I've always been a wordperfect fan, and it's nice to get back to it, although the fonts are really killing me. I think I'll wait for the suite to come out before I consider a purchase, though. More to the point, I highly reccommend that y'all check out the latest KOffice. Yeah, it can be a bitch to get Qt2 and compile everything, but DAMN! That baby (especially KWord) is really coming along. --JZ
Well, I agree with some of your principle idea -- linux could be a great choice for systems that heavily emphasize one app (say, CAD or a database), because the vendor could heavily customize it for support. However, that would be a huge deviation from the vendor's core competency in, say, CAD software. A better possibility might be to work closely with a Linux distributor and have them offer it as a special version. That, however, risks locking out other distros. A tough situation all around. Hopefully DEC and SGI will be able to contribute some of the necessary muscle to XFree86 as they begin to really embrace linux on their platforms.
Anybody have any experience with 9netave.com? They seem pretty large, and they certainly can afford a substantial advertising budget. But the price is REALLY low. I mean, they hit $130 a month including hardware, hosting, and 5 GB of transfer (which is all my site needs), with an up-front of like $400. There was another one that had been advertising on/. a bit about their $150/month linux special too that looked pretty reasonable. We're looking for our own server that really only needs to do 100-150MB a day of throughput and run some server apps, so I'm hoping to keep things under $200. Any advice? Thanks! --JZ
I've been surprised for a while that there is no open source equivalent to say, Dreamweaver. I haven't even found a project in progress (I'd love to help), except for WebMaker for KDE, which is great, but much more of the HomeSite variety. Does the Mozilla project include Composer? I guess that'd be the best bet. Considering the amount of web work that goes on on linux, it's a shame we don't have a decent graphical editor. As for the ongoing source code vs. WYSIWIG debate: I do everything by hand, but I work on an online college newspaper that incorporates artists, programmers, and random volunteers. Do you guys really expect us to say "Yeah, thanks for volunteering, we're glad to see that you're an incredible graphic artist and a wiz at page layout, but get the hell out of here unless you're willing to learn HTML"? The fact is that everybody has his or her own style, and even I can appreciate the benefit of a graphical editor when you have to do tons of menial cutting and pasting in the midst of complex tables, or the like.
Hey, I've been looking around at different options for managing an online newspaper, and Zope seems pretty promising, but I was wondering if anybody could give me their opinion of its suitability. Basically, we need to update a large number of articles daily, keep some features rolling continuously, maybe add a chat board, and do some searching. But my big concern is the difficulty of moving existing HTML back issues (several hundred issues and thus many thousands of articles) to Zope. Would this be really tough?
Well, I'm no Java pro, but I've been fairly impressed by the speed of a good JIT on Solaris IF AND ONLY IF you don't let it get anywhere near a GUI. They have a fair number of the key mathematical functions implemented and optimized in native instructions, so those can choke their way by, but both Swing and AWT, easy as they are to work with, just plain suck in terms of power. Ever used HotJava? OH MY LORD! On an old-ish SparcStation, HotJava can't even keep up with scrolling. My understanding of the "server side only" statement was twofold: 1st translation) "Sorry, our GUIs still suck guys, come back next year." 2nd impression) Well, the dynamic optimizer has to sit an run for a while before it gets up to speed, so it's better for continually-running server apps. I really wish I could write consumer-level desktop apps in Java, but it just doesn't fly right now.
Hmm... As a first-year potential CS major at Princeton, I'd have to disagree with those generalizations. Our first three required intro classes (which many non-CS folks, including almost all engineers, often take) are all Unix and C based (with a little bit of Java thrown in for fun). The intro class, taken by about 1/3 of the total student population, teaches almost exclusively theory and concepts: FSAs, trees, operating systems, boolean logic, abstract data types, etc. Our algorithms class actually allows you to do assignments in any language that you choose, as long as it can run on Solaris. To my knowledge, the only time you have to touch an Intel PC is for OpSys, and then you run your own bootloader and system fragments, definitely not Windows. I had the exact same impression of computer science before I started here, and someone actually told me, "Whatever you learn will be outdated in 5 years." Now, though, I look back at the two main textbooks I've used over 2 semesters: Algorithms in C, and Kernighan and Ritchie's white book, and I realize that they're been around for about 10 and 20 years respectively. Hmm... Older than my German book, whaddaya know?
But the act of being able to share your story with someone outside of your own immediate friends (many of whom may have been in that same high school situation) can be the most useful, cathartic experience that a person can have. Honest expression of oneself is the foundation for all modern psychotherapy. I absolutely applaud/. for these articles.
