I learned how to use a slide rule in tenth grade. My geometry class was one of those combined ones, where half the students have up to that point taken two years of math ("pre-algebra" and "elementary algebra"), and the other half have taken one year ("algebra 1") and learned roughly three times as much, so then the school district throws them all together for geometry, and the results are predictable: no C grades, lots of As and Bs on the one hand and Ds and Fs on the other. The latter category slow the course down to a real crawl, and since I was also taking Algebra II the same year (in order to get ahead so I could take the AP Calc as a senior), the Algebra II teacher had me in OHML (a competitive math team thing, semi-extracurricular). So in order to let me absorb the geometry faster, my geometry teacher halfway through the year agreed to let me study at my own pace and take the tests when I was ready. I finished the remaining two quarters in nine weeks, and after that he gave me other stuff to do. The slide rule was one of those things. Non-euclidean geometry was another, and at least as valuable. It was a great opportunity for me. The school district really should split that class in half and put the algebra 1 students into a geometry course that moves faster and covers more.
If the demand for XP is not enough to be termed "strong", then I hate to think what terms might apply to the demand for Vista at this point. The first word I can think of that means "less than ten percent as strong as something that isn't strong" is "weak", but I'm not sure that's really adequately severe.
NT-based versions of Windows have a long history of lukewarm receptions when they first come out, the general consensus for each new version usually being "it's not an improvement". But eventuall the service packs start coming out and sooner or later most users of the previous version end up upgrading, if only because the new version comes with new hardware.
It'll happen that way with Vista too. and Longhorn/Vienna/Seven for that matter. Give it time.
Of course the "Worst-Received OS Version Ever" prize (at least in terms of Microsoft products) goes to DOS 4, with Windows Me running a distant second, but neither of those was part of the NT/2K/XP/Vista product line.
> the majority of the world uses GSM, the US is strangely skewed towards cdma
The majority of the world is irrelevant when you're marketing a device as expensive as the iPhone. You're going to sell significant numbers of the thing in North America, Western Europe, Japan, Australia, South Korea, Israel, maybe China, _possibly_ South Africa... and a very small handful of other countries. It doesn't matter what kind of cellphone network they use in most of Africa, most of Asia, or Latin America, because US$300+ for a PDA/phone is too much for most cellphone users (already a fairly small percentage of the population) to pay. There are probably more potential iPhone buyers in Dayton Ohio than in all of central Africa between the Sahara and the Kalahari.
Consequently, North America is a larger percentage of the market for such a device than it would be for most things.
Yes, there will be people in Rwanda who would want to buy the thing, but not very many people, not enough to drive a decision about what kinds of cell-phone networks the thing should support.
Frankly, the reason to standardize on GSM for the iPhone can pretty much be summed up in the single word "Europe". American iPhone users on average are _substantially_ more likely than the average American to travel in Europe on business on a regular basis. (Most Americans never visit Europe, or go once in a lifetime for a special vacation. But most Americans don't have $300 to blow on an iPhone.) Otherwise it might've made sense to go with separate versions of the device, a US version if you will and a GSM version. But then the US version wouldn't work in Europe, and for people who can afford to drop $300 on a phone (or, at least, for a significant percentage of them), that could be a very real inconvenience.
Here I always thought MIT was full of really smart kids, and a relatively normal person like me could never get in. I always sorta figured MIT only took people with crazy-high SAT scores, like 1500 or more. But 1380, that's only sixty points higher than _my_ score, when I took it in 1993, and if that's the _average_... I'm probably within the range of what they accept, not even taking into account the College Board's "recentering" of the scale in '95. Not to mention, I only took the thing once and might've been able to squeeze out a few more points if I took it a couple more times.
I probably could've got into MIT if I wanted. Huh. Never would've guessed that.
Not that I necessarily would've chosen to go there (I'm actually quite happy with where I did go), but it's interesting to think about.
> I have 26.4 kbps dial-up access to the internet... it would be > nice to do some [work] stuff remotely when I am on 24/7 call.
Umm, so use ssh, what's the problem?
When I was on dialup, the modem was admittedly 33.6, but OTOH the connection was shared, so that's pretty comparable. I used ssh all the time to shell into work from home for various reasons, and occasionally into home from work. The only reason I'm not still on dialup is because Verizon appears to have completely stopped repairing land lines in my area and the line was no longer clear enough to sustain a dialup connection. So I had to break down and shell out for cable modem service (which, fortunately, is available here, albeit from one provider only). Otherwise I'd still be on dialup.
X11 forwarding is painfully slow over dialup, but I never experienced any significant problem with regular ssh (or tramp for editing remote files), and although it can be nice to have you don't actually _need_ X11; any network administrator who can't get work done without a GUI is in altogether the wrong line of work.
