All this legalise is too complicated for me to tell properly.
Short (and how I understand situation): you can read a book w/o anything. You do not need a license for using your own brain and eyes - the only things required to read a book. Nor you need a license to print a book.
Compare to DVD, which is unusable w/o licensed player, licensed recorder, etc.
Imagine a world (as it was actually several centuries ago), where you would have to have a license to have a printer at home.
It's like we are back in medieval cetruries with feudalism, with CEOs playing role of feudals, (ab)using USA corrupted legal & political systems to fight for their piece of turf of our money.
Reminds me that some [CENSORED] from RIAA wanted recently share of profits from Apple iPod sales.
Also, they wanted medical offices to pay for music they play for customers, citing the same reasons: more people can wait for doctor, so music does increase doctor's profits.
Very very silly. TheRegister once suggested that this guys would argue one day to perform lobotomy on people leaving concerts/movie theaters: thus people will not take away RIAA/MPAA's precious Intellectual Property with them.
Seems book sellers & writers guilds are going in the same direction. They even might argue that if I read book and learned a thing from it, I owe them money, since I could profit from the learned thing. After all I paid them for reading the book - not for profiting from it.
Let's hope, that books will continue to be sold - unlike DVDs which already are licensed, not sold.
P.S. To your remark that you bought more books after finding them with Google. I was always confronted by this dilema. You might by yourself notice that writers whose books we are buying, normally has nothing to do with writers who would argue on Google Print thing. What's more, the authors I like never ever raised the question about scanned books appearing on the net. I guess there are authors who like to reach more people - and authors who want to profit. No need to guess whose books are better: like with software, best stuff is made by people who want to make the stuff - not by people paid to do so.
I'm sorry, but none of the Sun hardware I have worked with - couple of WSs and couple of carrier grade servers - had serial consoles.
We used networked message logging and it worked perfectly Okay. I cannot give you guarantie that it was Sun's native solution but I'm pretty sure it was. To put it simply, serial console cannot be made redundant, so in my application fields it is rarely used for anything. (Thou everything is equipped with serial (rs232, hdlc) ports.)
P.S. Even if you need to keep terminal always attached, you can implement your own terminal application and add filters to remove non-error messages. + buy a additional blade/rack server (cheapest fitting one) and equip it with multiport serial card. I actually seen place where logging info from bunch of server racks was muxed into some wider serial connection (FR or hdlc), and then relayed to special server where it was demuxed, cleaned up and stored onto raid. People this way were keeping track of all messages for one day, so if critical error occured they have a chance to find anomalies preceding the error and thus localize the problem.
PC industry is filled with crippled hardware. Only if people knew the quality of what they buy...
It is quite showing that OF is adopted only by Sun & Apple: both well know for quality harware with good record. Companies not afraid to respond to customers claims. And frequently responding.
Additionally, from rumors I have heard, EFI is designed madularly, and modules basicly can be any kind of binary blobs. So EFI will improve nothing on side of BIOS - it will remain closed source, and will quite quickly end-up being fix for every hole system might potentially have. Just like BIOS.
After all, all those magic stuff they promise, can be as easily implemented on top of modern BIOS. There is no technical problems as it is. *NO* *TECHNICAL* *PROBLEMS*. Many companies have source code for BIOS. Just sit down and do it. Just like IBM did with Slimline Open Firmware - stripped down version of OF, sufficient to boot Linux on JS20 blades. http://www-128.ibm.com/developerworks/power/librar y/pa-dw-slof.html
P.S. From what I have heard about BIOS developement, it is low amount of flash which is problem: normally installed on PCs is only 64k. That's the problem - not BIOS itself. EFI I'm sure will mandate something like 1MB minimum of flash - after all the prices are now so much lower. But still as it stands: there is no un-solvable problem with BIOS.
The worst problem they have is their fab technology is a couple generations out of date. They are actively seeking suckers... err... fab equipment makers who want to partner with them while they steal.... err.... license their technology.
Semiconductor factory is not something what can be easily stolen. Modern factory costs about $2bn, and not much companies can build them - IBM, Intel, TI, STMC and alike.
IIRC the stories about Godson-I, China was implicitly looking into ways of improving performance cheaply. They do not need PCs, they do not need embedded - this markets are saturated and has no regulations. What they need is home made supercomputers - it is the market which still has reguilations. And Godson is one of the components of the plan.
Don't make mistakes: Godsons are SMP capable. Add here production capacity of China and its ability to drive prices down - and in next few years you might expect China's supercomputer loaded with zillions of Godsons.
As an example of such "disruptive innovations" check the top of the Top500 - those IBM's BlueGenes loaded with embedded PowerPC 440.
China's super computer won't be sophiisticated, but it will have low proice and raw power. That's for sure.
Sorry for breaking/. traditions. I have worked for three hardware companies. I'm working for one right now.
I'm not writing documentation, but seldom help with some technical details and help with translations.
Guys, just to understand the problem you have try first write something. Writing good documentation would require someone who can think straight and clearly. And as you can imaging if company got such guy/girl - s/he will be busy doing product itself, rather than documentation. That's first.
Second. On PC market change of technologies occurs every half of year or something like that. Writing documentation, getting thru all bureaucracy for confidential document to be released will be precisely about this half of year. In other words, when documentation is ready - it is already outdated. You can btw notice that industries with longer release cycles normally end-up with decent documentation - they have time and money. Having no money you cannot afford longer release cycle. Telcos are good example of that. Industrial automation is another one.
So PC industry is sort of its own worst enemy. Tight competition force producers to save where possible - e.g. optional component such as documentation. If you have some bright heads, producer can manage to update product line more often - edging against competitors - but again bright heads will be busy with product release, rather than its documentation.
