You can download an HTML validator for Firefox that builds it right into View Source. It will validate it within the browser and also provide accessibility warnings. It's based on Tidy.
While Composer is certainly handy, and one of the big reasons I use Mozilla rather than Firefox & Thunderbird (other than that Mozilla hasn't scrogged any mail files for years and even the latest T-bird sometimes does), it's hardly a serious tool for building web pages (even in its latest Nvu guise.)
I find it vital though for things like quickly correcting serious brain damage in pages that either won't print, will print with a zillion little adsd taking up half the page, or will use 2 lbs of color laser toner if you try.
For that, especially if you're building any sort of modern web standards site using CSS, you need tools designed for that - TopStyle is my favorite, and it also has a built-in validator based on Tidy.
I'll say this - I don't buy too much software anymore, but TopStyle is one of those rare programs that is actually worth significantly more than the selling price. I wish I'd bought it six months before I finally did. It's also now one more reason I'm not moving from a Windows desktop - there is certainly nothing approaching its power and utility on Unix/Linux (and I'm a 20-year Unix bigot...)
Pseudoscience appeals to false authority, to emotion, sentiment, or distrust of established fact.
Thank you so much for declaring on behalf of us all that these experts in their field are practicing "psuedoscience" - we will now rest easy in your certainty that "real science" is never, ever, wrong. (And of course, the multibillion dollar wireless industry has nothing at stake, either, now does it?)
Someone please alert the FCC, FDA, and the universities, so they can summarily reject as "pseudoscience" any proposed research to determine if there might really be a threat posed by a technology so new (only a few years old!) that it's hard for anyone to know much yet.
We are truly fortunate you've saved us all the trouble and expense of actually conducting real scientific research on a *potentially* significant heatlh issue.
Whatever happened to science as a quest for knowledge (which implies truth)? I guess our political and emotional worldviews trump concerns for truth these days...
GSM starts with the highest possible signal when a connection gets established. The phone then decreases the signal to a level, which is sufficient to maintain the connection.
IIRC newer standards do it the other way around - the phone increases the level until the connection is maintainable.
No, actually, because the CDMA signal is much lower-power and noise-like by design, it's much more susceptible to the "near-far" problem, and so needs to manage transmitter power far more closely. It was primarily the problem of getting suitably "power agile" radios developed that slowed CDMA's entry to the market - it's a hard problem.
As i have heard, GSM is a little less harmful. Can someone with a bit more knowledge provide some insight?
Actually, although the evidence is far from conclusive, there have been several studies that indicate that GSM/TDMA is significantly more likely to be harmful than CDMA.
If you look at the power distribution across both frequency and time, the research shows it's pretty apparent why - in GSM/TDMA, there's silence, followed by a very quick high-power set of RF spikes (the ones and zeroes you transmit during your time slot) usually concentrated at a particular frequency. Several research studies seem to point to high-power fast rise time and fast fall time signals (the square waves of bits) as having not-yet-fully-understood detrimental biological effects that are apparently not due to heating per se. The number of studies showing the existence of such harmful effects does seem to be growing over time, although the data is "noisy" at best.
(Remember that polarized molucules like water, sugars, and many proteins found in abundance in the brain will try to align themselves with each successive polarity reversal. At high powers, it is this molecular flipping that generates heat in microwave ovens. Apparently, just stirring those molecules around even gently may not be a very good idea, even if it's not at the very high powers required to generate significant heating.)
CDMA, on the other hand, has the distinct advantage of not having high-power RF spikes, but has a much lower amplitude signal (still square waves, but a whole bunch of much smaller ones spread out across both frequency and time.) The signal literally looks like noise, and it's designed to: the "chipping sequence" mixed with the data in CDMA type modulations is sometimes referred to as a "PN" (for "Pseudo Noise") code. (There are a set of these PN codes that are mathematically orthogonal to each other, these are the "gold codes" that are most useful for CDMA or other direct sequence spread spectrum communications. The math behind this cool, hairy, and above my pain threshold...)
I had to learn about this stuff about ten years ago when I was doing a lot of early wireless work (both voice and data) in hospitals, which are justifiably paranoid about RF interference, especially since medical devices are exempt from most FCC regulation and sadly, are therefor often much more susceptible to RF interference than commercial devices. The week after I learned what I've set out here, I traded my old phone in for a new Sprint CDMA PCS phone - I don't plan to ever have a GSM/TDMA "CancerPhone".
(BTW, Richard Branson, head of the Virgin Group that includes Virgin Wireless, will not use a cellphone at all, for whatever that's worth - I'm not sure what the availability of CDMA is in the UK, though; GSM/TDMA may be the only choice...)
I don't have all my bookmarks on the topic handy now, (and they're probably badly out of date anyway), but some persistent web searching should turn up much of the medical research that does exist - only those that deny reality can say there are no health effects of cellphone RF - we know those effects exist, but we don't know what all is behind them, or what risk is posed by those effects. Such research is made especially difficult by the rapid pace of technology in the wireless business - we may not really be able to generate meaningful staistical studies for decades yet. Until we do, I'm going to play it as safe as I can, and as far as I can tell, that means CDMA (Sprint and Verizon mostly, in the US.)
One bad traffic/train accident and half of Texas loses their internet connections. Yes, that's a simplification and/or exaggeration...
Actually, it's already not an exaggeration. There is a small town in east Texas (no names - I'd rather not tip off the Muslim terrorists) that is the terminal point for a surprisingly large portion of the fiber in this country. Taking out this otherwise rather ordinary town could put a very serious dent in the US voice/data network.
...in the United States, since before the Civil War, people are Americans first, well except for Texas.
Well, we tried to reassert our rights to self determination 140 years ago, but were forced at the bloody point of a bayonet to re-join "the Union"; this despite the fact that Texas was an independent nation before joining the US, and had (and still has) every right to reclaim its sovereignty. Of course, Abraham Lincoln was dead-set on doing *anything* to prevent real government by the people - such as, for instance, creating another state by decree (splitting Virginia) to ensure a cooperative majority in Congress.
Like more than a few Texans I know, I occasionally list "Texas" as my country on forms (this can be a challenging sport with web forms.) Texas - It's a state of mind.
BTW, for those posters earlier that didn't realize that Texas really is pretty darn big I have this story: Several years ago, I did some work for a firm in Rhode Island. Not knowing any better, I made reservations to fly into Providence and drove to Narragansett (beautiful drive BTW, I don't regret it a bit.) My hosts were shocked that I would consider "driving clear across the state" until I pointed out to them that my daily commute across Houston (which was admittedly quite long, but near work before taking a cross-town job) would not fit inside their entire state!
