Source code control only works with strong managment and as long as everyone is on the same page and aiming in the same direction. My former employer had that problem, people could use CVS, people knew how to add files, remove files, check them in (a 'asshole' or two didn't understand "update" and would work alone for 2 weeks and commit but I digress) they knew the tool well but shitty and broken code would routinely get checked in. It was one of the reasons I had to leave, you can give everybody a chance, you're allowed one broken build or patch digression but then you should learn and it was a practically a daily occurence. The bosses didn't seem to ever understand that they were supposed to bite the heads off of people who screw things up. The policy now is there is a build everyday at 3:30 and nobody can leave until it builds and regression tests. (which is insanity in itself, also note how everybody is punished and not the people who screw up the build)
Some people use source code control as a way to backup their system. They don't mind checking code in every hour, cutting labels, etc.. Some people see it as a step to doing the job (and often a painful one) and they will dump code in once a week or every two weeks and introduce radical changes (usually at times associated with scheduled code drops so they can get their stuff in.) And there are other ideas or theories still. Linux doens't really have management and so you have to deal with all of that stuff while trying to still move forward. I think you probably end up at the same place, a Linus tree that he approves, you just spend a painful year or two trying to do something different. The fundamental prereq for CVS or any source control is trust, trust that the people putting the code in are going to the same place with their code that you are. I also think that the last thing you'd ever want is for someone other than Linus to settle a dispute by virtue of having commit rights... that has happened when Alan was the patch hoover or at least it was accused that he wasn't passing some patches on the Linus for political reasons.
Second, I think there might be a lot of public misconception about Linux branching. I built an embedded Linux kernel and maintained it for a consumer device. I was a small scale Linus for close to 2 years. If you are using Linux in a production environment you're probably forked. (maybe in the head?;) Either the product you are building out of Linux demands forking (some custom embedded device or server or something) or you're running on Redhat or Mandrake and you're using their fork, and they do have unique branches. For the time being, that's how it is and that's how it's done and it's an awesome thing, it's not bad. Especially for the embedded world where you can really tweak the kernel to your device. I think that's the bigger problem Lineo, MontaVista and company have is catering to custom devices without customization, right now you pretty much have to break off from the Linus kernel if you're doing that kind of stuff. The other side of that is you have no idea how difficult it can be to maintain a custom fork, it's the same job Linus has and over the 2.3.90+ to 2.4.now kernels it has been especially difficult. The half life of patches for 2.4 isn't terribly long, particularly if they do anything with physical memory. I'd take about a week to roll from 2.4.3 to 2.4.8 and we really only use a handul of customizations and patches, that's just how long it would take to go through all of the changes by hand and make sure that the good ones get integrated and then fix them so they could integrate. In the mean time you might be hearing firealarms becuase the kernel someone is using is crashing. It's tough work, more so when you want to track Alan and Linus's kernels in case they fix bugs that might affect you. I have the highest respect for Linus he does a difficult job very very well.
As 2.5 simmers down some I think there will be a much stronger desire to not make the same mistakes, with 2.4 there were interfaces that changed during the 2.4 life span, a patch against 2.4.3 might not work against 2.4.8 without some modification. That makes Linus' job that much more difficult. Part of the reason is that maintainers who are supposed to be trusted came out with code that had some problems and wasn't trusted by other maintainers, so much for enough trust to use CVS.
Lastly (I'll close this ramble off) Linus and Linux have one strength that I really admire a lot having worked in the "mission critical" world of IBM and the banking industry. If something can be better they aren't afraid to make it better, even if it means changing code that isn't "broken." Right now in particular can be a radical time in the kernel, they're changing the block I/O layer! There are several VM layers in circulation. Fundamental parts are changing for the better, it's really hard to do that some times, you don't want to start crashing kernels or lose stability.
It's far better to do it than it is to try and hack around something that has outlived its design but "works." If Linus ignores all the new device drivers for 5 months while the VM, scheduler and block I/O stuff is sorted out, that's not a terrible thing.
The thing I would be thinking about would be low power, fanless, micro-atx or pc104 type form factor. So we could put Linux no them about build them in to little network routing and web serving bricks.
High priced reference platforms have never been in short supply. If I'm going to buy a $5000 system then getting an alpha, mips, sparc, or POWER/powerpc has never been difficult.
I own a couple of pieces of alpha hardware and it's fun in a geeky way to have. It's nice to test code on other platforms, it's nice to be able to learn assembly to other platforms and have something to work on, and a 21264 makes a hell of a web server. I'd love to have a newer PowerPC machine to work on but the prices just aren't there. If I could buy a motherboard and processor for $400-$500 maybe even $600 then I could easily see a little clique of people doing it. I can see real market value to it as well, I've seen 6 or 7 embedded jobs over the last month that were for PowerPC products.
I hope that they are interested in lowering the prices and ramping up some mass production of the hardware. I could also see a huge market for lower priced integrated PowerPC motherboards with G3s or even 60x processors on them; put 3 NICs, IDE and a PowerPC on a motherboard for $200 and you have a nice DIY home gateway/firewall/router box.
The desqview code itself isn't important. There might be pockets (no doubt there are) where this will matter but for the most part it doesn't.
I think the big thing is the shift in thinking that some big companies are starting to take on. A company cannot support a product forever, for the most part. It also cannot afford to not support products. At some point in time, and it's probably a lot earlier than it has been with the Borland products they've made free (beer) so far, it makes no sense for them not to make things free. The next logical step is for them to start releasing code to these older products. Something like desqview is special in that if people are using it and relying upon it, there is probably not anything that can really replace it. It's good to see, they aren't losing anything by giving it away now and the next step is the source code. It's part of the promise of opensource, your solutions will never die becuase you'll have to code if you need it; well it's about time that the software industry stepped up to that plate also because the competition is good and I've got hardware sitting around that might be interesting with some of that old software. Also, who's to say that something like WordStar might not have continued had it been opened up, it certainly had the community of users it was a support problem that did it in. Some of that older stuff could get migrated to more modern platforms. It does nobody any good to have something like that die.
