I felt that KDE was too ``german'' in it's looks and feels (no offence meant to germans, but I'm sure some of you know what I mean). So I preferred GNOME over KDE. However, GNOME was slow. Especially when switching desktops, I would have to wait for -too long- to get a netscape away and six terminals where I wanted them.
So I decided to just skip the desktop race, and go with enlightenment, straight, mean and lean. Much to my surprise, this didn't change the situation... After a little wondering, I started out on a new track.
Now, with GNOME and icewm as my window-manager, I have a lean and fairly mean desktop system going, and it's *MUCH* faster at desktop switching than enlightenment ever was.
Clue: If GNOME is too slow for you, try replacing the default window manager. Try icewm (has an *ugly* default theme, but has others which are nice and readily available from T.O.). Or try sawmill (haven't done that myself - yet).
For historical reasons, enlightenment is the default window manger in the GNOME releases done by redhat and others. This is changing. Enlightenment is - hands down - the most artistical window manager I've ever seen and used. But there's just (IMO) too much of artistical sophistication instead of lean code in that one, to perform well on ``old'' systems (my dual PPro with a Matrox I for example).
GNOME with icewm rocks. It seems faster on FreeBSD 3.4 than on Linux 2.2 though, but on the other hand, Linux wins when it gets to disk I/O etc. on the day-to-day workloads. GNOME is not slow in a sane configuration, it's just slow in it's default configuration.
Never ask a geek why, just nod your head and back away slowly;)
Yes it's bad that you'll be paying for a CE license when you won't be using it. However, if enough people request a BSD or ``blank'' version of the z50, I'm sure IBM will notice and offer that option.
Don't underestimate the value of BSD or Linux on this kind of machine. Fit it with a FE card, and you'll have a router/firewall/web-server/etc. Not that you would want to use it for production, but imagine coming to some customer as a networking consultant, technician, or whatever, then plugging in your handheld to the network - and instantly they have the transparant proxy and news server they needed temporarily;) It impresses the shit out of people. I've seen people running news servers and web servers on these pocket devices, as a temporary solution though, but still the ordinary unknowing people (induhviduals) around you are baffled. Their beliefs faint, they turn pale, they eat off you hand, they'd bend over and say thanks afterwards if you told them to:)
Think about it. Whenever you use a piece of software with those rediculous licences (if you do so at all), you just click ``accept'' and never think about the license again.
Basically the licences often say that ``you have no rights'' and ``it's your own fault'' etc. People ignore that now, and they think that if they don't bug the company, the company won't bug them.
If these licences were enforced, and if they could be so with backing in the law, people might start considering what they are actually agreeing to.
If this passes, I think licences might start to change. People might not accept a ridiculous licence if they know for sure that by accepting it they _will_ be breaking the law.
(funny example: licence on the drivers to the i810 motherboards on the accompanying CD say thay you may not use them commercially, eg. you cannot run windows on an i810 motherboard in your company)
This is another step towards wide use of Free Software and sane licences.
Re:Lack of software ? Where ?? - correction
on
Playboy And...Linux?
·
· Score: 2
"If I type in my shell"
That should - of course - have read:
"If I type TAB TAB in my shell"
But Slashdot ate my < and >s because they weren't - eh.
I keep hearing that from everyone who doesn't actually use (GNU/)Linux... blablabla but it lacks software blablabla
If I type in my shell, I get this question: "There are 1852 possibilities. Do you really wish to see them all? (y or n)" That's the number of single _programs_, or ``pieces of software'' that I have installed on a fairly standard desk-top developer machine.
I recently started porting some software to NT, it's been four years since I used that OS last... After installing IE4 (required for VC), Visual C++ 6.0, MS Win32 SDK, (and in order not to lose my mind completely: Cygwin too), I fired up this huge IDE GUI development environment. Especially VC++ is something that Win developers claim Linux lacks. I simply cannot imagine why. Hang on:
In order to do development on NT I depend on the IDE. If the IDE does not offer the combination of commands I wish to apply to my work, I'm shit out of luck. There's just no way to get boring repetitive work done easily, if it happens to not fit exactly into the provided dialog box. The IDE may be very nice, but it is _inherently_ limited because it has no way of executing scripts. It lacks a shell.
On GNU/Linux, I run Emacs. For editing. Then I run bash for CLI. Then I run make to build. I use X with KDE to *integrate* these xterms and editors into one large IDE. In short: NT *has* an IDE, GNU/Linux with the standard tools *IS* an IDE.
X+Emacs+GCC+Make+xterm+bash+... is quite a large program, if you look at all the parts as one. Anyone claiming that GNU/Linux is lacking software is someone who can't see the forest for trees.
Granted, there may be a lack of integrated office suites. They should be just around the corner though. A lot of people won't need them though. Again, they limit you somehow. As an example: Work four people on a report using either Word or LaTeX+CVS. Assuming you actually _know_ a little word or a little LaTeX. LaTeX has no support for workgroups, CVS has no support for typesetting, but they both keep things simple and work on files, there you have your workgroup-aware typesetting tool.
If you have to complain, say something clever. Like lack of huge but clumsy, limited, and hard-to-make-do-repetitive-work-for-you GUI applications.
That was pretty much about time someone took the wavelets into image compression.
I worked with wavelet transforms (Daubechies wavelets) a year and a half ago, and back then it was pretty clear that it would be possible to compress images and sound much harder using the wavelet domain rather than the Fourier domain.
For those who don't know, the trick is: JPEG/MPEG/MP3 uses Fourier transform to transform the image/sound data into their spectral components. But this spectral representation of the data does not say anything about the _locality_ of the frequency components. Therefore representing spikes/discontinuties will require a very large number of frequency components when using Fourier domain, which in turn leads to poor compression. You can see this problem by drawing a few sharp lines in an image and compressing it hard with JPEG.