Dialup really isn't all that bad, once you're accustomed to it. Really large items take a long time to download, but with a decent resume-where-it-left-off tool (e.g., wget), even that is not really a big deal, you just let it run while you sleep and/or are away from home. The largest thing I ever downloaded that way was a three-CD set of ISO images for a Linux distro (I do not now recall which one). It took a few days, but it worked.
You could possibly get ISDN, but it's probably not worth what it would cost. My advice is to learn to live with the dialup. Yes, you *can* do remote system administration over dialup.
This was back in the Mozilla 0.9.x era, roughly. I installed the Flash plugin, because a friend recommended it and claimed that it was useful. Then I did some regular browsing, and I think it was maybe twenty minutes later, maybe an hour at the outside, when I uninstalled Flash.
When the Flashblock plugin became available for Firefox, I thought maybe I'd give Flash another try, but IIRC the port (or one of its dependencies maybe) was broken with a note from the FreeBSD security team, something to the effect that there was a known unpatched security-relevant bug, and I kind of shrugged and didn't pursue the matter further, because I can think of a LOT of things I'd rather spend time on than getting Flash installed.
Consequently, Flash-based advertisements will not soon be making an appearance on my screen.
OTOH, I'm pretty much an edge case, in that my eyeballs are probably not that valuable to advertisers anyway.
> When will they focus on usability and speed rather than adding features.
They focus on usability whenever *specific* usability problems are pointed out, but when people just whine "it needs better usability", that's too vague to deserve any significant response. Usability is actually an issue of major importance to the OO.o people, and their software has in some ways lead the way for others. For example, dialog box buttons in OO were labelled with what they do ("Save", "Discard", and so on) while other office suites were still using traditional confusing labels ("Yes", "No", "Okay", and similar largely-meaningless words), forcing the user to read the entire text of the dialog box each and every time or just guess and hope. I'm sure there is plenty more room for improvement, but it's better than most of what's out there, and if you have specific complaints, you ought to enumerate them. Asking vague questions like "When are they going to focus on usability" is pointless.
As far as speed, I don't know how old your computer is, but I have not noticed any speed problems with OO on systems with a decent amount of RAM (say, 512MB or more if you're one of those single-tasking people who basically runs one app at a time, 2GB if you run multiple apps at the same time as OO). Even on five-year-old systems with more like 128MB of RAM, OO doesn't have noticeably worse performance than any other modern app (e.g., Firefox). And if a computer is much older than five years, it needs to stick with the software it has. New software on old computers, after about five years (three years if the computer was bargain-basement when new), is generally a fundamental mismatch, kind of like pouring new wine in old wineskins, or trying to run software designed for Windows XP under Windows 95. It kinda don't work pretty good. There are exceptions, but most of them involve repurposing the system, e.g., you can take a ten-year-old desktop PC and turn it into a router and firewall with current software, no problem, but if you try to run current *desktop* software on it, it's not going to perform so very well. It'll still run software from its own era.
The big case where I've seen OO perform poorly even on relatively decent hardware is when it's running under Windows with Symantec AV installed, but in that case *everything* crawls, so you can't blame OO for that.
Oh, there's one other case. If a document has really large images in it (like, four-digit numbers of pixels in each direction), performance can be somewhat degraded, particularly if your amount of RAM wasn't too much overkill in the first place. But that really shouldn't be surprising. And the correct solution is for OO to support embedding SVG images -- yep, a feature -- because then we wouldn't *need* enormo-huge bitmaps so often. (You're still occasionally going to want to put a big photo in, but that's more of an edge case.)
> More importantly, since this appears to be based on a 1.x OO fork
Okay, wow. Talk about being out of date.
I mean, I _remember_ OO 1.x, sure (heck, I remember when StarOffice was produced by StarDivision), but for that matter I also _remember_ Emacs 19, Netscape 2, and EDLIN. That doesn't mean I can imagine wanting to use any of the above today.
I would think, assuming your laptop has wireless internet capability, that you could set it up with a cron job to periodically wget an unusually-named non-existent file from some web server where you have access to the error log That gives you its IP address on a regular basis, and in the event that it's stolen traceroute will presumably tell you more or less where it's connected. Even if all it tells you is the ISP, that's a starting point.
Of course, if your laptop has a GPS receiver, then you could have the cron job actually transmit its coordinates somewhere -- say, email them to your gmail account every five minutes.
All of this relies on the thief taking the laptop someplace where it can get signal, but I don't see how that can be avoided in principle. Even if you give the laptop a full-time cellular connection, it's still gotta be able to get at least a cell-tower signal, in order to send any information anywhere.
> That said, while Truecrypt exists for Linux, I'm sure there is a native way to do encryption > without additional software.
The question of what constitutes "additional software" is a bit less straightforward in the open-source world. Linux technically is the kernel, but to my knowledge nobody runs just the Linux kernel on a laptop with no other software. A typical distribution, on the other hand (Ubuntu, for instance) includes a great deal more software than the pitiful little collection of accessories that comes with Windows out of the box. But the details of precisely what's included vary a bit from distro to distro (although of course there are a number of things that all the major distros include).