How hard could this be to experiment with?
on
UNIX Process Cryogenics?
·
· Score: 5, Interesting
I've thought about this for booting issues. I have a server that's all journaled and everything and it's periodically get's bumped. Boot time is still on the order of 2 to 4 minutes for a full Linux server install. With my current stats that means I'm probably going to miss a hit or two on one of the web pages, all things being equal. A good portion of that is just icing though, things that are there "just in case" or get used infrequently. (Okay, I can screw with the init order and the problem essentially goes away or I can switch hardware but we're nerds and geeks so let's just explore this)
I was thinking about this and here was my dirty hacky idea. You need kexec, lobos, or something similar (actually a fairly modified version of it) you'll need on the order of 8MB of disk space and some kernel mods, which might not be that extensive.
I was thinking we develop some driver or process that consumes all of the memory and CPU in a system. It forces all of the processes to swap out, it would probably need to be a driver of sorts on current linux systems. Then it could dump the kcore out to a file somewhere, sync it, and hibernate. Then when the kernel boots up, if the right arg is passed in it could either load this image back in to ram in place of the kernel and then jump into it (easier said than done) early in the boot (page tables are made long before you have access to the drives and such so the logistics of this would need to be figured out) or it could boot up and use a different swapper partition and then have some kind of tool like kexec to load that image back in to ram and start it up. Or something, some how you should be able to recover the state of the system. File handles and everything would be there.
The harder part would be hardware and network transparency. You'd need to modify all of your drivers to make sure that the hardware could be reset and they could deal with it. I think it's a little easier for the network side because it would be similar to simply unplugging the network cable, you have open sockets that are talking to nothing and some software can deal with that pretty well. There is also some kind of system integrity or robustness piece that is needed, if the system some how changes when you bring your old image back it could break things, munge files, etc..
I've seen a TOS from Sprint that essentially required you to run a firewall. Personally, I think anyone who can afford the hardware and isn't running a firewall is nuts.
What are they going to do about all those cool $200 hub/switch/firewall/router boxes Linksys and company are making?
I can sympathize with the few who are worried about things being taken out of context or used in the public minds to sway opinions. but that's rare and what's more if you have a lawyer picking a your documents trying to make you look dirty you're also going to have a lawyer poking holes at those theories and defending you. The market can be quite forgiving also (Texaco comes to mind)
I'm thinking long termish. First of all, look at the MS case. They are looking at documents from the DR-DOS/Win 3.0 days. These are 10 year old documents and they have an impact on whether justice will be served or not. That very concept is defined by those documents; personally I don't think Gates himself ordered destructive actions towards competitors (so he shouldn't go to jail) but it's clear that there were some and mgmt knew of it.
Further look at the FOIA. There are 50 year old documents that tell us new things about history that are released all the time. I'd hate to think that a major crime could happen and be covered up so easily that in 50 years nobody could figure it out or see the truth.
For the most part, most people don't have anything to hide and shouldn't have to worry. I really have a problem with information destruction by policy, it's too similar to the nazi's burning books. In the digital age there isn't any reason to do it, space isn't a problem.
I'm wondering why both VM's can't be included in a distro and allowing the end user to select the one he/she wishes to compile into the kernel? Are the two implementations THAT mutually exclusive?
I've been wondering that myself. It's just work to do but it shouldn't be rocket science. Of course a number of people are concerned about the code size as is, you can't just add a new branch of code every time there is a conflict and pack them all together. This does seem like a somewhat good case for it.
BTW, this kind of bashing between the high priests of Linux is not good. You can bet your bottom dollar that MS is going to use this conflict to fuel their propaganda machine, saying Linux is a fractious OS run by a bunch of young upstarts who can't agree on anything.
No this is a good thing. Most of the linux kernel hackers have egos the size of small countries, and that's a good thing because they take pride in their work. Most of them also work as professionally and egolessly as I have ever seen. They can get in fights and then deal with it and keep working. This is far better than the closed world where people get in fights and take it personally and then try to react in some way. I've seen project where people were trying to fail the project to get revenge on someone on the team or managment for something stupid they did in the past. In the linux world people get in fights and everyone can see and they react accordingly, sometimes being told that you're being an ass is a good thing when you're being an ass, sometimes having people stop talking to you for a few days is a good thing, and sometime people appologize and that's a good thing. It's not behind closed doors though and it's hard to undermine things when it's all in the open. There is also some insanely good discussion on certain things some times. There is also something to be said for defending ideas and the strength they have when you can defend them in public.
Look at what they are trying to do. For the last 10 years, or more, really, that have been trying to get in to the enterprise through the desktop. That's the gold, that's where the long term hugh contracts live, that's how a company can make IBM dollars. (I think they are on the order of 4x more gross than MS still)
Now companies like IBM, Sperry, EDS and the former DEC and even to some degree HP aren't lazy or slouchy groups. Those companies and the few others like them have built the infrastructure that the western world runs on. So when MS tries to sell a SQL server solution for pennies on the dollar compared to the similar Oracle or DB2 setup you have to wonder, are they that much better and can do it that much cheaper or are corners getting cut? Or does it not do everything as well? I'm not tryng to raise up one company over another and I think sqlserver is a hell of a product but I really have to wonder if everything it's built upon is as solid as a zServer running DB2 simply because of the economics involved. I think a company like IBM would find a way to do it cheaper if it could.
Now MS owns the desktop, they aren't fighting for it nearly as much as they were before. They are still doing radical development, NT4 to 2000 was a big changed. 2000 to XP was a big one as well. They still want to sell things cheaply and they want to get to that gold ring. Something has to give and it's the stuff that nobody outside the enterprise cared about for the first half of MS's existence. 5+ 9s reliability, which you almost can't do on Intel hardware. Rock solid security, it's hard to add that on to something already built, especially hard if you have a complex security model. The kind of reliability and security performance that once compromised by stories like that can take months and years to recover from in the minds of some people.