Having dedicated personnel to handle documentation is just expensive.
P.S. One more problem is when product is partially based on some licensed third party development. Most small producers have to license things, since they cannot develop everything in house. Than it happens that documentation ends up with hell a lot of copy'n'pasting from third party one. To release that you have to get a permission from third party: but in most situations you will find out that third party has licensed some parts itself. In the end no-one just want to risk releasing the documentation, especially in litigious U.S.: most companies when finding their parts in someone's else released products may start asking for fees. With all consequences. Hiring experienced attorney for going thru all this licensing mess - and getting clearance from all involved parties - will cost a lot money, most producers just not able to afford.
Memory fragmentation is well known. That's true that there is no solution. But on the other side, there are technics one can employ to resist memory fragmentation.
Check memory pools. Network equipment uses them (you expect router/gateway to run 24/7). Check Linux kernel for slab allocator. It is not ideal - but still works quite well.
The problem one cannot solve is allocation of huge memory peice of continous memory. But on the other side, this is problem more specific to kernel. User space applications are much less constrained, compared to kernel space tasks. It is just question of willingness of developers to solve the problem. It is just all things will have to be splitted into manageable pieces. Unfortunately, in some cases it is just impossible (e.g. second party libraries).
Web browser is quite complicated peice of software, with support of images, video, audio and various texts. Such changes might prove to be complicated to implement and maintain. With complications come bugs. I honestly rather relaunch Ff, rather have it crash by itself.
But then detecting things like memory leaks is quite complicated. You cannot be sure that item has to be decallocated and will not be deallocated some time later. Memory management is lazy. And you never can be sure about it completly. Saying that you have found memory leak, by just watching memory consumption going up is of no help to developers. And is not proof of application bug.
Honestly, I am, as a programmer, bemused by industrial automation people all the time I have to work with them.
Money squeezing is norm of life: complaining aloud about say analyzer with encrypted data format so now they cannot use their expensive software from 3rd party. On first turn. And then, on second turn, during meeting they'll be first to propose change in file format just to ensure that customers will not be able to use cheap tools to read them.
I'm seeing that shit for 2.5 years now. Norm of life over here.
JPI is by no mean not exception - I just hoped that aviation equipment has some regulations applied to prohibit the boneheads from screwing anything they don't like.
Thou my new company - Thanks God - seems to not have such attitude to customers. On my first industrial automation job, amount of ideas scrapped for simple reason that it will give too much breathing room to customers was near 100%. Some of the product managers did humorously advertising principle: do it worse, customers will have to pay us twice. And this finds sympathy in some people's minds over here. What is very sad.
Can they please release list of planes equipped with this stuff? - So I will have a chance to take another plane.
Hope they will not opt to DRMfy they data. Just try to imaging message on plane: "Dear guests our plane is crashing down due to expired license for motor control protocol encryption. Please festen your belts and pray."
Well, IIRC, this [CENSORED] was already tried in medicine and failed. No system is allowed to use encryption for sake of encryption when human life is at stake. Both medical life support equipment and aviation equipment generally classified as mission critical for that very reason - human lives depend on them.
I have no experience nor information about mission critical systems, but I have just started working for safe equipment producer - and already have clue as what "mission critical" means. Safe equipment is right one step below mission critical equipment: safe equipment has to shutdown system in case of alarm, while mission critical stuff has to work as long as it possibly could - whatever situation/alarms/malfunctions are - someone's arse depends on it.
Development of such equipment is kept as transparent as possible - any mistake might cost way too much. Adopting encryption doesn't seem to me as step in right direction.
For me it is important that Google has competition.
Google in the beggining was not that shiny as it is now. And not much people knew about Google, when it only started. MSN/etc have to challenge the king of search - and "it is always hard first hundred years" (c) russian proverb.
Google is doing terribly weel to date. But in long term Google needs competitor.
A9 & MSN IMNSHO are most welcome to challenge Google.
I can only hope, that another day, Internet will not die under load produced by bunch of search engines competing in indexing web;-)
Apple is about new cool things. It cannot compete with Dell. (*Now* we have complete antonym of "Apple" - "Dell", now it is easier to compare).
If you want to make widely used computer systems, you have to bow to many customers & companies and their needs.
Remember this: Apple bows no-one. It is on innovative edge. It costs much. And they do mistake too. But it pays off too.
And you cannot be innovaitve and large at the same time. Widely used technologies take long time to create market - by time technology has chance for wide-spread huge market, Apple is already on something new.
P.S. iPod being recent exception.
P.P.S. Think of Apple as of engineering company. They generally - and Jobs in particular - do things people use every day. And Apple engineers are first to use this technologies - they their own dog food. They don't do recket science - they do things person can/want to have at home. And they know it not from analyst's papers, but from real life - they do things for themselves. Nothing more. Bit like history of Porsche - they started like little company to do *good* car for themselves and their friends. They are small, vvvery pricy, but they are still around because they are still very innovative. Thou Apple likes more to parallel itself with BMW;-)
No, there will only be one company 'doing business'.
I hope it will not become soap opeara like with Digital heritage.
Digital (later DEC) was bought by Compaq to... seems to me to kill competitor. But stupid customers still used DEC stuff, since it worked and worked well, compared to Compaq stuff.
Then Compaq was bought by HP to... mating of dinos? Most likely. But still this bone-head DEC customers insisted on using software and hardware which just works, and doesn't enrich market with new services & jobs, like PC Repairman.
So after all this years, HP finally killed off DEC legacy stuff, since it is of no use to them - it just works.
Parallels with current situation are obvious. Oracle tries to sell something what sometimes works with market cost of $10 for $100 so no-one wants to buy it. And instead of building better cheaper product it buys PeopleSoft.