Anyway, the usual way to encrypt an entire filesystem on Linux is to use a loop device for the filesystem. See for instance encryptionhowto.sourceforge.net
> For about $200, the cost of a refrigerator, we are able to save someone's life
For that matter, people in rural sub-saharan Africa every day die for want of less than $200 worth of medical care, prevention, or treatment (or in some cases even food). Various groups (humanitarian aid groups, religious groups, NGOs,...) have been working on this problem for decades, but they can't really scratch the surface of the problem, due to its scope.
It sounds nice to say, "for only $200 we can save a person's life", but it really isn't that straightforward.
> MS DOS 5.0 was rock stable; many of the so called 'bugs' were in fact bugs in the > applications that were now showing up.
Apart from bugs in new features (which it's not fair to pick on too much, since DOS 3 didn't have the features in question at all), the key issue that I still remember is an extremely annoying bug in the handling of floppy diskettes, particularly when you changed diskettes, especially if you only had one floppy drive, or only one of the type in question (e.g., only one 5.25" drive), which was becomming increasingly common around that time. You'd issue the drive-letter directive (e.g., A: and hit enter) to signal the OS to re-read the diskette in case you'd inserted a new one, and the OS would get confused. My memory of the details has finally begun to fade (at long last), but IIRC the bug would not manifest the first time you changed floppies, but subsequently it would, especially if you were using more than two diskettes altogether in the same drive during the same session. I don't know if the bug was new in DOS 5 or a remaining bug left over from DOS 4, but I do know that it was *not* present in 3.3, *was* present in all 5.x versions and in 6.0, and was fixed in 6.2.
On the other hand, I also know that DOS 6 has no problem installing on, booting from, and using a SATA hard drive, even though Windows 95 won't touch it with a ten foot pole.
> I know I'm showing my age here but DOS 5 was GREAT.
My memory says it was buggy.
> Everyone knew that even numbered DOS releases were very poor. DOS 4 was a hugh piece of crap > that IBM force upon Microsoft.(When was te last time you heard that!) After DOS 4 bombed and > no one upgraded from their stable DOS 3.11, Microsoft fixed the problems and released v5.
DOS 4 was so bad, scarcely anybody ever even saw it. It's pretty much a footnote. But DOS 5 was fairly buggy itself. After 3.3 (which was rock solid), the next version of DOS that I'd consider reasonably stable was 6.2.
On the other hand, DOS 5 did introduce some useful features, not least dosshell, which provided task-swapping capability almost as good as Windows 3.x. (You had to do more setup, like telling it how much RAM to allocate for each process, but once you got things going it was really not bad.) Also, for those who had not already obtained a good third-party freeware text editor (yeah, all three of you), the text editor that came with DOS 5 was a quite major improvement over EDLIN. In fact, I don't think Microsoft has really materially improved on that text editor since, unless there's a new one in Vista I don't know about. For Windows 95 they separated it into its own executable (rather than having it embedded in QBasic), but other than that it's pretty much the same. And there's Notepad, which is better integrated with the OS, but it's actually considerably worse in several other ways.
> Column stores are great (better than a row store) if you're just reading > tons of data, but they're much more costly than a row store if you're > writing tons of data. Therefore, pick your method depending on your needs.
To me, this is really just performance tuning, which the RDBMS and/or the DB admin should handle behind the scenes. The application sitting on top of the database shouldn't need to know or care whether the underlying storage is organized by row or by column. The RDBMS should handle such things automagically, and the database administrator can tune such performance settings as necessary, but the application developer (to say nothing of the user) shouldn't have to be concerned with such matters.
Eight megs for an operating system is stinking enormous. PC-DOS 3.3 fits comfortably on a 360k floppy, with room left over for applications. The only application you really need is a text editor. I like UED, a full-screen editor that supports up to nine files at a time, split window, copy and paste, *and* search and replace. Plus UED comes in at only 38704 bytes, so there's still space left over for data, right there on the same floppy diskette with your OS *and* the text editor. Who needs a hard disk?
Okay, so the truth is I use Emacs these days, not to mention OpenOffice, Gimp, Inkscape, and about thirty Firefox extensions. What can I say? I like features.
But if non-bloatedness were a key goal, UED would by my editor of choice. It's better than a lot of MUCH larger text editors. For instance, ed is significantly larger and isn't even a full-screen editor, much less supports copy and paste between multiple files.
> Looking at the comments in this thread, I think we can safely assume that not one > of us really has a clue as to what is going on. I cheerfully admit I don't.
You have to know some background. Chiefly, you have to know what Eudora is. Eudora is (or was, at any rate) one of the major proprietary GUI-based mailreaders. A couple of years ago it was the second-oldest one still under development, and then the company behind it decided for whatever reason that they weren't going to maintain it any more. (FWIW, the dude behind the oldest one subsequently decided the same thing, which now leaves, if I'm not mistaken, MS Outlook as the oldest still-under-development GUI-based mailreader, which is just plain sad.)
