Mozilla Milestone 16 Released!Wednesday June 14th, 2000Come and get it! Visit the ftp site to get versions for Linux, Mac and Windows! Read the Release Notes for more info. Also, pick up the latest version of PSM (Personal Security Manager) for M16 at the iPlanet website. PSM gives M16 support for secure websites. Think it's finally worth becoming the default browser, especially startup is much quicker on my machine (NT4/SP4, old Athlon). Great work! This is my only complaint. Why doesn't Sun send over some people to help the Open Java Interface team? Or why don't they work with Blackdown? My impression is that the OJI team could use all the help they can get. ...here's a disturbingly easy way to crash the browser, and possibly your Mac (did the job on my iMac running Mac OS 9.04): Visit www.w3.org/DOM Obviously, DON'T do this unless you are willing to get burned. No, I have no idea why this crashes Mozilla. because of the way the header is formed. a bug was filed on this page and marked as a dupe of bug 34520. (No I didn't spend a lot of time looking it up, I remember it because It's one of the bugs I tested) This is worse than M15 for me, because I'm stuck behind a firewall and M16 doesn't handle the proxy correctly. I have entered the server address and port number, but I never get prompted for a password. The output on the terminal I started Mozilla from gives a "document done" message immediately for any page I try to go to. Using a standard RH Linux 6.2 distro. Any ideas? Join the club, pal. I guess if you're not on a M$ machine, you're at the bottom of the bug fix heap. They've been "fixing" this particular bug, on and off, with a success rate that varied with the phase of the moon, for about six months now. What works seems to work great on my RH 6.1 system. there are just so many interesting things on "localhost" tho. :) Maybe someday I'll be able to see the M16 relase notes with a Mozilla browser :( Hey, there's always M17... Please don't forget us Linux folx behind firewalls, guys! I'm behind a firewall, using an HTTP proxy and it works great. Socks doesn't work, see bug 16103. I'm not sure if Netscape plans to include that code in ther release version, but I hope so... otherwise I'll be banished to building my own. Okay, so maybe it wouldn't be so bad. :) This is a rather annoying. The installer crashed with a sig fault, but it seemed to have downloaded M16 Ok. Otherwize It Rocks! I've been waiting desperatly for this for some time now. I was skeptical at first, but they are closer to the alt on my keyboard, so I really like it better this way. Unfortunately, switching from alt-arrow to alt-bracket is a headache, because it's been the standard on both browsers since 2 or 3.0, any shortcut buttons on a keyboard or on a touchpad will be rendered useless. As an example, my touchpad has a feature where if I drag my finger to the left at the very top, it'll go back a page, if I drag it right, it'll go forward. It's very convenient. There's also plenty of "internet keyboards" and such that have back and forward buttons on them. Yes, it's only a convenience, but I see no real reason to make this change, and if it keeps even one person (me) from using the browser, why bother? As an aside, alt+[ doesn't work to make the browser go back, though alt+h finally allows me to go home. This is on the Win32 platform, other platforms may vary. The very least they could have done was leave both in there. Does an ALT-Arrows combination do anything at all now? I live with one hand on the mouse and the other on an ALT+arrow... #65 I agree, but there is another great new feature...by Silverthorn Saturday June 17th, 2000 2:40 PM According to the release notes, seems like the forward/back shortcuts have been temp. disabled, which is a pain, but oh well. As you said, the old shortcuts should be there too, though. Actually, I'd like it if you could easily select your own keys. I always liked the z and x keys for back and forward in opera... But anyway, here is something I didn't realize until I looked in the preferences. If you have a mousewheel, you can hold ALT and move the wheel to go back for forward. Can also hold control for changing font sizes. Especially for if you use the mouse a lot, it seems very convenient... I think the most annoying thing for me right now is that I can't use page up or down to move on a page without clicking the page first... I tried to install the PSM but everytime i try, my Moz segfaults. You have to install it as the user who owns the mozilla lib files. (Giving everyone wriate access to the directory doesn't help). Theres a wishlist bug on being able to install stuff as non-root, somewhere on bugzilla This is a very good and