I'd be laughing a lot if in couple of years if M$ or IBM will come up with bid to buy Oracle. It would be fun for sure. Especially for People Soft customers.
I got impression from people who develop TB, is their primary target is stability. What is very very right approach IMHO. I've being using The Bat! for quite long time on Windows. Then after switch to Linux, I started with mutt (and fetchmail). Then after HTML for e-mails became popular enough I went to Netscape Messenger. I'm just tired of switching e-mail cleints. I bet others are too. I want one, stable one, I can use anywhere I come. Web-mail doesn't cut in - I cannot take it with me anywhere I like or need it. Mail is something important to have all the time with you.
People used to switch e-mail readers - so they are less dependant on it. And fortunately, e-mail formats are not "forked" - like M$ did with HTML/DOM/CSS for IE - they are still basicly plain text.
Browsing experience was crippled much more by IE, than e-mail experience by Outlook. For some pople Outlook (not OE) is the only choice, even when they have no Exchage Server. Outlook is true point of M$ Desktop & Office integration, many times unfortunately overlooked and later poorly copied by Evolution. And as this integration point Outlook enjoys popularity among virus writers: it supports everything IE supports, it provides e-mails servies, it manages address book, it manages notifications, etc. This integration is very important for some people. But for others like me it is not.
Amongst many clients I have used before only mutt and Netscape Messenger (now Mozilla Thunderbird) are most stable for me. But it is probably only me. I switched back then from mutt after couple of months getting some html messages with pictures - I have couple of friends who love this feature and use it way too much.
Another day I have heard screams in computer room. I went there just to find my friends literally laughing to death. They were trying thru laughing point to the screen of WinXP with error message.
As soon I have taken a look at screen - I have joined them laughing to death under table.
"Invalid MS-DOS function"
For sure, we had over-reacted, due to couple of
M$ Zealot who tried to persuade development department that WinXP is complete rewrite of Windows from scratch. And it has nothing to do with MS DOS.
As a person who switch to Linux & Apple long time ago I find bit fuzzing insistence of some companies on using technology from 80s. If you haven't noticed, all external hard-drives are shipped formated with FAT.
No-one yet came out and proposed read-write file system for hard-drives supportable by all OSs. File systems are not standard - I'll love to see OpenGroup/POSIX/ISO having standardize some file system in order for interoperability between OSs. Just like it was done for CD/DVD media.
P.S. Message in our case was showed when one guy tried to delete file with name 'nul' with Explorer. Who remember DOS times - it is reserved name which is presumably impossible to give to a file. Some tools do allow to create/delete files with such name under WinNT/friends.
Sorry - you are not right about MySQL. AFAIK MySQL maintains GPL & non-GPL versions separately.
If you will contribute code to GPL's MySQL - it will not appear in proprietary one. MySQL people might choose to re-implement it for proprietary MySQL - but GPLed code will be held GPLed. On opposite side, MySQL AB makes some parts of their proprietary code GPLed - they are trying to keep both versions as close to each other as possible.
XFree86. Again you not right. One submitting code keeps all copyrights and what's more with her own license. Debian people did analysis of XFree86 and found about 40 variations of BSD license. Sometimes very different licenses. XFree86 folks admitted to be absolutely lax on licensing/copyrighting. Code with is genuinely XFree86 - is rather modest part of whole XFree86 (think applications, fonts, extensions, drivers, scripts, configs, etc).
I cannot be sure about Mozilla and Apache Software Foundation (ASF). IIRC both of them are pretty strict on licensing and copyrighting. I'm absolutely sure that they very strict on licensing: if you have submitted code under one license - you will be asked if project will like to change the license. That was a precedent with Mozilla going from MPL/GPL/LGPL -> GPL/LGPL - they have e-mailed every contributor with request for her/his code being transfered under different license. Apache didn't have had problems like that - I believe they are more to engineering, than to licensing - so I expect them to require copyright transfer, so if license will turns out to have a problem - they can address problem without years of getting "yes" from every contributor.
Relicensing process with its recent example in Mozilla have older cases: BSD's one I have heard about. RMS (Stallman of FSF/GNU) successfully argued to BSD folks that 4 clause BSD license (with acknowledgment clause) will only hurt in future - and since then BSD adopted 3-clauses BSD license, which is compatible with GPL.
No. Havoc leading no-one. He is assigned to this job now. Long time ago he may be was a leader. But now he has a job, corporate business plan and roadmap. Most of what he is doing - it is for his employer - RedHat, it is not his ideas - it is something his management told him to do. He is not anymore self-motivated (besides having his job at RedHat).
If you like see Steven Jobs in action - check recordings of his conferences with shareholders. To date shareholders are buying everything he is saying - they are sort of his employer, but still it is Jobs who is saying what Apple needs to do. Not his employers - shareholders.
Conflicts, where I got that bleak picture of Havoc Pennington, where window minization animation and gnome control center. Dig up archives. Have Jobs spoken this way with users once - no will ever buy Apple computers afterwards.
Making up another standard will not help to get GNOME out of those stupor it is right now.
Making another standard will imho *ensure* that GNOME will not get out of this stupor: poeple will spend another N (where N>0) years arguing about what needs to be in HIG2. And I do not see that GNOME people are uncontent with current HIG. (*)
GNOME is definitely over-rugulated. You have this impression of "dealing with children," because those corporate types I'm talking about are just out of any communication channels. At all. They do not care. Period.
You better be watching commited patches - and who commits them. And then judge by content of patches - is this child or some other type.
You have said "children" - GNOME/Gtk went some funny direction, so with its APIs you actually can frighten little children. Unfortunately for me - and this is why I always liked GNOME/Gtk - I used to do absolutely the same kind of APIs: overcumbered with functionality, un-intuitive and extremely hard to use. Well, they are reliable, complete and error-free. But just try to use them...
