Slashdot Mirror
Mozilla M8 Released
bergie writes "The Mozilla milestone release 8 is now available! Go check the coverage on MozillaZine. Go fetch it! " For
those interested, MozillaZine has a pseudo-changelog available. It seems blizzard's Xlib port
is coming along quite nicely. Anyone at OLS next week will
be able to attend Mike Shaver's "Inside the Lizzard" talk. Congrats to the Mozilla folk!
Not only that, there is a Solaris package available for download. That's what I've been using, because I've been unable to get it to run on Slackware, even with libc6.
Minor quibbles:
1. The fullcircle version creates almost 900 separate files, a lot of which are just 1-2 lines of configuration stuff. That's a lot of wasted space, even on an HFS+ drive. Perhaps some of those options could be combined?
2. It takes a bit of time to open, during which there's no perceptible activity. I almost Command-Control-Powered the machine because I thought it had hosed itself.
3. Double-clicking a word doesn't automatically select it. Yeah, it's a little thing, but after 15+ years you get used to it. :-] And yes, I would fix it myself if I had the knowledge (I'm working on getting it now).
If the finished version is as big an improvement over M8 as M8 was over M6, maybe it can replace IE as Steve's browser of choice. Now, if they can start supporting Mac OS Runtime for Java....
I use Macs for work, Linux for education, and Windows for cardplaying.
Or has the Mozilla team suddenly kicked everything into high gear. Two milestones in under a month, my lord.
:)
There are still things I'd love to see in Mozilla. I actually fixed a bug in the up/down key scrolling, Mozilla is amazingly easy to read and understand.
I'm not sure why people complain that it's too much, it's a very clean C++/C program and my lord LXR is useful
--
The world is neither black nor white nor good nor evil, only many shades of CowboyNeal.
10. Renders pages very quickly.
9. Free!
8. New features such
*segmentation fault*
There have been glitches in building on libc5 systems. In fact right now the configure process pretty much sucks, and misses a lot of requirements that it actually has but isn't programmed to look for. YMMV in compiling it yourself. Some people have had good luck with libc5 builds, as I seem to recall reading a while back in some of the mozilla newsgroups, others haven't. Personally, I haven't had much luck on anything but a pretty clean RedHat 5.2 or higher install.
:)
I build the new client every morning, automagically. Some days it works great, other days it doesn't. If you want to try building your own copy with libc5, I'd suggest two things -- if its not your first time building it, wipe the sources and repull them if you're using CVS. The configure process misses dependancies some times, and things don't always get built right. If everyone else seems to be able to get it to work but you, starting from a fresh pull is a good first start.
Personally I think the CVS method of building it using client.mk is the way to go, any idiot can do it without any problems. Make sure you actually grab the M8 branch if you do though, because they're starting to drop the Necko code into the tree today (I believe), and the whole thing is likely to be horked for a while. I've heard its going to be even faster with the Necko code. I haven't even gotten close to having a Necko build work though.
Either way, the i686 build will run fine on a i586.
(navigator!=msie)!=(navigator=bad)
No
(navigator!=msie)!=(navigator==bad)
I think proxy support (and basic things like a cache) are coming with the Necko code drop this week, which hopefully will be stable by M9.
Necko is the new networking code, replacing the current networking code. Promises to be more efficient, blah blah blah. I just hope it doesn't block on DNS lookups like Navigator does under Linux. That'd go a long ways towards making the program "feel" faster.
The Moz project is very important to the continuing success of Open Source platforms and while the code-donatin' heros know how much their work is appreciated I'd like to pass the team a few buck-equivalents from my long-ago-smashed piggybank.
Where is the website where I can get a secure connection, pull out my CC and make a symbolic financial contribution to the volunteer Mozzers?
Buy some hardware, get together for free beer or whatever - I'd just like to show my gratitude.
Should invading one's peaceful neighbours be opposed, or rewarded with trade deals?
I knew this would be the most controversial of my ideas, but if you think about it, bookmark management really does belong in a separate application...
I've felt this for several years now.
1) Navigator's "Edit bookmarks" thing looks like a separate application anyway, so there's no reason for it to be all in one. Netscape itself should provide the bookmark manager as a separate app, but part of the whole package.
2) people do have to use multiple browsers at times, and you could rig up the same bookmark manager to handle multiple browsers (Netscape, IE, kfm, Opera) instead of having each have its own bookmark list.
3) The bookmark manager could be opened separately, or even have a way to incorporate it into a menu; then rather than start the browser and then select the bookmark, you just click on the bookmark you want.
4) Navigator's bookmark management is woefully inferior to IE's. If I could fix one thing for Mozilla, that would be it. There's no reason to make me compile the whole app just to fix that manager.