I think it must drive them nuts. I think what also must drive them nuts is simply the fact that this stuff happens. They aren't stupid and I know that they can understand a market and I'm pretty sure they know what they need to do to win the enterprise over, I just don't think they can do it and I wouldn't be surprised if they were starting to think that when this kind of thing happens so regularly. I just don't see a company beating a world full of IBM, Oracle, SAP, EDS, and others at their game without doing something different. Something different isn't cutting corners and costs, it has to be radically different like Linux.
We already have a ton of this waste. That's the critical part. YM is only being used for commercial grade waste and there is a lot of it out there already just waiting for an accident or to be stolen by terrorists or something.
I think that it's far more likely that there could or would be an accident or "issue" at the dozens of sites around the country than at one storage facility.
I think that are some other issues of praciticality too. Yucca Mountain isn't exactly Malibu or Park City or Tampa. For the forseeable future people aren't going to be aching the build on it. It's a desert, a rough desert. They aren't tarnishing prestine wilderness or some ultradesirable place to visit or live. The waste needs to be stored and that's as good a place as there is. Throwing out a few of the more radical climate change theories, YM is going to substantially change for thousands and thousands of years and it hasn't for many thousands of years. I also kind of dismiss the "forget about it idea" it is marked and unless there is a catastrophic change to the human race it will be remembered for several thousand years.
More importantly, and more likely is that they aren't trying to convict you, they are trying to convict someone else and they believe you may have evidence. There is no 5th amendment in that case. You give up the keys or you obstruct justice or are found in contemp and go to jail.
To drive that to the logical conclusion, they try to convict you neighbor and they want all of you correspondance with him, do the saber ratteling routine and make you give up the keys. Then as they go through you computer they don't find evidence of wrong doing so much as they find evidence that you have unpopular views (perhaps you're a racist or a communist or something, not illegal to be but unpopular) and that get's entered in to the public record as well. You're an outcast and your public reputation is ruined and you have no recourse at all.
Now I'll give them the benefit of the doubt and assume that their PR firm screwed up and they're really talking about very specific types of data but this is most likely a scam. The counting argument is proven mathematically and you can't unprove that or circumvent it.
Let me explain. With a string of bits you can describe different types of data. Different combinations of bits can be used to describe different things, different strings of bits can be used in a multitude of ways but there is a limit to how much you can describe with a string of bits becuase it is fixed in length.
Say I take 1 bit. It's either 0 or 1. There are only two things I can describe with that one bit without extra information. (In compression, "extra information" implies bullshit, your idea doesn't work if it needs "extra information") It's impossible to make that one bit represent 3, 4 or more things without extra information.
With 2 bits I can represent 4 things. It's impossible to represent more. They are 00, 01, 10, and 11, there isn't a 5th possibility. From a compression standpoint that means that if I compress something down to 2 bits then with that particular compressor there are only 4 possible input that can compress to that size. One of those inputs might be the encyclopdedia britannica but that's a pretty specializae compressor then. Do the induction, 2**bits is the equation that defines the maximum number of possible inputs to produce that output of that length. This is encoding and what Shannon is most famous for, and his laws are still laws. There are a couple popular ways to do encoding with binary data in compression, arithmetic and Huffman are the most popular.
So where is this going? Well first, recursive compression. It's not compression becuase of that "extra information" problem. How many times do I run it to get my data back? Well that's simple, you compressed the linux kernel and you ran it a number of times x. x just happens to be about the same size as the linux kernel when represented in binary, gzip that number up and keep it in a safe place and you'll be able to restore your kernel that was compressed into 1 bit.
Second. Compressing random data. Say I compress a random string to 1/100th of it's size. To make the math easy, let's use a string that is 1024bits long, it is reduced to 10 bits (or 11 now and then.) 1024 bits can represent a lot of different things. 2**1024 of them. 10 bits can only represent 1024 different things, 2**10 = 1024. 2**1024 has 309 digits, 1024 is an teeny tiny fraction of it. Well under 1%, way way way under 1%. With that compressor that does 100:1 compression you can only compress, at the most, 1024 things out of 2**1024, without "extra information." That might be randomesque data but the fact is that if you pick a random string of 1024bit, you will very very rarely pick one of those 1024 strings that your compressor works on, it pretty much will never happen in practice because it's such a small percentage. This is a truth with all compressors. That's not to say that they can't compress a truley random string of bits 100:1, they just might want to take a few centuries to come up with the particular random string that they are going to use.
LZ77, LZ78, Burrows-Wheeler transforms, PPM modelers, Markov modelers, all modern generic lossless compressors exploit the fact that most "interesting data" that we want to compress has redundancy to it. If you allow for some degree of redundancy or perhaps a lot of it then you reduce the number of possible inputs to the "interesting inputs" and that's usually far smaller than 2**1024 for a 1024 bit string. It's still probably orders and orders of magnitude larger than 1024 though, it's quite easy to find more than 1024 interesting things you can represent with 1024 bits, the short of it is that a 100:1 compressor will only be able to compress 1024 things of that length. No matter how you model the data, you're limited by the possible number of inputs represented by the encoding.
Wine has been getting better and better and with the misc. binformat support I have it loading and running java and windows programs seemslessly, it's only a matter of time before someone tries to tie it all together.
Is this friend thing a start at copying advogato?
on
Slashdot Code Update
·
· Score: 2
If you had a few more degrees of friend or foe then and karma was applied accordingly then it starts to look more like the advogato trust metric which works pretty well.
It's not a well defined term. What exactly are you expecting? Some "problem solver" object you can just drop in to any program and reuse?