(*) After some time, I have developed impression that no-one really likes nor uses GNOME (besides making sexy screenshots). Journalists are paid better for positive reviews. Users on most of the lists after having spent 2 weeks compiling new GNOME just cannot say "it sucks" - it would be too hard after two weeks strugling to compile. Developers - being myself developer - I cannot belive that a developer can use GNOME. There were a release of GNOME, where you were not able to configure keyboard, and default keyboard shortcuts obviously configured for some mysterious end-user, who have problems telling keyboard from any other board... So who are really using it anyway???
Choice is something that the experience user wants.
This is precisely why Linux on Desktop sucks.
Look at Apple as a good example. What is Apple? - It is just cover company for ideas of Jobs. Why is Mac consistent? Because there is Steven Jobs - and there is no choice. People at Apple do not waste their time arguing on mailing lists about better desktop. Jobs has vision - and he drives company according to this vision.
KDE? KDE is made of people who enjoy desktop. Probably they are not greatest GUI programmers - but they like what they do. They are enthusiasts of what they do. Not consistent, Not polished - but with load of features. Great utility from people who have enjoyed doing KDE.
GNOME? GNOME is made up from pollitically correct corporate sponsored full time developers (RedHat, Sun, FSF). They are more to politics and to deliver on corporate business plan, than to listen to their users. It was absolutely funny how Havoc Pennington (of RedHat brainwashing fame) was arguing on list against end-users that he has statistics from end-users at hand, and every-one on the list is wrong, because some has given him statistics and - well - he doesn't care, he has a road map ha has to adhere since RedHat is planning release of next RedHat, etc, etc, etc.
IOW, Democracy - like one found in GNOME - is no substitute to leadership (Apple, Enlightenment). Republican structure of KDE performs here better too.
Linux will not get good Desktop until someone will step forward as a leader. And I see more chances for this to happend in KDE, rather than in GNOME project. Only if someone really disgusted will decide to fork off GNOME, what is IMHO not worth doing.
P.S. Best Linux desktop to date is Enlightenment (E). It is shiny and brilliant. It is finished, polished and complete. Why? Because guy who did it - Rasterman - really cares. And he is driving his project forward. Not fast, but GNOME as was two years behind of E - it is now the same two years behind of E in usability & eye candy.
P.P.S. Forking off GNOME. Well it might be not the worst idea. After all even XFree86 was sucessfully forked, pushing development of both - X.Org & XFree86 - ahead on new wave of competition. If someone will fork GNOME, providing good desktop - it might attract some independent developers working on GNOME, potentially making viable alternative to old fans of good old GNOME.
I've forgotten - and actually never knew this things very good. They usually go out of scope of school chemistry courses.
I remember that my god-father was using some derivative of spirit to improve quality of his home heating system. Bit his system was open one.
Another important property - as you have absulutely correctly pointed out - is density. So it seems that we are left only with open systems, since CPUs usually run with relatively low temperatures. (Trick with lithium is simple - it runs with very-very high temperature - so its density is pretty close to constant.)
P.S. Car engines are using oil. It is run inside of closed system with lots of redundant volume, since oil tends to decrease density (and increase in volume) with higher temperatures. Worst thing with oils - they are all easy to vapourize. But probably oil might be used as a secondary conductor - after aluminium/copper, which sits on top of CPU - to increase effective surface of heat exchange with water (think lots of water pipes dipped into oil bath standing on top of CPU - easier to produce than aluminium radiator for water cooling).
Well, in closed system you can definitely use something better than water.
Water is not the best coolant ever. I beleive it is possible to find some other chemical agent which will have higher heat capacity and heat conductivity. Improving this two properties will improve cooling for sure.
Another option - obviously - make water flow faster. But this can introduce noise and make system more expensive (given you will try to keep the same noise level).
For even more expensive option one can use trick used in electric plants: use two cooling circuits. First one is connected to CPU. You can ensure that this circuit doesn't have any leaks and use any toxic material to get heat out of CPU *very* *very* fast. Second curcuit obviously is used to cool first curcuit. It can be open, it can use water. It can be very very cheap, compared to first one;-) (FYI nuclear plants are using lithium for first cooling curcuit. Water in second curcuit is used to cool lithium. Hot water then is used to heat houses, produce electricity, whatever)
P.S. I'm not an expert - I might be wrong somewhere here. But gosh - it's/.;-)
Windows is sort'a nice, I like it much more than Linux. Really.
But it is just little things.
I'm using VIM very much. I hate VIM, I hate Emacs, I love MSDevStudio. But. But under VIM+Bash I'm at least three times more productive.
Windows GUI is good and consistent. To some degree. I'm as a person who designed for two years GUI applications for Windows and knowing every input/output/message/control available I can say that Windows GUI is most advanced GUI ever created. But. But M$ itself stopped following itsown GUI desing guidelines, and I'm not taliking about dumb so-called "VB Programmers" and other commercial software developers who have problems doing simple window with two buttons right. This is really sickening.
Error handing in Windows is just awful. It has nowthing comparable to/var/log/messages. Once I have spent 3 month being not able to run one of the my development tools. It was really bad situation and no one have ever had any clue what have happened. I have used other machine with devkit installed, where my tool worked Ok. But after sometime it stopped working there too. After two weeks of games with regedit/etc it turns out that this application was Win16 application (Win32 has no required system call - but Win16 subsystem does) and when and size of evironment was giong over some limit Win16 subsystem was just stopping to work. With no error message whatsoever.
I can go on and on. For a long time. I've being long-time M$ user and developer. But once I (actually bit forcefully) switched to Linux - I was really amased: some things didn't worked, but most of other things just worked. Without reboots, without crashes, without asking tons silly questions. Just worked. Breath of fresh air after 6 years of WinNT 4.0.