Wavelets on the other hand, represent both an equivalent of the frequency component, along with locality information. Spikes/discontinuities can be approximated well using only a few wavelets. This in turn leads to good compression.
Another nice thing about wavelet compression is, that wavelets tend to represent discontinuities well, even with hard compression (eg. a lot of missing or roughly approximated wavelet components). Therefore a very hard compressed image will still have fairly sharp edges, completely contrary to JPEG compression. This is pretty important if you compress a picture holding text.
Anyways, someone is now working on JPEG with wavelets... What about sound and video ?? There is no reason as to why wavelets should not provide equal improvements in both audio and video.
My personal conspiracy theory is, that there exists a *LOT* of expensive hardware that can do Fouries forwards and back to allow real-time encoding and decoding of MPEG movies in good quality. The companies producing these devices will lobby any standards-organisation to *NOT* consider wavelets and stick to good old Fourier. If this holds, it will take a few years until we see Wavelet compressed video:)
Nope. Sure it's a problem, but nobody gets hurt, planes don't fall out of the sky.
The report can be re-printed when the problem is fixed. But if the only problem really turns out to be exactly this, formatting of output intended for humans-only, then the problem is small. I know pretty darn well that a syslog entry from 1985 doesn't belong in a recently installed PII system, and my supervisor at university will also have a pretty good understanding of whether the report I hand in is written in jan. 2000 or jan. 1900.
I have already seen some ``Y2K problems'', which were caused by people trying to upgrade BIOS software on systems to ensure Y2K compatibility. And I'm pretty confident that that will be all of the Y2K bugs I get to see. Even at the university some machines (especially a web-server holding lecture material that I desparately need now) are shut down due to Y2K. I nearly pissed my pants when I found out. Sure, there are lots of Y2K problems out there, but I have yet to see one which is caused by a problem in handling dates.
But the bug you see is only in formatting of output right ? The systems still work, the little clock in the top right corner just prints the wrong year, or ?
My argument is, that since most systems don't use decimal numbers internally, they will keep on running. Perhaps the user gets confused when he/she sees the wrong year (especially after a tough new-year's eve you can sort of get in doubt as to what century you're waking up in), but it's only formatting and not critical (outside the financial sectors).
Or have you actually seen a Y2K bug that caused something to *break* except for formatting/prettyprinting ?
Sure there must have been problems in the banking/insurance/financial sector with the old Cobol code, but computers hasn't used the decimal numbers for keeping track of time since sometime in the late sixties. Y2K is an application-software-only problem, and it occurs mainly with either _very_ poorly written software or with software written in a language/environment that has the problem (Cobol).
All modern software (from early '70s and newer) uses other ways to keep track on the time (eg. seconds from 1.jan 1970), it's almost entirely written in C, and it will not have a problem. (Even software written in other languages have the language compilers/environments written in C, and therefore doesn't have the problem either).
So why were there BIOS updates out ? Well if enough customers believe there is a problem, the manufacturers better do _something_ (eg. flip a few bits, insert a nop here and there) to give the customer the impression that a problem present was fixed. Marketing and crowd-of-lemmings stuff.
I would be willing to bet a beer with anyone on this. However, there is the _slight_ chance * consequence equation that makes me sort of reluctant to bet with all of slashdot, despite the incredible amount of beer I would receive. (This is a small country, I wonder if it would sink in the ocean once I won;)
My 1996-BIOS dual PPro keeps on running with it's vintage BIOS until it gets replaced in a few years, or until the apocalypse caused by the aforementioned crows of lemmings makes computers a little concern.
Happy new year ! (Oh, and if anyone has stocked up a few too many of those cal. 50 with IR scope and they figure they won't need them for looting/rioting/protection/whateverthehellyouthoug htwouldhappen, I'd be happy to receive donations;)
You can be held liable for whatever you promise, which is why most open source software has a clause something like: >> This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
If you read the EULA from Microsoft, I'm pretty sure that they have a similar clause, much to most people's surprise. Then what are you actually paying for, you might ask. Well, that's the good question;)
The car-makers have a responsibility of making cars *reasonably* safe, according to government regulations. They are not required to stop your kids from driving into brick walls using your car. They are however required to make sure your car doesn't fall apart or stops breaking when you want it to etc.
There are no such rules (yet) for software. The vendors make the rules, and the vast majority of customers/consumers simply neglect this fact and *expect* that there is some sort of reasonable agreement behind it all, just like when they bought their car.
Open source licences are usually very cautios to ``warn'' people of the possible dangers that lie ahead when using the software. And some people may even pay attention because ``there's gotta be a catch with gratis software after all''. I think this is a pretty good way to handle things.
There could be some sort of either regulations or at least some rule that software vendors should state LOUD AND CLEAR what they promise and what they don't. Pretty much like the warning messages on cigarette boxes:) This would probably not change the promises or the software, but it would make the general public aware of the lack of promises they actually get from spending huge cash on closed source software.
>> My mother said there aren't real monsters. But there are - Newt, Aliens II
Unfortunately both mozilla and sendmail are monsters. It is going to take a lot of effort to make sendmail a lean and mean mailer, and I don't see why anyone would care to even try. Postfix is out there, which has the features of sendmail (almost) and is already lean and mean. Of course, postfix is not a big name yet, like sendmail, and they're already sponsored. I think sponsoring sendmail is nothing but a political decision. It really gets us nowhere we can't go already.
Mozilla, while being a monster, could be wort sponsoring. Personally, I'd rather see a Gtk+ port of Konqueror, the KDE browser, or some funding of Mnemonic (anyone remember that effort?). Those seem to be much more clean by design, and actually sponsoring them could well give us a browser that was both lean and mean, and had the features we hope to get in a usable and stable manner from the Mozilla effort.