Hmm... it seems like we've known the KDE thing for ages, but it didn't seem like Corel discussed the extent to which they will lead to a splinter in KDE. Will there be a 100% code similarity between the two desktops? It seems really unlikely (i.e. they really want to integrate the office suite into the desktop, while that would seem ridiculous as a part of the main distro). Any thoughts on whether we'll now have two development tracks or just a separate kcorel pack that works smoothly with existing stuff? Of course I'm hoping for the latter, so Corel can contribute to everybody's desktop, but who really knows?
As much as I think Hungarian notation is outdated and annoying in the C/C++ world, I think we've finally found a new application for it: Hungarian License Notation. You know, a "p" before the name indicates that modifications must be distributed as patches, "Csl" indicates that closed source programs may statically link to it, and "Dmf" means that you can distribute modified versions freely. So the LGPL would become something like the CslDmfLGPL instead. (If we're really cool, we can be like Visual C++ and have all sorts of prefixes that are no longer relevant. Don't they still have lp for long pointer in 6.0?)
Well, I can't speak to the installation procedure on RH 5.2, but for future releases Mandrake has been mumbling about X-based installation. Plus a default easy KDE setup. But I agree that Mandrake no longer has much of an edge, however, Bero has the key feature that it is Pentium optimized. Why the hell should we all be running code compiled for a 386? Is anyone actually going to be dumb enough to try and run GNOME on a 386? --JZ
Both of their sites have said for ages that the next release would be out in late February or MAYBE early March. Well, it's obviously March. Anybody heard updates/rumors? (I did hear the spiel about Netscape's failure to compile already.) Also, anyone have thoughts on Yellow Dog? They have yet to ship an actual product, but they seemed to get fairly good media coverage. How does it differentiate itself from a repackaged LinuxPPC distro (besides the high price)?
Good: Red Hat can hire some more folks full time to crunch bugs in GNOME and other projects. Bad: More news stories saying that Red Hat invented Linux. Worse yet, more vendors dying to install an "official" copy of RH Linux, jacking up the cost, rather than just installing a free copy of it. Why would I POSSIBLY want to pay for an installation manual and installation support when it's ALREADY INSTALLED?!? These people just don't get it. RH has always said that they'll get their revenue from support when things have taken off, but they're becoming just another software company. . . Future: Obviously a lot of shakeout in the market, but we'll continue to see more distros like BeroLinux and Linux Mandrake that build on Red Hat. Of course let's not forget that Corel's distrobution will shake things up a lot once they come in. Red Hat on all the servers, Corel on all the desktops? I don't think it'll be quite that stark, althogh I do think that KDE 2.0 (which I think we can assume Corel will use, although they're technically 'evaluating' GNOME too) will be REALLY good. Check out QT 2.0, on which it'll be based. Much slicker, plus support for the mouse wheel. Or don't. GNOME's buggy, but I do like the way gmc and gtk+ look. -JZ
RedHat outdated, but for a reason
on
Friday Quickies
·
· Score: 1
I agree that RedHat is completely outdated. They have their reasons, but I'm not very impressed by them: -Compiling on 386: because they like to advertise that linux runs on ancient hardware and it is already complicated enough to distribute separate Alpha and Sparc versions. -Slow pace of change: this is the most understandable. Linux changes so much faster than other OSes that it absolutely confuses regular resellers. I remember being impressed to see a couple dozen RH5.1 boxes in both Borders and CompUSA, but somewhat perplexed to see them sitting next to a dozen more RH5.2 boxes being sold for the same price! Many resellers asked RedHat to slow down the new releases, and I imagine that other distros had the same issue. -Fvwm. The only excuses here are the desire to make Linux look like win95 and the fact of RH's moral opposition to KDE. My opinion? For version 6.0, they need pentium optimization. Maybe they can ship out stickers for resellers to put on the old 5.2 boxes saying "Try 5.2 for older hardware!", or something to that effect. They should also go back to their old practice of adding updates within a vesion number. I hear they continued this up through a few version 5.1's, which have a slightly newer kernel than older 5.1's (although it might have been 5 when they stopped, I can't remember. ..). I know resellers wanted an end to this too, but the company should simply include very easy instructions for obtaining updates (how about the.deb update capability for the next version of RPM?) and say screw you to the resellers who complain, because consumers matter more than vendors. As for FVWM, they should have been investing more time in documenting and stabilizing WindowMaker (which NEVER crashes on me, I'd add, but I have heard rumors of stories of complaints) to include that as their default WM. It is far more usable than E or FVWM, and with help from RedHat it could have every bit as much GNOME integration as E. I don't mean to slander other distros (although SuSe stuck with libc5 for far too long and Debian, great though it is, should have scrapped Slink to work on 2.2), I just focused on RedHat because it has the most relevance to the topic at hand (BeroLinux, for those who are completely lost in my rambling). --JZ
Slashdot Mirror
I've always been a wordperfect fan, and it's nice to get back to it, although the fonts are really killing me. I think I'll wait for the suite to come out before I consider a purchase, though.
