Forgot your password?
typodupeerror
Netscape The Internet

Netscape 6 Vs. 4.7x 364

Posted by Hemos
from the taking-the-time-to-load-things-up dept.
rafa writes "Linuxworld has an informal comparison between Netscape 6, Mozilla, Opera and Netscape 4.7 with focus on resource usage. It reflects what I've been experiencing with Mozilla." A lot of this is well known, but the article does a good job of bringing it all together.
This discussion has been archived. No new comments can be posted.

Netscape 6 Vs 4.7x

Comments Filter:
  • Perhaps you should check for yourself before you correct others; VC++6's 'Process Viewer Application' (haven't got a better one at the moment) lists 10 threads (in the 'Num. Threads' column) for NETSCAPE.EXE.
    In fact, one of 'em appears to be running at 'Time Critical' priority...

    -- Sig (120 chars) --
    Your friendly neighborhood mIRC scripter.
  • Even if that is the case, you have to provide me with an alternative to IE before I can reasonably be expected to switch. Netscape 4.x is not an alternative, because it crashes far too frequently - I tried to use it as my main browser for several months and finally got fed up and switched to IE. I'm certainly not going back. Both Mozilla nightlies and Netscape 6 are not alternatives for me either, because they load and run incredibly slowly on my Pentium II 266 with 96 MB RAM. I'm left with basically IE and Opera, both of which I use, but Opera doesn't render some of the more eyecandy-heavy pages (or will crash on them occasionally) so I'm forced to use IE for those. Unless Mozilla gets less bloated or I get a faster computer IE it'll stay I suppose.
  • Actually, the client (or proxy) sends an If-Modified-Since: header in the http request with the datestamp of the object in the cache. The server compares the datestamps and returns an HTTP 304 response if the client (or proxy) already has a current copy of the object, otherwise it proceeds with a normal HTTP 200 response.

    With HTTP 1.1 this whole process goes pretty quickly since it doesn't need to create a new TCP connection for each object.

  • You're right. "user" could mean "manager", "client", "boss", "spouse", or even "user" ;)

    But I think I got my idea across.

    Dave

    Barclay family motto:
    Aut agere aut mori.
    (Either action or death.)
  • Fullscreen browsing would be nice too... I can't express my disappointment that despite all the bloat, such a simple feature isn't there.

    BTW, if you do a side-by-side comparison of the screen real-estate of NS4 with text-only menus to the graphical icons of NS6, 4.7 still takes up more space (including the shortcut bar).

    Although the stupid double-thickness bar running along the bottom with no information in it (and no method to turn it off) brings NS 6 back into the lead of space wastage. You can turn off that stupid bar in 4.7 with CTRL-ALT-S.

    I can't believe I'm defending Netscape 4. I hated it when it came out. It was a bloated peice of buggy crap. The only saving grace was that IE4 wsa so horrifically unstable and would shred the OS.

    Now IE is quite stable, and comparitively lightweight. I still can't bring myself to use it, but it is certianly technically superior in almost every way.

  • by amccall (24406) on Thursday November 30, 2000 @07:45PM (#590043) Homepage
    More features? Or bad programming combined with bad languages?

    Years ago the WordPerfect for Windows 5.1 was released. A WYSIWYG word processer that ran with 8 mb ram, and about 40 meg hard drive space. The fact is: there is no reason that WinME should take 550 Meg alone on my hard drive.

    I'll grant you that features creep in, and users demand more and more and more features. But, newer programming classes fail to teach students some very basic important things. And what I speak of, is the often repeated line, that I've heard spoken in the classroom, and by a great many professional programmers, "RAM is cheap."

    Coding is taking less and less effort, not more. Any fool with a copy of Visual Basic or Visual C++, can create a passible text editor. Compare this to the days of hand optimized assembly, where one must stretch the processor beyond its current capabilies, getting every ounce of RAM out as possible. Intelligent, well thought out designs, were the only way you could create a solution that would run well.

    Now, Linux is one of the few enviroments where talented programmers have joined together to create something nice. In terms of requirements, the Linux world is moving at a much slower pace than most other industries. It is actually possible to runt the latest version of slack on a 486DX4-100 with 24 mb of ram, use an older version of netscape(or mozilla), and have things feel a bit slow, but the system be usable. This was the configuration of my Compaq laptop, which I used until I sold a week ago. :P

    But their is much software where, the often repeated statement, "RAM is cheap", pops up. Even in Linux. I find the whole situation disgusting myself. One should not justify not thinking fully through a program with this qualification. Clever algorithems, thoughtfull code, and interesting tricks are no longer allowed. Coding has begun to become something for the braindead. And the sad thing is, that many corporations will hire these pimple faced teen VB programmers that have no knowledge of algorithem analysis, and have little to know experience writing anything else than yet another Visual Basic Calculator.

  • MS Excel for Windows 95 Performance Whitepaper [microsoft.com]

    Being the original poster of this meme, I quote:


    Performance Enhancement Techniques

    To increase speed, Microsoft Excel developers started by identifying the most commonly used areas of the product, and re-wrote the code for the recalculation engine in assembler. Most of other improvements were made by writing more efficient routines. This process allowed them to decrease redundancies in the code. As a result, there is a more efficient level of software coding. These code refinements had a drastic benefit to speed, as it produced speed enhancements of over 100%, in some instances.




    --
  • Maybe what he means is that "whoever produces these programs is lazy", and the thing he misses is that it's not programmers who produces them, but the organizations they work for. And those organizations, taken as a whole, have many faults.
  • It is nice to have one ported framework you can write your cross-platform application for. However Netscape released a less than perfectly working framework with their browser. Not only is it still in the dev stage but it is needless overkill for a single application. The scope of their API is way too large for a single application, which is what spurned my gripe. I think a good component framework and layout engine that worked over multiple platforms would be cool (a la KDE and GNOME) but not for a single application. Its way too much overhead and the added complexity makes it harder to find problems in specific elements of the system. They've put a lot of effort into the GUI and whatnot but then Navigator is up to its old habits of sucking. They make a browser that barely browses.
  • I'm more productive on linux because it's such a pain to surf the web using netscape.
  • by divec (48748) on Friday December 01, 2000 @12:24AM (#590062) Homepage

    This is not a troll, I'd genuinely like to know people's experiences. I have been comparing it to Galeon v0.8 [sourceforge.net] and Konqueror 1.9.8 [konqueror.org] on Debian GNU/Linux on a 200 Mhz pentium with 32 MB ram.

    • speed. Opera seems to be slower than Galeon or Konqueror. On simple pages, or w3c-conformant pages, they're all tolerably fast. On complex pages, Opera seems to fall behind. After about half an hour of browsing, Opera starts churning the hard disk (presumably the swap partition, since this is with disk cache turned off). Neither Galeon 0.8 nor Konqueror suffer from this (although Galeon 0.7.6 did).
    • stability. When given lots of complex pages in succession, Opera seems slightly less likely to crash than Galeon 0.8, and slightly more likely to crash than Konqueror 1.9.8. This is based on the pages I have tried, YMMV.
    • w3c conformancy. Can't comment much on this; I've heard all three are pretty good. Certainly, all three are probably better than most of the web pages out there.
    • Internationalization. Opera is *terrible*. I have international fonts installed, but Opera doesn't appear to be able to render non-roman text! (Or maybe I just haven't worked out how to configure it). It replaces the Japanese, Greek and Korean on my page [freeuk.net] with blank spaces. The least I would expect is a question-mark! Galeon and Konqueror are both fine at this. (BTW Lynx (2.8.3) is really cool at this - give it a try! - it transliterates the Japanese and the Greek and certain Polish letters into letters which it can display).

    If other people's experiences are anything like mine, I don't see how Opera 4 for Linux sells. (Ok, I wouldn't buy it anyway because it's non-free, I just wanted to know how it compared to Galeon and Konqueror; but I couldn't see any technical merit in it either).

    Is there something which Opera is good at which I haven't noticed from the pages I read?

  • by goingware (85213) on Friday December 01, 2000 @12:27AM (#590065) Homepage
    I want to assert very firmly that the above was not a troll. I meant it very seriously and it is something that I have been discussing and posting widely on newsgroups and mailing lists for years.

    I was close friends with a carpenter when I was younger, and he told me that he arrived at a new job site one day and found the following sign posted at the entrance:

    If you don't take pride in your work you have no reason to be here.
    This was back in my bad old days of being a college dropout, hungry with no idea what I was going to do for a career. I told him I thought that would be a terrible place to work, the boss would always be bugging you to work harder.

    But my friend thought it was great and said he wished more construction companies would hold such high standards. It happened that this friend took great pains to always learn new skills, and he spent a great deal of money on tools, and always did his best to always have, not just the right tool for the job, but the most obscure tools right on hand so there'd be no time wasted running to the hardware store or doing it a more difficult way.

    And guess what? My friend was consistently among the highest paid carpenters for his level of experience. I haven't spoken to him in years but last I heard he's gone back to school because he wants to be a high-energy physicist. (This same fellow taught himself to program in x86 assembly after he bought a 486. I think it says something about his intellect and style that he chose to program in such a low-level language from the very start because it would be the fastest.)

    I believe in having the best tools for the software job too, and by this I mean not the machine - a fast CPU is handy but doesn't help that much; what does help is my personal tools - the skills, experience and insight. To that end I work hard to study and sharpen my skills.