5) Some places will want to have a bookmark czar, who maintains a global set of useful bookmarks (say, to the company's key website pages, suppliers, and competitors) that should be accessible to some group of people as part of their menu. Having a separate app would allow building a manager that supports a global and a local list of bookmarks.
Also, there should be a way (if there isn't already) to have each new page submitted to an external app. This app could then keep track of the page marks, just like the back and forward menus, but it would also keep a tree view of all links traversed so that if you (for example) go to slashdot, go to freshmeat, hit "back" and then go to a Slashdot story, freshmeat would still be visible as a previous path.
BTW, I agree that responder to you was rather out of line, that there was no reason to be so antagonistic.
P.S. To the mozilla crew, good work! I'm acquiver with anticipation...
Ooh, a sarcasm detector. Oh, that's a real useful invention.
You didn't say what platform you're running, but one reason the Linux version is both less of a memory hog now, and less prone to crashing is, I believe, because there was a bunch of shared libraries actually being loaded a bunch of times each up until very recently, and for whatever reason I guess they were chewing up RAM, and causing problems with particularly shutting down Mozilla, causing coredumps.
:)
At least that was the explanation I read, I never actually noticed the problem myself.
I don't mind if it comes extra; the only reason I'm concered about it for Mozilla is that working on those features might distract the team from focusing on the #1 goal: being a web browser.
It is entirely possible that even if these features were stripped, and only added later, the code would take just as long to reach production quality, because there are already enough people on this, and throwing more bodies won't necessarily make it better. I can't help but think though that they could have moved along much faster if they had just focused on getting the pure web browser functionality first, then started worrying about plugging in extra things like email, news, and HTML authoring.
Can someone who is familiar with the code or the development team comment on this? I don't mean this as flamebait; I'm just echoing a previous poster's concerns along these lines. Would focusing on just web browsing have helped much, or are the real issues totally unrelated to this and adding the extra stuff doesn't really slow things down that much at all?
----------
In a real emergency, we would have all fled in terror, and you would not have been notified.
Oh yeah, it's fast. But half the pages on the net look awful. If this product shows the inherent limitations of the real HTML specifications, then sign me up for IE.
It's probably less of a limitation of the real HTML specs as a limitation of the people who didn't follow the real HTML specs when creating their pages. My experience is that the rendering engine is positively anal about compliance, and just plain doesn't like the kludges people have put in to make things look good on IE and NS.
Could someone who has played with it please report on what sorts of embedded scripting languages it supports? Are we still stuck with just javascript, or can we now say SCRIPT LANGUAGE=xxx for other interesting values of xxx?
Yeah... sad, isn't it? I've personally come to the decision that when a usable Mozilla is finally released, I'm going to go to pure HTML 4.0 with only CSS for formatting on my personal pages. If people don't want to make the choice between plain grey pages and screwed up CSS formatting, they can get a real Web browser. I'm not going to spend my personal time supporting their some-"standard"-we-just-made-up browser, especially if they're going to make me make a choice as to which non-standard browser to support.
My employer's page is another matter, but I'm really looking forward to the day when I can dump the layers and layers of proprietary extensions from that, too...
A couple things... To the best of my knowledge, the Windows version of Mozilla writes all of it's info to the file \windows\mozregistry.dat, NOT into the system.dat and user.dat files which make up the windows registry.
To your other question about a simple web browser, try viewer.exe. It is the browser, and nothing else. (Even so, it may insist on a profile.)
If a tree falls on an anonymous coward yelling 'first post' in the forest, does anybody hear?
This message is being posted from Mozilla M8.
It's long been my contention that the biggest problem with Linux these days is that there are no decent graphical web browsers for it. I'm looking to Mozilla to make that change.
The current standings:
This version seems to be okay for stability on Win98. It hasn't crashed yet, although the "back" feature is still a bit quirky so I had to quit once and restart Mozilla to make it work again. Please note that this is not a huge issue for me yet, as I know this is still alpha software, but it's good to know where it stands so far.
Speaking of the "back" button, it is still broken: like Netscape, it takes one to the TOP of the previous page, rather than to the link on the previous page that was used to proceed to another page. This, IMO, makes it unusable for browsing: if I click on a link at http://www.freshmeat.net/ and then return later to resume reading Freshmeat, I expect it to take me where I left off, not dump me at the top of the page to spend a minute finding my place again.
IE seems to have figured out this little feature. What's wrong with the Netscape guys? Is this in the blueprints for Mozilla, and simply unimplemented at this point?
(Note: I didn't throw that remark about IE in as flamebait; it has its own passle of problems that make it next to unusable too... although I suspect many of the multitasking problems are results of the underlying OS, but then, since IE doesn't run on Linux, and is made by the same people who make the OS, the blame still goes to the same people.)