More to the point, I highly reccommend that y'all check out the latest KOffice. Yeah, it can be a bitch to get Qt2 and compile everything, but DAMN! That baby (especially KWord) is really coming along.
--JZ
Well, I agree with some of your principle idea -- linux could be a great choice for systems that heavily emphasize one app (say, CAD or a database), because the vendor could heavily customize it for support. However, that would be a huge deviation from the vendor's core competency in, say, CAD software. A better possibility might be to work closely with a Linux distributor and have them offer it as a special version. That, however, risks locking out other distros. A tough situation all around. Hopefully DEC and SGI will be able to contribute some of the necessary muscle to XFree86 as they begin to really embrace linux on their platforms.
Anybody have any experience with 9netave.com? They seem pretty large, and they certainly can afford a substantial advertising budget. But the price is REALLY low. I mean, they hit $130 a month including hardware, hosting, and 5 GB of transfer (which is all my site needs), with an up-front of like $400. There was another one that had been advertising on /. a bit about their $150/month linux special too that looked pretty reasonable.
We're looking for our own server that really only needs to do 100-150MB a day of throughput and run some server apps, so I'm hoping to keep things under $200. Any advice?
Thanks!
--JZ
I've been surprised for a while that there is no open source equivalent to say, Dreamweaver. I haven't even found a project in progress (I'd love to help), except for WebMaker for KDE, which is great, but much more of the HomeSite variety. Does the Mozilla project include Composer? I guess that'd be the best bet. Considering the amount of web work that goes on on linux, it's a shame we don't have a decent graphical editor.
As for the ongoing source code vs. WYSIWIG debate:
I do everything by hand, but I work on an online college newspaper that incorporates artists, programmers, and random volunteers. Do you guys really expect us to say "Yeah, thanks for volunteering, we're glad to see that you're an incredible graphic artist and a wiz at page layout, but get the hell out of here unless you're willing to learn HTML"? The fact is that everybody has his or her own style, and even I can appreciate the benefit of a graphical editor when you have to do tons of menial cutting and pasting in the midst of complex tables, or the like.
Hey, I've been looking around at different options for managing an online newspaper, and Zope seems pretty promising, but I was wondering if anybody could give me their opinion of its suitability. Basically, we need to update a large number of articles daily, keep some features rolling continuously, maybe add a chat board, and do some searching. But my big concern is the difficulty of moving existing HTML back issues (several hundred issues and thus many thousands of articles) to Zope. Would this be really tough?
Thanks a lot!
Well, I'm no Java pro, but I've been fairly impressed by the speed of a good JIT on Solaris IF AND ONLY IF you don't let it get anywhere near a GUI. They have a fair number of the key mathematical functions implemented and optimized in native instructions, so those can choke their way by, but both Swing and AWT, easy as they are to work with, just plain suck in terms of power. Ever used HotJava? OH MY LORD! On an old-ish SparcStation, HotJava can't even keep up with scrolling.
My understanding of the "server side only" statement was twofold: 1st translation) "Sorry, our GUIs still suck guys, come back next year." 2nd impression) Well, the dynamic optimizer has to sit an run for a while before it gets up to speed, so it's better for continually-running server apps.
I really wish I could write consumer-level desktop apps in Java, but it just doesn't fly right now.
Hmm... As a first-year potential CS major at Princeton, I'd have to disagree with those generalizations. Our first three required intro classes (which many non-CS folks, including almost all engineers, often take) are all Unix and C based (with a little bit of Java thrown in for fun). The intro class, taken by about 1/3 of the total student population, teaches almost exclusively theory and concepts: FSAs, trees, operating systems, boolean logic, abstract data types, etc. Our algorithms class actually allows you to do assignments in any language that you choose, as long as it can run on Solaris. To my knowledge, the only time you have to touch an Intel PC is for OpSys, and then you run your own bootloader and system fragments, definitely not Windows.
I had the exact same impression of computer science before I started here, and someone actually told me, "Whatever you learn will be outdated in 5 years." Now, though, I look back at the two main textbooks I've used over 2 semesters: Algorithms in C, and Kernighan and Ritchie's white book, and I realize that they're been around for about 10 and 20 years respectively. Hmm... Older than my German book, whaddaya know?
But the act of being able to share your story with someone outside of your own immediate friends (many of whom may have been in that same high school situation) can be the most useful, cathartic experience that a person can have. Honest expression of oneself is the foundation for all modern psychotherapy. I absolutely applaud /. for these articles.