    I spoke about that here just a couple days ago in Self-Training is Vitally Important [slashdot.org] as part of the discussion on What's the Best Way to Retain Trained Employees? [slashdot.org]

    I also discuss it in my article Study Fundamentals Not APIs, OSes, or Tools [goingware.com]. The gist of that article is that while you must study particular apis or tools to get work done, you shouldn't concentrate on or dive deeply into them but work to improve basic skills that will serve you well on any job.

    Perhaps one of the problems these days is the overemphasis on APIs and the lack of emphasis on the basics, like good coding style and efficiency. Two people who know a given API equally well will get dramatically different results if one of them is well-grounded in algorithm analysis as well as having a good understanding of how computers actually work.

    My comment about assembly code wasn't meant to say we should all start implementing our products in it. Rather, we should all learn and write some, and do some work with hand-tuning assembly code so that we have a good grasp of what the computer is doing when we write higher level code. Two books that discuss this pretty well are Gary Kacmarcik's Optimizing PowerPC Code [fatbrain.com] and Michael L. Schmit's Pentium Processor Optimization Tools. [fatbrain.com]

    While they emphasize assembly code they should give you enough insight into the actual functioning of your computer that it should make your higher-level programming more efficient. And I do mean to say that your overall code will be more efficient on any processor, not that you should hand-tune it for one particular processor at the expense of another as someone here suggested would be the result.

    A lot of people in this thread say the reason things have gotten so bad is because of pressure from marketing, management, clients or customers to add features and ship in a hurry. Yes, I acknowledge that such pressure exists and while they share responsibility you cannot blame them because that is their nature, much like the alligator who ate the frog after offering it a ride across the stream. (Frog? But frogs can swim)?

    I've been in this business 13 years and there has always been marketing pressure but code quality has not always been so bad.

    The quality and efficiency of your product is ultimately your responsibility as an architect and implementor. This is the case whether you're working in a well-funded dot-com or you're writing free software when you get the spare time.

    At every step of the way in your software development process, you make choices. All too often we (and I do include myself) take the easy way out and write bad or inefficient code. It is a far better life to live if we strive for excellence in our products, and to do so we must strive for excellence with every choice we make in our software development.

    I hope very much for the success of Linux and Free Software in general, but I think that it suffers overall from a severe quality problem. You may find this tolerable because you are a developer, but I'm a developer who has used lots of systems and personally I think Linux sucks as a development environment. It is no where near where it could be taken seriously as a desktop environment.

    Now before you curse me for criticising, you should know that I run Linux on two Pentium III machines (Slackware) and I'm going to add LinuxPPC to my Mac soon. This is, in part, because I want to work to make it better. But part of the way I am going to work to make it better, isn't just fixing things directly but also advocating that everyone should take responsibility for their code and make it the very best that it can be.

    My final word in this post is that if you want to get a good start on improving the quality of your work, read the Forum on Risks to the Public in Computers and Related Systems [ncl.ac.uk] also available on the Usenet News as comp.risks [comp.risks]

    Risks is a very well-moderated list that is frequented by some very serious and experienced experts on computer reliability, safety, fault-tolerance and public policy. But it is also often funny as your just as likely to see the latest UI bug in Word next to a problem with the control system in some nuclear power plant. It will give you a great deal more respect for the problems with computer code but there is also a great deal of discussion as to what can be done about it.


    Michael D. Crawford
    GoingWare Inc

  • Netscape would have been better served by enhancing the Mozilla preject in the key areas it is lacking (speed, bugs) rather than adding tons of useless marketing features.

    Netscape put a lot of developer time into building Mozilla and fixing bugs before branching Mozilla. (After branching, they mainly focused on stability bugs, and the trunk was somewhat untended for a while.) I'll grant that you can argue that Netscape should have put higher priority on performance than stability, or should have left out certain marketing features in the first release to keep ram usage down, or should have fixed bug 36283 [mozilla.org] before rtm. Please don't assume that Netscape didn't make any effort in these areas, though, or that marketing features are "useless" (they're a large part of how Netscape makes money from the browser product).

    Netscape _was_ a "champion" of OSS and a leader in the anti-MS compaign.

    I don't see why any of this makes Netscape any less a "'champion' of OSS". They're still contributing to Mozilla as much as they have in the past, and they're still showing the world that it's possible to create open-source software and generate revenue at the same time.

    --

  • why not, give away another broswer to gain market share, wait a minute, they have the market share! Guess they could do it out of the kindness of their hearts.
  • by ghost. (85872) on Thursday November 30, 2000 @09:45PM (#590072)
    "Coding has begun to become something for the braindead. And the sad thing is, that many corporations will hire these pimple faced teen VB programmers that have no knowledge of algorithem analysis, and have little to know experience writing anything else than yet another Visual Basic Calculator."

    Oh, that old gag.

    C'mon, don't people ever get tired of blaming VB for all the world's problems? And where are all these corporations employing armies of teenage VB grinders? How many Northwind DB apps can a single company need? :)

    I thought the issue was software bloat, specifically in light of Netscape 6? No one's using VB to write web browsers, operating systems, or most of the other crap that makes your WinME install take up 550MB of real estate.

    VB app developers are cut and pasting If-Else blocks to script GUI widget events. They're not supposed to have to know about algorithm analysis. If they did, they wouldn't be VB coders, they'd be programmers.

    Now, the folks who coded VB itself, it'd be nice if they knew how to code tight, efficient software...

  • 2. the fact you can't minimize the file download window, despite the fact this has been in bugzilla for more than a year

    You could probably say the same for most projects that use private bug systems, if you could look at their bug lists. So what? For both open and closed projects, plenty of "newer" bugs have been fixed, and new bugs can just as easily be severe or easy to fix as old bugs can.

    Whining about that kind of thing discourages other companies from moving to open bug systems. It also discourages developers from documenting known bugs that they think they won't be able to fix in a given product cycle, which hampers outside source code contributors who want to find features and bugs that other programmers aren't working on. Whine about a severe bug if you want to, but please don't whine about an "old" bug.

    --

  • Judging from the comments that have surfaced, I seem to be an oddity:

    Okay, I have a pretty hefty machine (PIII/256M), but for me NS6 works perfectly: it is far more stable than NS4.7x, layouts pages faster, is able to handle fonts correctly - now I can actually SEE some pages, supports SSL without a hitch... I really only use NS4 to run some Java applets that seem to hang the NS6. As a browser, NS6 is FOR ME far better than NS4.

    I might use Mozilla, but I am too busy to keep up with the nightly builds... I only use the Milestone builds, and NS6 offers more functionality than the M18.

    One other thing which I like about NS6 is the fact that I can - if I want to - to run NS4 at the same time. Very nice!
  • Coding is taking less and less effort, not more.
    You're right. But schedules are more and more insane to keep up with the competition. Developers aren't lazier. We do far more with far smaller teams than anybody could have done 10 years ago. We can do this because we can use tools with large, prebuilt APIs like Java that aren't always the most efficient or the fastest.
    Now, anyone who has ever done any development will tell you that you have to treat performance as a feature in any application. You can demand more performance optimization or design for performance, but it will cost you other features or it will cost you schedule or both. Please see any elementary book or course material on software engineering or tradeoffs in the software development process to help understand this.
    The tools of today do allow for what you brand as "laziness". But I believe that developers just have more options these days to meet the absurd demands placed on them by people who have no clue about the technology (whether these are PHBs, clients, marketing morons or what have you) who demand more and more other features in shorter periods of time.
    Anyway, I'm not trying to defend the Netscape engineers... they've pretty much fucked up in every way possible. And Microsoft is habitually responsible for massive code and feature bloat. But please don't generalize this to include all software developers.
  • I'd have to agree with you. I have been running Netscape 6.0 at home and at work and on a windows box and on a linux box. The linux boxes are a PII 400 with 172Meg, and Dual P233 with 128Meg of RAM. On the PII 400 it does not run that bad. The windows box is a PII 5000 with 196Meg of RAM. Yes it starts up slow, but it renders pages pretty fast. yes it does need some work, but I like the fact that it now is MORE standards compliant than most other browsers and this includes Mozilla and konquereor and opera. At some style sheets to your pages, or better yet go view home.nbci.com and see what happens under Netscape 6 when you move your mouse over the links. Netscape 6 and IE whill change colors from blue to red on all the links, Mozilla M18 and konquereor WONT!

    It is not just that netscape supports a:hover as part of the style sheets it is that it supports better style sheets than most other browsers.

    What the author fails to tell you about the download is yes it is 30 megs but about 20 of them are the JVM.

    When you consider that today we have machines that are over 1GigHz, and that having 256 megs of RAM is not that impossible, then Netscape 6 not bad.

    Lets take a look at some modern programs. Mac OS X 64Meg to install. Windows 2k 64Meg to install (afaik). KDE was not a small download. Neither is X. Granted they do more, but Netscape 6.0 supports many standards. Netscape isnot targeting those with old 486 or old Pentiums, they are targeting the AOL 6.0 market. At some point maybe AOL 7.0 they will include this beast. Hey my sister just got an 800Mhz machine from Gateway. This is more of a modern day standard than a k6-2 with 32Meg. Who has 32Megs on a machine today and plays quakeIII?

    Get real and stop bnashing netscape.

    I don't want a lot, I just want it all!
    Flame away, I have a hose!