Anyway, I have very mixed feelings about this project, although I could probably get behind it pretty well if one of more of the passenger car lanes had no speed limit (or a "reasonable and prudent" limit that was recognized to be well in excess of 100 MPH in good weather.) Why are we the first generation in human history to travel slower in the aggregate than the previous generation?
No joke about permissions - I would give ANYTHING to have file / directory permissions set up the way Netware's NDS (Netware 4.x and newer) does it.
Come to think of it, now that Novell is hacking on the Suse code base - how about adding NDS as the file permission scheme?
Because that only gets you halfway to where we want to be? I agree completely that the winning setup is to combine the best of Netware with the best of Linux. What we really need is the best-in-class granularity of permissions of the Netware filesystem (which is what you see reflected in NDS) AND a network file server that uses NCP, all combined with the goodness of a Linux-based server environment.
For those that don't remember or know, NCP is Netware Core Protocol - a simple, lightweight, richly functioned file sharing protocol that was the heart of the old Netware systems that could easily support hundreds of simultaneous users on 66 MHz '486 hardware. NCP absolutely kicks the snot out of both NFS and SMB/Samba, and is *by far* the best of the widely-deployed network filesystems out there. Don't overlook the fact that it's super lightweight design can allow current and future embedded network devices to easily access filer server resources, something that's not really practical in most cases with SMB or NFS.
Why Novell hasn't added NCP server capabilities to SuSE is beyond me - it would be the single best thing they could do to reclaim relevance, sell new stuff into their still fairly large installed base, and take a chunk out of MS by providing a real server alternative that's not just different (like Samba), but actually *better*.
Although not trivial, it's also not that hard - and if done correctly, it would successfully blend the strengths of NCP-based file services and the Linux server environment. I know if I were running the show over at Novell, an crash skunkworks effort along these lines would be a top strategic priority...
This is radically different from all times past. Many modern "discoveries" were in fact known in the classical world, two thousand years ago--we have evidence of simple steam engines, clockworks, etc. The difference between then and now is that today we have a system for publishing and exploiting such discoveries.
Publishing makes most of the difference. Well, that and the fact that had there been patents to secure a proprietary right to an invention for some period of time, there would have been a profit incentive to propagate those technologies.
In both the examples you use, we have the evidence that those technologies were developed (although quite crudely), but not really used and certainly not propagated - Hero's aeleopile was just a steam-powered toy or curiosity not useful for anything. Raal steam power would wait centuries for Thomas Newcomen and James Watt to harness steam as power. The Antikythera mechanism (the only extant example of ancient gearworks), although a very impressive product of the superior Rhodian mathematics and culture of 80 BC, lacked some very fundamental understandings of mechanics - for instance, the gear teeth are triangular, showing that its makers were astute mathematicians, but had not got far enough to realize that same level of math could solve the wear and binding problems created by such crude gear teeth. The Antikythera mechanism is a true anomaly, though: There is no other known clockwork device using differential gearing until AD 1575, although possibly similar devices are referenced by Cicero during his studies in Rhodes in 79-77 BC. (I know a bit about this gadget because it happens to be a hobby of mine...)
A final note: Reverse engineering has *nothing* to do with the justification of patents. I have no idea where you got that idea. Patents and trade secrets are completely different beasts, and by definition, if you patent something, the way it works is no longer a trade secret.
Seems to me like these "reforms" will only serve to lock out the small guy even more. Extra hurdles and extra expense will mean that only those who can afford the best patent attorney can get patents.
This is *exactly* the problem - the patent reforms (read abolition) most people in the open source community favor would actually just ensure that the Microsofts and IBMs of the world could never be successfully challenged by two guys in a garage. And yes, such challenges actually happen (and succeed) much more often than you might expect.
For a few good examples, here is an excerpt from a letter to the editor I wrote to LWN a few years ago that outlines why most "patent reform" is a very bad thing: (original URL: http://lwn.net/2000/0420/backpage.phtml#backpage
From the "gararge-shop" POV, well, just off the top of my head, there are the
examples everyone is familiar with: Bill Hewlett and David Packard (HP, instruments), Steves Jobs and Wozniak (Apple, home computer), and outside the computer industry, folks like Edwin Land (Polaroid, polarized materials and instant camera), Chester Carlson (Xerox, xerography), Henry Ford (Ford, affordable automobiles), Thomas Edison (GE, light bulb, motion pictures, phonograph...), and Alexander Graham Bell (AT&T, telephone), all of whom profited greatly from their patented works. (One could argue for the inclusion of Jeff Bezos in that list, although around here, that's a bit like whacking a hornet's nest with a stick...)
But the classic twentieth century example of patents providing exactly the kind of protection I'm talking about is probably that of Philo T. Farnsworth, whom you may never have heard of, although you likely use his invention (electronic television) every day. Farnsworth was the prototypical individualist inventor who persevered against all odds and eventually defeated David Sarnoff and Vladimir Zworykin of the immensly powerful RCA. RCA was truly the Microsoft of its day in terms of control of the market and underlying technologies through acquisition - often under severe economic and other pressure. RCA had a policy of never paying royalties for any technology - a policy they managed to uphold until they met Philo Farnsworth, who just wouldn't give up.
Farnsworth fought virtually alone against all of RCA's power for seven years before the final court rulings that his patents had clear validity and precedence over Zworykin's, forcing a tearful RCA lawyer to sign a royalty payment agreement to Farnsworth. (Farnsworth publicly displayed television *five years* before Sarnoff unveiled RCA's infringing version to the world amidst great fanfare at the 1939 World's Fair, leading many to believe Sarnoff and RCA were the inventors of television - sound like anyone today?)
Farnsworth's experience is, if anything, a case study for the need to *strengthen* patents and either streamline patent appeals or extend the length of patents when thier commercial utility is impacted by unsuccessful challenges. (World War II intervened, and the government outlawed television for the duration of the war (the technology was needed for radar, night vision and other inventions Farnsworth then worked on), and so Farnsworth's patents expired before he could profit from them.
Do you still think patents are a bad idea? I'd argue experience shows that patents should be strengthened and perhaps that the duration of Farnsworth's patent should have been extended, due to RCA's clear abuse of the patent system and the courts. (I also think the government should have been upright enough to grant extensions in the name of fair play to all inventors whose inventions were commandeered for the war effort, but that's another issue entirely.)
