Slashdot Mirror
Mozilla: News from the front
Point_Blank
pointed us to an update on Mozilla.org regarding the
state of mozilla
written by Mike Shaver. Mainly it refutes some of the arguments that
the project isn't "Open" because @netscape.com developers outnumber
outside developers. I agree with him- the fact that there are /any/
outside developers is a great thing. Anyway, some interesting
stats regarding download numbers and bug submissions and stuff. A
nifty piece if you're following the project.
In *nix builds of Communicator, you can already
/home/kurt/3.xpm
change its colors and map an xpm imge to the
window by putting this in your ~/.Xdefaults file:
! Some Netscape hacks
*nsMotifFSBHacks: true
Netscape*background: grey20
Netscape*backgroundPixmap:
Netscape*foreground: grey80
! End Netscape hacks
...now I haven't been able to get the Pixmap
option to work but I've seen screenshots from
someone who has. No details on how they got
it to work.
test
-
Mozilla is the way it is- and has taken as long as it has- because it has been driven by the demands of those who use it. You could argue that the end users have not had that much imput, but the truth is that the real 'users' of any browser software are web page developers.
Mozilla's choice to go with Raptor (a good idea of their own), and to fully support DOM1/CSS1/HTML4.0/ECMAScript are a godsend to developers.
This came about because these decisions were made in the open. The initial idea was to do a release on the 4.x codebase, but with the community (and the WSG) clamoring for standards compatibility, the correct design decision was made to go for the next generation layout engine instead of the heavily patched 4.x codebase.
As TCaTB[1] mentions,
"Plan to throw one away; you will, anyhow" (Fred Brooks, "the Mythical Man-Month".
~mindlace
We have more users, so we're more popular. (At least that's what I'm able to parse)
You're certainly right; it is a bit verbose.
--
I don't know that having per-URL JavaScript and Cookie preferences is that "unreasonable". IE already has a limited version of this in it's security zones feature. (Unfortunately, you can't create new zones, so you are stuck with 'Internet' and 'Trusted Sites'.)
--
Business. Numbers. Money. People. Computer World.
Don't forget that 30% of 'the Internet' is on AOL - and AOL only uses IE.
Subtract the AOL user base (I wish we could), and it's more like 50/50. Once AOL starts using Mozilla, Netscape will have the markt lead.
(Despite this post, I think browser market share is one of the most stupid concepts of all time. Who cares about the market share of $0 products. The intention of both Netscape Nav and MSIE from the beginning was mearly free advertising and standards embrace+extend for Netscape's and MS's server products. Which is why the iPlanet brandname is so odd. Oh well...)
--
Business. Numbers. Money. People. Computer World.
If netscape/AOL paid 100 programmers to work on Mozilla, and the source were closed, nobody would call Mozilla a failure.
However, that is not the situation; the source is open, and an additonal twenty people, NOT paid by Netscape are contributing code. Furthermore, an untold number of people are regularly submitting bug reports/ideas for enhancement. For some reason beyond my understanding, the popular press deems this as a refutation of open source, simply because there are more people from Netscape contributing code than those unaffiliated with Netscape.
Netscape has not lost any brainpower by opening the code, and they are not paying their programmers any more or less. Rather, they have a 20% enhancement in contributers, and people say this is a failure. I DON't GET IT! Could some wise person please explain this logic to me? Please?
Humbug! Its open because anyone can contribute. Just because a specific collective of developers outnumbers the independant ones, doesnt mean it isnt open. This is 'Open/Free software' politics getting daft again...
It is my observation that on most open-source projects there are a very small number of core developers (often friends in real life), but alot of users who submit bug reports and pester about the next release. Often the process of delegation can be more time-consuming than just doing it yourself. They say that human brains can only really cope with working in groups of up to 7 people anyway. Having to work over the Internet makes this even more difficult.
--
Another thing is that the package is made up of thousands of little files. Whatever happened to the convention that an application is made up of one application and maybe a shared library or data file. I have never seen an application with so many separate data files. Have they ever heard of a resource fork?!
p.s. I would run linux but my mac is one of the few that has unsupported hardware. I do have a pc sitting next to me with the case open that will run linux as soon as I can get a cd drive for it and a bugger hd.
I understand that a total re-write was necessary,. However, the more successful open-source projects all started with working code, and only used open-source to extend it, or to replace bits at a time. All the abandoned projects are usually the ones that went open-source when all they had was the idea. Freeping Creaturism is also probably to blame, since there was no reality (code).
I think Mozilla 5 will succeed, but that for version 6, there'll be way more non-netscape people involved, as people add little bits here and there, optimize this and that, and generally mutate Mozilla into a more advanced life form than a Sea Monkey.
"There's so much left to know/ and I'm on the road to find out." -Cat Stevens
Unfortunately for us, there has been little-to-no support from people who have been blessed with CVS checkin authority to allow us to get the build patches in that we need to stay current with the latest changes on the tree. Without getting these patches in (regardless of whether we boxcar or not), our ability to address bugs is virtually nil.
What good is a bug fix metric when we can't even build the latest tree?