P.S. w2k/xp really didn't changed this balance much since early year 2000 - the time I switched to Linux completely.
Protocol used to transfer data and commands between computer and device (e.g. harddrive) is just better. It is asynchronous - so OS do not need to use delays, it can be pipe-lined, it can be (and usually is) optimized on the fly by SCSI controller to.
Usually SCSI vs. IDE is noticable in CPU intensive operations: CPU useage of SCSI hard drives is less by about 20-25%.
Slashdot Mirror
All this legalise is too complicated for me to tell properly.
Short (and how I understand situation): you can read a book w/o anything. You do not need a license for using your own brain and eyes - the only things required to read a book. Nor you need a license to print a book.
Compare to DVD, which is unusable w/o licensed player, licensed recorder, etc.
Imagine a world (as it was actually several centuries ago), where you would have to have a license to have a printer at home.
It's like we are back in medieval cetruries with feudalism, with CEOs playing role of feudals, (ab)using USA corrupted legal & political systems to fight for their piece of turf of our money.
Reminds me that some [CENSORED] from RIAA wanted recently share of profits from Apple iPod sales.
Also, they wanted medical offices to pay for music they play for customers, citing the same reasons: more people can wait for doctor, so music does increase doctor's profits.
Very very silly. TheRegister once suggested that this guys would argue one day to perform lobotomy on people leaving concerts/movie theaters: thus people will not take away RIAA/MPAA's precious Intellectual Property with them.
Seems book sellers & writers guilds are going in the same direction. They even might argue that if I read book and learned a thing from it, I owe them money, since I could profit from the learned thing. After all I paid them for reading the book - not for profiting from it.
Let's hope, that books will continue to be sold - unlike DVDs which already are licensed, not sold.
P.S. To your remark that you bought more books after finding them with Google. I was always confronted by this dilema. You might by yourself notice that writers whose books we are buying, normally has nothing to do with writers who would argue on Google Print thing. What's more, the authors I like never ever raised the question about scanned books appearing on the net. I guess there are authors who like to reach more people - and authors who want to profit. No need to guess whose books are better: like with software, best stuff is made by people who want to make the stuff - not by people paid to do so.
We used networked message logging and it worked perfectly Okay. I cannot give you guarantie that it was Sun's native solution but I'm pretty sure it was. To put it simply, serial console cannot be made redundant, so in my application fields it is rarely used for anything. (Thou everything is equipped with serial (rs232, hdlc) ports.)
P.S. Even if you need to keep terminal always attached, you can implement your own terminal application and add filters to remove non-error messages. + buy a additional blade/rack server (cheapest fitting one) and equip it with multiport serial card. I actually seen place where logging info from bunch of server racks was muxed into some wider serial connection (FR or hdlc), and then relayed to special server where it was demuxed, cleaned up and stored onto raid. People this way were keeping track of all messages for one day, so if critical error occured they have a chance to find anomalies preceding the error and thus localize the problem.
IMNSHO, OpenFirmware is way too open.
r y/pa-dw-slof.html
PC industry is filled with crippled hardware. Only if people knew the quality of what they buy...
It is quite showing that OF is adopted only by Sun & Apple: both well know for quality harware with good record. Companies not afraid to respond to customers claims. And frequently responding.
Additionally, from rumors I have heard, EFI is designed madularly, and modules basicly can be any kind of binary blobs. So EFI will improve nothing on side of BIOS - it will remain closed source, and will quite quickly end-up being fix for every hole system might potentially have. Just like BIOS.
After all, all those magic stuff they promise, can be as easily implemented on top of modern BIOS. There is no technical problems as it is. *NO* *TECHNICAL* *PROBLEMS*. Many companies have source code for BIOS. Just sit down and do it. Just like IBM did with Slimline Open Firmware - stripped down version of OF, sufficient to boot Linux on JS20 blades.
http://www-128.ibm.com/developerworks/power/libra
P.S. From what I have heard about BIOS developement, it is low amount of flash which is problem: normally installed on PCs is only 64k. That's the problem - not BIOS itself. EFI I'm sure will mandate something like 1MB minimum of flash - after all the prices are now so much lower. But still as it stands: there is no un-solvable problem with BIOS.
Semiconductor factory is not something what can be easily stolen. Modern factory costs about $2bn, and not much companies can build them - IBM, Intel, TI, STMC and alike.
IIRC the stories about Godson-I, China was implicitly looking into ways of improving performance cheaply. They do not need PCs, they do not need embedded - this markets are saturated and has no regulations. What they need is home made supercomputers - it is the market which still has reguilations. And Godson is one of the components of the plan.
Don't make mistakes: Godsons are SMP capable. Add here production capacity of China and its ability to drive prices down - and in next few years you might expect China's supercomputer loaded with zillions of Godsons.
As an example of such "disruptive innovations" check the top of the Top500 - those IBM's BlueGenes loaded with embedded PowerPC 440.
China's super computer won't be sophiisticated, but it will have low proice and raw power. That's for sure.
Gosh. That's great idea for TLD!
. sucksk s
;-)
".sucks"!
ms.sucks
compaq.sucks
apple.sucks
everything
it.sucks
gov.sucks
mil.sucks
sommer.suc
winter.sucks
binladen.sucks
lukashenko.sucks
and so on.
I was first. Where can I apply for the domain?
P.S. I have had about twenty websites with names "www.${brand}sucks.com". Why not to unite user communities under one roof?
I'm not writing documentation, but seldom help with some technical details and help with translations.
Guys, just to understand the problem you have try first write something. Writing good documentation would require someone who can think straight and clearly. And as you can imaging if company got such guy/girl - s/he will be busy doing product itself, rather than documentation. That's first.
