, I like it Smile) - so I don't make use of that scrolling

_________________
... a professor saying: "use this proprietary software to learn computer science" is the same as English professor handing you a copy of Shakespeare and saying: "use this book to learn Shakespeare without opening the book itself.
- Bradley Kuhn
Post 30 Dec 2004, 15:09
View user's profile Send private message Visit poster's website Reply with quote
f0dder



Joined: 19 Feb 2004
Posts: 3170
Location: Denmark
f0dder
Quote:

I think that means IE isn't fully unloaded (just killing the windows, kleeping it in memory - probably not swapped out...

It's not there, task manager does not lie. Of course it's kept in the filesystem cache, but every OS does that.

Quote:

ok, so a feature for speed trade off, well, i guess it depends on the needs of the applets one uses. btw, "because of suns bitching" - pardon me, but didn't Microsoft violate an agreement

The speed is bad enough that an Athlon700 can't play a simple java-based card game... "oh fear, it has animation!" Smile. As for the MS vs. SUN case I don't know exactly what happened, but my understanding is that Microsoft simply added some extensions and this pissed sun off. But I'm probably wrong and I didn't follow the debate very closely - all I know is that the end result was that windows users only have a crappy VM available.

Quote:

so I don't make use of that scrolling

Try setting it to the normal "fancy scrolling" temporarily, you should be able to see what I mean Smile
Post 30 Dec 2004, 15:58
View user's profile Send private message Visit poster's website Reply with quote
rea



Joined: 14 Nov 2004
Posts: 92
rea
mmm, I have a comment about java Wink, I hava actually a Athlon K7 650Mhz with 128Mb and actually can run the applications that I have finded of java, included the netbeans the last one... also it was running on mandrake Wink that you know have Linux installed is bloat bla bla and suchs things, having open other things and net beans was opened, sure there where a little retard, but after that no problems..., where is the app that you say that a athlon 700 cant open? Let me try it Smile.

What I can say is that Flash applications they really consum a lot of processor for the things that they do... you know normally only put some fun graphics there...
Post 30 Dec 2004, 17:45
View user's profile Send private message Reply with quote
Display posts from previous:
Post new topic Reply to topic

Jump to:  


< Last Thread | Next Thread >
Forum Rules:
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You can attach files in this forum
You can download files in this forum


Copyright © 1999-2020, Tomasz Grysztar. Also on YouTube, Twitter.

Website powered by rwasa.