  • There are three states.
    Eating, Sleeping, Neither

    So two bools are almost sensible. Since the Eating&&Sleeping state is illegal, it would be better to represent it with an enum anyway - the lack of enums is one of Java's less debateable bad design decisions. [Yes, I know there workarounds to this ommision but they suck].
  • Well, every time any of these three are mentioned, comments start about all three, so here's a few comments.

    First of all, IE is built into Windows nowadays. It loads faster. Period. If you think I'm wrong, why can't you uninstall it? Why is Microsoft in court? Some of the DLLs are used by regular Explorer as well as IE, so it is loaded at boot, and it's in your system's memory...

    With that in mind, opening the program is one thing, _*rendering*_ speed is another. As near as I can tell, they render differently, but loading a page seems close enough time-wise.

    Netscape 6 is based on an earlier build of Mozilla. Basically if you want all the Netscape features, use Netscape 6 and have a nice day. If you like the browser but don't care about the extra features (or you just hate AOL), go use some of the nightly builds of Mozilla, or at least the Milestones (which are gone and we're now looking at version numbers like .6 instead of Milestones for the future).

    Something else you might want to keep in mind is that Mozilla has that nice little Debug menu. It's got extra code, unless I'm mistaken, to help further development. Eventually the extra code will be removed, and at the same time things like memory footprint, load times, etc will be worked on making it *gasp* faster. Are we at v1 yet? No, we are not. Mozilla 1.0 won't be out 'til next year sometime. Which means it's lovely beta software. Beta software means bugs, beta software means slower than it could/should be. Netscape 6 is based pretty darn close to beta software.

    And while I'm rambling on here, does anyone know of any commonly used program that gets a new version but takes up less resources? Or that is faster to load? The more things the program can do, chances are it will take more resources than the previous version. The point releases address things like speed, memory, and of course bugs...

    Currently I use IE5.5 and some Mozilla nightlies at home. At work it's IE5 and Netscape 4.x. IIRC, Netscape 4.x still has some advantages over IE5.x: JavaScript. Are Netscape 4.x's strengths enough for me to use it regularly? No. Netscape 6? Not when I can use Mozilla... :)

    Bottom line (not literally): Use what works for you.

    That's enough ranting for one night. :)

    R

  • by QuantumG (50515)
    bah.. it's not a "good" browser. It's a bloated crash ridden piece of shit.. but god damn it, it's the best their is ('cept possibly for opera which is great if you don't mind 5% of sites not working).
  • Yeah, I agree with you. IE is a really good browser. I just wish we'll have as good a browser someday under Linux. Even if it would be Microsoft stuff, I would gladly use IE under Linux if there was such a thing.

    ---
    Guillaume

  • Hello,

    I am aware of the threading issue. If you read the article it does not actually add the memory from every thread. It only adds the the actual usage of the program plus the usage of the Java JVM.

    If I were to add the entire memory set, it would falsely show Netscape 6 taking over 80 megs of RAM all buy itself. That would not include the JVM.

  • Nah, why bother.

    If NS loses just a few more percent of the browser share (to IE or Mozilla or whatever) No Webdesigner will bother to make a special version for Netscape users. They will make pages, test them with IE and leave it at that.

    NS/Mozilla/Opera/whatever-users will have three choises, accept being shut out from more and more sites, Get IE as a backup and start it up everytime a site doesn't display or surrender to the power of MSFT.

    Soon only hard core geeks will use anything but IE. But as more and more services move exclusively to the web (and use IE only technology) even geeks will need IE.
    Oops IE doesn't work with "alternative" systems. Too bad. Pehaps you'll better get a copy of Windows after all?

  • "but why exactly did netscape make Java for M18 a separate module?"

    Netscape used to include their own vm. Doing so, they practically killed the whole concept of an applet because their VM was so crappy it would invariably crash after a certain amount of time. On linux, all you have to do to crash netscape 4 is load a few applets.

    Luckily netscape is no longer in the business of making java virtual machines. Instead they use an API that allows third parties to plug in their VM. Netscape 6 optionally installs jre 1.3.0 (from Sun), which is probably one of the best JVM's available. However, mozilla does not do so. To run Java anyway on mozilla you have to install it manually and *gasp* read some documentation that tells you how to do so. M18 and the mozilla nightlies are not intended for end users so I don't think that's a big problem.

    If you want a shrinkwrapped product, don't use the development version, wait for a release or download netscape 6. I wouldn't recommend the latter since it was released way to early and contains many bugs (many of which have been fixed in the nightlies already).

    That the development versions are usable and indeed much better than the released netscape 6, is nice. I think it is a sign of some good work being done by the mozilla developers. However, you shouldn't make the mistake of treating the development versions as release versions by expecting documentation and shrinkwrapped plugins and stuff like that.

    As a mozilla enthousiast I regret it that netscape released 6.0. I think they should have waited. You could argue that they had to release at some point but on the other hand they pissed off a lot of users by releasing this crappy excuse for a browser.
  • Reply I found IE is often the cause of netscape crashes. If you can install an original Win95 and apply service pack 1. This will not include IE. The run netscape 4.76 it should be very stable. After running that for a while install a new IE and watch netscape start crashing. You can not do this test on a new version of windows because they include IE. I think much of the stability problems with netscape on windows are IE. Under linux I have seen a different problems. Under linux netscape seems to do synchronous dns. So when a dns request locks up the whole app does. What I found fixed that and many other problems while making things run faster and use less memory is to use squid. I turn off the caches on all browsers I have netscape, konqueror, mozilla, netscape 6, and opera then set them to run through the squid cache. That gives me one big cache to share between all browers which is more efficient and it is very good at what it does. So in 12 megs of ram it does better at caching then netscape does in 16 or more. Also squid and reiserfs are very sweet together. There has been a lot of optimization between them. Thus browsers of any kind very rarely crash on my linux boxes and they run very fast.
  • What I do is set memory and disk cache to 0 on all browsers and use squid. That has worked a lot faster then any of the browsers caches every has. Since I have a raid I just made it a 400 meg cache. I have not ever seen a problem of loading old pages. It runs very fast. Also strangely enough netscape with no cache running through a proxy is more stable. One reason I think is that 4.x can't do async dns requests which squid will do for it.
  • yes.. but it's KDE! I officially announce the competition: if you can deliver HTML/CSS2/Javascript/Cookies (and any of the "essentials" that I have left out) in a browser under 1.4 meg that uses less than 8 meg of memory (total) and doesn't crash every 15 minutes (read IE) I will give you a few million for it. Oh.. and if it starts up and renders web pages faster than IE, that would be nice too. If you say it's impossible then I will have truely lost faith in humanity.
  • ...and you still haven't told us how you know that Excel's spreadsheet execution engine is written in assembly language. (Other than we all have to be idiots not to believe you.)

    People don't use assembly language for everything for a reason: it offers a low level of abstraction. This means the programmer has to keep more in his or her head. People--even skilled, uber-programmers--are not perfect, and the more you make them keep in their head, the more likely they are to screw up. The increased cost of screwing up more has to be worth less than the gain of speeding up your app with assembly language in parts.

    Cell computation in Excel is non-trivial, and implementing it in assembly would be tough. That doesn't mean Microsoft didn't do it, but it does mean that it isn't obvious that they did do it either.

    Nice diatribe though.

  • vectus there is right, that article wasn't fair. It wasn't fair in the least. This guy didn't bother to look at all the fun things that happen to memory when you open up E-Mail. He didn't look at what happens when you get into browsing newsgroups. How about starting up that composer huh?

    A fair comparison of these products would have included speed and bulk for all of it's key components. Heck, open up 12 screens of NS 4.7, get E-Mail going with a seperate window for Usenet, and a stack of pages opened for editing in it's composer. Approach the same with NS 6.0, if you can.

    3 years of development later, the last best hope for not allowing Microsoft to dictate the standards to be used on the web has resulted in this? 3 years of, "quit bagging on it, it's still in work" and this is it? Is this fair?

    In short, the article was not fair in the least. It took the one portion of 6.0 that actually doesn't suck gobs of memory, and completely ignored the rest of it. Hell, at least if they had managed to sell a few copies of it they could have helped paid for lawyers, guns, and money to fight MS with. At this point Mozilla sure as hell isn't going to win any battles based on product quality, regardless of the platform.
  • Both Mozilla nightlies and Netscape 6 are not alternatives for me either, because they load and run incredibly slowly on my Pentium II 266 with 96 MB RAM.

    Slow load I'll grant. Slow run is bullshit. (Posted from a Celeron 300 with 96 MB RAM)
  • That's what themes are for, of course. Download one of the themes out there with smaller buttons. Create your own theme with tiny buttons (grab a copy of Chameleon for this, perhaps...) Use the classic theme if you find that any better.
  • bahahaha.. now THAT is funny. A software company actually improving their product to get you to buy the upgrade? As if! That's not the way it is done in the industry baby. If you want people to buy the upgrade you shove in more features and more features and more features! Even if they don't use em they're gunna want em and that's the only way you can get em to shell out their cash! But don't take my word for it, just have a read of Microsoft's first brief to their appeal. It actually spells it out as a good reason why Jackson didn't know what he was talking about. Improve the product, heh, next thing you'll be asking for an operating system that doesn't crash.
  • Netscape, on all platforms, is "threaded."

    In Linux, on the 4.x versions, threads are "faked" by the Netscape Portable Runtime, and thus there really only is one "real" process.

    On WinX, threads are not "faked" (or rather, they are calls to Windows' threading system, which... well, could be called "fake").