stable release! AND it's fast. But two things: 1. In File -> Offline there are no entrys 2. on home.netscape.com the left part and the toolbar at the top doesn't exist on Moz. I think that's a custom version of the netsape page for Netscape 6. It's intended to be that way. I'm just wondering what it takes for a build to get a "thumbs down"... we've had "avoid mail-news" and "layout issues on many sites", both with thumbs up on all platforms. I don't think *I* particularly want to try either of those builds... Since I'm offtopic anyway, I'll also ask: Why does the buildbar have no discussion forum? Wouldn't it be better if *everyone* could comment on builds? Just my 2c. Stuart. Actually, that's a really good idea... it could pick up the slack for Asa when he's busy. mozineadmin? I know you are around somewhere ;) ~luge We don't want to discourage people from downloading builds, even if there are some problems, because that's how bugs get teased out. We only discourage downloading when you wouldn't reasonably be able to do any testing. It means "there are bugs that would make it impossible to test". There should be fewer and fewer builds like that as we move forward. There should be a discussion forum for each nightly. But I do like the loose requirements for a thumbs-up (does it run at all? thumbs up!) because otherwise no one would do any debugging where it's needed. Does anypone know where the source tarball lies ? I'd like to contribute a Solaris 7 SPARC build (again :-), but I need the sources... any hints ? The current source can always be found at ftp://ftp.mozilla.org/pub/mozilla/nightly/latest/ in a file called mozilla-source.tar.gz -- but you'll want to read http://www.mozilla.org/source.html and http://www.mozilla.org/hacking/ before downloading it. If you want to contribute code beyond trivial patches you'll want to use CVS: http://www.mozilla.org/cvs.html I was testing some DHTML with the M16 build, and found that the navigator.appName returns "Netscape" instead of "Mozilla" with this milestone release. My impression was that Mozilla would stay a separate navigator app from Netscape. If this is so, it could be a problem to have the same application name for both apps. Perhaps this is just a typo you'd better detect objects rather than appNames because, because appName is not reliable... for example : if ( document.implementation ) { // code for W3C DOM compliant browser // Mozilla 5+, IE6 ?, IE5Mac ?, Opera 4 ? } else if ( document.all ) { // code for IE4/5 } else if ( document.layers ) { // code for Navigator 4 } else { // code for old browsers } -- Hervé you'd better detect objects rather than appNames because, because appName is not reliable... for example : if ( document.implementation ) { // code for W3C DOM compliant browser // Mozilla 5+, IE6 ?, IE5Mac ?, Opera 4 ? } else if ( document.all ) { // code for IE4/5 } else if ( document.layers ) { // code for Navigator 4 } else { // code for old browsers } -- Hervé The browser still takes too long to startup. Like on the order of 5 times the loading time of IE. That's unacceptable and hopefully will be fixed. The reason that IE starts so fast has nothing to do with its "great design". It's because Windoze loads the libraries that IE uses when you start Windoze, so that they are always in memory. Comparing the startup time of Mozilla to IE is not even remotely fair. It would be interesting to see which starts faster if those libraries weren't already in memory. I think it's unacceptable that IE loads browser resources at bootup, which slows down my entire machine. There is also plenty of anecdotal evidence that shows that removing that browser integration from the OS drastically speeds *the entire OS*. So, I'm not too concerned about Mozilla's lag on startup. Fine, let's compare it to MS Word or Wordperfect on Linux. They also load at a fraction of Mozilla. Also, I've removed the exe that loads the IE libs from the startup sequence, and IE is still way faster to startup. So I would be definetely concerned. Maybe getting rid of superflous things like splash screen would speed it up. I don't think the splash screen is superfluous, but I *wish* Mozilla would actually use the black "status bar" at the bottom of the splash graphic to indicate what's it's doing (e.g. "Initialising <feature>"...-type messages) - at least you've have something to look at whilst it loads. I had two items making startup rockin': 1) I deleted localstore.rdf in my profile dir. esp for international users, that file can grow enormously due to a bug with int'l characters. If the file is bigger than 15 to 20 KB, you should kill it. You do not lose any important information, it just resets startup window size, sidebar open/close status etc to default values. 