Second. On PC market change of technologies occurs every half of year or something like that. Writing documentation, getting thru all bureaucracy for confidential document to be released will be precisely about this half of year. In other words, when documentation is ready - it is already outdated. You can btw notice that industries with longer release cycles normally end-up with decent documentation - they have time and money. Having no money you cannot afford longer release cycle. Telcos are good example of that. Industrial automation is another one.
So PC industry is sort of its own worst enemy. Tight competition force producers to save where possible - e.g. optional component such as documentation. If you have some bright heads, producer can manage to update product line more often - edging against competitors - but again bright heads will be busy with product release, rather than its documentation.
Having dedicated personnel to handle documentation is just expensive.
P.S. One more problem is when product is partially based on some licensed third party development. Most small producers have to license things, since they cannot develop everything in house. Than it happens that documentation ends up with hell a lot of copy'n'pasting from third party one. To release that you have to get a permission from third party: but in most situations you will find out that third party has licensed some parts itself. In the end no-one just want to risk releasing the documentation, especially in litigious U.S.: most companies when finding their parts in someone's else released products may start asking for fees. With all consequences. Hiring experienced attorney for going thru all this licensing mess - and getting clearance from all involved parties - will cost a lot money, most producers just not able to afford.
Memory fragmentation is well known. That's true that there is no solution. But on the other side, there are technics one can employ to resist memory fragmentation.
Check memory pools. Network equipment uses them (you expect router/gateway to run 24/7). Check Linux kernel for slab allocator. It is not ideal - but still works quite well.
The problem one cannot solve is allocation of huge memory peice of continous memory. But on the other side, this is problem more specific to kernel.
User space applications are much less constrained, compared to kernel space tasks. It is just question of willingness of developers to solve the problem. It is just all things will have to be splitted into manageable pieces. Unfortunately, in some cases it is just impossible (e.g. second party libraries).
Web browser is quite complicated peice of software, with support of images, video, audio and various texts. Such changes might prove to be complicated to implement and maintain. With complications come bugs. I honestly rather relaunch Ff, rather have it crash by itself.
But then detecting things like memory leaks is quite complicated. You cannot be sure that item has to be decallocated and will not be deallocated some time later. Memory management is lazy. And you never can be sure about it completly. Saying that you have found memory leak, by just watching memory consumption going up is of no help to developers. And is not proof of application bug.
Honestly, I am, as a programmer, bemused by industrial automation people all the time I have to work with them.
Money squeezing is norm of life: complaining aloud about say analyzer with encrypted data format so now they cannot use their expensive software from 3rd party. On first turn. And then, on second turn, during meeting they'll be first to propose change in file format just to ensure that customers will not be able to use cheap tools to read them.
I'm seeing that shit for 2.5 years now. Norm of life over here. JPI is by no mean not exception - I just hoped that aviation equipment has some regulations applied to prohibit the boneheads from screwing anything they don't like.
Thou my new company - Thanks God - seems to not have such attitude to customers. On my first industrial automation job, amount of ideas scrapped for simple reason that it will give too much breathing room to customers was near 100%. Some of the product managers did humorously advertising principle: do it worse, customers will have to pay us twice. And this finds sympathy in some people's minds over here. What is very sad.
Hope they will not opt to DRMfy they data. Just try to imaging message on plane: "Dear guests our plane is crashing down due to expired license for motor control protocol encryption. Please festen your belts and pray."
Well, IIRC, this [CENSORED] was already tried in medicine and failed. No system is allowed to use encryption for sake of encryption when human life is at stake. Both medical life support equipment and aviation equipment generally classified as mission critical for that very reason - human lives depend on them.
I have no experience nor information about mission critical systems, but I have just started working for safe equipment producer - and already have clue as what "mission critical" means. Safe equipment is right one step below mission critical equipment: safe equipment has to shutdown system in case of alarm, while mission critical stuff has to work as long as it possibly could - whatever situation/alarms/malfunctions are - someone's arse depends on it.
Development of such equipment is kept as transparent as possible - any mistake might cost way too much. Adopting encryption doesn't seem to me as step in right direction.
For me it is important that Google has competition.
;-)
Google in the beggining was not that shiny as it is now. And not much people knew about Google, when it only started. MSN/etc have to challenge the king of search - and "it is always hard first hundred years" (c) russian proverb.
Google is doing terribly weel to date. But in long term Google needs competitor.
A9 & MSN IMNSHO are most welcome to challenge Google.
I can only hope, that another day, Internet will not die under load produced by bunch of search engines competing in indexing web
Apple is about new cool things. It cannot compete with Dell. (*Now* we have complete antonym of "Apple" - "Dell", now it is easier to compare).
If you want to make widely used computer systems, you have to bow to many customers & companies and their needs.
Remember this: Apple bows no-one. It is on innovative edge. It costs much. And they do mistake too. But it pays off too.
And you cannot be innovaitve and large at the same time. Widely used technologies take long time to create market - by time technology has chance for wide-spread huge market, Apple is already on something new.
P.S. iPod being recent exception.
P.P.S. Think of Apple as of engineering company. They generally - and Jobs in particular - do things people use every day. And Apple engineers are first to use this technologies - they their own dog food. They don't do recket science - they do things person can/want to have at home. And they know it not from analyst's papers, but from real life - they do things for themselves. Nothing more. Bit like history of Porsche - they started like little company to do *good* car for themselves and their friends. They are small, vvvery pricy, but they are still around because they are still very innovative. Thou Apple likes more to parallel itself with BMW ;-)
I hope it will not become soap opeara like with Digital heritage.
Digital (later DEC) was bought by Compaq to ... seems to me to kill competitor. But stupid customers still used DEC stuff, since it worked and worked well, compared to Compaq stuff.