    Mozilla now uses Linux/Pthreads (whatever they're called... multiple processes via the clone() call) instead of creating "lightweight"/"fake" threads in NSPR.
  • Its a total joke, slow slow slow slow. I browse the web weird, I like to run with 8 browsers at once. Not gonna happen with 6. And I have a dual celeron 400 and 256 megs of ram. If I keep under 3 windows, its fine, but past that, yuck. My ex-girlfriend's computer, which is a 466 with 64 megs of ram, can't even handle one window fo the bitch. It sends her load average up to like 2.3+ with nothing but that, afterstep, gaim, and X running, and a very minimal number of daemons.

    Did I also mention it doesn't want to let me see the edge of the ext window I am typing this in, and its handling of style sheets is so "random" that it is completely useless.
  • no.. I think it is a perfectly valid term.. I often build hotels on the top right corner of my screen and whenever IE lands there it looses wads of cash and very often is knocked out of the game.
  • Last I heard it was in Mozilla, but Netscape decided to leave it out for what ever reason.

    As the fella that reported that bug in the first place, let me clarify that point. It was a decision made many months ago at both Mozilla and Netscape to not invest time into LDAP. No point in just blaming big bad NS doing nasty things to the wonderful open source Moz.

    Initially the plan was to include LDAP, and every other feature that was present in 4.x. It was after all supposed to be an 'UP'grade. Apparently later on as folks were attempting to reinvent everything that appeared round developer interest just wasn't there for it. Of course, nobody outside of Netscape or Mozilla got wind of this. No, not on the newsgroups, the web page, or even the application itself. It was only about a month ago that they actually removed "New Directory" from that Address Book menu.

    I had first heard about LDAP not going in after reading an interview with some folks at NS following the first beta release of 6.0. At first I thought I must have read it wrong, so I got to posting on the mail-news newsgroup asking about this. Sure enough, complete no go.

    There was, and still is, a single Moz developer working on LDAP now, but only as a browser component. Nobody has worked in the hooks between what he's doing and the Address Book, which is put together with what is apparently a completely undocumented db format called Mork.

    Oh, and this is the really fun part. As I got to expressing my concern to the newsgroups and bugzilla it turns out that it's not being there is MY fault! Yes, apparently I needed to stop my life in it's tracks and learn to program in C, and how to interface with an undocumented db format. That, or try to convince my employer (who is not in the IT biz) to invest funds for an AOL project. All the while nobody at Mozilla was exactly advertising the fact that LDAP wasn't going in.

    Yeah, I'm still steaming on this one. Unfortunately, darn near every argument I presented to folks at Moz has turned out to be true. NS 6.0 has hit the streets, and corporations have left it sitting on the curb. Worse still, home users have either not noticed, or have been highly negative of it. If it wasn't for AOL press releases, is there any good press out there for it?
  • I can confirm that... I don't know where the performance "breaking point" is exactly with respect to speed/memory but I've got a celeron 333 w/64mb ram here, using dec 30's nightly of mozilla.

    It takes about the same amount of time to load as netscape 4 did (approx 10 seconds, probably spent loading all the XUL for the interface). Pages display as soon as they are received and the images just pop right in without delay as fast as my modem can download them.

    Even with norton systemworks, IRC and jabber in the background soaking up my meager memory, I don't get HD thrash from mozilla unless I open up over a half dozen windows with image and plugin-heavy pages.

    I have noticed that if you leave mozilla alone for a while, you'll get some HD thrash when coming back to one of its windows even if you haven't increased your memory usage through other apps enough to make the OS swap mozilla out. Mozilla seems to swap itself out on its own when it notices it's not being used.
    ---
    Where can the word be found, where can the word resound? Not here, there is not enough silence.
  • Netscape 6 is a piece of crap. But that doesn't mean that Mozilla is.

    Try a Mozilla nightly build. Seriously. It's not as unreliable as it sounds. I have only found a few nightlies recently that I wouldn't want to use, and in each case the next night's fixed the problems. Or if you're worried about the memory footprint, use Galeon.
    --
    Obfuscated e-mail addresses won't stop sadistic 12-year-old ACs.

  • IE is fast because it's already loaded on boot.

    It's not on NT 4.0, and definitely not when you specify IE to run in it's own memory space to avoid OS conflicts. Still pops up a heck of a lot fast than Mozilla nightly builds.

    DISCLAIMER: on NT I'm still using NS 4.7x as my primary browser. Just to be clear, this isn't an opinion of an IE zealot. Oh, and posted with Konqeror on FreeBSD.
  • bah.. IE is the standard, no matter what the UN wannabe's at W3C say. When you can't release a browser or publish a web page until it is W3C compliant, then you'll have something to hold onto, but until then we're stuck with Microsoft. And W3C sitting in their ivory towers essentially saying "if the web page is not compliant with this spec the browser should display an error and refuse to display the page" is not helping!

  • why don't you use Flash? At least that standard is controlled. You wont see anyone writing non-standard flash code, and why? Because it is generated from an editor. Oh wait, maybe it's because most people hate flash with a passion.
  • I actually do a large amount of development, all of it is inhouse stuff that no one will ever see outside the company, but I do program. And I've experienced deadlines breathing down my own throught.

    My point was not against Microsoft or Netscape, its a general trend that I see in the software industry that concerns me: the complete lack of disregard for good design. This doesn't necesarily mean performance. "Ram is cheap" - is only one particular, and well spread symptom of this. And your right, performance is a feature. But a good design, will include the consideration of performance. Many leave it to the hardware now.

    A good design will save you much time, compared to the typical high school hax0r who feels he can do anything on the fly without any planning, and again any elementary book or engineering course material will tell you that. I'm refering to programmers who have no understanding of what happens behind the scenes and use the tools with a complete disregard for how it will really affect the entire system in stability, performance, features, you name it.

    "Schedules are killing us" is no excuse for the disregard of engineering principles being shown in a great percentage of the software industry, if any thing, it should be more a reason for them.

  • Hmm, perhaps it's not slow run in the usual hard-drive-thrashing swapping of memory sense, but it definitely "feels" slow. At the very least it's less responsive than most Windows programs are - moving your mouse around menus, clicking, dragging, etc., should feel like you're performing the actions instantly, not having them lag behind your mouse slightly.
  • Actually, Mozilla/NS6 use floating point operations A LOT in the layout and rendering engine. Look at the source code if you don't believe me.
  • Konqueror (KDE2) is nice, but once you take the KDE bloat into account, using it just for the web browser is a bit much. And I've still had crashes from it.

    Just for fun, load up BlackBox for a window manager. Mainly just cause it's pretty darn small and all. Keep an eye on top or ps and load up NS 4.7. Okay, now close it down and start up Konqueror with all it's KDE bloat.

    On my FreeBSD box this takes up a wee bit more memory for Konq than NS. Not enough to really matter real world.

    As for the crashes, I've had a couple of those as well. Thing is, this browser is showing one hell of a lot more promise than Mozilla ever did, in far less time.
  • COnsidering that last stats from statmarket.com, Netscape only has 11.2% of the market share (IE has 88.5%)and is going down quickly... so I'd probably go out on a limb and say that most of the Netscape users are hard core geeks, or users of older systems which always had Netscape and they can't be bothered with upgrading yet... It's sad, but it's true ... IE won... and now with all the bad press Netscape 6 is getting, there's no chance in hell for NS to ever come back... Ah well, to be honest, it's not that big of a loss... *shrug* How many Linux users actually would use NS6 to Mozilla anyhow? eek...

  • Mozilla is, in a way, a very succesful open source project, which attracted lots of really talented outside contributors. And indeed, a lot of attention is being paid to reducing bloat at the moment.

    Unfortunately, a number of design decisions were errrr less than optimal. The XUL user interface language seems to have a big impact on performance. And leaving aside whether one likes the UI or not, the fact that it behaves different than other apps on any given platform also leaves a lot to be desired


    Ok, once and for all: If you want to fix all of the above AND have a wickedly fast, standards compliant
    browser with a tiny memory footprint, smooth scrolling, etc, etc- ie, if you want IE for gnome - do the following:

    (1) Download build 2000-11-27 (the best linux build so far) from http://people.redhat.com/blizzard/software/RH7/RPM S/i386/ [redhat.com]. Dont download it from mozilla.org because the installers there dont fetch libgtkembedmoz.so.

    (2) rpm -vvi what you downloaded in step (1)

    (3) Goto http://galeon.sourceforge.net/ [sourceforge.net] and download the latest galeon. Its a small download.

    (4) rpm -vvi what you downloaded in (3)

    (5) add the following to your .bash_profile: MOZILLA_FIVE_HOME=/usr/lib/mozilla ; export MOZILLA_FIVE_HOME

    You now have 2 browsers: mozilla in /usr/lib/mozilla/ and galeon in /usr/bin/galeon. Use galeon and never look back. Not only do you get gecko which is certainly the one thing mozilla did right., you get superior bookmark management, lovely themeable gtk+ and last but not least, you get a reason to install the only internet app that matters, the mutt email client.

    (Whatever you do, dont rewrite mutt as a gtk app that hooks into libgtkembedmoz.so to render html email. That's what microsoft does.)

    --

  • by srichman (231122) on Thursday November 30, 2000 @05:52PM (#590139)
    Uh, doesn't adding memory totals from top give an inaccurate picture of real memory usage because it counts shared pages multiple times?