Before I rant for too long, it's important to note that support has changed very recently (Henry is working with someone to get the OS/2 build patches checked in), but it took WAY too long before that support materialized.
"But always she's the spectre of uncertainty I first endured, then faded, then embraced..."
I think that the Mozilla project is making great progress. I try out the builds daily, often on Linux and NT.
However, I feel that it would be useful if they woeked towards implementing extra functionality so that it could replace Netscape 4.6 for general webbrowsing. For example, there is still no right click on links and the Preferences dialog hasn't been hooked up yet. However, we already have features like Translate which are not used that regularly.
By adding a few basic features, I would use Mozilla as my main browser, I would find more bugs, and contribute more bug reports. If I had more time, I'd look at improving any features that annoyed me and so on.
I think that the number of outside developers contributing to Mozilla will snowball in a few milestones when more features are added. It is already shaping up to be a great product, and I miss its many of its features when I return to Netscape.
One of the perhaps smaller, but of a certainty significant, aspects of the Mozilla project is apparent to those of us who browse the Bugzilla database. Ergo, it has lain unnoticed by the silent majority, the flamedot minority, and the ha-ha-Netscape-fools gawkers.
Users and programmers have traditionally been the poles of a divide (if I may carelessly mix my metaphors), kinda like boys and girls. (Which of the pairs is analogous to which I leave as an exercise to the reader >:+} ). While other companies or groups have been renowned for their attention to user interface or responsiveness to users, Mozilla, through Bugzilla, more so even than through the newsgroups, stewards a new user/coder frontier: The blessed enhancement request. Pssst - Rob - your code won't permit me to include the necessarily long CGI URL.
Here, in this well-mannered and efficient forum, users make unreasonable requests - and watch with astonishment as they are sometimes granted! The Netscape engineers are for the most part tolerant and polite - even enduring unwarranted abuse - and are open to luser suggestion. If indeed lusers they be. And most proposals are at the very least discussed, for the greater number.
The seeding of this hitherto untapped and rather mangy range of the noosphere (to use your beloved but limited vernacular), the (*scoffing*) user base, is an advanced, or rather advancing, inclusion that makes our trumpeted Open Source method more of a societal, a popular?, phenomenon than before. (*Leaving further such analysis to the grandiose*)
Needless to say, these words apply only to those members of society who are sufficiently interested to linger circa such domains. So should it be. We (or, perhaps, I) mad bastards who think to shape the next Netscape browser toward our ends and in reflection of our method-minds rather like the lack of company >:+)
And there is another aspect of appeal in the Bugzilla milieu. (Milieu being a browsable web database, an ongoing discussion with engineers, a devoted newsgroup set, a sense of comradeship against hostile outside, media, forces, &c) The satisfaction of submitting a bug and awaiting it's speedy repair soon becomes a quite forthright expectation, something akin almost to a human instinct, undiscovered alas until this late march of the Industrial age. It is the desire and expectation that, finding a bug, one reports it, and will soon be using a fresh copy of the software that is bereft of the very flaw. If such a cycle were established in all public domains, many corporations would be afflicted, and many consumers would rejoice. And lo!, the yobbers would owe we "computer hackers". It nearly calls to mind the fabled customer service and quality of vendors such as the Eaton's of the 1960s (to those non-Canadians who do not recognize the reference, *nyyahh* to ye).
Or perhaps I'm foaming verbose again - there was the Great Overboard some time ago, as I recall - but we'll see when it ships, won't we, kiddies?
*+]Strange moods are the validation of the universe.[+*
1) The size of the source: If you want to contribute you have to get into the source - this means read/read/read some some source to just try to get how the whole thing is working this takes times. Many Netscape dev are doeing this full time, they have the time to learn and can ask directly to the coder what he inteded to do when he wrote the code. While your learning everything is changing fast - Ok not so fast but fast enough so that it takes time to catch up. Ok but you might argue that the linux kernel is also Big and that it changes also constantly - right but the basic behind a kernel are thautght in any decent Computer school - it helps a lot, plenty of books are available on the subjects - so it you'll learn faster what the code does and how it does it
2) Time : this factor is only applicable to non US, nor Canada resident. In many Other Countries you have to pay for local communications so dialing your ISP and staying online cost some Phone-money , dowloading 25 or so MB really isn'yt cheap and really is painfull with a 33.6 even with a 56 k modem ...
none Yet.
I never said that.
The practical upshot is that the larger the team, the more productivity is lost to the overhead of dealing with other people. Therefore, the best software is typically written by very, very small teams (or often, single people).
It's not hopeless, however, to have large teams working effectively. If the project is modular and parallel enough, the team is broken in several smaller teams (and further subdivided, as necessary), each responsible for a subset of the system. The sub-team leads are responsible for coordinating with the other sub-team leads to keep the project coherent. This requires the sub-team leads to be a manager as well as a developer.
Every day my system pulls the tree and builds it for me. I've seen Mozilla progress so much in the last six months, and particularly in the last month.
Used it for several hours today for general browsing, without any crashes. I haven't had that happen in several months, and this was after the Necko code landing (the new fancy-pants networking code... noticably faster, IMHO).