As a fairly old and mature product, Eudora had quite a long lead on Thunderbird, in terms of functionality. Eudora wasn't Pegasus, not by a wide margin, but nonetheless Thunderbird still has a ways to go to catch up to where Eudora was five or ten years ago.
So no doubt a lot of Eudora users are not eager to move away from Eudora. (Indeed, why should they? Nothing currently under active development is better.) But the old Eudora codebase contains proprietary components, so even though Qualcomm is willing to release what they can do the open-source community, they can't really release a working codebase that can be easily turned into a working Eudora. Even when an app is quite mature, people still want to see active development.
So what's going on is an attempt to make a "Eudora substitute" based on the Thunderbird codebase. What they'll probably end up with is something along the lines of the Advanced Tea Substitute that the Heart of Gold created for Arthur Dent, which he noted was "almost, but not quite, entirely unlike tea". Nonetheless, they're trying.
The meaning of the word "Eudora" means different things at different points in the article.
Previously, Eudora was a proprietary email application, with a long history and a pretty significant user base. At some point the company behind it decided for whatever reason that they didn't really want to maintain it anymore, so it's no longer being actively developed.
However, some of the original developers have teamed up with some Mozilla developers to create a new, open-source version of Eudora, based on the Thunderbird codebase.
In order to make it similar to the old Eudora, they're using a two-pronged approach.
Where possible, changes from Thunderbird's behavior are being made in an extension. Some of the changes, e.g., cosmetic stuff like where the buttons are located, can be implemented as an extension.
Where that's not possible (because the way the old Eudora did things is too fundamentally different from the way Thunderbird does them) they're making changes to the underlying Thunderbird codebase. Some of these changes may get accepted into the main Thunderbird code, and others may not and thus may have to be maintained as patches.
Personally, I'd rather see mozilla.org try to clone Pegasus Mail. Eudora was always harder for new users to learn than Pegasus, while simultaneously managing to have less functionality for experienced users.
> Will EMACS finally be getting a decent editor added to its functionality?
The functionality is pretty much there, you just have to set various options to turn on the specific bits of editor functionality you need, and set up the keybindings you want, then you're good to go. I know it seems daunting at first, having to pore over all those options, figure out what key bindings you want, and so on and so forth, but it really only takes thirty or forty hours, maybe fifty or sixty at the outside, and it's well worth it in the long run.
There are technical solutions to some of the technical problems (e.g., malware), but there are no technical solutions to the human problems. I can tell you this for free: there's nothing you can do that will stop people who need constant attention from needing constant attention.
I'm The Computer Guy at a small public library, so I have a fairly good idea what kinds of things people will need you for. They will need you because the website they are looking at says they need Adobe Acrobat (or Macromedia Shockwave, or Real Audio, or Quick Time, or Bob's Obscure Plugin, or whatever), and they don't know remember, from the last time they asked you, whether they have the plugin or not. Also, in many cases the get-the-plugin link is significantly more prominent on the site than the link you should click if you already have the plugin, so they will click the "click here" link and be taken to the download site for the plugin, and they won't know why or how to get the content they wanted. They will need you because the website they are looking at doesn't answer their questions, or have the information they want. They will need you because the website they are looking at doesn't have the photo on it that it had last week, that they want to look at. They will need you because the website they are looking at does contain the picture they want to look at, but it's shrunk down really small, and they want it to be bigger, and they tried clicking on the picture to make it bigger, but it didn't work. They will need you because the website they are looking at was designed by an idiot and is virtually impossible to navigate. They will need you because the website they are looking at contains a factoid, and they aren't sure whether it's really true. They will need you because the website they are looking at contains an advertisement designed to look like a scary system error dialog containing the word "illegal", and now they are afraid they will be in trouble with the law. They will need you because the website they are looking at has informed them that their computer monitor is emitting radiation. But perhaps the biggest one is, they will need you just because they are bored and lonely and want to talk to someone, so any excuse will do. Technology cannot solve all these problems.
Getting rid of the option to print will help quite a lot, though. Printers are a major source of various strange annoyances in any case, but printing content from arbitrary websites is generally an exercise in frustration even for a quite knowledgeable user, and it really REALLY does not mix well with end users. They will whine and plead and beg, but absolutely do not let them talk you into installing a printer. You *will* regret it if you do. The printer will be the bane of your existence.
Slashdot Mirror
Carrier pigeons have reasonable bandwidth, but the latency is very bad. Smoke signals have better latency but much less bandwidth. HTH.HAND.