At it's most primitive layer, OOP is a coding thing, a syntactic sugar as so many professors like to say. A way to keep code clean. It's quite nice in that capacity, I don't mind writing "c = a + b;" and under the covers mpz_add is getting called and adding the arbitrary precision number a to the arbitrary precision number b and storing the out put in c. This isn't true OOP, I only include it becuase it's the OOPLs that have made it popularly available.
The next level is encapsulation and reuse. It's pretty damn nice to be able to write "d = {}" in python and create a hashtable dictionary. I do that is literally dozens of python programs. I could easily produce the same thing with list primitives or vectors and what have you and I might even be able to make it better for my particular use but for the most part the dictionary object in python or any of the STL objects are insanely useful, they are easy to use, quick and they are building blocks you can use to build bigger programs. You're already using this kind of code whether you know it or now, the C library is the same thing, it doesn't have the same syntactic sugar though and the things it does are generally more simple where some of these objects are higher level components. I wouldn't call the C library object oriented, per se, but might call a component that does the job of half a dozen C library calls an object.
The final layer of OOP as I see it is modeling. Again it's not well defined outside of a few circles and even then Grady Booch and Bertrand Meyer and the other gurus spend countless hours and books bickering about it. The modeling part is how we break a problem down and implement a solution for it. This I would say is a universal area where OOP has won and will continue to. When programs reach certain sizes, the human mind can really only deal with the complexity by building heirarchies and mapping the problems to solvable isomorphs and then trying to reuse ideas as you move through the heirarchies. If you're trying to multiply a few matrices and then hit them with the Gauss stick to tell if the bridge is going to work, it's probably not a good example of a problem to model that way. A GUI on the other hand is a perfect example where you can have a very clean hierarchy of windows of different types with different functions that behave differently. In my experience, I haven't seen any large scale projects started in the last 15 years that weren't object modeled, even if they were implemented in C or some other non-OOPL, I have seen a fair amount of code from the 1960s and 1970s that was very very complex because it didn't have any real object modeling. The linux kernel has a fair amount of object modeling in it, there are abstract block devices and character devices and they can all be treated the same way, buffering is an abstract concept regardless of the device. Even functionally implemented programs have some degree of object modeling in them.
Now what kind of OOP are you looking for? If you're looking for some kind of advanced object model with reflection and a complex object hierarchy, then maybe it's not being used to solve some engineering problems, it takes a certain size problem before that kind of effort really pays off in the implementation. If you're looking for code reuse then there are numerics packages and probably tons of other packages the will do that.
I've long though that this would happen and that it's overdue. Within IBM the PCCo is seen as a money loser, they are always let slide when other groups are expected to profit and given bonuses and raises depending on how much they actually do. They sabotaged the PSP group and many people believe they have blood on their hands for helping to fail OS/2.
They are also an integral part of IBM's business as a solution provider.
Back in the early 90's and late 80's PS/2 cost, literally 2 to 3 times more than a clone of the same vintage. They had MCA, to boot. There are also a fair amount of those machines still in service. It's not terribly uncommon in certain enterprises to look in the back room and see an old PS/2 486 class machine serving up files for a work group or driving a printer. While certainly not a speed demon, it does the job. With MCA bus mastering in some situations it still makes an accpetable server for certain tasks. Part of the reason is becuase the machines are about as reliable as PCs get. IBM at that time was holding their hardware to much higher relibility standards than just about anyone else and they were more expensive because of it. At that time I thought they were doomed and would be out of the business. If you're selling solutions to people and those solutions will last decades and PCs are an integral part of the user interface you need good PCs.
The same can be said of RS/6000. Outside of a few markets they are under performers. I've also seen RS/6000 machines fall off of loading docs and get run over by a forklift and have cases that are all dented to hell and still run flawlessly. Is it the fastest workstation? No. Does it do the job IBM needs to sell a $10million contract? Yep. and because of that they keep making them.
I keep thinking that they need to get out of it or just focus on netfinity and thinkpad but the truth is, when they are selling an industrial solution, who are they going to buy PCs from? Compaq? Hell no, they are a big competitor. Dell? Possibly but I just don't see optiplexes running for 10 years like an IBM PC. Gateway? You've got to be kidding. If you think of a PC as the world's most expensive peripheral and you need those peripherals to run mainframes, minis and super computers then IBM will keep selling them.
I agree. There is a lot of history and interesting events other than the crimes and busts. PC relay and FIDO are interesting in and of themselves. Much of the social interactions is also fascinating.
There are people I've lost touch with that I only ever knew through the BBSes. At times I wonder where they are.
I doubt that you could find a case of a school keeping it's doors open when it otherwise would have closed because of free software. Personally, I would hope that a school would stop using computers before it stopped teaching, but that's just my opinion.
Realistically though, free software has made a huge impact. I think the most obvious exmaple to me is the use of GCC in college classes. Hundreds of colleges use it that otherwise might not be able to teach courses behind computers. (note: you don't need a computer to teach C or C++ or to learn it, plenty of people have done it that way, I think it's a bit more enjoyable with a computer though) Compilers on multiuser UNIX systems are traditionally very expensive, as are site licenses to compilers under Windows.
I also think that there is a behind the scenes factor that has always been very hard to measure with linux. I know that my old school district, Boulder Valley Public Schools, has several Linux machines in various capacities. A couple are used as lan servers in some schools, a couple are used as firewalls and proxys and email and web servers. I'm certain that some act as bridges and routers. That's stuff that makes their life easier, serves a purpose and it's really hard to measure. Off the shelf firewalls can cost thousands of dollars. I have no idea how much it costs to buy the hardware, software and then hire someone to build you an exchange server for email or setup an email server with something non-linux.
As for teaching software and that kind of thing, I think it's still in the infancy.
If you're doing fairly low sample rate stuff, you might be able to use a soundcard, parallel port, or custom PCI card.