2) I upgraded my RAM from 64 to 192 MB - now it rocks! When RAM useage will be cleaned up (scheduled for between M17 and M18 afaik), I'm sure there will be a great impact on startup speed because it will need less memory swapping :) Fine, let's compare it to MS Word or Wordperfect on Linux. They also load at a fraction of Mozilla. Also, I've removed the exe that loads the IE libs from the startup sequence, and IE is still way faster to startup. So I would be definetely concerned. Maybe getting rid of superflous things like splash screen would speed it up. "Fine, let's compare it to MS Word or Wordperfect on Linux. They also load at a fraction of Mozilla." Okay, if you're going to do that, then I will. You know what? Mozilla loads a *lot* faster than Photoshop. Both are totally irrelevant to the situation at hand. "Also, I've removed the exe that loads the IE libs from the startup sequence, and IE is still way faster to startup." What, you removed c:\windows\explorer.exe? Then your system wouldn't function. IE DLLs are still being loaded. "Maybe getting rid of superflous things like splash screen would speed it up." The splash screen itself is a bitmap - there is no slowdown as a result of the splash screen. If you think a smaller bitmap would load faster, why don't you just make a little 16x16 image with MSPaint and put it in with the mozilla.exe directory and call it mozilla.bmp. No difference. There is an exe that loads the IE dlls. If you install Win98, you'll see what I am talking about. Explorer.exe only loads IE dlls if you have the active desktop enabled (which I don't). So I am not comparing apples and oranges. I just don't want to wait half a minute for a browser to load up. People here keep on saying that this is a usable release. Frankly, it is not even close. Friend, i don't think you quite understand exactly what IE is built into. It's NOT just an .exe that loads some DLL's; IE is a _intergral_ part of certain system DLL's. Please, do not take my word for it. Here's a link to a CNN story (from an _Infoworld_ story) on removing IE: http://www.cnn.com/TECH/computing/9903/09/removeie.idg/ And if you go here: http://www.idg.net/go.cgi?id=20213 -- read the other IE colums by this guy from the Mar. 1999 period. He asks Mirosoft folks it the programs at http://www.98lite.net referenced in his eariler column remove IE, and they say NO. It's very, VERY intergrated. So, when people talk about browser bloat, I know they don't know what they are talking about. That we have a HTML render that is, in its essence, only 2-3 megs WITH debugging code is beyond amzing. If it's slow for you, I'm sorry. But it is PRE-BETA. Please, stop judging it on a PRE-BETA, on your personal macine. I ran M10-15 on my old machine, a P120!! I upgraded...to a P233, and it's not as fast as IE, but I also choose to acknoweldge it for the pre-beta it is, not as finished code like IE. Yes, there are some IE parts that the OS uses, BUT, there is still a separate IEXPLORER.EXE, and a bunch of DLLs that IE uses apart from the OS. Using nothing but switches available in Explorer's prefs dialog, and TWEAKUI (a Control Panel plugin WRITTEN BY MS employees, and distributed on the Win98 CD! [but not 98SE, somone wised up it made them look bad]) IE is extremely removed from my day-to day OS routines, even with Web View enabled. IE now takes about 6 to 7 seconds to load on my machine. Mozilla only take 10-11. You have been misled by the MS corporate propaganda. The technical realities, if you actually do a little exploring on your own, are quite different. The IE renderer is wrapped up in an ActiveX Component. It and the key dependence DLLs are not loaded on my system at startup. Grab a memory walker program and check it out for yourself. True, MS has added many miscallaneous parts of IE to DLLs like comdlg32.dll and olepro32.dll, etc... But those are only contain the supporting functions, like look-ahead typing, etc... The renderer is loaded, when you click on iexplore.exe I have a pretty good understanding of how IE works. I also have a fairly good understanding of how Moz works. That is why I am concerned that having removed the debug code (there isn't that much of it there), Mozilla will still have a pretty long startup time. Hey let's not compare it to anything! Just load the browser it's damn slow to load! Netscape 4.X was too (After Netscape 4's newness factor ran off I went back to IE 3.0. That was fast to start up on my p120 w/16 megs ram!). Yeah sure, complain that the browser in the OS slows the system, but when I'm on my ethernet connection and using the net to 1. check weather 10 minutes later 2. Hey I wonder who won the game 