    E.g., if I fork() (which is pretty much the same as making a new thread in the Linux world), all the text pages that are shared will be listed under the memory total for both the parent and child process in top. If I sum the memory usages, then, I'm counting the shared pages twice.

    Which can give an unfair appraisal of the memory hunger of a multithreaded program.
  • It's about time someone said that!

    I've been using NS6 since it was released and even though I hadn't used Mozilla for more than 20 minutes before then, I've downloaded nightly builds at least twice a month since... well, since nightly builds started, I suppose.

    other than what the author mentioned, 3 things bother me a lot:

    1. the amount of time it takes for a new window to open up, be that a navigator, composer or "new message" window.
    2. the fact you can't minimize the file download window, despite the fact this has been in bugzilla for more than a year
    3. the window positioning. new windows don't get put into the right places in Linux, and in Windows a new window is never maximized.

    Well, those are my complaints.

    Flavio
  • by vsync64 (155958) <vsync@quadium.net> on Thursday November 30, 2000 @05:53PM (#590143) Homepage
    Okay, now that my First Post! is out of the way...

    Quite honestly, I've recently decided that NCSA Mosaic is still the best browser out there. I've got an x86 box at work, so I downloaded the (statically linked, thank goodness) binary and messed around with it.

    Whoa. Nostalgia trip.

    Thing is, though, it's got so many things that newer browsers don't even bother with. Like making clicked links dashed instead of solid underlined. (Kinda relevant for those color-blind users mentioned a while back.) And allowing you to select fonts for each heading level (don't think it's in that version, but I remember doing that in Mosaic on Solaris). And letting you quickly flip between a fontset for the whole document.

    Oh yeah, and Mosaic is fast. Maybe that's because it focuses all its resources toward actually displaying HTML, and not trying to turn itself into some kind of sick Turing machine/security hole.

    A few updates would be nice... Cookies, SSL, maybe style sheets. But overall, I just wish people would bother to consider all the "accessability features" and "performance enhancements" that always existed, but were simply forgotten.

  • > * No right-click menu's on forms / the URL bar
    > this just happens to be very handy for copying /
    > pasting data and URLs.

    Fixed in the latest Mozilla nightly builds.

    > I was hoping to be able to give NS6 the URL for
    > a RSS file, and have the news channel displayed
    > in that oh-so-fancy side bar.

    There's a Web site that automatically creates a Sidebar panel from an RSS file. I can't remember the URL offhand. However, since you can do a lot more in a sidebar panel than you can with RSS (better layout, active controls), you're usually better off using one of the 700+ Sidebar panels or building your own.

    I know plenty of people who have SSL working in Linux.

    > * Tons of advertisements / plugs / commercial
    > junk. Even in the FILE menu!

    Not in Mozilla.

    > * Poor handling of DHTML / Javascript.

    Mozilla's support for DOM1 and DOM2 *standards* (not Microsoft's "extensions") blows away all other browsers, including IE5.5. There's a pointer to an independent test site in a recent Mozillazine article.
  • I disagree here. There's a piece of code where I work that has many repetitive code blocks and very poorly designed algorithms. One of the better programmers is currently completely cleaning it up. Each change he makes increases the readability of the code while reducing its size and memory requirements.

    The kind of programmers who write horribly obfuscated, but supposedly very tight code are usually ignoring better algorithms, the compiler's optimizer, or both. I, personally, would not let one of them be a part of an organization where I had some control who was working there.

  • by goingware (85213) on Thursday November 30, 2000 @10:30PM (#590150) Homepage
    Well, are you a developer?

    I'm typing this from my workstation at my client's office, a web application company (I've clocked out). I'm using whatever Netscape came from the Debian site when I updated my software on Monday when I started.

    Read about my laptop [goingware.com] which was my main development machine for most of the last year. It boots NT, Slackware and BeOS.

    You can read my resume [goingware.com] - note the long list of products I've shipped, and keep in mind I haven't been keeping that list up to date. See the long list of projects I've done in the two and a half years I've been a consultant.

    Note that among the jobs I've held was Senior Engineer at A Big Fruit Company [apple.com] where I held the role of "Debug Meister" - I did low-level debugging and in fact performance tuning of the Mac operating system.

    When a tester found that an application would crash under a new system build and they didn't understand what component was at fault, it came to my team (Traditional OS Integration, formerly known as the Blue Meanies). We would track down the bug and assign it to the right engineer or fix it ourselves.

    Note that sometimes, probably half the time, the bug was due to a third-party app bug, and we determined this purely by running MacsBug, an assembly level debugger, inside the app and system software. We had the entire Mac OS source code at our disposal but this wasn't usually readily available when you were visiting a crashed Mac at a tester's cube so you just had to know your MacBug.

    I use and contribute to open source. My latest effort was aiding the author of the ZooLib [sourceforge.net] cross-platform application framework in releasing his library under the MIT License; I worked with Andy Green for a year to test his code by developing a product with it and led a beta test of developers who also developed products with it.

    I found ZooLib to be an incredibly enlightening example of well-architected, efficient and compact code for what it does. Just using it and reading the source code increased my own abilities as a programmer and architect tremendously.

    You can read some of my thoughts on the business and technique of programming at GoingWare's Bag of Programming Tips [goingware.com]

    Linux is better than most as far as efficiency is concerned, but don't get me started about reliability and ease of use.

    As for what I think is a well engineered OS, try the BeOS [be.com] but you don't want to get involved with the company.

    Read why I think developers need to take back control of their lives from operating systems vendors [sourceforge.net] and how I think they ought to do it. If you really want the full-bore opinion, read The Cross-Platform Manifesto [goingware.com]

    So yes, I am in fact a developer, thank you. It's just that I maintain high standards and I like to encourage others to do so as well.


    Michael D. Crawford
    GoingWare Inc

  • I agree with some of your post, but you are making it seem like using RAM is evil. When you come down to a true time-space trade-off, I think it's fair to say "RAM is cheap" and go in favour of more space instead of more time.

    Granted, I wonder how many programmers actually know when true time-space trade-offs should occur. When I write something, I always think about how I could do it by streaming the data (if applicable). I tend to go from there.

    You should answer the "RAM is cheap" excuse with "RAM may be cheap, but it's also slow". Face it, with processor speeds up in the GHz and RAM bus speeds at best around 200-400 MHz (effective, not actual!), you should convince yourself of the hit going to memory is going to cost in the execution of your program unless you have decent caching. Of course, there will be many times that this a) does not matter all that much, b) cannot really be optimized.

    Woz
  • well I dont know.. when someone writes something from scratch the first version they release is usually a beta (and possibly an alpha) and even if that isn't the case it is either a 1.0 or a 0.x.. but here we have a product that is obviously beta and yet it has a version number of 6! Wow, when I see a product with a version number of six I generally expect it to not crash, not suck, not have a LOT of features that no one wants (yes, those are supposed to be removed when you're customers say "take out the sidebar, we hate the sidebar") and so the cry "it's beta" just doesn't cut it with Netscape 6 because it is version 6 of their browser. Crying beta is specifically the right of Mozilla.
  • by Anonymous Coward
    You can do the stuff you mentioned with a user-set style sheet. At least, you can in IE - not sure about NS or Moz.
  • He isn't adding memory totals. There's no way he could get two listings in top that add to only 25 megs for netscape. If you have only one window open, mozilla (M18, I think) shows up as one or more identical entries in top. Even opening other windows, I can't get mozilla to show non-identical sizes in top. Threads don't create a new process space, and no way is NS6 using less that 18 megs, so the reviewer is *not* adding entries in top. What I'm saying is that there isn't going to be somethign that says "mozilla: 10 megs; mozilla: 25 megs" and someone adds them. It's going to say, "mozilla: 35megs; mozilla: 35megs", and you say, "mozilla is using 35 megs".

    Mozilla is multithreaded, but it doesn't seem to create child processes, except for Java, and almost certainly for plugins. The 20 megs he reports for the Java component probably overlaps a little with the main executable, but I doubt more than a meg or two.
  • by god, did I say that (253932) on Thursday November 30, 2000 @10:39PM (#590161)
    Processes _also_ share memory except for (a) the insignificant amount consumed by a process slot; (b) copy on write data pages before they've been written, if they ever get written,.

    This is why you can have hundreds of apache processes (news readers, whatever) on a busy server without ever breaking a sweat or touching swap. Now, most large programs allocate a huge chunk of memory that they will manipulate for themselves in lieu of malloc(). That doesnt mean they're going to use (more accuately, write to) all the pages in that chunk! It does mean top will report those pages but so what, that isnt a measure of actual memory consumption.

    I'm not even considering pages swapped to disk which is a further "optimization" when you have 100 browser windows, all of them having seen active use over a 24 hour period but only 2 of them currently not minimized as icons.

    Modern VMs are a wonderful thing - stop being so stingy with imaginary memory.Open as many mozilla windows as you need, you wont be any wiser for it.

    (If memory request sizes matter to your os, your os is broken and or has some other serious issue that will probably require an immediate reboot.)

    This review is bogus. If you're running Linux, download a recent build (follow the notes link at mozillazine.org) and ket rid of Netscape 4.7. Mozilla renders pages almost as well as IE. I make heavy use of css on my site and all its pages render exactly alike on IE, Mozilla for Windows, Netscape 6 for Windows, and mozilla for Unix/XFree.