I've done a fair amount of debugging with logic analyzers and digital osciliscopes (software stuff, I'm not a hardware jock) and you couldn't debug an IDE interface or PCI interface with something on a parallel port or sound card, at least not easily. If you're doing simple analog stuff you could get away with it. Realistically you're going to want a real scope at some point in time though.
Hard to say though. I've seen them at hamfests and stuff like that for relatively cheap and any decent hardware shop is going to have them for their people to use or "check out."
It's pretty cool. Not the longest game in the world but the graphics are amazingly fluid and it's entertaining. It's not your run of the mill sports game or shoot 'em up, you've got to do a lot of exploring to beat it. I highly recommend it if you havne't played it yet. Check out Ico.
Also that scooter has two wheels, side-by-side, I've been wondering why it hasn't been posted on the front page yet.
It depends. In my place of work there are a lot of us doing Linux development running Linux on our desktops. We can get our email off of exchange servers with IMAP but we're locked out of the calendar system. As a result more than half of the team have windows machines on their desktops just for that reason. You're looking at a PC, a copy of windows, a copy of outlook, in our case you get the standard install with Office and a few dozen other apps on it.
In our case, if half of the people can get away with only Linux then when you start looking at the licenses and the fees it starts looking like maybe we can drop the site license and switch to something at about half the cost.
It would be better if Evolution did more but what it does is probably enough to be interesting in a lot of places
Unless you're manually adding and removing them what does it matter?
I'd like to keep things as clean as possible but that's kind of the nature of UNIX. Just the fileutils and textutils dump around 100+ files there and you're not doing anything interesting yet./usr has become "stuff that comes on the CD" with lib and bin being full of thousands and thousands of files./usr/local/ and/opt are where I put my crap. Since KDE, GNOME, etc.. all come with most Linuxi now then shouldn't they go in/usr/bin? Or should we add/usr/gnome/usr/kde and 600 other paths for every project out there?
I agree. I hope he described his system poorly and really meant something else.
The fact is, lot's of large projects are multi-language based right now. Not always in ways you might expect but for example, a large project I worked on for a very large and successful computer company had a large C++ code base. It has a set of no less than 35 support "scripts" written in Korn shell, it has an elaborate build process written in "make" which has portions that break out into Korn or Bourne shell to read environment variables and do various things. If you want to be pedantic you could easily point to pure C in the code base as well... And that doesn't include the java based client that was written later. Honestly I don't know of an easier or better way to run it, not for something like that. It was a very large "system" and not really an "app" or a "program" there were dozens of programs in the system that were sold as a package.
With web based apps I don't think it's terribly uncommon to have C or C++ doing some kind or service oriented task and then have something like perl or php interface to it and format some HTML. I could easily see a C++ server with some support scripts in Bourne shell (perhaps entries in/etc to start it up) feeding XML type data to a PHP or Perl script running on apache or something like that. As long as things are kept sane I could see it being a very easy way to develop an app and probably a sane one. I don't see someone writing C or C++ code to do what PHP does any better than PHP does it or producing the product very quickly. As a consultant, most of my clients would gladly pay for a few lines of php if it means I get done in 2 months instead of 4 months.
Slashdot Mirror
Some people use source code control as a way to backup their system. They don't mind checking code in every hour, cutting labels, etc.. Some people see it as a step to doing the job (and often a painful one) and they will dump code in once a week or every two weeks and introduce radical changes (usually at times associated with scheduled code drops so they can get their stuff in.) And there are other ideas or theories still. Linux doens't really have management and so you have to deal with all of that stuff while trying to still move forward. I think you probably end up at the same place, a Linus tree that he approves, you just spend a painful year or two trying to do something different. The fundamental prereq for CVS or any source control is trust, trust that the people putting the code in are going to the same place with their code that you are. I also think that the last thing you'd ever want is for someone other than Linus to settle a dispute by virtue of having commit rights... that has happened when Alan was the patch hoover or at least it was accused that he wasn't passing some patches on the Linus for political reasons.
Second, I think there might be a lot of public misconception about Linux branching. I built an embedded Linux kernel and maintained it for a consumer device. I was a small scale Linus for close to 2 years. If you are using Linux in a production environment you're probably forked. (maybe in the head?
As 2.5 simmers down some I think there will be a much stronger desire to not make the same mistakes, with 2.4 there were interfaces that changed during the 2.4 life span, a patch against 2.4.3 might not work against 2.4.8 without some modification. That makes Linus' job that much more difficult. Part of the reason is that maintainers who are supposed to be trusted came out with code that had some problems and wasn't trusted by other maintainers, so much for enough trust to use CVS.
Lastly (I'll close this ramble off) Linus and Linux have one strength that I really admire a lot having worked in the "mission critical" world of IBM and the banking industry. If something can be better they aren't afraid to make it better, even if it means changing code that isn't "broken." Right now in particular can be a radical time in the kernel, they're changing the block I/O layer! There are several VM layers in circulation. Fundamental parts are changing for the better, it's really hard to do that some times, you don't want to start crashing kernels or lose stability.
It's far better to do it than it is to try and hack around something that has outlived its design but "works." If Linus ignores all the new device drivers for 5 months while the VM, scheduler and block I/O stuff is sorted out, that's not a terrible thing.
The thing I would be thinking about would be low power, fanless, micro-atx or pc104 type form factor. So we could put Linux no them about build them in to little network routing and web serving bricks.
I own a couple of pieces of alpha hardware and it's fun in a geeky way to have. It's nice to test code on other platforms, it's nice to be able to learn assembly to other platforms and have something to work on, and a 21264 makes a hell of a web server. I'd love to have a newer PowerPC machine to work on but the prices just aren't there. If I could buy a motherboard and processor for $400-$500 maybe even $600 then I could easily see a little clique of people doing it. I can see real market value to it as well, I've seen 6 or 7 embedded jobs over the last month that were for PowerPC products.