Then Compaq was bought by HP to ... mating of dinos? Most likely. But still this bone-head DEC customers insisted on using software and hardware which just works, and doesn't enrich market with new services & jobs, like PC Repairman.
So after all this years, HP finally killed off DEC legacy stuff, since it is of no use to them - it just works.
Parallels with current situation are obvious. Oracle tries to sell something what sometimes works with market cost of $10 for $100 so no-one wants to buy it. And instead of building better cheaper product it buys PeopleSoft.
I'd be laughing a lot if in couple of years if M$ or IBM will come up with bid to buy Oracle. It would be fun for sure. Especially for People Soft customers.
People used to switch e-mail readers - so they are less dependant on it. And fortunately, e-mail formats are not "forked" - like M$ did with HTML/DOM/CSS for IE - they are still basicly plain text.
Browsing experience was crippled much more by IE, than e-mail experience by Outlook. For some pople Outlook (not OE) is the only choice, even when they have no Exchage Server. Outlook is true point of M$ Desktop & Office integration, many times unfortunately overlooked and later poorly copied by Evolution. And as this integration point Outlook enjoys popularity among virus writers: it supports everything IE supports, it provides e-mails servies, it manages address book, it manages notifications, etc. This integration is very important for some people. But for others like me it is not.
Amongst many clients I have used before only mutt and Netscape Messenger (now Mozilla Thunderbird) are most stable for me. But it is probably only me. I switched back then from mutt after couple of months getting some html messages with pictures - I have couple of friends who love this feature and use it way too much.
Another day I have heard screams in computer room. I went there just to find my friends literally laughing to death. They were trying thru laughing point to the screen of WinXP with error message.
As soon I have taken a look at screen - I have joined them laughing to death under table.
"Invalid MS-DOS function"
For sure, we had over-reacted, due to couple of M$ Zealot who tried to persuade development department that WinXP is complete rewrite of Windows from scratch. And it has nothing to do with MS DOS.
As a person who switch to Linux & Apple long time ago I find bit fuzzing insistence of some companies on using technology from 80s. If you haven't noticed, all external hard-drives are shipped formated with FAT.
No-one yet came out and proposed read-write file system for hard-drives supportable by all OSs. File systems are not standard - I'll love to see OpenGroup/POSIX/ISO having standardize some file system in order for interoperability between OSs. Just like it was done for CD/DVD media.
P.S. Message in our case was showed when one guy tried to delete file with name 'nul' with Explorer. Who remember DOS times - it is reserved name which is presumably impossible to give to a file. Some tools do allow to create/delete files with such name under WinNT/friends.
Recipe for mirror in five minutes:
Install BitTorrent, download FireFox, leave it for uploaders.
You are done!
If you will contribute code to GPL's MySQL - it will not appear in proprietary one. MySQL people might choose to re-implement it for proprietary MySQL - but GPLed code will be held GPLed. On opposite side, MySQL AB makes some parts of their proprietary code GPLed - they are trying to keep both versions as close to each other as possible.
XFree86. Again you not right. One submitting code keeps all copyrights and what's more with her own license. Debian people did analysis of XFree86 and found about 40 variations of BSD license. Sometimes very different licenses. XFree86 folks admitted to be absolutely lax on licensing/copyrighting. Code with is genuinely XFree86 - is rather modest part of whole XFree86 (think applications, fonts, extensions, drivers, scripts, configs, etc).
I cannot be sure about Mozilla and Apache Software Foundation (ASF). IIRC both of them are pretty strict on licensing and copyrighting. I'm absolutely sure that they very strict on licensing: if you have submitted code under one license - you will be asked if project will like to change the license. That was a precedent with Mozilla going from MPL/GPL/LGPL -> GPL/LGPL - they have e-mailed every contributor with request for her/his code being transfered under different license. Apache didn't have had problems like that - I believe they are more to engineering, than to licensing - so I expect them to require copyright transfer, so if license will turns out to have a problem - they can address problem without years of getting "yes" from every contributor.
Relicensing process with its recent example in Mozilla have older cases: BSD's one I have heard about. RMS (Stallman of FSF/GNU) successfully argued to BSD folks that 4 clause BSD license (with acknowledgment clause) will only hurt in future - and since then BSD adopted 3-clauses BSD license, which is compatible with GPL.
If you like see Steven Jobs in action - check recordings of his conferences with shareholders. To date shareholders are buying everything he is saying - they are sort of his employer, but still it is Jobs who is saying what Apple needs to do. Not his employers - shareholders.
Conflicts, where I got that bleak picture of Havoc Pennington, where window minization animation and gnome control center. Dig up archives. Have Jobs spoken this way with users once - no will ever buy Apple computers afterwards.
Making another standard will imho *ensure* that GNOME will not get out of this stupor: poeple will spend another N (where N>0) years arguing about what needs to be in HIG2. And I do not see that GNOME people are uncontent with current HIG. (*)
GNOME is definitely over-rugulated. You have this impression of "dealing with children," because those corporate types I'm talking about are just out of any communication channels. At all. They do not care. Period.
You better be watching commited patches - and who commits them. And then judge by content of patches - is this child or some other type.
You have said "children" - GNOME/Gtk went some funny direction, so with its APIs you actually can frighten little children. Unfortunately for me - and this is why I always liked GNOME/Gtk - I used to do absolutely the same kind of APIs: overcumbered with functionality, un-intuitive and extremely hard to use. Well, they are reliable, complete and error-free. But just try to use them...