History clearly shows that often patents are all that stands between real progress and innova
Tried both RealVNC and TightVNC and the "original" won at least in a LAN environment.
This is NOT surprising - Konstantin Kaplinsky, the author of TightVNC, made it clear in his original announcemnt of the product that the Tight protocol was designed to speed up VNC over high latency, low-bandwidth (dialup) connections - he even went out of his way to point out that the Tight protocol is usually slower than the plain old VNC protocol when running across a fast, low latency connection like a LAN.
I don't have a link handy, but it you look at the archives of the old vnc-list@uk.research.att.com list (assuming it's still out there somewhere...), you'll find his postings pretty much as I've described...
TS is superior to all VNC products, period. No one can develop something like TS without full access to a lot of windows code.
That's because the RDP protocol just flat kicks the VNC protocol's butt. If all (or even most) of the issue was hooking the video driver, then UltraVNC would be as fast as RDP implementations - but it's nowhere even close.
Big Question: the RDP protocol is (IIRC) publicly documented by MS - why hasn't anyone written an RDP client/server set for Unix/Linux, so we can all get the benefit of the superior protocol performance of RDP. (And even it the protocol is not publicly documented, what it's doing is relatively straightforward and would be *much* easier than what the Samba team's done to figure out and document the SMB protocol...
(For that matter, now that Novell's got open source religion, why can't we get them to open-source NCP, so we can all use a network filesystem that clearly outclasses both SMB and NFS? Is there still an NCP client included in newer versions of Windows? (It was there as recently as NT4 and I think W2K...))
RealVNC: the original. TightVNC: optimized for low-bandwidth Ultra: tons of extras - file transfer, chat, video driver, NT/AD security Tridia: get around firewalls, more management features
I miss anything?
Win2VNC: One virtual desktop across two computers, using kbd/mouse of Win2VNC machine to run both(even does copy/paste correctly between screens if you use a good modern server like Ultra.) BTW: I've been told that the Win2VNC link I posted above has been superseded by a new version that adds support for mouse wheels, Alt-Tab, and other useful stuff. You can find it at Sourceforge: http://sourceforge.net/projects/win2vnc/ Caveat: I haven't tried this one yet, but you can bet I will when I get a chance...
X2VNC: Same thing, but "master" desktop (the one with the shared kbd/mouse) is on a Unix box.
It really is one of the most useful pieces of software I use, and I'm *really* surprised it hasn't shown up in any other post in this entire discussion on VNC!
(Of course, there's X2VNC if the "master" desktop is X-based rather than Windows-based. Personally, I'm *way* too addicted to a few dozen critical apps that only exist in the Windows world.)
The RFB protocol (which VNC uses) has specified ServerCutText and ClientCutText messages since protocol version 3.3... You just need to find server and client implementations that actually support it.:)
Several of them do - I've been using VNC for all kinds of things and across all kinds of platforms for many years - Until last year, I used Konstantin Kaplinsky's TightVNC, but it's gotten a bit dated and UltraVNC is more stable, plus the file transfer feature is very handy.
My current config is UltraVNC as the remote server and WinIIvnc on the viewer end - this produces a nearly perfect multiple computer/multiple screen envinronment, with a single shared keyboard, mouse, and copy/paste clipboard. (Although I used to use Linux and BSD on the desktop, I've moved back to an all-Windows desktop environment, so this is really the thing I care about most...)
This is the standard dekstop setup for me whenever I'm in my office - I use the laptop as a second screen, which is especially handy for keeping a documentation window (web, pdf, etc.) in view while working on the other screen, and having copy/paste between them is often quite handy, for fairly obvious reasons. Other than having two Start buttons and taskbars, it's just like having a nice multiple-screen computer...
1992: 48% voted R and won 30%(9) of the US House seats 50% voted D and won 70%(21) of the US House seats
1994: 56% voted R and won 37%(11) of the US House seats 42% voted D and won 63%(19) of the US House seats
1996: 54% voted R and won 43%(13) of the US House seats 44% voted D and won 57%(17) of the US House seats
1998: 52% voted R and won 43%(13) of the US House seats 44% voted D and won 57%(17) of the US House seats
2000: 48.0% voted R and won 43%(13) of the US House seats 46.7% voted D and won 57%(17) of the US House seats
2002: 53% voted R and won 47%(15) of the US House seats 44% voted D and won 53%(17) of the US House seats
Someone please mod uppackman97's parent post. As a native Texan myself, the amount and underhandedness of the Democrat's cheating in Texas politics is beyond staggering - there is literally no depth to which they will not stoop to retain thier seats, legally or illegally. They still have their panties in a wad about the fact that W managed to somehow become governor in spite of their best efforts - turns out they just can't seem to cheat enough (yet) to outwiegh the opinions of legitimate voters.
If you doubt the Democrat's adoption of cheating as an actual strategy here in Texas, remember that Travis County was Ground Zero for the recent low in sleazy presidential politics: Ben Barnes and the Democrats using CBS news to broadcast forged documents in an effort to defame a sitting president and influence a national election. That sort of underhanded crap ought to infuriate people regardless of party affiliation.
Designed and implemented Wget. Personally, I feel wget is the greatest software every to hit the GNU/Linux desktop!
I'd have to go with Daniel Stenberg of cURL fame in this category. If you are still using wget, then try cURL. A lot of people only know wget, and that's a shame, because cURL is better in almost every possible dimension: see the table comparing cURL to wget and others to see for yourself. Not only that, but cURL is much more actively maintained and improved than wget.
While wget isn't a bad place to start, it's good to know there is a far more powerful alternative out there.
And, of course, it's part of far more than just Linux desktops - Apple even saw fit to make it part of OS X, and I routinely use it on XP, my own desktop OS of choice, as well several Linux and BSD-based servers.
It's an incredibly useful and valuable piece of code, and will become even more so in the future...
What sort of a shop are you running anyway, a sweat shop?
You're only going to be able to keep the mundane workers with that attitude. Anyone with any spark in them will be gone yesterday.
Doesn't matter, though, since all non-customer-contact positions will be outsourced to Bangalore, anyway. (A near quote from a Trilogy spokesman in today's TechMonday section of the Austin American Statesman...)
On the other hand, what percentage of "Alice"'s office data bases ever get anywhere close to getting scaled up to the whole enterprise? In most cases, I'd guess that the number is pretty darn small, tending toward zero.