    (There's marginal differences with fonts across platforms but Mozilla on unix does a lot better with fonts than Netscape 4.7, that's for fscking certain. You cant specify font size in pt measure with Netscape 4.7 under XFree and not have the font look like it was rendered by an Apple II.)

    --

  • other than what the author mentioned, 3 things bother me a lot:

    <Bastard> Its open source, if these things bug you so much why don't you help fix them. Isn't that the way the model is supposed to work. </Bastard>

    Seriously, since I started using mozilla there have been a couple of times that I've come pretty damn close to getting the code to play with it myself. Haven't had those feelings much since I started to more or less live at work though.

    --locust

  • Netscape6 is just Mozilla with a few doodads. Thus, most of its source IS open. Also, since 90% of Mozilla development is done by the Netscape engineers, it's not fair to bash AOL-the-company and praise Mozilla in the same breath.

    Use Mozilla, be happy, but don't put the boot into Netscape. If AOL cut them off then Mozilla would be in dire straits.
  • Is it possible that pages being stored in memory (as suggested by the memory use going up as more pages are browsed) happened because the memory was available? Might it do something sensible on a memory starved system (like tossing the cached pages out of memory)?

    Of course, even if this is the case it is still not good - disk caching would be far better.

    Or better yet, stick with a less bloated browser.
  • Lets put it this way: not only has the barrier to entry of programming lowered, allowing those who don't know enough to care about the quality of their code to get jobs, accountability for software companies has also dropped. And will keep dropping if the UCITA becomes more widespread. So not only do you wind up with bloated, cruddy software, but (in some cases) its bloated cruddy software that could do serious damage (maybe even kill someone) without accountability.


    -RickHunter
  • That being said, Mozilla still sucks on Linux compared to W2K and solaris.
    You think Mozilla is better on Solaris? I have the opposite impression - though 95% of my experience with Mozila is on Solaris. Of course, my biggest complaint about Mozilla on Solaris right now is that PSM has disappeared [mozilla.org], so there's no way to look at https sites. Also, it seems that there are more problems with integration with CDE than with Linux window managers.

    --
  • by Alatar (227876)
    While I love top to death, isn't it a pretty crappy way to measure memory usage? I mean, top is just a quick-and-dirty tool, not one for serious analysis. Of course, the article writer claims that his tests are 'in no way scientific', and yet has no problem concluding his article by recommending against the use of Netscape 6.
  • From my experiences with Netscape/Mozilla in the Windows world, the results (aka Netscape 6 being huge and getting bigger and bigger and bigger) are consistant with what it does in Windows. After an hour of browsing, I can run a report on a 300k record database in Access 2000, chart something from a remote odbc source in Excel 2000, and be typing this very message in Opera (or IE5, I have a lot of ram leeway right now), and still use less ram then Netscape 6 takes to show the slashdot frontpage (and doing nothing else).

    It really is quite pathetic, Netscape must be in bed with a Ram maker.


  • MSHTML DLL 2,359,568 03-18-99 12:00a MSHTML.DLL


    I rest my case.

  • The thing I find most interesting is that the
    Mozilla nightlies are significantly better than
    Netscape 6. Even the ones from around the release date.

    NS4.75 is faster and smaller, but (for me) much
    less stable and it can't render pages at all well.

    I don't care how quick it is when it still crashes several times a day. (Under Win98 and Mandrake7.2)

    NS6 is pretty but sooo slow. I find it quite useable under Windows, but the Linux version just crawls.

    However, I'm finding the Mozilla nightlies quite wonderful - haven't had a crash for months, and although the startup time is terrible, once it's loaded it's quite useable.

    I still think there's room for a lightweight standards compliant browser *only* though. Galeon seems promising for that. Maybe a Galeon based on M19 will be what I want.

    Konqueror (KDE2) is nice, but once you take the KDE bloat into account, using it just for the web browser is a bit much. And I've still had crashes from it.

    - Muggins the Mad
  • by _xeno_ (155264) on Thursday November 30, 2000 @10:52PM (#590190) Homepage Journal
    The point here is that IE loads faster because part of it's load time is moved to boot time. It renders faster because it does render faster. (And this has nothing to do with kernel-space issues, it has everything to do with partial rendering support and Netscape's hatred of <TABLE>s.)

    Given the integration between IE and so many Windows things, it's impossible to say exactly how long IE really takes to load. (The UI libraries are loaded at startup, for example, because they're also used in a lot of other programs. As far as I can tell (not really knowing the internals) all the IE program you use does is create the widgets around the HTML render. I'd imagine the HTML renderer is loaded with the program, but I may be wrong - if you use Active Desktop, for example, it's used in that. It's also available as a COM object/ActiveX control, making it handy to use whenever as a Windows developer you need HTML support. Overall, I'm not sure exactly what IE preloads - but given that IE isn't also a newsreader, editor, chat client, and kitchen sink cleaner, it probably would load faster anyway.)

    Find a site with a large text or html index file, press back and then forward in netscape, wait as the status meter goes from 0-100% in the lower left. Try it with IE, no wait. The page loads instantly.

    Sounds like a caching issue - possibly because Netscape checks the server to see if the page has changed and IE doesn't. I dunno, it's late, and I can't test it right now. (Need to create a page and a server to test it - maybe IE keeps the page in memory and Netscape writes it out? Compare memory cache sizes if possible.)

  • As I followed the Mozilla Milestones upward I was truly hopeful that when Netscape 6 final came around, it would be a good product. Sadly, as is the case far too often with software today, it should still be in its beta stages. I mean lets be serious. This thing is an attrocious resource hog. It has many VERY annoying bugs. (Locks up a lot on the least bit challenging pages, like the ones at netscape.com(!!!), shows phantom new messages in the inbox, etc.) I truly like the format of the software, the mail and news interface is nice, but the bugs are more than I can handle. My main gripe is just the fact that they consider this a major release. Come on, you've waited this long since the last version, why distribute this now?
  • It's open source, if these things bug you so much why don't you help fix them. Isn't that the way the model is supposed to work.

    The open source movement should have got beyond the 'you have to be a programmer to offer an opinion' stage. It's impossible to be involved with everything that you might find in principle worthwhile.

  • I'm sorry, but I have to disagree with your basic premise. The focus of software development has changed radically in the last decade, and with good reason.

    Within reason, I don't waste my time (and I really did mean to say "waste") with hand-optimizing blocks of code in the applications I'm writing. Why? Because my employeer needs my code yesterday, and they couldn't care less about the relative beauty as long as it will do its job without eating the system.

    You see, my coworkers and I write the software that makes the company run. If a database query is sub-optimal but working, then it will still allow the business to make money. I'm not saying that this is something to brag about, but that's the reality of the New Economy, as far as I can tell. Time-to-market can make or break an entire corporation, so I guarantee you that my boss would rather have something that gets the job done then a shining paragon of hand-rolled assembler.

    Now, this isn't to say that we don't optimize later. In fact, every release of our major internal-use product is much faster, memory efficient, and CPU-friendly than its predecessor. Once our projects have hit the "it works!" milestone, we usually have the opportunity to go back and fix stuff that we knew was kludgey even as we wrote it.

  • This leads to the ridiculous situation that an old computer runs slower and slower as new software is loaded on it, until you finally have to buy a new one just to run at all.

    Oh, please. There is an extremely simple somution to your problem. If you want your computing experience to fully benefit from the latest and greatest hardware upgrades, the answer is painfully simple.

    Run older software on newer hardware.

    There is no law or mandate anywhere that says you can't run an older, stable OS on your blazing new Thunderbird system. You'll be absolutely blown away by the performance, I assure you. You can read e-mail, surf the web, serve web pages and write and compile code on a wide variety of older, rock-solid systems. If you want a fast, virtually bulletproof web graphical web browser, use Netscape 3; even better, stick to Lynx, if you don't mind seeing the Web in all ASCII. There's nothing that says you must use the latest C compiler; there's no real reason to upgrade beyond Pine for reading one's mail, as attachments can be saved and viewed externally. It is perfectly sensible to run older software on newer hardware, and I have seen it done quite a number of times.

    I'll say it again: run older software on newer hardware.

    Now, if for some reason, you want to take advantage of some of the extra features that more modern operating systems and programs provide, I advise you swallow your rhetoric and buy a modern computer for the job. Just as there is no reason for you not to run older systems on newer hardware, there is absolutely no reason for software developers to not take advantage of the power available to them. If, for some reason, you feel compelled to use a newer program because it provides functionality above and beyond what you can achieve with older counterparts, you can bet that program uses some section of the "thick layers of software bureaucracy" you pan. (Fire BAD! bureacracy BAD!)

    And finally a lot of people just don't know how to architect or code. I think we could all benefit from learning and writing some assembly, so we could really understand what our software is doing.

    ...you think that the answer to a lack of understanding of code architecture is simply to code in assembly? I assure you, merely coding in assembly will not somehow divinely imbue the ability to analyze and optimize code. If you have an idiot who writes a sort algorithm in C that takes N! time to complete, what magical aspect of assembly is going to impart the necessary skills to write it properly? Sound education in computational theory and systems architecture teaches people how to code well; all assembler does is let already bad coders take ten times as long to write code that's fifteen times less readable. Your archetypal pizza-faced, brainless VB coder will crank out absoulte crap for assembler, I assure you; in doing so, he/she will have taken several weeks to write the equivalent a single line worth of VB script, their program will crash and fail most gracelessly, memory will leak like a sieve, and the code itself will be so poorly written and organized that any attempt at fixing or upgrading it will be utterly hopeless. Assembly is not a cure-all.