I learned how to use a slide rule in tenth grade. My geometry class was one of those combined ones, where half the students have up to that point taken two years of math ("pre-algebra" and "elementary algebra"), and the other half have taken one year ("algebra 1") and learned roughly three times as much, so then the school district throws them all together for geometry, and the results are predictable: no C grades, lots of As and Bs on the one hand and Ds and Fs on the other. The latter category slow the course down to a real crawl, and since I was also taking Algebra II the same year (in order to get ahead so I could take the AP Calc as a senior), the Algebra II teacher had me in OHML (a competitive math team thing, semi-extracurricular). So in order to let me absorb the geometry faster, my geometry teacher halfway through the year agreed to let me study at my own pace and take the tests when I was ready. I finished the remaining two quarters in nine weeks, and after that he gave me other stuff to do. The slide rule was one of those things. Non-euclidean geometry was another, and at least as valuable. It was a great opportunity for me. The school district really should split that class in half and put the algebra 1 students into a geometry course that moves faster and covers more.
If the demand for XP is not enough to be termed "strong", then I hate to think what terms might apply to the demand for Vista at this point. The first word I can think of that means "less than ten percent as strong as something that isn't strong" is "weak", but I'm not sure that's really adequately severe.
NT-based versions of Windows have a long history of lukewarm receptions when they first come out, the general consensus for each new version usually being "it's not an improvement". But eventuall the service packs start coming out and sooner or later most users of the previous version end up upgrading, if only because the new version comes with new hardware.
It'll happen that way with Vista too. and Longhorn/Vienna/Seven for that matter. Give it time.
Of course the "Worst-Received OS Version Ever" prize (at least in terms of Microsoft products) goes to DOS 4, with Windows Me running a distant second, but neither of those was part of the NT/2K/XP/Vista product line.
> the majority of the world uses GSM, the US is strangely skewed towards cdma
The majority of the world is irrelevant when you're marketing a device as expensive as the iPhone. You're going to sell significant numbers of the thing in North America, Western Europe, Japan, Australia, South Korea, Israel, maybe China, _possibly_ South Africa... and a very small handful of other countries. It doesn't matter what kind of cellphone network they use in most of Africa, most of Asia, or Latin America, because US$300+ for a PDA/phone is too much for most cellphone users (already a fairly small percentage of the population) to pay. There are probably more potential iPhone buyers in Dayton Ohio than in all of central Africa between the Sahara and the Kalahari.
Consequently, North America is a larger percentage of the market for such a device than it would be for most things.
Yes, there will be people in Rwanda who would want to buy the thing, but not very many people, not enough to drive a decision about what kinds of cell-phone networks the thing should support.
Frankly, the reason to standardize on GSM for the iPhone can pretty much be summed up in the single word "Europe". American iPhone users on average are _substantially_ more likely than the average American to travel in Europe on business on a regular basis. (Most Americans never visit Europe, or go once in a lifetime for a special vacation. But most Americans don't have $300 to blow on an iPhone.) Otherwise it might've made sense to go with separate versions of the device, a US version if you will and a GSM version. But then the US version wouldn't work in Europe, and for people who can afford to drop $300 on a phone (or, at least, for a significant percentage of them), that could be a very real inconvenience.
Here I always thought MIT was full of really smart kids, and a relatively normal person like me could never get in. I always sorta figured MIT only took people with crazy-high SAT scores, like 1500 or more. But 1380, that's only sixty points higher than _my_ score, when I took it in 1993, and if that's the _average_... I'm probably within the range of what they accept, not even taking into account the College Board's "recentering" of the scale in '95. Not to mention, I only took the thing once and might've been able to squeeze out a few more points if I took it a couple more times.
I probably could've got into MIT if I wanted. Huh. Never would've guessed that.
Not that I necessarily would've chosen to go there (I'm actually quite happy with where I did go), but it's interesting to think about.
> I have 26.4 kbps dial-up access to the internet ... it would be
> nice to do some [work] stuff remotely when I am on 24/7 call.
Umm, so use ssh, what's the problem?
When I was on dialup, the modem was admittedly 33.6, but OTOH the connection was shared, so that's pretty comparable. I used ssh all the time to shell into work from home for various reasons, and occasionally into home from work. The only reason I'm not still on dialup is because Verizon appears to have completely stopped repairing land lines in my area and the line was no longer clear enough to sustain a dialup connection. So I had to break down and shell out for cable modem service (which, fortunately, is available here, albeit from one provider only). Otherwise I'd still be on dialup.
X11 forwarding is painfully slow over dialup, but I never experienced any significant problem with regular ssh (or tramp for editing remote files), and although it can be nice to have you don't actually _need_ X11; any network administrator who can't get work done without a GUI is in altogether the wrong line of work.
Dialup really isn't all that bad, once you're accustomed to it. Really large items take a long time to download, but with a decent resume-where-it-left-off tool (e.g., wget), even that is not really a big deal, you just let it run while you sleep and/or are away from home. The largest thing I ever downloaded that way was a three-CD set of ISO images for a Linux distro (I do not now recall which one). It took a few days, but it worked.