I offer a real-world example above proving it isn't zero. I've seen just such a system developed in just that way being used to run a serious enterprise. It can be done, although it's admittedly uncommon.
It's not FileMaker, but it looks pretty good, is cross platform, plays well with others, and seems to have the major features required.
From the link above:
Currently, Rekall supports the following database formats:
* MySQL
* PostgreSQL
* XBase with XBSQL (an SQL wrapper library for the XBase access library)
* IBM DB2
* ODBC
The above list will be expanded later. We plan to add drivers for Oracle, MS SQL Server/Sybase and Interbase/Firebird. Please note the ODBC and DB2 drivers are not included in the standard edition of Rekall; they have to be purchased separately. Also you will need a fully licensed copy of the Database Server for the selected driver.
...
Rekall can do all the things that you would expect of a database front-end (or if it can't, let us know!). You can design and use forms and reports, construct database queries, and import and export data in several different formats (actually, you can copy data -- import is just copy file to table, and export is just copy table to file).
This is an excellent question, and one that you should really think out before changing at all. Don't buy into all the snobbery that you have to have a relational database in order to do anything useful. There's nothing magic about the relational model, and some significant drawbacks.
Over 10 years ago (1989-90), I worked for Hughes Diamond Products, at that time the diamond bit division of Hughes Tool Company, the largest maker of oil and gas driling bits in the world. (Yes, this is the same company that was making Howard Hughes over a million dollars a day in the depths of the Depression, back when a million dollars was a lot of money:-))
All of Diamond Product's ordering, puchasing, inventory, invoicing, and shop floor paperwork packets were in FileMaker. The entire system was built directly by the people that used it: secretaries, admins, production planners, engineers and production supervisors - not a DBA or database expert in the crowd. It worked extraordinarily well - well enough to run a company doing tens of millions of dollars a year, and well enough to convince the Hughes brass that investing in diamond bits was smart, so they paid $800 million for Eastman Christensen, the largest diamond bit firm in the world. Unfortunately, the SEC made Hughes divest themselves of Diamond Products, so I worked myself out of a job, and Dresser absorbed Hughes and replaced the remarkably effective Filemaker system that helped the company reshape the entire industry.
I certainly had a *lot* of respect for FileMaker after that. As is usually the case, it's not the tools, it's what you do with them. You can do very serious work with FileMaker - don't change unless you have a good reason...
Perhaps one of the strongest recommendations for FileMaker was that I never had to understand how that system worked, or even deal with it very much, even though my job was to coordinate CAD, CAM, and this home-grown manufacturing system. There were a few flat file interfaces written and the rest just worked. It really was amazing.
BTW, if you're determined to change, though, you might want to consider The Kompany's Rekall as a modern cross-platform DB front-end. I'm looking at it now to simplify building and maintaining databases to generate web sites while avoiding platform and database lock-in. (It works with almost any SQL DB, whether you run Windows or Linux.) Check it out.
Using files are both files and directories is really nice - throw ACLs, metadata, whatever in a directory the same name as the file: access it as a file and it is the file, access it as a directory and it provides access to the metadata. It doesn't break things.
Wow! I'm going to have to read up on this (yes I admit to not having read the article yet...)
If it really works that way, we may well be witnessing the dawning of a very interesting new age for tools like Blosxom, NoSQL, and Starbase that really leverage the filesystem to help provide structure... The fact that all of those tools are tiny and offer awesome leverage is directly related to the ways they leverage filesystems structures to replace much more cumbersome conventional methods...
In my setup WEP offers no advantages whatsoever so I never bothered with it, but I guess that makes me just another dumb newbie in their survey.
The real problem isn't that people aren't using WEP (since any blackhat with a web browser to download the tools can crack WEP in a few hours at most.)
The REAL problem is that ALL low-cost "wireless gateway" appliances treat wireless nodes as part of the LOCAL network, when, of course, the wireless segment should be treated as another WAN (Internet) link, where the bad guys live, and where you have to authenticate yourself before connecting to the LAN. As long as this remains true, wireless will continue to be a huge security hole in most networks.
Unfortunately, the "business" networking vendors are more than happy with this arrangement, since it keeps savvy business users from buying their network gear at CompUSA or Fry's. The sad fact is that security comes at a very serious cost premium today - it shouldn't, but the factis that companies that value security will pay *much* more for it, so the vendors simply "de-feature" the mass market products to help justify "enterprise" capabilities such as this common-sense approach to wireless networks.
This won't change until one of the SoHo/Home market vendors gets a clue and decides that their buyers might actually like a wireless router that can protect the rest of their network. Why that hasn't happened yet is a mystery.
BTW: If anyone knows of a low-cost wirless router device that *can* treat wireless as an "outside" network, post a reply and let us know...
Maybe, but no ads. Remember, corporate, not ads to distract the lusers [sic]
Netscape has NEVER had ads. That's Opera. It's getting tiresome reading the ~10% of the comments in this thread perpetrating this LIE.
Anyone who says the open source crowd isn't against business and commerce has never seen Slashdot savage a better-tested version of the open source browser they claim to love, just because its's distributed by - gasp- a corporation!
Does.... not.. compute... NS 4 was the reason I landed on Mozilla back in those rough pre 1.0 days. Anything * was better than having a crash on every other page.
No, sorry, but *your* stement doesn't compute - From a stability point of view, it took Mozilla until 1.4 to get anywhere near as stable as stable and functional as 4.x. This is why I didn't use a Mozilla-based browser until Netscape 7.0 came out. Thankfully, Mozilla has improved a lot, but traditionally, the Netscape versions were faster, cleaner, and significantly more stable than the Mozilla equivalents. I only moved off of Netscape 7.1 earlier this year, to get security fixes.
Further, there was significant functionality in 4.7, (especially mail, bookmarks, and Palm sync) that is still missing in the Mozilla-based browsers to this day. (Similarly, Firefox still offers only a subset of the Mozilla browser's functionality - I'm not sure I agree that Firefox is the better direction for the future, and apprently, niether do a lot of the Mozilla developers, since they've announced that the regular Mozilla is NOT going away as the Moz foundation originally planned...)
It's just fashionable to bash 4.x these days because its horrendous support for CSS and web standards makes it a real PITA for web site builders. Fortunately, there are now so few 4.x browsers out there that they can finally be ignored.