    New features should not come at the expense of performance

    Do you ever actually read what you write? that's like saying education should not come at the expense of time! I challenge you--I bet you five hundred dollars--that you cannot write a new feature into any program that doesn't come at the expense of performance. How can somebody who purportedly codes in assembly spout such utter nonsense? Do you have any understanding whatsoever of the necessity behind the complexity of modern systems? Do you really believe that, if you just sat down and worked at it long enough, you could redesign the latest version of WordPerfect to be as computationally efficient as vi? Good freakin' luck.

    Why does this happen? One thing is because programmers are lazy, and if their code runs slow they assume the user will just get a faster machine.

    Programmers are lazy, and write slow, lousy code because of it? Well, who the hell wrote your fast, stable Slackware distribution? Wandering minstrels? Yes, there are more bad programmers in the world today; the demand for programmers is so great that even if ten times as many "skilled programmers" existed today, we still couldn't come close to filling the need. Does that mean, as you so blithely put it, that programmers are lazy? No. Some programmers are lazy, just as some Slashdot posters are trolls. There do exist a good many highly talented and educated computer programmers in the world developing very advanced, modern systems and applications; they understand the need for the complexity and abstraction of modern systems in relation to real-world concerns.

    If the needs of the computing world were performance and performance alone, your post would strike a resounding chord. As it stands, though, performance is but one factor amongst a sea of others, including things such as code readability, portability, robustness, maintainability, modularity, production time and production cost. If all you need out of your computer is performance, performance, and more performance, stick to older software and the hottest hardware. You can run a system at god-like speeds if you do this. Just remember, though, that the only reason you ever have to run that processor-sucking, machine killing modern system is if you want to use any of the wonderfully time and energy saving conveniences these disgustingly bloated travesties put at your fingertips. If you're unwilling to accept the fact that code becomes larger and slower in the name of ease of use and abstraction of complexity, though, well, you're out of luck.

    Do tell me, though, when you finish that port of WordPerfect.

    $ man reality

  • I suggest you take a look here [kde.com]

    Although it is not ready for prime time - it shows good signs of progress, and I expect the next release to have IMAP support.
  • Is Mozilla now the only thing standing between IE and total world domination!?

    Well, maybe not quite, but it's looking more and more that way. I would rather not use a Microsoft product, but I'm not rabbidly opposed to it. What I am opposed to is a browser monopoly.

    If Microsoft is allowed to hold on to a near monopoly on web browsers, they will hold too much influence on the future course of the web - influence that no single company should have. The Mozilla project is more important than ever, not to screw over Microsoft, but to keep competition on the web, forcing company's to turn to standards bodies and not simply do as they please.
  • just firing a shot over the fence :)
  • did you read my tag line? I know all about shockwave .exes in email. I must say I don't agree with that one bit. Just distribute a file format and make people get the player. All the time I hear people say windoze viruses don't spread because "people don't copy binaries anymore", and email attachments is what I always point at. But flash built into a web page is a common as java applets and becoming as common as javascript.
  • Heh, just in case anyone wants to experience the nostalgia...
    ftp://ftp.ncsa.uiuc.edu/Mosaic/Unix/binaries/2.7b/ [uiuc.edu]
  • by eam (192101) on Friday December 01, 2000 @03:50AM (#590218)
    You should never blame an application for crashing an OS. An OS should never crash as a result of some bug in an application. If it does, the problem is with the OS.
  • That could be worse ; I've seen a fresh engineer from a school which shall remain unnamed (okay, it's the ENSIMAG in Grenoble, France), who did implement (seriously !) bitfield operations.... using STRINGS !

    Yes ; that guy was converting an int into a string consisting of '1' and '0', then did things like
    if (a[i]='1') and not (b[i]='1') then
    r[i] = '1' else r[i] = '0';

    The worst part was that this guy applied something he said he had learned in school (okay, he had a much, much more competent colleague from the same place), and didn't see anything wrong there.

    Oh, yeah, that was a couple years ago, the language was Delphi.

    Needless to say, that guy was fired at once.
  • So what? You cant type fast enough or have enough different program names to type for this to be an issue. The incidence of new programs is an incidence measured in slow plodding human time.

    New programs are started infrequently in part because fork and exec is slow. Many things that would have been shell scripts have been done as small C/C++ programs, or perl scripts because "shell scripts are slow". Why are shell scripts slow? Because fork n' exec is slow.

    Now fork isn't hugly slow, and vfork is pretty fast (where it isn't just fork), even if it is a glatic scale kludge.

    exec , or execve is the slow one. Dog slow. Namei lookups, tons of MMU diddling, normally a lot of disk I/O. Oh, and we can't forget the cost of ld.so on a modern system. The dyanmic linking takes it's toll on process startup as well (of corse it can save time faulting in libc and libfoo, and it definitly helps memory starved systems).

    Personally, I find most software doesnt need threads and that introducing threads raises program complexity and bugs. Threads are a *difficult* abstraction.

    Same here. A lot of times select/pool/kevent, and sometimes an extra process or three are way simpler to debug. Most of the time even. But it is (many times) simpler to write threaded code then event loop code, but you can pay for it for a long time in debugging. Other times the threded code can run a whole whole whole lot faster.

  • Yeah... MSIE has a much smaller footprint then Mozilla ...

    According to MS, since the browser is integrated with the OS, I guess its footprint is, what?, 150 meg or so?

    note1: this applies only to the Windows version, YMMV w/Mac&Solaris
    note2: this only applies to an initial install, after several windows 'updates' this size has been known to grow to ~250 meg
  • heh.. they took out 6k of security bugs ;)
  • Another sub-optimal thing about Mozilla under the X Windows system is that Mozilla seems to set the backing store setting on all windows to NotUseful. This forces Mozilla to re-render any window that has had an Expose event (i.e. resorting the window stacking order, which means if you are a "open many windows and click around" type of guy like me you do a lot of re-rendering). If I weren't so damn busy at work, I'd pull down the sources and see if changing the settings would make Mozilla better or worse.
  • Not on linux you wont. IE may work okay with wine, but you'll take a performace hit with wine anyway. If you are using Linux then IE is NOT a choice.

    I don't want a lot, I just want it all!
    Flame away, I have a hose!

  • by {LF}Ceres (115933) on Thursday November 30, 2000 @11:33PM (#590241)
    But their is much software where, the often repeated statement, "RAM is cheap", pops up. Even in Linux. I find the whole situation disgusting myself. One should not justify not thinking fully through a program with this qualification. Clever algorithems, thoughtfull code, and interesting tricks are no longer allowed.

    Well answer me this: Which is more expensive nowadays... an experienced programmers time and a quicker time to market or RAM and CPU? It's not that "thoughtful code, and interesting tricks are no longer allowed" it's more "is it necessary to squeeze every bit of performance out of a piece of hardware anymore?" and "how much more will it cost to get a 5% increase in performance?". Back then when performance optimization and "tricks" were used more often, i'm sure programmers were REQUIRED to squeeze the performance out of the hardware because hardware cost so much. Now? How much money does it cost per meg or ram or HD space? How much power can u get for a $200 CPU?

    Performance optimization to a certain extent is necessary, but as another poster said "good performance is treated as a feature" and it's a feature that is (unfortunately) pretty low on the list in todays world.

    Coding has begun to become something for the braindead.

    How is this a bad thing? I mean really, doesn't having more programmers around allow more opportunities for really good programmers to come about? The barrier of entry for programming has lowered, but to be really good at your job (and to get paid accordingly) you still have to be very good at what you do.

    Ceres
  • by srichman (231122) on Thursday November 30, 2000 @06:11PM (#590244)
    In general, this is true: threads share a memory space and other things that processes do not. A process is an independently executing instance of a program; a thread is a sequence of execution within a process.

    In the Linux world, though, process creation (fork) and thread creation are both built on top of a Linux-specific system call called clone() that has arguments that specify what you resources (file descriptors, data pages, etc.) you want shared between the new and parent process/thread and what you want duplicated. This is why, in Linux, threads have their own process IDs and show up separately in process listings. And in top.
  • by goingware (85213) on Thursday November 30, 2000 @06:16PM (#590277) Homepage
    Even while our friends at Intel, Motorola and IBM do the most amazing things to speed up computer hardware (and don't forget our friends at Adaptec with the blazing 29160 SCSI Ultra160 Host Bus Adapter), programmers consistently work harder year after year to steal from the end user the gains that they might otherwise have from purchasing new hardware.

    This leads to the ridiculous situation that an old computer runs slower and slower as new software is loaded on it, until you finally have to buy a new one just to run at all.

    It's not just that you have the perception that your computer of old is running slower than the new computers because it was less zippy when you bought it, but because the regressive performance dehancements of operating systems and bloated applications really do make your computers run slower.

    Note that I used to run SlackWare and Apache on a 100 MHz 486, serving up web pages (admittedly with a light load) while I used X at its console - and it worked fine. But when I loaded Windows 95 on it it was dog slow. There's no question of running Windows 98.

    I had a 233 MHz Pentium II with 32 MB of Ram that I ran BeOS 3 for Pentium on. It worked great - I shipped Spellswell for BeOS Intel with this. But when BeOS 4 came out and they switched to Elf format, I had to upgrade to 96 MB because I couldn't run a compile and read my email at the same time.