You could possibly get ISDN, but it's probably not worth what it would cost. My advice is to learn to live with the dialup. Yes, you *can* do remote system administration over dialup.
This was back in the Mozilla 0.9.x era, roughly. I installed the Flash plugin, because a friend recommended it and claimed that it was useful. Then I did some regular browsing, and I think it was maybe twenty minutes later, maybe an hour at the outside, when I uninstalled Flash.
When the Flashblock plugin became available for Firefox, I thought maybe I'd give Flash another try, but IIRC the port (or one of its dependencies maybe) was broken with a note from the FreeBSD security team, something to the effect that there was a known unpatched security-relevant bug, and I kind of shrugged and didn't pursue the matter further, because I can think of a LOT of things I'd rather spend time on than getting Flash installed.
Consequently, Flash-based advertisements will not soon be making an appearance on my screen.
OTOH, I'm pretty much an edge case, in that my eyeballs are probably not that valuable to advertisers anyway.
> When will they focus on usability and speed rather than adding features.
They focus on usability whenever *specific* usability problems are pointed out, but when people just whine "it needs better usability", that's too vague to deserve any significant response. Usability is actually an issue of major importance to the OO.o people, and their software has in some ways lead the way for others. For example, dialog box buttons in OO were labelled with what they do ("Save", "Discard", and so on) while other office suites were still using traditional confusing labels ("Yes", "No", "Okay", and similar largely-meaningless words), forcing the user to read the entire text of the dialog box each and every time or just guess and hope. I'm sure there is plenty more room for improvement, but it's better than most of what's out there, and if you have specific complaints, you ought to enumerate them. Asking vague questions like "When are they going to focus on usability" is pointless.
As far as speed, I don't know how old your computer is, but I have not noticed any speed problems with OO on systems with a decent amount of RAM (say, 512MB or more if you're one of those single-tasking people who basically runs one app at a time, 2GB if you run multiple apps at the same time as OO). Even on five-year-old systems with more like 128MB of RAM, OO doesn't have noticeably worse performance than any other modern app (e.g., Firefox). And if a computer is much older than five years, it needs to stick with the software it has. New software on old computers, after about five years (three years if the computer was bargain-basement when new), is generally a fundamental mismatch, kind of like pouring new wine in old wineskins, or trying to run software designed for Windows XP under Windows 95. It kinda don't work pretty good. There are exceptions, but most of them involve repurposing the system, e.g., you can take a ten-year-old desktop PC and turn it into a router and firewall with current software, no problem, but if you try to run current *desktop* software on it, it's not going to perform so very well. It'll still run software from its own era.
The big case where I've seen OO perform poorly even on relatively decent hardware is when it's running under Windows with Symantec AV installed, but in that case *everything* crawls, so you can't blame OO for that.
Oh, there's one other case. If a document has really large images in it (like, four-digit numbers of pixels in each direction), performance can be somewhat degraded, particularly if your amount of RAM wasn't too much overkill in the first place. But that really shouldn't be surprising. And the correct solution is for OO to support embedding SVG images -- yep, a feature -- because then we wouldn't *need* enormo-huge bitmaps so often. (You're still occasionally going to want to put a big photo in, but that's more of an edge case.)
I don't want your music, Mr. Reznor.
Anyway, most of the music I *do* like is in the public domain. Something about the composers having died by the middle of the eighteenth century...
> More importantly, since this appears to be based on a 1.x OO fork
Okay, wow. Talk about being out of date.
I mean, I _remember_ OO 1.x, sure (heck, I remember when StarOffice was produced by StarDivision), but for that matter I also _remember_ Emacs 19, Netscape 2, and EDLIN. That doesn't mean I can imagine wanting to use any of the above today.
I would think, assuming your laptop has wireless internet capability, that you could set it up with a cron job to periodically wget an unusually-named non-existent file from some web server where you have access to the error log That gives you its IP address on a regular basis, and in the event that it's stolen traceroute will presumably tell you more or less where it's connected. Even if all it tells you is the ISP, that's a starting point.
Of course, if your laptop has a GPS receiver, then you could have the cron job actually transmit its coordinates somewhere -- say, email them to your gmail account every five minutes.
All of this relies on the thief taking the laptop someplace where it can get signal, but I don't see how that can be avoided in principle. Even if you give the laptop a full-time cellular connection, it's still gotta be able to get at least a cell-tower signal, in order to send any information anywhere.
> That said, while Truecrypt exists for Linux, I'm sure there is a native way to do encryption
> without additional software.
The question of what constitutes "additional software" is a bit less straightforward in the open-source world. Linux technically is the kernel, but to my knowledge nobody runs just the Linux kernel on a laptop with no other software. A typical distribution, on the other hand (Ubuntu, for instance) includes a great deal more software than the pitiful little collection of accessories that comes with Windows out of the box. But the details of precisely what's included vary a bit from distro to distro (although of course there are a number of things that all the major distros include).