10 minutes later 3. Let's check Mozillazine/slashdot/ok i'm addicted 4. I need to look that up 5. Need a phone number etc. etc. etc. Guess what It turns out opening my browser so many times takes my time and diminishes the time i would have savd from the fast rendering of gecko! I could leave it open, but i just hit the [x] and netscape is gone (cause i don't wanna see it). ANYWAY we're all just saying it's slow and that we appreciate sometimes that a browser opens quickly! M16 and the last few nightlies consistently die when I try to load http://www.mozillazine.org. Anyone else have this problem or know why? I using the Linux version on RedHat 6.2. This http://www.yomiuri.co.jp/newse/0614ec18.htm is to an article aboutJim Clark. According to the article, Netscape isn't a browser, it's a Search Engine! We're all so mistaken! How could we have been so silly! We can stop with all this Mozilla Broswer stuff now, since it's supposed to be a search engine! Oh, what a relief. Would it be possible to write a little systray program that loads Mozilla in the backround and keeps it running, so when I double click on the icon, the startup is fast? I realize most people wouldn't want a browser just running (not that we have that choice now), but I open and close browsers all the time, i would go for that. This was bandied about before, in several forums, and I'm thinking a pre-loader would be a good idea on the Windows platform. The speed difference when starting Mozilla from scratch as opposed to closing it then launching it (and thus pulling it out of memory cache) is HUGE. On an athlon 700, 196MB RAM, and an ATA-UDMA/66 drive, it takes about 11 seconds. If I close it then relaunch it (mem cache again) it takes 3 or 4 seconds. Make it an optional component. If they can reduce the amount of system resources that Moz needs to run, I'd use a preloader and wouldn't mind having the browser hanging around in memory. Comparing Moz load time to the load time of IE is a little misleading though, because IE is tied to the Windows shell. I think we should wait till we see what performance optimizations happen in the next milestones before entertaining the thought of creating a preloader. Yes, but two suggestions 1. Make optional/default OFF! (maybe ask in installer. Perhaps even say "Would you like Mozilla to load at start up. This will make Mozilla load faster, but will slow down your system. This is why IE loads so fast etc. etc." 2. If the choice is a. cut load time by 3/4 or b. develop "load moz at startup" we all know which we'd rather have!! ;-) If someone would send me the list of DLLs to load into memory, I'd be glad to write such a quickie program. My email address is RobertGelb@zdnetonebox.com For now, whenever we want to update Mozilla, the best thing to do is to delete Moz and reinstall. I'm wondering if there are any plans to have "update" packages for bugfixes. With the large number of changes that can occur between milestones and nightlies, I don't think "patches" would be practical and effective. It will be for updates/fixes to a final release, but not now. <:3~~ I understand that it is currently not practical to patch Mozilla, but I'm wondering if there might be any possible issues that might prevent Mozilla from being "patchable" like some other browser we know of. If you look in the preferences dialog, there will be a feature to update Mozilla on the fly. It's not being done yet because there is no product to update yet. Remember, we're still pre-beta. Come to think of it, is this related to the "silent download" featured that was once found on the blue sky page on Mozilla.org? Speaking of "silent download", is anyone interested in a "noisy download"? Something like a download manager thing... Hi, I thought that Mozilla runs on Linux. The latest installer required RedHat Linux. Shame.... my options are being narrowed once again. Thats only for the installer, an installer is not required though. You can build from source. from ftp://ftp.mozilla.org/pub/mozilla/releases/m16/src/ And compile it yourself http://www.mozilla.org/source.html should be able to help you. You may also want to read http://www.mozilla.org/hacking/ When you shwitch to native windows scrollbars mozilla crashes on exit. Dr. Watson pops up. (WinNT4/SP6a/Athlon750/196M). I'm using Mozilla (binaries!) on SuSE 6.3 and I'm sure you can run it on other distributions as well. You may have to create a symlink for, I think, it was libjpeg.so.6.2 or something which pints to your existing .6.0 version - or really update the lib... I don't remember exactly though, 'cause it's a loing time since I did that on my box :) Check out one of the following links: http://opac.nebis.ch:4505/ALEPH/-/start/nebis-eng/new