Although it hurts, sometimes it's best to let die what cannot live. I don't see a point in funding sendmail. I'm not so sure about mozilla, but with my impressions of it, I'd say there are better designs out there which are already posing nice features worthy of sponsorship.
What do you think ? Is sponsoring old monsters really the way to go ? IMHO it's not.
Of course you can detect a stealth plane. You can detect a stealth anything. It exists, therefore it interferes.
I believe it's the Typhoon class submarines you can detect by listening for silence in the waters. The submarine is so ``stealthy'' that it's more silent than the water that surrounds it. Blammo!
I would have thought that satellites would have been the first to be used to track stealth aircraft. If you track how the earth reflects a radio signal (radar), and then suddenly something breaks the usual reflection (a stealth plane will break the reflection, but it will not reflect much back itself) you know where to aim.
Isn't it just fantastic what can be done with technology today;)
Huh ? Anything network based is prior art. This is the way message-passing systems work. Someone sends a message, and gets a processed reply. This is hardly new technology.
I as well as thousands/millions others who has written networked applications has prior art for this patent.
Besides, I more and more get the feeling that patents should be neglected in the field of computer science. At least the patents which everyone with basic knowledge in the field can see are merely another formulation of some well known problem or application of technology/theory.
I'm not going to check the patent office when I write apps. I'm convinced that I'm infringing a handfull of patents every time I write something, and thats good enough for me. This madness can be stopped in one way; not by changing the rules (because too many people with too much influence seem to like the rules), but by ignoring the rules.
It is not in general a viable solution to a bad rule, to just ignore it. Society can't build on a strategy like that. But in this case, I think it's justified. This could be thought of as a ``virtual civil war'', where the oppressed fight back but noone gets hurt.
I'm not afraid, and you shouldn't be either. If we live in a world where people can get arrested for using common knowledge in a way that helps society, then so be it. Go ahead and make my day.
Seems like quite a few places use domino. Still, all I hear is Exchange this and Exchange that. I'd like to exchange Exchange with something that runs on a fairly stable OS, like XXX (be it domino or whatever) on Linux (or *-BSD).
At least in my country people seem to be fairly MS oriented. What's the hazzles of moving on to domino ?
Anyone with experience please reply. I'm sick of power-cycling NT ``servers'' every two days...
This guy doesn't have terabytes of data. I'm sure you have some very nice hardware and I'm sure your solution is the best one possible. But please, we're talking mailboxes here, and this guy needs something faster and eventually safer than the _single_ disk he's using now.
About simultaneous operations... Well, why do we put a kernel underneath the applications ? IDE performs well in the real world. Besides, how many places can the heads in your SCSI drives be in simultaneously ? I would guess, just as many places as the heads can be in, in IDE drives. One.
And yes, the major problem with IDE is that you need one controller for every two disks, to keep the performance good. That's a problem which SCSI doesn't have to the same extent (you'll need one controller for every 6 drives or so, on U2W with fast drives). But considering that this is a small-scale setup, it's outright stupid to outrule IDE because you can't put 1000 disks in the system.
Remember ``Hardware RAID'' is just a smaller processor running software as well. The PII+ in most modern systems is way faster than the i960 or m68X in a hardware RAID controller.
I've seen quite a few people finding in disbelief that they surely didn't get what they thought they paid for when buying HW RAID solutions.
Back in the old days I'm sure letting an i960 do parity calculations was a boost. Well, times change.
The _only_ thing I've seen HW raid controllers being better at, is large setups (10+ disks) where a pure SW solution will load the memory and PCI busses of the system heavily. Especially RAID-1 where a SW solution will have to duplicate data to all disks, the HW solution will have an edge moving this duplication off the main memory / PCI bus.
For smaller setups, like the one in question here, software RAID is absolutely both a viable solution, and probably offers by far the best price/performance.
First of all you might want to check out other MTAs, as well as other methods for storing the user's mails. If all mailboxes reside in the same directory, you're spending all your time in the kernel doing _linear_ searches thru the mailbox directory. You could spend millions on EMC hardware without seeing _any_ performance increase.
I'd recommend using the Postfix MTA, as it has almost all features of Sendmail, and it's secure, and (hold on) it's even faster than QMail. Eventually you could use it with the Cyrus IMAP/POP services. You definitely want to make sure that you don't have all mailboxes in the same directory. Build a hierarchial structure where you never have more than say 30-50 subdirectories/files in one directory.
Ok, if disks are still your problem, consider: 1) Software RAID is usually a lot faster than hardware RAID. And for the money you save on the HW controller you could buy faster/more disks. 2) An IDE disk is identical to an SCSI one, except of course for the interface and the warranty. The price difference is mainly due to the warranty. 3) UDMA/ATA-{33,66} IDE interfaces are as fast as any SCSI solution if you keep _one_ disk per channel. The main problems with IDE solutions is the short cable length allowed (a problem for 10+ disks) and the number of controllers you must have (one controller for each two disks)
You can spend $50K on a SCSI/HW-RAID solution easily. And you won't know if you'll even get the speed of one single UDMA drive from it (yes people actually get 15MB/s both from their single UDMA drives, and from their expensive DPT RAID solutions). At least consider a software-RAID and eventually IDE solution before rushing out to spend the next 10 years budget on the shiny HW-RAID solution.
Your setup is fairly small, eg. you would probably do just fine with a four-disk RAID-5/10 for spool and mailboxes. This is where SW RAID is worth considering. Granted, for 20+ disk systems, HW RAID may well be a better way to go, eventually combined with SW RAID.