Anyway, the usual way to encrypt an entire filesystem on Linux is to use a loop device for the filesystem. See for instance encryptionhowto.sourceforge.net
> For about $200, the cost of a refrigerator, we are able to save someone's life
...) have been working on this problem for decades, but they can't really scratch the surface of the problem, due to its scope.
For that matter, people in rural sub-saharan Africa every day die for want of less than $200 worth of medical care, prevention, or treatment (or in some cases even food). Various groups (humanitarian aid groups, religious groups, NGOs,
It sounds nice to say, "for only $200 we can save a person's life", but it really isn't that straightforward.
> MS DOS 5.0 was rock stable; many of the so called 'bugs' were in fact bugs in the
> applications that were now showing up.
Apart from bugs in new features (which it's not fair to pick on too much, since DOS 3 didn't have the features in question at all), the key issue that I still remember is an extremely annoying bug in the handling of floppy diskettes, particularly when you changed diskettes, especially if you only had one floppy drive, or only one of the type in question (e.g., only one 5.25" drive), which was becomming increasingly common around that time. You'd issue the drive-letter directive (e.g., A: and hit enter) to signal the OS to re-read the diskette in case you'd inserted a new one, and the OS would get confused. My memory of the details has finally begun to fade (at long last), but IIRC the bug would not manifest the first time you changed floppies, but subsequently it would, especially if you were using more than two diskettes altogether in the same drive during the same session. I don't know if the bug was new in DOS 5 or a remaining bug left over from DOS 4, but I do know that it was *not* present in 3.3, *was* present in all 5.x versions and in 6.0, and was fixed in 6.2.
On the other hand, I also know that DOS 6 has no problem installing on, booting from, and using a SATA hard drive, even though Windows 95 won't touch it with a ten foot pole.
> I know I'm showing my age here but DOS 5 was GREAT.
My memory says it was buggy.
> Everyone knew that even numbered DOS releases were very poor. DOS 4 was a hugh piece of crap
> that IBM force upon Microsoft.(When was te last time you heard that!) After DOS 4 bombed and
> no one upgraded from their stable DOS 3.11, Microsoft fixed the problems and released v5.
DOS 4 was so bad, scarcely anybody ever even saw it. It's pretty much a footnote. But DOS 5 was fairly buggy itself. After 3.3 (which was rock solid), the next version of DOS that I'd consider reasonably stable was 6.2.
On the other hand, DOS 5 did introduce some useful features, not least dosshell, which provided task-swapping capability almost as good as Windows 3.x. (You had to do more setup, like telling it how much RAM to allocate for each process, but once you got things going it was really not bad.) Also, for those who had not already obtained a good third-party freeware text editor (yeah, all three of you), the text editor that came with DOS 5 was a quite major improvement over EDLIN. In fact, I don't think Microsoft has really materially improved on that text editor since, unless there's a new one in Vista I don't know about. For Windows 95 they separated it into its own executable (rather than having it embedded in QBasic), but other than that it's pretty much the same. And there's Notepad, which is better integrated with the OS, but it's actually considerably worse in several other ways.
> liberals 'could be expected to more readily accept new social, scientific or religious ideas.'
Stated another way, they're less sceptical, less discerning, more gullible.
But more likely the study is just bunk.
I guess I'm an old coot, because I still remember fondly the days when saying that a device was "off" meant that it was not on.
Yeah, but DOS 1.0 doesn't support directories, which are an important tool for keeping my data organized.
Besides, I don't have any 180K floppy drives. I do have a 360K floppy drive.
> Column stores are great (better than a row store) if you're just reading
> tons of data, but they're much more costly than a row store if you're
> writing tons of data. Therefore, pick your method depending on your needs.
To me, this is really just performance tuning, which the RDBMS and/or the DB admin should handle behind the scenes. The application sitting on top of the database shouldn't need to know or care whether the underlying storage is organized by row or by column. The RDBMS should handle such things automagically, and the database administrator can tune such performance settings as necessary, but the application developer (to say nothing of the user) shouldn't have to be concerned with such matters.
Eight megs for an operating system is stinking enormous. PC-DOS 3.3 fits comfortably on a 360k floppy, with room left over for applications. The only application you really need is a text editor. I like UED, a full-screen editor that supports up to nine files at a time, split window, copy and paste, *and* search and replace. Plus UED comes in at only 38704 bytes, so there's still space left over for data, right there on the same floppy diskette with your OS *and* the text editor. Who needs a hard disk?
Okay, so the truth is I use Emacs these days, not to mention OpenOffice, Gimp, Inkscape, and about thirty Firefox extensions. What can I say? I like features.
But if non-bloatedness were a key goal, UED would by my editor of choice. It's better than a lot of MUCH larger text editors. For instance, ed is significantly larger and isn't even a full-screen editor, much less supports copy and paste between multiple files.