I hope that they are interested in lowering the prices and ramping up some mass production of the hardware. I could also see a huge market for lower priced integrated PowerPC motherboards with G3s or even 60x processors on them; put 3 NICs, IDE and a PowerPC on a motherboard for $200 and you have a nice DIY home gateway/firewall/router box.
I think the big thing is the shift in thinking that some big companies are starting to take on. A company cannot support a product forever, for the most part. It also cannot afford to not support products. At some point in time, and it's probably a lot earlier than it has been with the Borland products they've made free (beer) so far, it makes no sense for them not to make things free. The next logical step is for them to start releasing code to these older products. Something like desqview is special in that if people are using it and relying upon it, there is probably not anything that can really replace it. It's good to see, they aren't losing anything by giving it away now and the next step is the source code. It's part of the promise of opensource, your solutions will never die becuase you'll have to code if you need it; well it's about time that the software industry stepped up to that plate also because the competition is good and I've got hardware sitting around that might be interesting with some of that old software. Also, who's to say that something like WordStar might not have continued had it been opened up, it certainly had the community of users it was a support problem that did it in. Some of that older stuff could get migrated to more modern platforms. It does nobody any good to have something like that die.
I was thinking about this and here was my dirty hacky idea. You need kexec, lobos, or something similar (actually a fairly modified version of it) you'll need on the order of 8MB of disk space and some kernel mods, which might not be that extensive.
I was thinking we develop some driver or process that consumes all of the memory and CPU in a system. It forces all of the processes to swap out, it would probably need to be a driver of sorts on current linux systems. Then it could dump the kcore out to a file somewhere, sync it, and hibernate. Then when the kernel boots up, if the right arg is passed in it could either load this image back in to ram in place of the kernel and then jump into it (easier said than done) early in the boot (page tables are made long before you have access to the drives and such so the logistics of this would need to be figured out) or it could boot up and use a different swapper partition and then have some kind of tool like kexec to load that image back in to ram and start it up. Or something, some how you should be able to recover the state of the system. File handles and everything would be there.
The harder part would be hardware and network transparency. You'd need to modify all of your drivers to make sure that the hardware could be reset and they could deal with it. I think it's a little easier for the network side because it would be similar to simply unplugging the network cable, you have open sockets that are talking to nothing and some software can deal with that pretty well. There is also some kind of system integrity or robustness piece that is needed, if the system some how changes when you bring your old image back it could break things, munge files, etc..
What are they going to do about all those cool $200 hub/switch/firewall/router boxes Linksys and company are making?
I'm thinking long termish. First of all, look at the MS case. They are looking at documents from the DR-DOS/Win 3.0 days. These are 10 year old documents and they have an impact on whether justice will be served or not. That very concept is defined by those documents; personally I don't think Gates himself ordered destructive actions towards competitors (so he shouldn't go to jail) but it's clear that there were some and mgmt knew of it.
Further look at the FOIA. There are 50 year old documents that tell us new things about history that are released all the time. I'd hate to think that a major crime could happen and be covered up so easily that in 50 years nobody could figure it out or see the truth.
For the most part, most people don't have anything to hide and shouldn't have to worry. I really have a problem with information destruction by policy, it's too similar to the nazi's burning books. In the digital age there isn't any reason to do it, space isn't a problem.
I've been wondering that myself. It's just work to do but it shouldn't be rocket science. Of course a number of people are concerned about the code size as is, you can't just add a new branch of code every time there is a conflict and pack them all together. This does seem like a somewhat good case for it.
BTW, this kind of bashing between the high priests of Linux is not good. You can bet your bottom dollar that MS is going to use this conflict to fuel their propaganda machine, saying Linux is a fractious OS run by a bunch of young upstarts who can't agree on anything.
No this is a good thing. Most of the linux kernel hackers have egos the size of small countries, and that's a good thing because they take pride in their work. Most of them also work as professionally and egolessly as I have ever seen. They can get in fights and then deal with it and keep working. This is far better than the closed world where people get in fights and take it personally and then try to react in some way. I've seen project where people were trying to fail the project to get revenge on someone on the team or managment for something stupid they did in the past. In the linux world people get in fights and everyone can see and they react accordingly, sometimes being told that you're being an ass is a good thing when you're being an ass, sometimes having people stop talking to you for a few days is a good thing, and sometime people appologize and that's a good thing. It's not behind closed doors though and it's hard to undermine things when it's all in the open. There is also some insanely good discussion on certain things some times. There is also something to be said for defending ideas and the strength they have when you can defend them in public.
Now companies like IBM, Sperry, EDS and the former DEC and even to some degree HP aren't lazy or slouchy groups. Those companies and the few others like them have built the infrastructure that the western world runs on. So when MS tries to sell a SQL server solution for pennies on the dollar compared to the similar Oracle or DB2 setup you have to wonder, are they that much better and can do it that much cheaper or are corners getting cut? Or does it not do everything as well? I'm not tryng to raise up one company over another and I think sqlserver is a hell of a product but I really have to wonder if everything it's built upon is as solid as a zServer running DB2 simply because of the economics involved. I think a company like IBM would find a way to do it cheaper if it could.
Now MS owns the desktop, they aren't fighting for it nearly as much as they were before. They are still doing radical development, NT4 to 2000 was a big changed. 2000 to XP was a big one as well. They still want to sell things cheaply and they want to get to that gold ring. Something has to give and it's the stuff that nobody outside the enterprise cared about for the first half of MS's existence. 5+ 9s reliability, which you almost can't do on Intel hardware. Rock solid security, it's hard to add that on to something already built, especially hard if you have a complex security model. The kind of reliability and security performance that once compromised by stories like that can take months and years to recover from in the minds of some people.
I think it must drive them nuts. I think what also must drive them nuts is simply the fact that this stuff happens. They aren't stupid and I know that they can understand a market and I'm pretty sure they know what they need to do to win the enterprise over, I just don't think they can do it and I wouldn't be surprised if they were starting to think that when this kind of thing happens so regularly. I just don't see a company beating a world full of IBM, Oracle, SAP, EDS, and others at their game without doing something different. Something different isn't cutting corners and costs, it has to be radically different like Linux.