All other threatening diseases seem to have been either cured, or controlled. Now that we might see cancer being ``cured'', it makes me wonder what the next great disease will be.
Ok, there's still HIV. Although we don't have a cure for it, we know how we can be reasonably safe not to get it (become a geek, don't have sex - (just kidding);)
Will it be a virus or bacteria ? Or do we really have to start a nuclear war to finally put an end to it all ?
New diseases has always come up. Another one has to, once cancer is out of the picture.
Well, if you asked a group of embedded systems engineers what language they used for fine tuning time-critical sections in drivers, you would see that Visual Basic is not popular, while 90% of the developers seem to be using asm for their ``applications''.
I'm sure VB is hot for business apps. But I can't say I care. What about real-world apps ?
One heck of a lot of programmers use Pascal (because of Delphi), and we're a lot out there using C and C++ because of run-time efficiency and portability, and because of language sophistication (well, that holds for C++ at least).
The main problem I guess is, that people tend to see the most popular language (or, the one that requires the most programmers to work with it) as the ``winning'' language. In real-world large applications however, a number of languages are often used. LISP/prolog for the extensibility and AI parts, C for the APIs, C or C++ for the core parts, Pascal/VB/objc for the front-ends etc. etc.
The fact that a language has a lot of programmers either reflects that it is a language which is good, or a language which require many people working with it to actually get results.
The thing that makes me wonder about these ``open sourcing windows/ce/nt/...'' is, that people seem to think that it would be a great thing.
Well, even if they GPL'ed or BSD'ed (or whatever) one or more of the windows variants, what would we have ? We would have a pretty crappy kernel, a GUI API that's mixed 16/32 bit and not going to 64. We would have one hell of a hard work in front of us, if we'd want to redesign windows to be a capable and stable OS.
It's much easier (and already happening) to simply stick with the OSS systems of today, and fix the one thing that's lacking there, namely larger applications (for business, CAD, etc.).
We already have a much better development environment, and games are coming too. We have the kernels, we have the development environments, and they're both far superior to anything windows can offer.
We're missing the _easy_ parts, the userspace stuff. If we where to fix an open-sourced windows, we would have a problem on our hands larger than we can possibly imagine. Fixing (redesigning) a broken kernel will break the one thing windows has as an advantage, the larger applications. We would have to convince the vendors to rewrite their apps too.
It's just so much easier for us, and for the ISVs to just write software on a well designed platform once and for all.
MS can open-source whatever they please. I doubt it would really help.
Sure, they had the aforementioned commercial all over the back of my last LJ issue. I read it, and though ``hell, these guys better buckle for the bumpy ride that's about to hit them hard, in the face, twice''. Sure, it's really good old mudslinging.
But keep in mind, that he who slings mud, looses ground.
I don't use their X server, and I'm not planning to. But if I was planning to, I might reconsider. It's not that a clueless marketing department will necessarily make their X server a bad one. Not at all. In fact i'm pretty sure their X server is a fine one.
But if their internal communication is so vague, that the marketing people have no clue what-so-ever that that is Not The Way (TM) to write an advertisment for the community, I seriously doubt that the executives know what product those ``techie'' guys are brewing, that the techies hear what customer support knows, and in the end, I doubt that the company as a whole is a company I could trust. Trust for support, and trust to be around in two years from now.
Yes, it's a little exaggerated, I know. You see marketing-, executive- and tech- departments missing eachothers points and goals even small companies. There are probably few companies out there where people in different departments actually have a clue about not only the goals, but also the ethics and methods used by other departments.
(Ok, code of ethics for marketing departments, that's easy.:) But the problem is that marketing were marketing for techies, and should have applied techie-ethics (read: community-ethics) not marketing ethics). Marketing ethics are fine when marketing for marketing people. Not when marketing for people who know words such as ``fair'', ``correct'' and ``respect''.
One reason for avoiding AMD microprocessors for SMP is that there seem to be no motherboards available.
Another good reason is, that people going for multiple CPUs are often people who need floating-point calculations going fast, and AMD has been unable to deliver this for long.
However with the Athlon, AMD seem to be past this. For sure, my next box at home will be a dual Athlon, if the boards show up, and the price is somewhere near the dual inel P today.
I can well imagine my next box at work being a SMP AMD box. But we need to see motherboards first. It would also mean a lot to me, if Asus shipped a SMP Athlon board. They've been shipping some rock solit intel based SMP boards so far, and seeing them shipping Athlon SMP boards _would_ make a difference, at least for me and my employers.
As you said, it depends on the problem at hand. Emacs is not really threaded, so your typing will still lack on a 550MHz if that's the kind of problems you're having:)
Still: *) Many systems run several serial compute jobs. Those will run on each their CPU, and thus you have close to 100% speedup (give and take some) *) Even if you run only one CPU intensive serial job, if it's a workstation it will feel as though nothing was running on it at all. That's nice. *) For completely serial tasks, such as compilation, make will start a number of jobs for you, and again, you have good speedup.
You will most often see below 100% speedup, because the CPUs share memory bandwith and disk I/O. But in some cases (where problems fit in L1/L2 cache) you see superlinear speedup because both CPU intensive tasks fit in the (not shared) CPU cache, and those other maintenance jobs will only destroy half as much of your total L1/L2 cache as it would with half the CPUs.
I have a dual at home and at work, because C++ compilation is slow.
You don't want titanium. It breaks when used with nikkel, which means almost any alloy used in computers.
That's why they don't use standard steel tools for the blackbird planes. They'd ``rust'' (or something similar to rusting, I'm not a material physics major) apart if repaired with tools made of nikkel-holding alloys, which is almost any kind of normal steel used in tools.
