Sections: Main page Linux in the news Security Kernel Distributions Development Commerce Announcements Back page All in one big page See also: last week's Back page page. |
Linux links of the weekThe Linux Music Station is another site dedicated to Linux and musical applications, with an emphasis on MP3 and recorded music. Looking for a Linux presentation? LinuxSlides.com contains pointers to all of the online presentations about Linux they are able to find. Talks are available in both English and French (with one each in Spanish and Japanese as well). Section Editor: Jon Corbet |
September 2, 1999 |
|
Letters to the editorLetters to the editor should be sent to letters@lwn.net. Preference will be given to letters which are short, to the point, and well written. If you want your email address "anti-spammed" in some way please be sure to let us know. We do not have a policy against anonymous letters, but we will be reluctant to include them. | |
Date: Wed, 1 Sep 1999 12:33:05 -0400 (EDT) From: Derek Glidden <dglidden@illusionary.com> To: <letters@lwn.net> Subject: "Star Office Portal" is a bad idea At first I was worried that Sun's acquisition of StarOffice would be a bad thing for the Linux community. Sun doesn't have the best track record at being supportive of Linux. (Witness the difficulties the Blackdown team have had dealing with the JDK port, at least partially because of the restrictiveness of Sun's "Community Source License"; the announcement that the previously forthcoming support for Linux for Netscape servers was no more.) The fact that Sun has decided to release the source for StarOffice via the SCSL alleviates some of my fears, but only some, and raised new ones because I think their "Star Office Portal" idea is a huge mistake that may cost the application entirely. Let's face it, Larry Ellison has been predicting the imminent death of the PC and the rise and domination of server-driven "Network Computers" and "Internet Terminal" devices for years, and it just hasn't happened. Java is proving itself in behind-the-scenes server-side application progarmming, but it's still slow for full-function interactive user applications. People still for the most part have 28.8K, 33.6K or 56K modems at home and are not going to want to have to wait for their computer to connect to their ISP and then still have to wait for a couple megs of Java code to come down the pipe just so they can type a letter to Mom. There are all sorts of technical *and* social problems trying to work with a system that keeps your important personal documents on someone else's server and not on your local workstation where you have instant, easy access to them. (The recent Hotmail security fiasco is probably going to only highlight these problems.) And probably the most important reason (IMO) that this "Star Office Portal" is a bad idea is because it's not Microsoft Office. "Everyone uses Office" is still the way the world works and as long as it works that way, it will be a constant uphill battle just to _keep up_ with the changes Microsoft make in the way Office files are saved so you can read those documents from work. In a recent online article about the "Star Office Portal" idea, [can't find the URL anymore...] Sun's Scott McNealy's made the comment that, no matter where he goes, he never brings his desktop software with him -- it always stays conveniently on some server someplace where it can be easily accessed through his browser. I think this just shows how out of touch with the real world he is. That may work for Scott, but the vast majority of the "real" computer world have no idea what he's talking about. All they know is they click this icon and their word processor pops up and now they can load that file from the floppy disk where they've saved it. Keeping Scott connected to the internet is probably not a problem, but as a consultant, I need to be very mobile and I still encounter meeting rooms, offices and cubicles that are not wired, or use a different networking topology than I have or are stuck behind a firewall, any of which problems, if I were relying on my ISP or a server back at my office for access to my word processor and documents, would cut me off at the knees and leave me non-functional. Technically, the idea of a platform-independent, run-anywhere, works-the-same-anywhere, follow-me-around-the-world desktop complete with applications is a good idea. Practically, we've seen the industry attempt it half a dozen times and fail. (Anyone remember Netscape's "Atlas" universal desktop?) It's hard to get people to change a fundamental mode of working such as moving from a locally-stored application/data model to a server-stored application/data model. (It's even harder to get Microsoft to "allow" the industry to move in that direction.) There are people who argue this is just a return to mainframe life, only with smarter dumb terminals. I don't think it matters what the change is to or from, people just plain don't like changing the way they