But really, 4.x was quite stable unless your underlying Win9X platform had problems. I wouldn't want to go back, but it was by far the best browser available at the time. (The instability some people experienced was sometimes due to incompatibilities in the user profiles, a problem that bedevils Mozilla-based browsers even today - I spent half a day finding and fixing bizarre profile and extension incompatibilities when upgrading from Mozilla 1.6 to Mozilla 1.7. That is NOT a user-friendly experience. WAnting to avoid that crap is one reason I may switch back to the more tested and stable Netscape version of the Mozilla code. (Tossing out a few AOL icons is a small price to pay for productivity...)
what does netscape bring to the table that Mozilla does not?
TESTING! SMALL FIXES! EXTRA STABILITY!
I second this. I've used Netscape and Mozilla exclusively since 1993 (barring several short trials of IE, always dropped in a day or two), and my experience is that the Netscape-branded versions are significantly more bug-free than Mozilla's releases. The Moz releases generally include the latest half-finished features at the expense of stability, while the Netscape versions just work better. (Of course, you do have to toss all the AOL and chat crap - why the heck does anyone think any chat client is a good idea except as a virus hose anyway? To be fair, you have to hack Maim (or is it chatzilla?) out of Mozilla, too, so it's not really that different. Despite the misinformation here, there are NO ADS in Netscape's releases. (I haven't yet tried the new 7.2 yet.)
The Mozilla team apparently thinks it's a good idea to significantly expand functionality, but force users to configure that functionality by hand with error-prone config file hacking. (Multiple mail profiles is a good recent example - nice functionality with NO user interface!) This is not the way to be considered a serious alternative to IE...
Slashdot Mirror
I would like to see a build in page validator.
You can download an HTML validator for Firefox that builds it right into View Source. It will validate it within the browser and also provide accessibility warnings. It's based on Tidy.
While Composer is certainly handy, and one of the big reasons I use Mozilla rather than Firefox & Thunderbird (other than that Mozilla hasn't scrogged any mail files for years and even the latest T-bird sometimes does), it's hardly a serious tool for building web pages (even in its latest Nvu guise.)
I find it vital though for things like quickly correcting serious brain damage in pages that either won't print, will print with a zillion little adsd taking up half the page, or will use 2 lbs of color laser toner if you try.
For that, especially if you're building any sort of modern web standards site using CSS, you need tools designed for that - TopStyle is my favorite, and it also has a built-in validator based on Tidy.
I'll say this - I don't buy too much software anymore, but TopStyle is one of those rare programs that is actually worth significantly more than the selling price. I wish I'd bought it six months before I finally did. It's also now one more reason I'm not moving from a Windows desktop - there is certainly nothing approaching its power and utility on Unix/Linux (and I'm a 20-year Unix bigot...)
Pseudoscience appeals to false authority, to emotion, sentiment, or distrust of established fact.
Thank you so much for declaring on behalf of us all that these experts in their field are practicing "psuedoscience" - we will now rest easy in your certainty that "real science" is never, ever, wrong. (And of course, the multibillion dollar wireless industry has nothing at stake, either, now does it?)
Someone please alert the FCC, FDA, and the universities, so they can summarily reject as "pseudoscience" any proposed research to determine if there might really be a threat posed by a technology so new (only a few years old!) that it's hard for anyone to know much yet.
We are truly fortunate you've saved us all the trouble and expense of actually conducting real scientific research on a *potentially* significant heatlh issue.
Whatever happened to science as a quest for knowledge (which implies truth)? I guess our political and emotional worldviews trump concerns for truth these days...
GSM starts with the highest possible signal when a connection gets established. The phone then decreases the signal to a level, which is sufficient to maintain the connection.
IIRC newer standards do it the other way around - the phone increases the level until the connection is maintainable.
No, actually, because the CDMA signal is much lower-power and noise-like by design, it's much more susceptible to the "near-far" problem, and so needs to manage transmitter power far more closely. It was primarily the problem of getting suitably "power agile" radios developed that slowed CDMA's entry to the market - it's a hard problem.
As i have heard, GSM is a little less harmful. Can someone with a bit more knowledge provide some insight?
Actually, although the evidence is far from conclusive, there have been several studies that indicate that GSM/TDMA is significantly more likely to be harmful than CDMA.
If you look at the power distribution across both frequency and time, the research shows it's pretty apparent why - in GSM/TDMA, there's silence, followed by a very quick high-power set of RF spikes (the ones and zeroes you transmit during your time slot) usually concentrated at a particular frequency. Several research studies seem to point to high-power fast rise time and fast fall time signals (the square waves of bits) as having not-yet-fully-understood detrimental biological effects that are apparently not due to heating per se. The number of studies showing the existence of such harmful effects does seem to be growing over time, although the data is "noisy" at best.
(Remember that polarized molucules like water, sugars, and many proteins found in abundance in the brain will try to align themselves with each successive polarity reversal. At high powers, it is this molecular flipping that generates heat in microwave ovens. Apparently, just stirring those molecules around even gently may not be a very good idea, even if it's not at the very high powers required to generate significant heating.)
CDMA, on the other hand, has the distinct advantage of not having high-power RF spikes, but has a much lower amplitude signal (still square waves, but a whole bunch of much smaller ones spread out across both frequency and time.) The signal literally looks like noise, and it's designed to: the "chipping sequence" mixed with the data in CDMA type modulations is sometimes referred to as a "PN" (for "Pseudo Noise") code. (There are a set of these PN codes that are mathematically orthogonal to each other, these are the "gold codes" that are most useful for CDMA or other direct sequence spread spectrum communications. The math behind this cool, hairy, and above my pain threshold...)
I had to learn about this stuff about ten years ago when I was doing a lot of early wireless work (both voice and data) in hospitals, which are justifiably paranoid about RF interference, especially since medical devices are exempt from most FCC regulation and sadly, are therefor often much more susceptible to RF interference than commercial devices. The week after I learned what I've set out here, I traded my old phone in for a new Sprint CDMA PCS phone - I don't plan to ever have a GSM/TDMA "CancerPhone".
(BTW, Richard Branson, head of the Virgin Group that includes Virgin Wireless, will not use a cellphone at all, for whatever that's worth - I'm not sure what the availability of CDMA is in the UK, though; GSM/TDMA may be the only choice...)