There are still significant bugs, and its important if you're going to pull the CVS tree you know what to expect, because what works and doesn't work depends on the time of day. If it builds and the tests run, then tinderbox will be green even if some glaring feature is missing. (Like menus in the mailreader under Linux on the tree I pulled at 2pm EST today...)
There are people talking on here about how its not beta quality, hardly works, etc. The fact of the matter is ITS NOT beta softare. No one claimed it was. The milestones seem to work relatively well, although I thouht M6 and M7 weren't too good under Linux -- half the systems I tried on wouldn't build them. M8 was great. M9 (real soon now, I think they were mostly waiting on the Necko code becoming the default and stabilizing the problems from that...) should be even better. its definately a useable browser right now for most things, even if some stuff is flaky. (I can't log in on Slashdot for example)
I've said this a few times before on here, but its really worth saying over and over. Mozilla is really coming along. Its running suprisingly well, rendered pages with proper HTML look great on it. It certainly separates the good HTML coders from the bad in that regard. Its *fast*. I noticed that today's build is much faster -- both networking and rendering -- than the last one I actually tried which was on Monday or so.
The better rendering engine, and GTK widgets/menus make it MUCH nicer to look at than Communicator.
I'd suggest the complainers stop complaining and start submitting bug reports if they're having such problems, but people like that aren't likely to give useful bug reports anyway.
Any user that has found a bug and gotten rapid response from developers will never want to go back to the dreaded tech support line to wait an hour to get to the person that knows enough to tell you that it is a known bug and _may_ be fixed in the next release.
...and then the developers get so bogged down in handling bug reports, many of which turn out not to be bugs. I speak from (proprietary software development) experience.
Open source projects do have the advantage of an open bug list, which reduces some of the volume. But just reading Slashdot for a little bit will show you how people, even with the most powerful research engine in history at their fingertips (the web), will still ask someone else what "Echelon" is.
Ooh, a sarcasm detector. Oh, that's a real useful invention.
Well, considering that IE does not run on Linux, of course we still need Netscape, Mozilla, and other projects. Mozilla is not perfect. In fact, it's quite flawed. But, it is a step in the right direction.
As the Open Source community, we need to promote and applaud such efforts. Every bit helps. Especially one that has gotton such coverage. We have to be careful about predicting the failure of Mozilla. It will be seen as a blow to Open Source whether anyone really cares or not.
I believe that it started out flawed: Here have ALL THIS REALLY UGLY code and go wild!
However, since then, the processes has really shaped up, and almost all of Mozilla is new code, especially the layout engine.
Whee!
jf
Basically, he's saying that this not-yet-used section of ESR's metaphorical "noosphere" (the bugfix database) is something that makes Open Source more popular than before.
As for the post's diction, I find it rather refreshing that a few people still know how to use words which are greater than two syllables.
10 PRINT CHR$(205.5+RND(1)); : GOTO 10
Don't forget that Netscape once had "75% plus share of the browser market" and was dictating internet standards, and was charging for corporate use of their product. Of course, people consider Netscape 1.x - 3.x as the greatest thing to ever happen to the Internet. That didn't stop millions of folks of going out of their way to download and install IE on Win95 and NT4, and there's nothing stopping users from switching back to some future superior product such as Netscape 5.
Microsoft is just reading from the Netscape Embrace and Extend playbook here, although they rightfully have a bad guy reputation. As far as I'm concerned, I'm happy a ~50/50 breakdown and public sites that aren't 'optimized' for any specific browser. Anything goes on your intranet.
(People who don't bother to change their homepage hardly seems like a very ripe market, and it hasn't got either Netscape or MSN rich. Yahoo rightfully wins without 'herding' because they've got superior content. However, the only portal I visit daily is the toliet, so what do I know.)
--
Business. Numbers. Money. People. Computer World.
Doh! (Thanks for the correction)
--
Business. Numbers. Money. People. Computer World.
I will make one point. When Netscape released the browser source, it was unusable and uncompilable. There were large sections of code that were licensed from other companies and had to be ripped out before the source could be released. Mozilla started with a massive, hairy code base that was far from working order. That initial release was an overwhelming chunk of broken code, and it turned away many potential developers. JWZ has stated that it would have been better just to have started over from scratch. People wouldn't have had as much to learn before they became useful.
What I am saying is that Mozilla had problems acquiring developers because Netscape botched that initial release. In the Open Source world, you should always have a working chunk of code before announcing your project and looking to sign up developers.
A web browser is a large project, but is it fundamentally more difficult than a kernel, or a compiler? I do not believe it is. The initial reaction was problematic, but perhaps as the milestones roll on, more developers will get their hands dirty in the project. As for "flocks of developers", I don't think Netscape or the community realistically expected that. What *is* expected is a community willing to fill out bug reports.
You can by cynical about the project if you like, but I'm not sure why you are smearing it around Slashdot. Does it offend you that some are optimistic about the project? If you don't think its going anywhere, just ignore it -- you can filter out Netscape stories from your Slashdot account. I, for one, really *want* it to succeed, because once we have it, I can eliminate Netscape which is the last proprietary chunk of code on my box.
--Lenny, who is going to download the latest build right now.