(*) After some time, I have developed impression that no-one really likes nor uses GNOME (besides making sexy screenshots). Journalists are paid better for positive reviews. Users on most of the lists after having spent 2 weeks compiling new GNOME just cannot say "it sucks" - it would be too hard after two weeks strugling to compile. Developers - being myself developer - I cannot belive that a developer can use GNOME. There were a release of GNOME, where you were not able to configure keyboard, and default keyboard shortcuts obviously configured for some mysterious end-user, who have problems telling keyboard from any other board... So who are really using it anyway???
Choice is something that the experience user wants.
This is precisely why Linux on Desktop sucks.
Look at Apple as a good example. What is Apple? - It is just cover company for ideas of Jobs. Why is Mac consistent? Because there is Steven Jobs - and there is no choice. People at Apple do not waste their time arguing on mailing lists about better desktop. Jobs has vision - and he drives company according to this vision.
KDE? KDE is made of people who enjoy desktop. Probably they are not greatest GUI programmers - but they like what they do. They are enthusiasts of what they do. Not consistent, Not polished - but with load of features. Great utility from people who have enjoyed doing KDE.
GNOME? GNOME is made up from pollitically correct corporate sponsored full time developers (RedHat, Sun, FSF). They are more to politics and to deliver on corporate business plan, than to listen to their users. It was absolutely funny how Havoc Pennington (of RedHat brainwashing fame) was arguing on list against end-users that he has statistics from end-users at hand, and every-one on the list is wrong, because some has given him statistics and - well - he doesn't care, he has a road map ha has to adhere since RedHat is planning release of next RedHat, etc, etc, etc.
IOW, Democracy - like one found in GNOME - is no substitute to leadership (Apple, Enlightenment). Republican structure of KDE performs here better too.
Linux will not get good Desktop until someone will step forward as a leader. And I see more chances for this to happend in KDE, rather than in GNOME project. Only if someone really disgusted will decide to fork off GNOME, what is IMHO not worth doing.
P.S. Best Linux desktop to date is Enlightenment (E). It is shiny and brilliant. It is finished, polished and complete. Why? Because guy who did it - Rasterman - really cares. And he is driving his project forward. Not fast, but GNOME as was two years behind of E - it is now the same two years behind of E in usability & eye candy.
P.P.S. Forking off GNOME. Well it might be not the worst idea. After all even XFree86 was sucessfully forked, pushing development of both - X.Org & XFree86 - ahead on new wave of competition. If someone will fork GNOME, providing good desktop - it might attract some independent developers working on GNOME, potentially making viable alternative to old fans of good old GNOME.
I've forgotten - and actually never knew this things very good. They usually go out of scope of school chemistry courses.
I remember that my god-father was using some derivative of spirit to improve quality of his home heating system. Bit his system was open one.
Another important property - as you have absulutely correctly pointed out - is density. So it seems that we are left only with open systems, since CPUs usually run with relatively low temperatures. (Trick with lithium is simple - it runs with very-very high temperature - so its density is pretty close to constant.)
P.S. Car engines are using oil. It is run inside of closed system with lots of redundant volume, since oil tends to decrease density (and increase in volume) with higher temperatures. Worst thing with oils - they are all easy to vapourize. But probably oil might be used as a secondary conductor - after aluminium/copper, which sits on top of CPU - to increase effective surface of heat exchange with water (think lots of water pipes dipped into oil bath standing on top of CPU - easier to produce than aluminium radiator for water cooling).
Well, in closed system you can definitely use something better than water.
;-) (FYI nuclear plants are using lithium for first cooling curcuit. Water in second curcuit is used to cool lithium. Hot water then is used to heat houses, produce electricity, whatever)
/. ;-)
Water is not the best coolant ever. I beleive it is possible to find some other chemical agent which will have higher heat capacity and heat conductivity. Improving this two properties will improve cooling for sure.
Another option - obviously - make water flow faster. But this can introduce noise and make system more expensive (given you will try to keep the same noise level).
For even more expensive option one can use trick used in electric plants: use two cooling circuits. First one is connected to CPU. You can ensure that this circuit doesn't have any leaks and use any toxic material to get heat out of CPU *very* *very* fast. Second curcuit obviously is used to cool first curcuit. It can be open, it can use water. It can be very very cheap, compared to first one
P.S. I'm not an expert - I might be wrong somewhere here. But gosh - it's
But it is just little things.
I'm using VIM very much. I hate VIM, I hate Emacs, I love MSDevStudio. But. But under VIM+Bash I'm at least three times more productive.
Windows GUI is good and consistent. To some degree. I'm as a person who designed for two years GUI applications for Windows and knowing every input/output/message/control available I can say that Windows GUI is most advanced GUI ever created. But. But M$ itself stopped following itsown GUI desing guidelines, and I'm not taliking about dumb so-called "VB Programmers" and other commercial software developers who have problems doing simple window with two buttons right. This is really sickening.
Error handing in Windows is just awful. It has nowthing comparable to /var/log/messages. Once I have spent 3 month being not able to run one of the my development tools. It was really bad situation and no one have ever had any clue what have happened. I have used other machine with devkit installed, where my tool worked Ok. But after sometime it stopped working there too. After two weeks of games with regedit/etc it turns out that this application was Win16 application (Win32 has no required system call - but Win16 subsystem does) and when and size of evironment was giong over some limit Win16 subsystem was just stopping to work. With no error message whatsoever.
I can go on and on. For a long time. I've being long-time M$ user and developer. But once I (actually bit forcefully) switched to Linux - I was really amased: some things didn't worked, but most of other things just worked. Without reboots, without crashes, without asking tons silly questions. Just worked. Breath of fresh air after 6 years of WinNT 4.0.
P.S. w2k/xp really didn't changed this balance much since early year 2000 - the time I switched to Linux completely.
Usually SCSI vs. IDE is noticable in CPU intensive operations: CPU useage of SCSI hard drives is less by about 20-25%.