Slashdot Mirror
I felt that KDE was too ``german'' in it's looks and feels (no offence meant to germans, but I'm sure some of you know what I mean). So I preferred GNOME over KDE. However, GNOME was slow. Especially when switching desktops, I would have to wait for -too long- to get a netscape away and six terminals where I wanted them.
So I decided to just skip the desktop race, and go with enlightenment, straight, mean and lean. Much to my surprise, this didn't change the situation... After a little wondering, I started out on a new track.
Now, with GNOME and icewm as my window-manager, I have a lean and fairly mean desktop system going, and it's *MUCH* faster at desktop switching than enlightenment ever was.
Clue: If GNOME is too slow for you, try replacing the default window manager. Try icewm (has an *ugly* default theme, but has others which are nice and readily available from T.O.). Or try sawmill (haven't done that myself - yet).
For historical reasons, enlightenment is the default window manger in the GNOME releases done by redhat and others. This is changing. Enlightenment is - hands down - the most artistical window manager I've ever seen and used. But there's just (IMO) too much of artistical sophistication instead of lean code in that one, to perform well on ``old'' systems (my dual PPro with a Matrox I for example).
GNOME with icewm rocks. It seems faster on FreeBSD 3.4 than on Linux 2.2 though, but on the other hand, Linux wins when it gets to disk I/O etc. on the day-to-day workloads. GNOME is not slow in a sane configuration, it's just slow in it's default configuration.
Never ask a geek why, just nod your head and back away slowly ;)
;) It impresses the shit out of people. I've seen people running news servers and web servers on these pocket devices, as a temporary solution though, but still the ordinary unknowing people (induhviduals) around you are baffled. Their beliefs faint, they turn pale, they eat off you hand, they'd bend over and say thanks afterwards if you told them to :)
Yes it's bad that you'll be paying for a CE license when you won't be using it. However, if enough people request a BSD or ``blank'' version of the z50, I'm sure IBM will notice and offer that option.
Don't underestimate the value of BSD or Linux on this kind of machine. Fit it with a FE card, and you'll have a router/firewall/web-server/etc. Not that you would want to use it for production, but imagine coming to some customer as a networking consultant, technician, or whatever, then plugging in your handheld to the network - and instantly they have the transparant proxy and news server they needed temporarily
The OS is the computer.
Eric, can we please have the JF as a context diff ?
:)
Of course, it's not because we want to save bandwidth, it's only to make it more easily readable for those who read earlier versions
Think about it. Whenever you use a piece of software with those rediculous licences (if you do so at all), you just click ``accept'' and never think about the license again.
Basically the licences often say that ``you have no rights'' and ``it's your own fault'' etc. People ignore that now, and they think that if they don't bug the company, the company won't bug them.
If these licences were enforced, and if they could be so with backing in the law, people might start considering what they are actually agreeing to.
If this passes, I think licences might start to change. People might not accept a ridiculous licence if they know for sure that by accepting it they _will_ be breaking the law.
(funny example: licence on the drivers to the i810 motherboards on the accompanying CD say thay you may not use them commercially, eg. you cannot run windows on an i810 motherboard in your company)
This is another step towards wide use of Free Software and sane licences.
"If I type in my shell"
That should - of course - have read:
"If I type TAB TAB in my shell"
But Slashdot ate my < and >s because they weren't - eh.
I keep hearing that from everyone who doesn't actually use (GNU/)Linux... blablabla but it lacks software blablabla
If I type in my shell, I get this question: "There are 1852 possibilities. Do you really wish to see them all? (y or n)" That's the number of single _programs_, or ``pieces of software'' that I have installed on a fairly standard desk-top developer machine.
I recently started porting some software to NT, it's been four years since I used that OS last... After installing IE4 (required for VC), Visual C++ 6.0, MS Win32 SDK, (and in order not to lose my mind completely: Cygwin too), I fired up this huge IDE GUI development environment. Especially VC++ is something that Win developers claim Linux lacks. I simply cannot imagine why. Hang on:
In order to do development on NT I depend on the IDE. If the IDE does not offer the combination of commands I wish to apply to my work, I'm shit out of luck. There's just no way to get boring repetitive work done easily, if it happens to not fit exactly into the provided dialog box. The IDE may be very nice, but it is _inherently_ limited because it has no way of executing scripts. It lacks a shell.
On GNU/Linux, I run Emacs. For editing. Then I run bash for CLI. Then I run make to build. I use X with KDE to *integrate* these xterms and editors into one large IDE. In short: NT *has* an IDE, GNU/Linux with the standard tools *IS* an IDE.
X+Emacs+GCC+Make+xterm+bash+... is quite a large program, if you look at all the parts as one. Anyone claiming that GNU/Linux is lacking software is someone who can't see the forest for trees.
Granted, there may be a lack of integrated office suites. They should be just around the corner though. A lot of people won't need them though. Again, they limit you somehow. As an example: Work four people on a report using either Word or LaTeX+CVS. Assuming you actually _know_ a little word or a little LaTeX. LaTeX has no support for workgroups, CVS has no support for typesetting, but they both keep things simple and work on files, there you have your workgroup-aware typesetting tool.
If you have to complain, say something clever. Like lack of huge but clumsy, limited, and hard-to-make-do-repetitive-work-for-you GUI applications.
That was pretty much about time someone took the wavelets into image compression.
:)
I worked with wavelet transforms (Daubechies wavelets) a year and a half ago, and back then it was pretty clear that it would be possible to compress images and sound much harder using the wavelet domain rather than the Fourier domain.
For those who don't know, the trick is:
JPEG/MPEG/MP3 uses Fourier transform to transform the image/sound data into their spectral components. But this spectral representation of the data does not say anything about the _locality_ of the frequency components. Therefore representing spikes/discontinuties will require a very large number of frequency components when using Fourier domain, which in turn leads to poor compression. You can see this problem by drawing a few sharp lines in an image and compressing it hard with JPEG.