http://www.icab.de/test.html http://www.simifilm.ch Mozilla screws them up completely. I thought this should be standard compliant browser... I downloaded aphrodite and placed it in the chrome directory, and I changed my startup string as documented, but mozilla doesn't start up. Is aphrodite no compatible with the latest milestone? Look for Aphrodite 0.04 to be released this coming week. Look for new features such as in place skin switching, Chris Sears action menus, skin persistance, crash recovery etc and at least 4 skins to choose from include the curent ruling skin Sullivan. All this stuff is finished and running real good. I am just doing some debugging. Posted with Aphrodite 0.04! pete What JRE plugin do I have to download in order to use Java with Moz ? JRE 1.2.2, as it is said in the M16 release notes that JRE 1.3 is not supported ? Any idea of if/when JRE 1.3 will work with Moz ? On my more pessimistic days I say never. There is a special patch to JRE 1.3 that you can download with the Netscape preview release, ONLY FOR WINDOZE. http://www.mozilla.org/oji/ Who knows if it works with M16. No signs of anyone even attempting to get it working on other platforms. Ok, I know all the Mozilla developers are working very hard. I wish there was more I could do to help. I just wish there was more information about the progress that the OJI team has made, there has been no report form them in over 6 months. I would like to see even preliminary java support on Linux so I could begin testing it and send in bug reports. Ok, i'm done ranting now. Yet another disappointing release and yet more condescending remarks at those who still care to report on the problems facing Mozilla. Perhaps more disturbing than the countless UI glitches and the lackluster performance is the communal attitude that has developed, where reasonable criticism is stonewalled by invoking some variant of the "if you're not part of the solution then you're part of the problem" canned response. If being "part of the problem" means taking a critical look at where Mozilla is headed, then I wonder what it means to be "part of the solution" -- because Mozilla itself is nowhere near being any kind of solution at all. See you at M17. Bugtesters do not have the time to go seeking out bugs. If you post them here they will go unheeded. And if you get flak for complaining about a build, it's likely because (on my part at least) I'm not expecting anything close to usability until at least M18. Netscape employees joke that NS PR3 will be the first usable build of Netscape 6, and that I think would be the time to really get worried. For the last month we've been getting all new features and the tons of regressions that go with them, and are only now starting to get into the optimizing mode. I just don't think now is the time to be disappointed. If the final product sucks, well, then that's different. Honestly I think he wasn't interested in reporting bugs. You are expecting something usable at about M18? I'm thinking more of Moz beta: Whenever that would be. But I'm already using the nightlies on a regular basis already, so it doesn't really matter to me. Just that I wouldn't recommend it to anyone who isn't interested in bug testing until Moz beta. For anyone interested in making Moz quicker, and who has winNT or UNIX, you can go to http://www.steveheller.com/visual.htm/ This is a page on Steve Heller's site. He writes books on programming, some geared towards novices (a he seems to be really popular). I saw this page on his site. It says he used this profiler called Rational Quantify on his company's character recognition software and increased performance by 300%, in about 6 hours! Well, anyway, the program cost upwards of $700 but you can get it free for 14 days if you have WinNt or UNIX. I thought maybe this could help with the optimization of Mozilla, which, in my opinion and on my machine (32Mb RAM) is desperately in need of, if it is going to be accepted by public. Yeah, yeah, I know...Optimization doesn't begin till M18 or 19 or so, but if anyone has WinNt or UNIX (I don't) they can probably do SOMEthing! Share results if you get any! PS: This may only be useful for very good C++ programmers, but I don't know, as I can't try it. PPS: It is a very large download. ~100 Mb I think, if that is an issue for you (is for me). Yours truly, Sam Kelly Sorry, a slash at the end. Go here. http://www.steveheller.com/visual.htm Okay, I'm running M16 on windos and about the only thing I can do to help out (since I'm just a good user not even a bad programmer) is to report problems. I have found a page which I cannot load using M16. It is at www.mtbreview.com. Who should I tell |