I can't wait!
I think that it's far more likely that there could or would be an accident or "issue" at the dozens of sites around the country than at one storage facility.
I think that are some other issues of praciticality too. Yucca Mountain isn't exactly Malibu or Park City or Tampa. For the forseeable future people aren't going to be aching the build on it. It's a desert, a rough desert. They aren't tarnishing prestine wilderness or some ultradesirable place to visit or live. The waste needs to be stored and that's as good a place as there is. Throwing out a few of the more radical climate change theories, YM is going to substantially change for thousands and thousands of years and it hasn't for many thousands of years. I also kind of dismiss the "forget about it idea" it is marked and unless there is a catastrophic change to the human race it will be remembered for several thousand years.
To drive that to the logical conclusion, they try to convict you neighbor and they want all of you correspondance with him, do the saber ratteling routine and make you give up the keys. Then as they go through you computer they don't find evidence of wrong doing so much as they find evidence that you have unpopular views (perhaps you're a racist or a communist or something, not illegal to be but unpopular) and that get's entered in to the public record as well. You're an outcast and your public reputation is ruined and you have no recourse at all.
Now I'll give them the benefit of the doubt and assume that their PR firm screwed up and they're really talking about very specific types of data but this is most likely a scam. The counting argument is proven mathematically and you can't unprove that or circumvent it.
Let me explain. With a string of bits you can describe different types of data. Different combinations of bits can be used to describe different things, different strings of bits can be used in a multitude of ways but there is a limit to how much you can describe with a string of bits becuase it is fixed in length.
Say I take 1 bit. It's either 0 or 1. There are only two things I can describe with that one bit without extra information. (In compression, "extra information" implies bullshit, your idea doesn't work if it needs "extra information") It's impossible to make that one bit represent 3, 4 or more things without extra information.
With 2 bits I can represent 4 things. It's impossible to represent more. They are 00, 01, 10, and 11, there isn't a 5th possibility. From a compression standpoint that means that if I compress something down to 2 bits then with that particular compressor there are only 4 possible input that can compress to that size. One of those inputs might be the encyclopdedia britannica but that's a pretty specializae compressor then. Do the induction, 2**bits is the equation that defines the maximum number of possible inputs to produce that output of that length. This is encoding and what Shannon is most famous for, and his laws are still laws. There are a couple popular ways to do encoding with binary data in compression, arithmetic and Huffman are the most popular.
So where is this going? Well first, recursive compression. It's not compression becuase of that "extra information" problem. How many times do I run it to get my data back? Well that's simple, you compressed the linux kernel and you ran it a number of times x. x just happens to be about the same size as the linux kernel when represented in binary, gzip that number up and keep it in a safe place and you'll be able to restore your kernel that was compressed into 1 bit.
Second. Compressing random data. Say I compress a random string to 1/100th of it's size. To make the math easy, let's use a string that is 1024bits long, it is reduced to 10 bits (or 11 now and then.) 1024 bits can represent a lot of different things. 2**1024 of them. 10 bits can only represent 1024 different things, 2**10 = 1024. 2**1024 has 309 digits, 1024 is an teeny tiny fraction of it. Well under 1%, way way way under 1%. With that compressor that does 100:1 compression you can only compress, at the most, 1024 things out of 2**1024, without "extra information." That might be randomesque data but the fact is that if you pick a random string of 1024bit, you will very very rarely pick one of those 1024 strings that your compressor works on, it pretty much will never happen in practice because it's such a small percentage. This is a truth with all compressors. That's not to say that they can't compress a truley random string of bits 100:1, they just might want to take a few centuries to come up with the particular random string that they are going to use.
LZ77, LZ78, Burrows-Wheeler transforms, PPM modelers, Markov modelers, all modern generic lossless compressors exploit the fact that most "interesting data" that we want to compress has redundancy to it. If you allow for some degree of redundancy or perhaps a lot of it then you reduce the number of possible inputs to the "interesting inputs" and that's usually far smaller than 2**1024 for a 1024 bit string. It's still probably orders and orders of magnitude larger than 1024 though, it's quite easy to find more than 1024 interesting things you can represent with 1024 bits, the short of it is that a 100:1 compressor will only be able to compress 1024 things of that length. No matter how you model the data, you're limited by the possible number of inputs represented by the encoding.
Wine has been getting better and better and with the misc. binformat support I have it loading and running java and windows programs seemslessly, it's only a matter of time before someone tries to tie it all together.
If you had a few more degrees of friend or foe then and karma was applied accordingly then it starts to look more like the advogato trust metric which works pretty well.
At it's most primitive layer, OOP is a coding thing, a syntactic sugar as so many professors like to say. A way to keep code clean. It's quite nice in that capacity, I don't mind writing "c = a + b;" and under the covers mpz_add is getting called and adding the arbitrary precision number a to the arbitrary precision number b and storing the out put in c. This isn't true OOP, I only include it becuase it's the OOPLs that have made it popularly available.
The next level is encapsulation and reuse. It's pretty damn nice to be able to write "d = {}" in python and create a hashtable dictionary. I do that is literally dozens of python programs. I could easily produce the same thing with list primitives or vectors and what have you and I might even be able to make it better for my particular use but for the most part the dictionary object in python or any of the STL objects are insanely useful, they are easy to use, quick and they are building blocks you can use to build bigger programs. You're already using this kind of code whether you know it or now, the C library is the same thing, it doesn't have the same syntactic sugar though and the things it does are generally more simple where some of these objects are higher level components. I wouldn't call the C library object oriented, per se, but might call a component that does the job of half a dozen C library calls an object.