Wavelets on the other hand, represent both an equivalent of the frequency component, along with locality information. Spikes/discontinuities can be approximated well using only a few wavelets. This in turn leads to good compression.
Another nice thing about wavelet compression is, that wavelets tend to represent discontinuities well, even with hard compression (eg. a lot of missing or roughly approximated wavelet components). Therefore a very hard compressed image will still have fairly sharp edges, completely contrary to JPEG compression. This is pretty important if you compress a picture holding text.
Anyways, someone is now working on JPEG with wavelets... What about sound and video ?? There is no reason as to why wavelets should not provide equal improvements in both audio and video.
My personal conspiracy theory is, that there exists a *LOT* of expensive hardware that can do Fouries forwards and back to allow real-time encoding and decoding of MPEG movies in good quality. The companies producing these devices will lobby any standards-organisation to *NOT* consider wavelets and stick to good old Fourier. If this holds, it will take a few years until we see Wavelet compressed video
Nope. Sure it's a problem, but nobody gets hurt, planes don't fall out of the sky.
The report can be re-printed when the problem is fixed. But if the only problem really turns out to be exactly this, formatting of output intended for humans-only, then the problem is small. I know pretty darn well that a syslog entry from 1985 doesn't belong in a recently installed PII system, and my supervisor at university will also have a pretty good understanding of whether the report I hand in is written in jan. 2000 or jan. 1900.
I have already seen some ``Y2K problems'', which were caused by people trying to upgrade BIOS software on systems to ensure Y2K compatibility. And I'm pretty confident that that will be all of the Y2K bugs I get to see. Even at the university some machines (especially a web-server holding lecture material that I desparately need now) are shut down due to Y2K. I nearly pissed my pants when I found out. Sure, there are lots of Y2K problems out there, but I have yet to see one which is caused by a problem in handling dates.
Oh, it was never my show anyway :)
But the bug you see is only in formatting of output right ? The systems still work, the little clock in the top right corner just prints the wrong year, or ?
My argument is, that since most systems don't use decimal numbers internally, they will keep on running. Perhaps the user gets confused when he/she sees the wrong year (especially after a tough new-year's eve you can sort of get in doubt as to what century you're waking up in), but it's only formatting and not critical (outside the financial sectors).
Or have you actually seen a Y2K bug that caused something to *break* except for formatting/prettyprinting ?
Y2K is a scam.
;)
g htwouldhappen, I'd be happy to receive donations ;)
Sure there must have been problems in the banking/insurance/financial sector with the old Cobol code, but computers hasn't used the decimal numbers for keeping track of time since sometime in the late sixties. Y2K is an application-software-only problem, and it occurs mainly with either _very_ poorly written software or with software written in a language/environment that has the problem (Cobol).
All modern software (from early '70s and newer) uses other ways to keep track on the time (eg. seconds from 1.jan 1970), it's almost entirely written in C, and it will not have a problem. (Even software written in other languages have the language compilers/environments written in C, and therefore doesn't have the problem either).
So why were there BIOS updates out ? Well if enough customers believe there is a problem, the manufacturers better do _something_ (eg. flip a few bits, insert a nop here and there) to give the customer the impression that a problem present was fixed. Marketing and crowd-of-lemmings stuff.
I would be willing to bet a beer with anyone on this. However, there is the _slight_ chance * consequence equation that makes me sort of reluctant to bet with all of slashdot, despite the incredible amount of beer I would receive. (This is a small country, I wonder if it would sink in the ocean once I won
My 1996-BIOS dual PPro keeps on running with it's vintage BIOS until it gets replaced in a few years, or until the apocalypse caused by the aforementioned crows of lemmings makes computers a little concern.
Happy new year !
(Oh, and if anyone has stocked up a few too many of those cal. 50 with IR scope and they figure they won't need them for looting/rioting/protection/whateverthehellyouthou
You can be held liable for whatever you promise, which is why most open source software has a clause something like:
;)
:) This would probably not change the promises or the software, but it would make the general public aware of the lack of promises they actually get from spending huge cash on closed source software.
>> This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
If you read the EULA from Microsoft, I'm pretty sure that they have a similar clause, much to most people's surprise. Then what are you actually paying for, you might ask. Well, that's the good question
The car-makers have a responsibility of making cars *reasonably* safe, according to government regulations. They are not required to stop your kids from driving into brick walls using your car. They are however required to make sure your car doesn't fall apart or stops breaking when you want it to etc.
There are no such rules (yet) for software. The vendors make the rules, and the vast majority of customers/consumers simply neglect this fact and *expect* that there is some sort of reasonable agreement behind it all, just like when they bought their car.
Open source licences are usually very cautios to ``warn'' people of the possible dangers that lie ahead when using the software. And some people may even pay attention because ``there's gotta be a catch with gratis software after all''. I think this is a pretty good way to handle things.
There could be some sort of either regulations or at least some rule that software vendors should state LOUD AND CLEAR what they promise and what they don't. Pretty much like the warning messages on cigarette boxes
>> My mother said there aren't real monsters. But there are - Newt, Aliens II
Unfortunately both mozilla and sendmail are monsters. It is going to take a lot of effort to make sendmail a lean and mean mailer, and I don't see why anyone would care to even try. Postfix is out there, which has the features of sendmail (almost) and is already lean and mean. Of course, postfix is not a big name yet, like sendmail, and they're already sponsored. I think sponsoring sendmail is nothing but a political decision. It really gets us nowhere we can't go already.