I don't have all my bookmarks on the topic handy now, (and they're probably badly out of date anyway), but some persistent web searching should turn up much of the medical research that does exist - only those that deny reality can say there are no health effects of cellphone RF - we know those effects exist, but we don't know what all is behind them, or what risk is posed by those effects. Such research is made especially difficult by the rapid pace of technology in the wireless business - we may not really be able to generate meaningful staistical studies for decades yet. Until we do, I'm going to play it as safe as I can, and as far as I can tell, that means CDMA (Sprint and Verizon mostly, in the US.)
One bad traffic/train accident and half of Texas loses their internet connections. Yes, that's a simplification and/or exaggeration...
Actually, it's already not an exaggeration. There is a small town in east Texas (no names - I'd rather not tip off the Muslim terrorists) that is the terminal point for a surprisingly large portion of the fiber in this country. Taking out this otherwise rather ordinary town could put a very serious dent in the US voice/data network.
...in the United States, since before the Civil War, people are Americans first, well except for Texas.
Well, we tried to reassert our rights to self determination 140 years ago, but were forced at the bloody point of a bayonet to re-join "the Union"; this despite the fact that Texas was an independent nation before joining the US, and had (and still has) every right to reclaim its sovereignty. Of course, Abraham Lincoln was dead-set on doing *anything* to prevent real government by the people - such as, for instance, creating another state by decree (splitting Virginia) to ensure a cooperative majority in Congress.
Like more than a few Texans I know, I occasionally list "Texas" as my country on forms (this can be a challenging sport with web forms.) Texas - It's a state of mind.
BTW, for those posters earlier that didn't realize that Texas really is pretty darn big I have this story: Several years ago, I did some work for a firm in Rhode Island. Not knowing any better, I made reservations to fly into Providence and drove to Narragansett (beautiful drive BTW, I don't regret it a bit.) My hosts were shocked that I would consider "driving clear across the state" until I pointed out to them that my daily commute across Houston (which was admittedly quite long, but near work before taking a cross-town job) would not fit inside their entire state!
Anyway, I have very mixed feelings about this project, although I could probably get behind it pretty well if one of more of the passenger car lanes had no speed limit (or a "reasonable and prudent" limit that was recognized to be well in excess of 100 MPH in good weather.) Why are we the first generation in human history to travel slower in the aggregate than the previous generation?
No joke about permissions - I would give ANYTHING to have file / directory permissions set up the way Netware's NDS (Netware 4.x and newer) does it.
Come to think of it, now that Novell is hacking on the Suse code base - how about adding NDS as the file permission scheme?
Because that only gets you halfway to where we want to be? I agree completely that the winning setup is to combine the best of Netware with the best of Linux. What we really need is the best-in-class granularity of permissions of the Netware filesystem (which is what you see reflected in NDS) AND a network file server that uses NCP, all combined with the goodness of a Linux-based server environment.
For those that don't remember or know, NCP is Netware Core Protocol - a simple, lightweight, richly functioned file sharing protocol that was the heart of the old Netware systems that could easily support hundreds of simultaneous users on 66 MHz '486 hardware. NCP absolutely kicks the snot out of both NFS and SMB/Samba, and is *by far* the best of the widely-deployed network filesystems out there. Don't overlook the fact that it's super lightweight design can allow current and future embedded network devices to easily access filer server resources, something that's not really practical in most cases with SMB or NFS.
Why Novell hasn't added NCP server capabilities to SuSE is beyond me - it would be the single best thing they could do to reclaim relevance, sell new stuff into their still fairly large installed base, and take a chunk out of MS by providing a real server alternative that's not just different (like Samba), but actually *better*.
Although not trivial, it's also not that hard - and if done correctly, it would successfully blend the strengths of NCP-based file services and the Linux server environment. I know if I were running the show over at Novell, an crash skunkworks effort along these lines would be a top strategic priority...
This is radically different from all times past. Many modern "discoveries" were in fact known in the classical world, two thousand years ago--we have evidence of simple steam engines, clockworks, etc. The difference between then and now is that today we have a system for publishing and exploiting such discoveries.
Publishing makes most of the difference. Well, that and the fact that had there been patents to secure a proprietary right to an invention for some period of time, there would have been a profit incentive to propagate those technologies.
In both the examples you use, we have the evidence that those technologies were developed (although quite crudely), but not really used and certainly not propagated - Hero's aeleopile was just a steam-powered toy or curiosity not useful for anything. Raal steam power would wait centuries for Thomas Newcomen and James Watt to harness steam as power. The Antikythera mechanism (the only extant example of ancient gearworks), although a very impressive product of the superior Rhodian mathematics and culture of 80 BC, lacked some very fundamental understandings of mechanics - for instance, the gear teeth are triangular, showing that its makers were astute mathematicians, but had not got far enough to realize that same level of math could solve the wear and binding problems created by such crude gear teeth. The Antikythera mechanism is a true anomaly, though: There is no other known clockwork device using differential gearing until AD 1575, although possibly similar devices are referenced by Cicero during his studies in Rhodes in 79-77 BC. (I know a bit about this gadget because it happens to be a hobby of mine...)
A final note: Reverse engineering has *nothing* to do with the justification of patents. I have no idea where you got that idea. Patents and trade secrets are completely different beasts, and by definition, if you patent something, the way it works is no longer a trade secret.
This is *exactly* the problem - the patent reforms (read abolition) most people in the open source community favor would actually just ensure that the Microsofts and IBMs of the world could never be successfully challenged by two guys in a garage. And yes, such challenges actually happen (and succeed) much more often than you might expect.
For a few good examples, here is an excerpt from a letter to the editor I wrote to LWN a few years ago that outlines why most "patent reform" is a very bad thing: (original URL: http://lwn.net/2000/0420/backpage.phtml#backpage
Tried both RealVNC and TightVNC and the "original" won at least in a LAN environment.
This is NOT surprising - Konstantin Kaplinsky, the author of TightVNC, made it clear in his original announcemnt of the product that the Tight protocol was designed to speed up VNC over high latency, low-bandwidth (dialup) connections - he even went out of his way to point out that the Tight protocol is usually slower than the plain old VNC protocol when running across a fast, low latency connection like a LAN.
I don't have a link handy, but it you look at the archives of the old vnc-list@uk.research.att.com list (assuming it's still out there somewhere...), you'll find his postings pretty much as I've described...
TS is superior to all VNC products, period. No one can develop something like TS without full access to a lot of windows code.
That's because the RDP protocol just flat kicks the VNC protocol's butt. If all (or even most) of the issue was hooking the video driver, then UltraVNC would be as fast as RDP implementations - but it's nowhere even close.