> Looking at the comments in this thread, I think we can safely assume that not one
> of us really has a clue as to what is going on. I cheerfully admit I don't.
You have to know some background. Chiefly, you have to know what Eudora is. Eudora is (or was, at any rate) one of the major proprietary GUI-based mailreaders. A couple of years ago it was the second-oldest one still under development, and then the company behind it decided for whatever reason that they weren't going to maintain it any more. (FWIW, the dude behind the oldest one subsequently decided the same thing, which now leaves, if I'm not mistaken, MS Outlook as the oldest still-under-development GUI-based mailreader, which is just plain sad.)
As a fairly old and mature product, Eudora had quite a long lead on Thunderbird, in terms of functionality. Eudora wasn't Pegasus, not by a wide margin, but nonetheless Thunderbird still has a ways to go to catch up to where Eudora was five or ten years ago.
So no doubt a lot of Eudora users are not eager to move away from Eudora. (Indeed, why should they? Nothing currently under active development is better.) But the old Eudora codebase contains proprietary components, so even though Qualcomm is willing to release what they can do the open-source community, they can't really release a working codebase that can be easily turned into a working Eudora. Even when an app is quite mature, people still want to see active development.
So what's going on is an attempt to make a "Eudora substitute" based on the Thunderbird codebase. What they'll probably end up with is something along the lines of the Advanced Tea Substitute that the Heart of Gold created for Arthur Dent, which he noted was "almost, but not quite, entirely unlike tea". Nonetheless, they're trying.
HTH.HAND.
The meaning of the word "Eudora" means different things at different points in the article.
Previously, Eudora was a proprietary email application, with a long history and a pretty significant user base. At some point the company behind it decided for whatever reason that they didn't really want to maintain it anymore, so it's no longer being actively developed.
However, some of the original developers have teamed up with some Mozilla developers to create a new, open-source version of Eudora, based on the Thunderbird codebase.
In order to make it similar to the old Eudora, they're using a two-pronged approach.
Where possible, changes from Thunderbird's behavior are being made in an extension. Some of the changes, e.g., cosmetic stuff like where the buttons are located, can be implemented as an extension.
Where that's not possible (because the way the old Eudora did things is too fundamentally different from the way Thunderbird does them) they're making changes to the underlying Thunderbird codebase. Some of these changes may get accepted into the main Thunderbird code, and others may not and thus may have to be maintained as patches.
Personally, I'd rather see mozilla.org try to clone Pegasus Mail. Eudora was always harder for new users to learn than Pegasus, while simultaneously managing to have less functionality for experienced users.
> Will EMACS finally be getting a decent editor added to its functionality?
The functionality is pretty much there, you just have to set various options to turn on the specific bits of editor functionality you need, and set up the keybindings you want, then you're good to go. I know it seems daunting at first, having to pore over all those options, figure out what key bindings you want, and so on and so forth, but it really only takes thirty or forty hours, maybe fifty or sixty at the outside, and it's well worth it in the long run.
There are technical solutions to some of the technical problems (e.g., malware), but there are no technical solutions to the human problems. I can tell you this for free: there's nothing you can do that will stop people who need constant attention from needing constant attention.
I'm The Computer Guy at a small public library, so I have a fairly good idea what kinds of things people will need you for. They will need you because the website they are looking at says they need Adobe Acrobat (or Macromedia Shockwave, or Real Audio, or Quick Time, or Bob's Obscure Plugin, or whatever), and they don't know remember, from the last time they asked you, whether they have the plugin or not. Also, in many cases the get-the-plugin link is significantly more prominent on the site than the link you should click if you already have the plugin, so they will click the "click here" link and be taken to the download site for the plugin, and they won't know why or how to get the content they wanted. They will need you because the website they are looking at doesn't answer their questions, or have the information they want. They will need you because the website they are looking at doesn't have the photo on it that it had last week, that they want to look at. They will need you because the website they are looking at does contain the picture they want to look at, but it's shrunk down really small, and they want it to be bigger, and they tried clicking on the picture to make it bigger, but it didn't work. They will need you because the website they are looking at was designed by an idiot and is virtually impossible to navigate. They will need you because the website they are looking at contains a factoid, and they aren't sure whether it's really true. They will need you because the website they are looking at contains an advertisement designed to look like a scary system error dialog containing the word "illegal", and now they are afraid they will be in trouble with the law. They will need you because the website they are looking at has informed them that their computer monitor is emitting radiation. But perhaps the biggest one is, they will need you just because they are bored and lonely and want to talk to someone, so any excuse will do. Technology cannot solve all these problems.
Getting rid of the option to print will help quite a lot, though. Printers are a major source of various strange annoyances in any case, but printing content from arbitrary websites is generally an exercise in frustration even for a quite knowledgeable user, and it really REALLY does not mix well with end users. They will whine and plead and beg, but absolutely do not let them talk you into installing a printer. You *will* regret it if you do. The printer will be the bane of your existence.