Mozilla, while being a monster, could be wort sponsoring. Personally, I'd rather see a Gtk+ port of Konqueror, the KDE browser, or some funding of Mnemonic (anyone remember that effort?). Those seem to be much more clean by design, and actually sponsoring them could well give us a browser that was both lean and mean, and had the features we hope to get in a usable and stable manner from the Mozilla effort.
Although it hurts, sometimes it's best to let die what cannot live. I don't see a point in funding sendmail. I'm not so sure about mozilla, but with my impressions of it, I'd say there are better designs out there which are already posing nice features worthy of sponsorship.
What do you think ? Is sponsoring old monsters really the way to go ? IMHO it's not.
Of course you can detect a stealth plane. You can detect a stealth anything. It exists, therefore it interferes.
;)
I believe it's the Typhoon class submarines you can detect by listening for silence in the waters. The submarine is so ``stealthy'' that it's more silent than the water that surrounds it. Blammo!
I would have thought that satellites would have been the first to be used to track stealth aircraft. If you track how the earth reflects a radio signal (radar), and then suddenly something breaks the usual reflection (a stealth plane will break the reflection, but it will not reflect much back itself) you know where to aim.
Isn't it just fantastic what can be done with technology today
Huh ? Anything network based is prior art. This is the way message-passing systems work. Someone sends a message, and gets a processed reply. This is hardly new technology.
I as well as thousands/millions others who has written networked applications has prior art for this patent.
Besides, I more and more get the feeling that patents should be neglected in the field of computer science. At least the patents which everyone with basic knowledge in the field can see are merely another formulation of some well known problem or application of technology/theory.
I'm not going to check the patent office when I write apps. I'm convinced that I'm infringing a handfull of patents every time I write something, and thats good enough for me. This madness can be stopped in one way; not by changing the rules (because too many people with too much influence seem to like the rules), but by ignoring the rules.
It is not in general a viable solution to a bad rule, to just ignore it. Society can't build on a strategy like that. But in this case, I think it's justified. This could be thought of as a ``virtual civil war'', where the oppressed fight back but noone gets hurt.
I'm not afraid, and you shouldn't be either. If we live in a world where people can get arrested for using common knowledge in a way that helps society, then so be it. Go ahead and make my day.
Seems like quite a few places use domino. Still, all I hear is Exchange this and Exchange that. I'd like to exchange Exchange with something that runs on a fairly stable OS, like XXX (be it domino or whatever) on Linux (or *-BSD).
At least in my country people seem to be fairly MS oriented. What's the hazzles of moving on to domino ?
Anyone with experience please reply. I'm sick of power-cycling NT ``servers'' every two days...
This guy doesn't have terabytes of data. I'm sure you have some very nice hardware and I'm sure your solution is the best one possible. But please, we're talking mailboxes here, and this guy needs something faster and eventually safer than the _single_ disk he's using now.
About simultaneous operations... Well, why do we put a kernel underneath the applications ? IDE performs well in the real world. Besides, how many places can the heads in your SCSI drives be in simultaneously ? I would guess, just as many places as the heads can be in, in IDE drives. One.
And yes, the major problem with IDE is that you need one controller for every two disks, to keep the performance good. That's a problem which SCSI doesn't have to the same extent (you'll need one controller for every 6 drives or so, on U2W with fast drives). But considering that this is a small-scale setup, it's outright stupid to outrule IDE because you can't put 1000 disks in the system.
Remember ``Hardware RAID'' is just a smaller processor running software as well. The PII+ in most modern systems is way faster than the i960 or m68X in a hardware RAID controller.
I've seen quite a few people finding in disbelief that they surely didn't get what they thought they paid for when buying HW RAID solutions.
Back in the old days I'm sure letting an i960 do parity calculations was a boost. Well, times change.
The _only_ thing I've seen HW raid controllers being better at, is large setups (10+ disks) where a pure SW solution will load the memory and PCI busses of the system heavily. Especially RAID-1 where a SW solution will have to duplicate data to all disks, the HW solution will have an edge moving this duplication off the main memory / PCI bus.
For smaller setups, like the one in question here, software RAID is absolutely both a viable solution, and probably offers by far the best price/performance.
First of all you might want to check out other MTAs, as well as other methods for storing the user's mails. If all mailboxes reside in the same directory, you're spending all your time in the kernel doing _linear_ searches thru the mailbox directory. You could spend millions on EMC hardware without seeing _any_ performance increase.
I'd recommend using the Postfix MTA, as it has almost all features of Sendmail, and it's secure, and (hold on) it's even faster than QMail. Eventually you could use it with the Cyrus IMAP/POP services. You definitely want to make sure that you don't have all mailboxes in the same directory. Build a hierarchial structure where you never have more than say 30-50 subdirectories/files in one directory.
Ok, if disks are still your problem, consider:
1) Software RAID is usually a lot faster than hardware RAID. And for the money you save on the HW controller you could buy faster/more disks.
2) An IDE disk is identical to an SCSI one, except of course for the interface and the warranty. The price difference is mainly due to the warranty.
3) UDMA/ATA-{33,66} IDE interfaces are as fast as any SCSI solution if you keep _one_ disk per channel. The main problems with IDE solutions is the short cable length allowed (a problem for 10+ disks) and the number of controllers you must have (one controller for each two disks)
You can spend $50K on a SCSI/HW-RAID solution easily. And you won't know if you'll even get the speed of one single UDMA drive from it (yes people actually get 15MB/s both from their single UDMA drives, and from their expensive DPT RAID solutions). At least consider a software-RAID and eventually IDE solution before rushing out to spend the next 10 years budget on the shiny HW-RAID solution.
Your setup is fairly small, eg. you would probably do just fine with a four-disk RAID-5/10 for spool and mailboxes. This is where SW RAID is worth considering. Granted, for 20+ disk systems, HW RAID may well be a better way to go, eventually combined with SW RAID.