Big Question: the RDP protocol is (IIRC) publicly documented by MS - why hasn't anyone written an RDP client/server set for Unix/Linux, so we can all get the benefit of the superior protocol performance of RDP. (And even it the protocol is not publicly documented, what it's doing is relatively straightforward and would be *much* easier than what the Samba team's done to figure out and document the SMB protocol...
(For that matter, now that Novell's got open source religion, why can't we get them to open-source NCP, so we can all use a network filesystem that clearly outclasses both SMB and NFS? Is there still an NCP client included in newer versions of Windows? (It was there as recently as NT4 and I think W2K...))
RealVNC: the original.
TightVNC: optimized for low-bandwidth
Ultra: tons of extras - file transfer, chat, video driver, NT/AD security
Tridia: get around firewalls, more management features
I miss anything?
Win2VNC: One virtual desktop across two computers, using kbd/mouse of Win2VNC machine to run both(even does copy/paste correctly between screens if you use a good modern server like Ultra.) BTW: I've been told that the Win2VNC link I posted above has been superseded by a new version that adds support for mouse wheels, Alt-Tab, and other useful stuff. You can find it at Sourceforge: http://sourceforge.net/projects/win2vnc/ Caveat: I haven't tried this one yet, but you can bet I will when I get a chance...
X2VNC: Same thing, but "master" desktop (the one with the shared kbd/mouse) is on a Unix box.
Sorry, that's Win2VNC, not winIIVNC: Here's the link:http://fredrik.hubbe.net/win2vnc.html
It really is one of the most useful pieces of software I use, and I'm *really* surprised it hasn't shown up in any other post in this entire discussion on VNC!
(Of course, there's X2VNC if the "master" desktop is X-based rather than Windows-based. Personally, I'm *way* too addicted to a few dozen critical apps that only exist in the Windows world.)
The RFB protocol (which VNC uses) has specified ServerCutText and ClientCutText messages since protocol version 3.3... You just need to find server and client implementations that actually support it. :)
Several of them do - I've been using VNC for all kinds of things and across all kinds of platforms for many years - Until last year, I used Konstantin Kaplinsky's TightVNC, but it's gotten a bit dated and UltraVNC is more stable, plus the file transfer feature is very handy.
My current config is UltraVNC as the remote server and WinIIvnc on the viewer end - this produces a nearly perfect multiple computer/multiple screen envinronment, with a single shared keyboard, mouse, and copy/paste clipboard. (Although I used to use Linux and BSD on the desktop, I've moved back to an all-Windows desktop environment, so this is really the thing I care about most...)
This is the standard dekstop setup for me whenever I'm in my office - I use the laptop as a second screen, which is especially handy for keeping a documentation window (web, pdf, etc.) in view while working on the other screen, and having copy/paste between them is often quite handy, for fairly obvious reasons. Other than having two Start buttons and taskbars, it's just like having a nice multiple-screen computer...
http://www.fairvote.org/dubdem/tx.htm
1992:
48% voted R and won 30%(9) of the US House seats
50% voted D and won 70%(21) of the US House seats
1994:
56% voted R and won 37%(11) of the US House seats
42% voted D and won 63%(19) of the US House seats
1996:
54% voted R and won 43%(13) of the US House seats
44% voted D and won 57%(17) of the US House seats
1998:
52% voted R and won 43%(13) of the US House seats
44% voted D and won 57%(17) of the US House seats
2000:
48.0% voted R and won 43%(13) of the US House seats
46.7% voted D and won 57%(17) of the US House seats
2002:
53% voted R and won 47%(15) of the US House seats
44% voted D and won 53%(17) of the US House seats
Someone please mod uppackman97's parent post. As a native Texan myself, the amount and underhandedness of the Democrat's cheating in Texas politics is beyond staggering - there is literally no depth to which they will not stoop to retain thier seats, legally or illegally. They still have their panties in a wad about the fact that W managed to somehow become governor in spite of their best efforts - turns out they just can't seem to cheat enough (yet) to outwiegh the opinions of legitimate voters.
If you doubt the Democrat's adoption of cheating as an actual strategy here in Texas, remember that Travis County was Ground Zero for the recent low in sleazy presidential politics: Ben Barnes and the Democrats using CBS news to broadcast forged documents in an effort to defame a sitting president and influence a national election. That sort of underhanded crap ought to infuriate people regardless of party affiliation.
Designed and implemented Wget.
Personally, I feel wget is the greatest software every to hit the GNU/Linux desktop!
I'd have to go with Daniel Stenberg of cURL fame in this category. If you are still using wget, then try cURL. A lot of people only know wget, and that's a shame, because cURL is better in almost every possible dimension: see the table comparing cURL to wget and others to see for yourself. Not only that, but cURL is much more actively maintained and improved than wget.
While wget isn't a bad place to start, it's good to know there is a far more powerful alternative out there.
And, of course, it's part of far more than just Linux desktops - Apple even saw fit to make it part of OS X, and I routinely use it on XP, my own desktop OS of choice, as well several Linux and BSD-based servers.
It's an incredibly useful and valuable piece of code, and will become even more so in the future...
What sort of a shop are you running anyway, a sweat shop?
You're only going to be able to keep the mundane workers with that attitude. Anyone with any spark in them will be gone yesterday.
Doesn't matter, though, since all non-customer-contact positions will be outsourced to Bangalore, anyway. (A near quote from a Trilogy spokesman in today's TechMonday section of the Austin American Statesman...)
On the other hand, what percentage of "Alice"'s office data bases ever get anywhere close to getting scaled up to the whole enterprise? In most cases, I'd guess that the number is pretty darn small, tending toward zero.
I offer a real-world example above proving it isn't zero. I've seen just such a system developed in just that way being used to run a serious enterprise. It can be done, although it's admittedly uncommon.
It's not FileMaker, but it looks pretty good, is cross platform, plays well with others, and seems to have the major features required.
From the link above:
More to the point, why even upgrade to 7 ?
:-))
This is an excellent question, and one that you should really think out before changing at all. Don't buy into all the snobbery that you have to have a relational database in order to do anything useful. There's nothing magic about the relational model, and some significant drawbacks.