work. So, how does this relate to Linux, and not just leaving Sun and Scott with egg on their faces? Well, it sounds like Sun is putting a *lot* of resources into trying to make this "star Office Portal" thing work, and if it fails, as I think it will, that's going to make the whole Star Division acquisition look pretty bad and Sun is a commercial operation out to make money. Currently, StarOffice is only one of a small handful of complete office suites available for Linux and is only one of two or three with any sort of version history on the platform. If Sun decides that it's a bigger drain on their resources than it is worth and dumps the whole thing, Linux could lose it. Because of the restrictions of the SCSL, this doesn't mean some other group can pick up the code and fork it into their own version the way the GPL or a BSD license would allow - it means if Sun says bye-bye, it's gone. Sure, there are other office suites under development (KOffice, Corel Office for Linux, to name just a couple) that could conceivably replace it, but losing Star Office on Linux would make the Linux and Open Source community look bad ("See, another 'open source' project that failed, just like Mozilla!" is what you'll hear.) and, personally, I would rather rely on the software I have now, that works now, than have to bet my productivity on something that may be available and functional "in a few months." | ||
Date: Wed, 01 Sep 1999 23:56:27 -0400 From: "Jay R. Ashworth" <use-reply-to-address@gte.net> To: ewt@redhat.com CC: letters@lwn.net Subject: A rock and a hard place Just another penguin head, weighing in... I gather there's a fuss over the trademark. Lawyers can be such a pain, can't they? My opinion (backed by that of Mark McCormick, see _What_They_Didn't_Teach_Me_at_Yale_Law_School) is that you need to be very cautious of letting the lawyers run the company. It's not their job. Yes, you have a duty to protect your trademarks. Waiting to act until you have had time to make sure that your reaction will not impair your _primary reason_ for your current market cap -- that is: the goodwill of the community -- would _not_ have killed you. Legally, I mean. Next time, ask the lawyer who's rousing the rabble what, exactly, he thinks it _is_ that makes RHAT worth 5.3 _billion_ dollars. And see if he read the parable about golden eggs when he was a kid. _If_ he was a kid. :-) Tread lightly, guys; we love you, and there hasn't ever been a company whose 'goodwill' item carried more weight on the balance sheet. Cheers, -- jra | ||
Date: Sun, 29 Aug 1999 10:14:43 +0900 (JST) From: David Moles <deivu@tomigaya.shibuya.tokyo.jp> To: letters@lwn.net Subject: LASER5's divorce from RedHat One of the reasons cited in LASER5's press release is the high cost of license fees paid to RedHat. I wonder if that means that now they'll be charging a more reasonable price for their distribution? I was in a shop in Akihabara (Tokyo's main hardware, software, and electronic goodies area) and saw that LASER5's "RedHat Linux v6.0J Server" priced at about 45,000 yen -- which is to say about $400. (Yes, you read that right.) As far as I could tell the only thing it offered to distinguish itself from their stock version (8100 yen, or about $70) was SSH. I'm not really sure what's up with these Japanese distributions -- TurboLinux "Pro" was going for 24,000 yen (about $200), while the non- pro edition was about the same as the stock LASER5. It's not support -- LASER5's $400 gets you just three incidents, within the first 180 days. I suppose they might be paying license fees for commercial Japanese input methods -- certainly the free ones out there are kind of lame, though I don't expect that to last -- but if so, why is it the server edition that needs the fancy input? But then again, as a friend of mine pointed out, this is a country where people will pay $15 to see a movie, and then come out saying they would have paid $30 if they had to. :) --D | ||
To: letters@lwn.net Subject: Emacs vulnerability to macro viruses. From: Alan Shutko <ashutko@itms.com> Date: 26 Aug 1999 10:39:33 -0400 Although Emacs can evaluate arbitrary lisp code upon loading a file, there are significant differences between MS Word's and Emacs' openness to viruses. * Only variable setting is enabled by default. By default, attempting to evaluate lisp causes Emacs to query you whether to evaluate the given lisp. Word, OTOH, defaults to vulnerability. * Emacs lisp is human-readable. When Emacs asks whether to evaluate embedded lisp, it will show the lisp to you, so that the user can decide whether it's safe. MS Word will simply ask whether you want to run unspecified macros. MS considers it a feature that embedded code can be non-readable. * Embedded Emacs lisp is visible. You can see it at the bottom of the file, so it is less likely that a virus could exist unseen. So, while Emacs _could_ be vulnerable to viruses, it would take conscious effort on the part of a user to make it so. -- Alan Shutko <ashutko@itms.com> | ||
|