My 0.02 Euro.
All other threatening diseases seem to have been either cured, or controlled. Now that we might see cancer being ``cured'', it makes me wonder what the next great disease will be.
;)
Ok, there's still HIV. Although we don't have a cure for it, we know how we can be reasonably safe not to get it (become a geek, don't have sex - (just kidding)
Will it be a virus or bacteria ? Or do we really have to start a nuclear war to finally put an end to it all ?
New diseases has always come up. Another one has to, once cancer is out of the picture.
Well, if you asked a group of embedded systems engineers what language they used for fine tuning time-critical sections in drivers, you would see that Visual Basic is not popular, while 90% of the developers seem to be using asm for their ``applications''.
I'm sure VB is hot for business apps. But I can't say I care. What about real-world apps ?
One heck of a lot of programmers use Pascal (because of Delphi), and we're a lot out there using C and C++ because of run-time efficiency and portability, and because of language sophistication (well, that holds for C++ at least).
The main problem I guess is, that people tend to see the most popular language (or, the one that requires the most programmers to work with it) as the ``winning'' language. In real-world large applications however, a number of languages are often used. LISP/prolog for the extensibility and AI parts, C for the APIs, C or C++ for the core parts, Pascal/VB/objc for the front-ends etc. etc.
The fact that a language has a lot of programmers either reflects that it is a language which is good, or a language which require many people working with it to actually get results.
You be the judge.
The thing that makes me wonder about these ``open sourcing windows/ce/nt/...'' is, that people seem to think that it would be a great thing.
Well, even if they GPL'ed or BSD'ed (or whatever) one or more of the windows variants, what would we have ? We would have a pretty crappy kernel, a GUI API that's mixed 16/32 bit and not going to 64. We would have one hell of a hard work in front of us, if we'd want to redesign windows to be a capable and stable OS.
It's much easier (and already happening) to simply stick with the OSS systems of today, and fix the one thing that's lacking there, namely larger applications (for business, CAD, etc.).
We already have a much better development environment, and games are coming too. We have the kernels, we have the development environments, and they're both far superior to anything windows can offer.
We're missing the _easy_ parts, the userspace stuff. If we where to fix an open-sourced windows, we would have a problem on our hands larger than we can possibly imagine. Fixing (redesigning) a broken kernel will break the one thing windows has as an advantage, the larger applications. We would have to convince the vendors to rewrite their apps too.
It's just so much easier for us, and for the ISVs to just write software on a well designed platform once and for all.
MS can open-source whatever they please. I doubt it would really help.
Sure, they had the aforementioned commercial all over the back of my last LJ issue. I read it, and though ``hell, these guys better buckle for the bumpy ride that's about to hit them hard, in the face, twice''. Sure, it's really good old mudslinging.
:) But the problem is that marketing were marketing for techies, and should have applied techie-ethics (read: community-ethics) not marketing ethics). Marketing ethics are fine when marketing for marketing people. Not when marketing for people who know words such as ``fair'', ``correct'' and ``respect''.
But keep in mind, that he who slings mud, looses ground.
I don't use their X server, and I'm not planning to. But if I was planning to, I might reconsider. It's not that a clueless marketing department will necessarily make their X server a bad one. Not at all. In fact i'm pretty sure their X server is a fine one.
But if their internal communication is so vague, that the marketing people have no clue what-so-ever that that is Not The Way (TM) to write an advertisment for the community, I seriously doubt that the executives know what product those ``techie'' guys are brewing, that the techies hear what customer support knows, and in the end, I doubt that the company as a whole is a company I could trust. Trust for support, and trust to be around in two years from now.
Yes, it's a little exaggerated, I know. You see marketing-, executive- and tech- departments missing eachothers points and goals even small companies. There are probably few companies out there where people in different departments actually have a clue about not only the goals, but also the ethics and methods used by other departments.
(Ok, code of ethics for marketing departments, that's easy.
One reason for avoiding AMD microprocessors for SMP is that there seem to be no motherboards available.
Another good reason is, that people going for multiple CPUs are often people who need floating-point calculations going fast, and AMD has been unable to deliver this for long.
However with the Athlon, AMD seem to be past this. For sure, my next box at home will be a dual Athlon, if the boards show up, and the price is somewhere near the dual inel P today.
I can well imagine my next box at work being a SMP AMD box. But we need to see motherboards first. It would also mean a lot to me, if Asus shipped a SMP Athlon board. They've been shipping some rock solit intel based SMP boards so far, and seeing them shipping Athlon SMP boards _would_ make a difference, at least for me and my employers.
As you said, it depends on the problem at hand. Emacs is not really threaded, so your typing will still lack on a 550MHz if that's the kind of problems you're having :)
Still:
*) Many systems run several serial compute jobs.
Those will run on each their CPU, and thus you have close to 100% speedup (give and take some)
*) Even if you run only one CPU intensive serial job, if it's a workstation it will feel as though nothing was running on it at all. That's nice.
*) For completely serial tasks, such as compilation, make will start a number of jobs for you, and again, you have good speedup.
You will most often see below 100% speedup, because the CPUs share memory bandwith and disk I/O. But in some cases (where problems fit in L1/L2 cache) you see superlinear speedup because both CPU intensive tasks fit in the (not shared) CPU cache, and those other maintenance jobs will only destroy half as much of your total L1/L2 cache as it would with half the CPUs.
I have a dual at home and at work, because C++ compilation is slow.
You don't want titanium. It breaks when used with nikkel, which means almost any alloy used in computers.
That's why they don't use standard steel tools for the blackbird planes. They'd ``rust'' (or something similar to rusting, I'm not a material physics major) apart if repaired with tools made of nikkel-holding alloys, which is almost any kind of normal steel used in tools.