Over 10 years ago (1989-90), I worked for Hughes Diamond Products, at that time the diamond bit division of Hughes Tool Company, the largest maker of oil and gas driling bits in the world. (Yes, this is the same company that was making Howard Hughes over a million dollars a day in the depths of the Depression, back when a million dollars was a lot of money
All of Diamond Product's ordering, puchasing, inventory, invoicing, and shop floor paperwork packets were in FileMaker. The entire system was built directly by the people that used it: secretaries, admins, production planners, engineers and production supervisors - not a DBA or database expert in the crowd. It worked extraordinarily well - well enough to run a company doing tens of millions of dollars a year, and well enough to convince the Hughes brass that investing in diamond bits was smart, so they paid $800 million for Eastman Christensen, the largest diamond bit firm in the world. Unfortunately, the SEC made Hughes divest themselves of Diamond Products, so I worked myself out of a job, and Dresser absorbed Hughes and replaced the remarkably effective Filemaker system that helped the company reshape the entire industry.
I certainly had a *lot* of respect for FileMaker after that. As is usually the case, it's not the tools, it's what you do with them. You can do very serious work with FileMaker - don't change unless you have a good reason...
Perhaps one of the strongest recommendations for FileMaker was that I never had to understand how that system worked, or even deal with it very much, even though my job was to coordinate CAD, CAM, and this home-grown manufacturing system. There were a few flat file interfaces written and the rest just worked. It really was amazing.
BTW, if you're determined to change, though, you might want to consider The Kompany's Rekall as a modern cross-platform DB front-end. I'm looking at it now to simplify building and maintaining databases to generate web sites while avoiding platform and database lock-in. (It works with almost any SQL DB, whether you run Windows or Linux.) Check it out.
Using files are both files and directories is really nice - throw ACLs, metadata, whatever in a directory the same name as the file: access it as a file and it is the file, access it as a directory and it provides access to the metadata. It doesn't break things.
Wow! I'm going to have to read up on this (yes I admit to not having read the article yet...)
If it really works that way, we may well be witnessing the dawning of a very interesting new age for tools like Blosxom, NoSQL, and Starbase that really leverage the filesystem to help provide structure... The fact that all of those tools are tiny and offer awesome leverage is directly related to the ways they leverage filesystems structures to replace much more cumbersome conventional methods...
In my setup WEP offers no advantages whatsoever so I never bothered with it, but I guess that makes me just another dumb newbie in their survey.
The real problem isn't that people aren't using WEP (since any blackhat with a web browser to download the tools can crack WEP in a few hours at most.)
The REAL problem is that ALL low-cost "wireless gateway" appliances treat wireless nodes as part of the LOCAL network, when, of course, the wireless segment should be treated as another WAN (Internet) link, where the bad guys live, and where you have to authenticate yourself before connecting to the LAN. As long as this remains true, wireless will continue to be a huge security hole in most networks.
Unfortunately, the "business" networking vendors are more than happy with this arrangement, since it keeps savvy business users from buying their network gear at CompUSA or Fry's. The sad fact is that security comes at a very serious cost premium today - it shouldn't, but the factis that companies that value security will pay *much* more for it, so the vendors simply "de-feature" the mass market products to help justify "enterprise" capabilities such as this common-sense approach to wireless networks.
This won't change until one of the SoHo/Home market vendors gets a clue and decides that their buyers might actually like a wireless router that can protect the rest of their network. Why that hasn't happened yet is a mystery.
BTW: If anyone knows of a low-cost wirless router device that *can* treat wireless as an "outside" network, post a reply and let us know...
Maybe, but no ads. Remember, corporate, not ads to distract the lusers [sic]
Netscape has NEVER had ads. That's Opera. It's getting tiresome reading the ~10% of the comments in this thread perpetrating this LIE.
Anyone who says the open source crowd isn't against business and commerce has never seen Slashdot savage a better-tested version of the open source browser they claim to love, just because its's distributed by - gasp- a corporation!
Netscape 4.x? Decent? In same sentance??
Does.... not.. compute...
NS 4 was the reason I landed on Mozilla back in those rough pre 1.0 days. Anything * was better than having a crash on every other page.
No, sorry, but *your* stement doesn't compute - From a stability point of view, it took Mozilla until 1.4 to get anywhere near as stable as stable and functional as 4.x. This is why I didn't use a Mozilla-based browser until Netscape 7.0 came out. Thankfully, Mozilla has improved a lot, but traditionally, the Netscape versions were faster, cleaner, and significantly more stable than the Mozilla equivalents. I only moved off of Netscape 7.1 earlier this year, to get security fixes.
Further, there was significant functionality in 4.7, (especially mail, bookmarks, and Palm sync) that is still missing in the Mozilla-based browsers to this day. (Similarly, Firefox still offers only a subset of the Mozilla browser's functionality - I'm not sure I agree that Firefox is the better direction for the future, and apprently, niether do a lot of the Mozilla developers, since they've announced that the regular Mozilla is NOT going away as the Moz foundation originally planned...)
It's just fashionable to bash 4.x these days because its horrendous support for CSS and web standards makes it a real PITA for web site builders. Fortunately, there are now so few 4.x browsers out there that they can finally be ignored.
But really, 4.x was quite stable unless your underlying Win9X platform had problems. I wouldn't want to go back, but it was by far the best browser available at the time. (The instability some people experienced was sometimes due to incompatibilities in the user profiles, a problem that bedevils Mozilla-based browsers even today - I spent half a day finding and fixing bizarre profile and extension incompatibilities when upgrading from Mozilla 1.6 to Mozilla 1.7. That is NOT a user-friendly experience. WAnting to avoid that crap is one reason I may switch back to the more tested and stable Netscape version of the Mozilla code. (Tossing out a few AOL icons is a small price to pay for productivity...)
what does netscape bring to the table that Mozilla does not?
TESTING! SMALL FIXES! EXTRA STABILITY!
I second this. I've used Netscape and Mozilla exclusively since 1993 (barring several short trials of IE, always dropped in a day or two), and my experience is that the Netscape-branded versions are significantly more bug-free than Mozilla's releases. The Moz releases generally include the latest half-finished features at the expense of stability, while the Netscape versions just work better. (Of course, you do have to toss all the AOL and chat crap - why the heck does anyone think any chat client is a good idea except as a virus hose anyway? To be fair, you have to hack Maim (or is it chatzilla?) out of Mozilla, too, so it's not really that different. Despite the misinformation here, there are NO ADS in Netscape's releases. (I haven't yet tried the new 7.2 yet.)
The Mozilla team apparently thinks it's a good idea to significantly expand functionality, but force users to configure that functionality by hand with error-prone config file hacking. (Multiple mail profiles is a good recent example - nice functionality with NO user interface!) This is not the way to be considered a serious alternative to IE...