As for style sheets, I must say I am impressed by Mozilla's renderer, much improved over Netscape. This has been a sticky point with me, considering that serious web designers (many of whom I work with) only laugh at Netscape and barely (and grudgingly) bother to throw in a little extra code on their CSS-enhanced pages to make it readable in Netscape. And they're right: the CSS design is a good one, and Netscape's non-conformance to the W3C's standard is a serious detriment to the growth of the 'web and structured document development and acceptance. That is, people are afraid to develop content using W3C standards since 50% of the popular browser market (Netscape) doesn't support them. The fact that Mozilla renders HTML "correctly" according to the W3C is a saving grace.
Other than these issues, little else in M8 is particularly notable or worth its bloat. When it becomes a little more stable and fixes the "back" button, I'll try my hand at compiling it without the myriad of consumer eye-candy schlock that is handled better by external programs (like mail, news, and bookmark management). If I discover anything else worth noting, I'll try to remember to drop a note here for anyone interested.
How do one "install"/run this beast? Which script to run?
Just open extract the tarball, go to the 'package' directory and run ./mozilla-apprunner.sh
--
Midgard Project - Open Source CMS
Does this mean Mozilla will not require GTK+?
From the Mozilla Xlib Project page:
"The fact of the matter is that there will always be more than one toolkit for Unix. Right now the big players include Motif, GTK+ and QT. Each of these toolkits unfortunately has its own look and feel for many things including menus and scrollbars.. None of these toolkits is going away anytime soon and it is inevitable that there will be ports of mozilla to each of these toolkits.
One of the goals of the Xlib toolkit project is to create a common base, written in Xlib, that will include all of the functionality that is common across the Unix toolkits. The majority of the code will include simple drawing and image handling in the gfx module. It will also include some of the widget side of the toolkit. Some candidates include the nsWindow class which is the simple drawing surface used by the html layout and the XP widgets.
One of the project goals should be to make it very easy to plug in your toolkit of choice. This means that you will be able to use Mozilla in your Motif, GTK+ or QT application. Until the world chooses the One True Toolkit or dies in the process, Mozilla should be able to run on all the toolkits without hardship to people writing applications.
It should also be possible to have a stand alone Mozilla browser that is based entirely in Xlib. While this compromises one of the primary goals of the many Desktop Projects for Unix and Unix-like systems, it allows for a lightweight browser implementation that could be used on a Kiosk system, or on very old hardware. The fact is that Mozilla contains most of the functionality of the modern Unix widget toolkits and it is a very small stretch to turn it into a complete toolkit system."
So the answer to your question is yes, altough I'm not sure if the Xlib port is that far along yet.
--
Midgard Project - Open Source CMS
Most of the problems on the Linux version that I have seen (in M6 and M7, haven't done M8 yet) is versioning problems with GTK and friends.
Fresh compiles seem to operate MUCH cleaner.
jf
I have read most of the comments here, and although most are pretty positive, there are enough 'odd' ones that require I make the following statement:
.9) to NT 3.0. Guess who would have won that one? The actual beta won't happen until about M12 (October 1), and by Jan 1 the Seamonkey browser should be finished.
MOZILLA IS CURRENTLY ALPHA SOFTWARE!!
Please remember this when you download the Milestone releases. Feedback on problems is very good, go to http://www.mozilla.org/bugs/ for more info. But griping and comparing the Mx releases to Communicator or IE is really counter productive. It's a lot like comparing early Linux versions (pre
Netscape will probably start tweaking Mozilla into Communicator 5 about this time (M12), if not before.
Also please note that Mozilla is not Netscape! the Milestone releases are actually Seamonkey, which is the reference browser for Mozilla, and will not be the same as Netscape 5.0, although Netscape 5.0 will be almost completely Mozilla components (including much of Seamonkey), with some 3rd party additions.
The Mozilla project develops code, design, and modules that may be used by anyone (under the Mozilla License) to create their own browsers or app that requires HTML, CSS, or XML rendering.
Personally, I think it is going great, and the Mozilla guys are still right on track.
jf
Actually I'd just like to see Eudora Pro for Linux, but I can't see Qualcomm being that progressive... still the best mailreader I've used.
these guys are working on it http://www.NexwareCorp.com/
---
I dunno, but this seems like a huge improvement. I'm running with only 32 megs of RAM so any memory sucking is very apparent to me... This release seems a whole lot lighter than the older versions. I'm starting to get excited - Our little baby Mozilla is all grown up!
There appear to be some focusing issues. The scrollbars tend to flash terribly when moving over any of the widgets. Its only an annoyance though.. and it seems that Mozilla is on the road to becoming usable.
...but I've not seen the answer anywhere; or maybe I've just been blind or something. anyway...
when Moz is finally released, are they going to continue to release the standalone client alongside the communicator? I, for one, have no need for the massive disk bloat of an html editor (still use pico, vi), mail reader (pine), or newsreader (tin). Thus, it would be nice if I could download just the neccessary component...
thanks.