    Later I installed a near-final beta of Windows 2000 server on this machine. I intended to use it to develop a Java GUI app under CodeWarrior for Windows. To get the machine to run at all - not even running CodeWarrior - I had to add another 128 MB of RAM for a total of 224 MB. The machine was dog slow even after the memory upgrade.

    There is no excuse for this. New features should not come at the expense of performance, and each new release of both operating systems and applications should be both faster and take up less space, not more. If substantial new features have been added then there may be cause for a little more code size but certainly not what we see in practice, such as what was listed in the Netscape 6 review.

    Why does this happen?

    One thing is because programmers are lazy, and if their code runs slow they assume the user will just get a faster machine. But friends, the user wants to buy fast hardware so they can actually run fast, not just so they can run at all.

    Pressure to ship a commercial product makes managers fail to support efforts to do substantial performance tuning, especially tuning that is not localized but would require substantial rearchitecture.

    And finally a lot of people just don't know how to architect or code. I think we could all benefit from learning and writing some assembly, so we could really understand what our software is doing.

    Maybe then we could strip out some of the thick layers of software bureaucracy that lies between the user and his cpu.


    Michael D. Crawford
    GoingWare Inc

  • by MobyDisk (75490) on Thursday November 30, 2000 @06:18PM (#590282) Homepage
    Netscape would have been better served by enhancing the Mozilla preject in the key areas it is lacking (speed, bugs) rather than adding tons of useless marketing features.

    Netscape _was_ a "champion" of OSS and a leader in the anti-MS compaign. Their key followers held their torch because of these things. Too bad they spend their effort alienating their strongest supporters.
  • by driehuis (138692) on Thursday November 30, 2000 @06:20PM (#590293)
    Mozilla is, in a way, a very succesful open source project, which attracted lots of really talented outside contributors. And indeed, a lot of attention is being paid to reducing bloat at the moment.

    Unfortunately, a number of design decisions were errrr less than optimal. The XUL user interface language seems to have a big impact on performance. And leaving aside whether one likes the UI or not, the fact that it behaves different than other apps on any given platform also leaves a lot to be desired (to Mozilla's defense, both Microsoft and Apple have gone down the tubes in this area as well, for example, I'm abhorred by Windows 2000's "browser like" clicking, where a single click will open a file rather than selecting it, and it isn't particularly obvious when it behaves the old way or the new).

    Anyway, to be able to fix something in Mozilla requires a significant investment of time on the part of the contributor. In the first Mozilla releases, which were based on the 4.x user interface, I could usually locate something I want to fix in an acceptable amount of time. Now, with the overhaul and the complete switch to C++, I spend hours grovelling through the thing, usually without coming up with an answer.

    So, open source or not, Mozilla's improvement still hinges in great part on the full time developers, who live and breath that code base. For me, and I think for a lot of other contributors, it has just become too complex.

    I have high hopes for its evolution over time, but it won't be soon that it will be as fast as 4.x. It will be interesting to see how spinoffs like Galeon will handle leveraging functionality from Mozilla. Time will tell whether XUL will become a boon to browser extension development, or remain a drag on the UI performance...

  • by vanguard (102038) on Thursday November 30, 2000 @06:34PM (#590321)
    Okay, I've read your comments twice and I see where you're coming from. Sometimes software is slow because developers have coded or designed it poorly because they were lazy or incompetant.

    However, that's almost never the case where I'm working (where I'd rather not mention). I work with some absolutely gifted technical peers who are sometimes forced to release crap because of deadlines. I also work with guys who have become drones who pump out crappy software that does it's job because they are only measured on (1) did you hit your deadline (2) did it meet the functional spec.

    It's not lazy coders (usually). It's a misguided reward system built by managers that don't know the first thing about software development. They fail to grasp that maintaining this terrible software will cost a fortune in the future.

    So before you blame the programmer for being lazy, consider what he's working against.

    Vanguard
  • by DrXym (126579) on Thursday November 30, 2000 @06:36PM (#590335)
    Netscape != Mozilla. The number 1 priority with Mozilla developers (and that includes Netscape engineers) at the moment is reducing the bloat and improving performance.

    Most of the problems boil down to:

    1. Packaging. The ~100 Mozilla DLLs should be condensed to 30 or so to reduce the per-DLL memory overhead.
    2. Loading unecessary services. A lot of XPCOM objects are created at startup when their creation could be deferred until they are actually needed.
    3. Boundless and untuned caches. Mozilla caches a lot of stuff and tweaks to the cache settings can dramatically affect memory consumption.
    4. Memory leaks. Leakage is pretty flat (in the browser anyway) but there is still work to be done here, especially tracking down refcounting problems.
    5. Inefficient structures. Certain structures hold onto more data than is absolutely necessary such as those to do with stylesheets.
    All of these things are being worked on but don't represent anything that can't be fixed. Mozilla (and therefore the next version of Netscape) will benefit from these changes.
  • by dbarclay10 (70443) on Thursday November 30, 2000 @06:37PM (#590337)
    I understand what you mean, but you also don't understand a lot of things. I'll try to correct you a point for point basis.

    programmers consistently work harder year after year to steal from the end user

    I think we both know that developers arn't laughing maniacally, saying, "Haha! They'll never be able to run this!"

    Note that I used to run SlackWare and Apache on a 100 MHz 486, serving up web pages (admittedly with a light load) while I used X at its console - and it worked fine. But when I loaded Windows 95 on it it was dog slow. There's no question of running Windows 98.

    Don't compare apples to oranges. Windows 95/98 had a lot more things running that Slackware did, even if you didn't see them running. *You* may not have used them, but a lot of people would have.

    Okay, anyways, I'm just going to skip ahead a bit. We all know that when you install the latest versions of software, it usually slows your computer down. Now on to the meat.

    One thing is because programmers are lazy, and if their code runs slow they assume the user will just get a faster machine.

    Well, are you a developer? If so, then you know what you say isn't true. If you arn't a developer, you must not know many of them. Users literally scream for new features, and the developer has to implement them, and FAST. It's not laziness as you should well know, it's priorities. Most users would rather a burgeoning web browser support cookies, rather than run 10% faster. Just go ahead and ask any user(who knows what cookies are), and they'll agree.

    The fact is, that you're partially right. A hundred new features shouldn't significantly slow down a program if those features are not used. However, it WILL take up more space on your hard drive - there is absolutely no way around it, short of having a CD custom-made with the software compiled to exactly your specifications.

    Now, unless you're going to go around to software projects and modularize their code(which was probably never meant to be modularized in the first place), I suggest you speak better of programmers. Especially Open Source/Free Software programmers who have graciously donated their time and effort to bring you, the ever so poor end user, a usable product.

    Dave


    Barclay family motto:
    Aut agere aut mori.
    (Either action or death.)
  • by atrowe (209484) on Thursday November 30, 2000 @07:06PM (#590343)
    That (-1, Troll) moderation is not fair. Just because my post goes against the slashbot groupthink, doesn't mean it's not a completely valid argument that deserves merit. I'm not exactly a MS fan either. They have commited some selfish, devious acts in the past, and their bundling IE with Windows was almost surely intended to defeat competition, but the fact is: MSIE is the best browser out there. It has a smaller footprint, a better user interface, and displays complex sites better than Netscape Mozilla. I used to use Netscape as my default browser as well, but haven't since around 4.0. It is simply not logical to waste my time downloading an inferior browser through my 56k pipe.
  • by WillAffleck (42386) on Thursday November 30, 2000 @07:09PM (#590345)
    Seriously, if those guys at MSFT didn't hate Linux, they'd just crank out a (slower, less capable) version of IE for Linux, and stomp all over Netscape.

    Yeah, I know, many /. hate MSFT, but not all Linux users hate MSFT, it's just an OS to them.

  • by Anonymous Coward on Thursday November 30, 2000 @07:09PM (#590347)
    Uh, doesn't adding memory totals from top give an inaccurate picture of real memory usage

    Yes, absolutely it does. 4.7 is not multithreaded. The reviewer has no clue and the review is utterly pointless since it focuses almost exclusively on the memory question.

    The reviewer must be one of those web "designer" types.

    This is only a story because organizations, Linuxworld and Slashdot, are run by technologically illiterates. Really disgusted with the level of professionalism around here.

    Anyway, I use nightly builds of mozilla and prefer them to Netscape 4.7 My machine will thrash a lot sooner with 4.7 than it will with mozilla.

    That being said, Mozilla still sucks on Linux compared to W2K and solaris. I suspect it has something to do with silly linuxthreads which, of course, Linus will never change because he wrote them and they are therefore ...

    B E S T T H R E A D S E V E R.
  • by ucblockhead (63650) on Thursday November 30, 2000 @07:11PM (#590351) Homepage Journal
    Netscape dropped LiveConnect, which is the way Netscape plugins comminicated with Javascript. They say that old plugins that require LiveConnect will "fail silently". They are correct, assumming that your definition of "fail silently" under Windows NT is "attempt to access a NULL pointer and crash the application".

    They have a replacement for LiveConnect. It is almost completely undocumented. There is no SDK.

    Geez, if I were a plugin developer, I'd be way pissed, and would hesitate to bother supporting Netscape 6.

    Oh wait, I am a plugin developer.
  • IE is fast because it's already loaded on boot. That's (part of) the reason why everything is slower in windows.

"A mind is a terrible thing to have leaking out your ears." -- The League of Sadistic Telepaths

Working...