Hmm... it seems like we've known the KDE thing for ages, but it didn't seem like Corel discussed the extent to which they will lead to a splinter in KDE. Will there be a 100% code similarity between the two desktops? It seems really unlikely (i.e. they really want to integrate the office suite into the desktop, while that would seem ridiculous as a part of the main distro). Any thoughts on whether we'll now have two development tracks or just a separate kcorel pack that works smoothly with existing stuff? Of course I'm hoping for the latter, so Corel can contribute to everybody's desktop, but who really knows?
--JZ
As much as I think Hungarian notation is outdated and annoying in the C/C++ world, I think we've finally found a new application for it: Hungarian License Notation. You know, a "p" before the name indicates that modifications must be distributed as patches, "Csl" indicates that closed source programs may statically link to it, and "Dmf" means that you can distribute modified versions freely. So the LGPL would become something like the CslDmfLGPL instead. (If we're really cool, we can be like Visual C++ and have all sorts of prefixes that are no longer relevant. Don't they still have lp for long pointer in 6.0?)
--JZ
Well, I can't speak to the installation procedure on RH 5.2, but for future releases Mandrake has been mumbling about X-based installation. Plus a default easy KDE setup. But I agree that Mandrake no longer has much of an edge, however, Bero has the key feature that it is Pentium optimized. Why the hell should we all be running code compiled for a 386? Is anyone actually going to be dumb enough to try and run GNOME on a 386?
--JZ
Both of their sites have said for ages that the next release would be out in late February or MAYBE early March. Well, it's obviously March. Anybody heard updates/rumors? (I did hear the spiel about Netscape's failure to compile already.) Also, anyone have thoughts on Yellow Dog? They have yet to ship an actual product, but they seemed to get fairly good media coverage. How does it differentiate itself from a repackaged LinuxPPC distro (besides the high price)?
Good: Red Hat can hire some more folks full time to crunch bugs in GNOME and other projects.
Bad: More news stories saying that Red Hat invented Linux. Worse yet, more vendors dying to install an "official" copy of RH Linux, jacking up the cost, rather than just installing a free copy of it. Why would I POSSIBLY want to pay for an installation manual and installation support when it's ALREADY INSTALLED?!? These people just don't get it. RH has always said that they'll get their revenue from support when things have taken off, but they're becoming just another software company. . .
Future: Obviously a lot of shakeout in the market, but we'll continue to see more distros like BeroLinux and Linux Mandrake that build on Red Hat. Of course let's not forget that Corel's distrobution will shake things up a lot once they come in. Red Hat on all the servers, Corel on all the desktops? I don't think it'll be quite that stark, althogh I do think that KDE 2.0 (which I think we can assume Corel will use, although they're technically 'evaluating' GNOME too) will be REALLY good. Check out QT 2.0, on which it'll be based. Much slicker, plus support for the mouse wheel. Or don't. GNOME's buggy, but I do like the way gmc and gtk+ look.
-JZ
I agree that RedHat is completely outdated. They have their reasons, but I'm not very impressed by them: .). I know resellers wanted an end to this too, but the company should simply include very easy instructions for obtaining updates (how about the .deb update capability for the next version of RPM?) and say screw you to the resellers who complain, because consumers matter more than vendors.
-Compiling on 386: because they like to advertise that linux runs on ancient hardware and it is already complicated enough to distribute separate Alpha and Sparc versions.
-Slow pace of change: this is the most understandable. Linux changes so much faster than other OSes that it absolutely confuses regular resellers. I remember being impressed to see a couple dozen RH5.1 boxes in both Borders and CompUSA, but somewhat perplexed to see them sitting next to a dozen more RH5.2 boxes being sold for the same price! Many resellers asked RedHat to slow down the new releases, and I imagine that other distros had the same issue.
-Fvwm. The only excuses here are the desire to make Linux look like win95 and the fact of RH's moral opposition to KDE.
My opinion? For version 6.0, they need pentium optimization. Maybe they can ship out stickers for resellers to put on the old 5.2 boxes saying "Try 5.2 for older hardware!", or something to that effect. They should also go back to their old practice of adding updates within a vesion number. I hear they continued this up through a few version 5.1's, which have a slightly newer kernel than older 5.1's (although it might have been 5 when they stopped, I can't remember. .
As for FVWM, they should have been investing more time in documenting and stabilizing WindowMaker (which NEVER crashes on me, I'd add, but I have heard rumors of stories of complaints) to include that as their default WM. It is far more usable than E or FVWM, and with help from RedHat it could have every bit as much GNOME integration as E.
I don't mean to slander other distros (although SuSe stuck with libc5 for far too long and Debian, great though it is, should have scrapped Slink to work on 2.2), I just focused on RedHat because it has the most relevance to the topic at hand (BeroLinux, for those who are completely lost in my rambling).
--JZ