The final layer of OOP as I see it is modeling. Again it's not well defined outside of a few circles and even then Grady Booch and Bertrand Meyer and the other gurus spend countless hours and books bickering about it. The modeling part is how we break a problem down and implement a solution for it. This I would say is a universal area where OOP has won and will continue to. When programs reach certain sizes, the human mind can really only deal with the complexity by building heirarchies and mapping the problems to solvable isomorphs and then trying to reuse ideas as you move through the heirarchies. If you're trying to multiply a few matrices and then hit them with the Gauss stick to tell if the bridge is going to work, it's probably not a good example of a problem to model that way. A GUI on the other hand is a perfect example where you can have a very clean hierarchy of windows of different types with different functions that behave differently. In my experience, I haven't seen any large scale projects started in the last 15 years that weren't object modeled, even if they were implemented in C or some other non-OOPL, I have seen a fair amount of code from the 1960s and 1970s that was very very complex because it didn't have any real object modeling. The linux kernel has a fair amount of object modeling in it, there are abstract block devices and character devices and they can all be treated the same way, buffering is an abstract concept regardless of the device. Even functionally implemented programs have some degree of object modeling in them.
Now what kind of OOP are you looking for? If you're looking for some kind of advanced object model with reflection and a complex object hierarchy, then maybe it's not being used to solve some engineering problems, it takes a certain size problem before that kind of effort really pays off in the implementation. If you're looking for code reuse then there are numerics packages and probably tons of other packages the will do that.
Has a bank's data security been compromised lately?
That's how I'd temper the worthiness of something like this.
They are also an integral part of IBM's business as a solution provider.
Back in the early 90's and late 80's PS/2 cost, literally 2 to 3 times more than a clone of the same vintage. They had MCA, to boot. There are also a fair amount of those machines still in service. It's not terribly uncommon in certain enterprises to look in the back room and see an old PS/2 486 class machine serving up files for a work group or driving a printer. While certainly not a speed demon, it does the job. With MCA bus mastering in some situations it still makes an accpetable server for certain tasks. Part of the reason is becuase the machines are about as reliable as PCs get. IBM at that time was holding their hardware to much higher relibility standards than just about anyone else and they were more expensive because of it. At that time I thought they were doomed and would be out of the business. If you're selling solutions to people and those solutions will last decades and PCs are an integral part of the user interface you need good PCs.
The same can be said of RS/6000. Outside of a few markets they are under performers. I've also seen RS/6000 machines fall off of loading docs and get run over by a forklift and have cases that are all dented to hell and still run flawlessly. Is it the fastest workstation? No. Does it do the job IBM needs to sell a $10million contract? Yep. and because of that they keep making them.
I keep thinking that they need to get out of it or just focus on netfinity and thinkpad but the truth is, when they are selling an industrial solution, who are they going to buy PCs from? Compaq? Hell no, they are a big competitor. Dell? Possibly but I just don't see optiplexes running for 10 years like an IBM PC. Gateway? You've got to be kidding. If you think of a PC as the world's most expensive peripheral and you need those peripherals to run mainframes, minis and super computers then IBM will keep selling them.
There are people I've lost touch with that I only ever knew through the BBSes. At times I wonder where they are.
Realistically though, free software has made a huge impact. I think the most obvious exmaple to me is the use of GCC in college classes. Hundreds of colleges use it that otherwise might not be able to teach courses behind computers. (note: you don't need a computer to teach C or C++ or to learn it, plenty of people have done it that way, I think it's a bit more enjoyable with a computer though) Compilers on multiuser UNIX systems are traditionally very expensive, as are site licenses to compilers under Windows.
I also think that there is a behind the scenes factor that has always been very hard to measure with linux. I know that my old school district, Boulder Valley Public Schools, has several Linux machines in various capacities. A couple are used as lan servers in some schools, a couple are used as firewalls and proxys and email and web servers. I'm certain that some act as bridges and routers. That's stuff that makes their life easier, serves a purpose and it's really hard to measure. Off the shelf firewalls can cost thousands of dollars. I have no idea how much it costs to buy the hardware, software and then hire someone to build you an exchange server for email or setup an email server with something non-linux.
As for teaching software and that kind of thing, I think it's still in the infancy.
I've done a fair amount of debugging with logic analyzers and digital osciliscopes (software stuff, I'm not a hardware jock) and you couldn't debug an IDE interface or PCI interface with something on a parallel port or sound card, at least not easily. If you're doing simple analog stuff you could get away with it. Realistically you're going to want a real scope at some point in time though.
Hard to say though. I've seen them at hamfests and stuff like that for relatively cheap and any decent hardware shop is going to have them for their people to use or "check out."
Also that scooter has two wheels, side-by-side, I've been wondering why it hasn't been posted on the front page yet.
In our case, if half of the people can get away with only Linux then when you start looking at the licenses and the fees it starts looking like maybe we can drop the site license and switch to something at about half the cost.
It would be better if Evolution did more but what it does is probably enough to be interesting in a lot of places
I'd like to keep things as clean as possible but that's kind of the nature of UNIX. Just the fileutils and textutils dump around 100+ files there and you're not doing anything interesting yet.
The fact is, lot's of large projects are multi-language based right now. Not always in ways you might expect but for example, a large project I worked on for a very large and successful computer company had a large C++ code base. It has a set of no less than 35 support "scripts" written in Korn shell, it has an elaborate build process written in "make" which has portions that break out into Korn or Bourne shell to read environment variables and do various things. If you want to be pedantic you could easily point to pure C in the code base as well... And that doesn't include the java based client that was written later. Honestly I don't know of an easier or better way to run it, not for something like that. It was a very large "system" and not really an "app" or a "program" there were dozens of programs in the system that were sold as a package.
With web based apps I don't think it's terribly uncommon to have C or C++ doing some kind or service oriented task and then have something like perl or php interface to it and format some HTML. I could easily see a C++ server with some support scripts in Bourne shell (perhaps entries in