Part of the problem here is we can't tell the review writer and his point of view from the book author.
A second problem is, the reviewer didn't address Who is this book addressed to? This reviewer appears to be caught up in the book, and the review is largely a retelling of what the book asserts.
A third problem is the review doesn't match the quality usually associated with the publsher: O'Reilly. O'Reilly books are generally in the Stewart Brand "access to tools" cultural-philosophical penumbra. This year incidentially is the 35th anniversary of the 1972 Whole Earth Catalog.
The reviewer hasn't distinguished "What is the question?". Is the book about computers, education, culture, philosophy, existential anxiety or the emptiness of the unexamined life?
I have been recently thinking back to the Energy Crisis of 1972 and 1979 and asking "What have we learned?" "What did we change about our society as a result of those crises?" "What myths have we wrapped ourselves in as a society in the last 28 years?". The choice of questions drastically conditions the audience and assertions.
This was a quality business scanner & printer for it's time.
I disassembled one ink cartridge to understand how the ink feed works. (Hacksaw around seam in top cover plate, wash pieces in sink ). Now that I have seen how the cartridge is arranged, I use bulk refill ink in a bottle with a stainless steel needle tip. I leave the cartridge in the printer, lift up a piece of tape, squeeze in 20 ml, run clean heads three times.
The flatbed scanner is pretty good. Scans of 2 1/4" b&w negatives and documents are decent.
Well I'd suggest your problem is not choosing between two different brands of conventional computers.
The answer to "what kind of new computer?" is "It must be low energy consumption now."
The first point I would suggest is: Now is the time to begin aggressively moving to low energy consumption computing for general purpose office machines.
The conventional mass market machinery that I see is not dramatically better than my 700 mhz 1998 Athlon. My computer burns 176 watts ( about 90 watts for the CRT monitor, and 90 watts for the cpu with 2 disk drives, My machine also wastes 4 minutes starting ).
The typical $800 business box with a CPU running at 2% load for 99% of every 8 to 10 hour workday is way too much energy consumption for typical office work.
One way of arguing for low energy computers is look at the total cost of the electricity used over the lifetime of the computer. Then reprice the electricity from $.15 per KWH for carbon based generation to the price of solar electric generation, which is currently perhaps $.40 to $1.20 per KWH, When electricity is priced as if it were solar electricity, low energy computers quickly become the lower total lifetime cost device.
The problem is to deliver a way of doing the work of your organization with dramatically lower carbon burden. If you set a goal of replacing individual machines with machines that use 1/3 to 1/5 the energy of present office machines you will be moving towards a low energy business profile that will look good for 5 to 7 years.
Energy reductions in the 20% to 33% range mean you need cutting edge low energy systems. "Cutting edge" means as consumers we need quality comparative reviews. Some low energy systems may not be right, other configurations may do what is needed elegantly.
Right now, low energy office computing is practical. For instance there is the Wyse terminal server client product that I think burns around 35 watts with a display. Lashups like a gumstix, an LCD, a keyboard, and a mouse are almost office computers in the 5 to 20 watt range.
Patents screwed up the first 20 years of airplane development (Wrigh), patents delayed the development of safe mass production cars (Durea, hydraulic brakes, ignition systems), patents delayed the widespread availability of hetrodyne radio receivers in the 30's, patents delayed the widespread availability of copying machines (Xerox), patents are currently blocking millions of people from cheap drugs for Aids.
The patent game is tightly connected to the merchantile theories of "what do high labor cost highly educated countries create and sell in order to buy what low labor cost countries make? They make "intellectual property", stuff protected by patent an copyright.
So that is the mercantilist clinch that prevents the United States from relaxing patent law. Neither of the major politicial parties can vote against "Keeping America competitive in world trade."
What we need is an economic analysis of the "monopoly costs to the world of unrestrained patent law" and a comparison with the global benefit of "automatic licensing of intellectual property at a sufficiently low user cost that patents enhance the quality and value of all production of the entire world."
The benefits of patents and patent revenue are the dream drug of a capitalist "Ahh, if I have enough patents I will not have any competitive pressures." Xerox rode a bunch of patents for 20 years.
But the reality of patents is the small guy is almost always screwed. The middle sized guys are gobbled up, and the big players play long, slow expensive games tying up entire industries with higher prices and inferior ideas.
It looks to me like the most successful open source software projects are basic tools that stick fairly close to the subject of computer science.
An interesting group of open source projects are in what seems to me to be a "stalemate". Stalemate means there is a proprietary program holding a large audience, and the open source solution is not gaining user share. I wish that if we think hard about specific problems with open source software we can open a gateway to more applications becoming the leaders in their subject area.
Development stalemate programs: More application users and fewer programmers. Often have no test framework to prove accuracy. Stalemate programs often have user interface problems, not programming problems.
I feel the Linux Documentation Project is a particularly painful case of "stalemate". After "Linux Undercover" I think the LDP tightened their copyright license. The result is I never see the LDP reprinted in bookstores, Here is an interesting problem; how could the licensing of LDP material be improved to make it the place of choice for authors, editors and program creators to document their application and it's usage?
---------- A short list of open source projects by "success level"
Open source projects that seem to be "successful".
the kernel, GNU utilities, Perl, Python, Ruby, Rails, the LAMP stack,Template Toolkit, MySQL
Common elements: Basic tools, high quality grammar and syntax, test suites.
Open source projects that are really great but are in some kind of stalemate:
Blender and Gimp - high centered for weak user metaphor and vocabulary
GnuCash, extraordinary for being ignored by CPA's who hold their noses and say "Quickbuhx"
Various project management software, need a test suite to match outputs against the MS product.
LinuxDocumentationProject hasn't solved problem of getting into bookstores with dated reference quality editing and indexing.
The open source alternative to various Computer Assisted Drawing CAD programs.
The open source alternative to Mathematica.
The open source alternative to circuit simulators like Spice.
Generic open source systems for common businesses: Restaurant point of sale, grocery store, wholesale business. The problem here seems to be that the free reference object never receives back the refinements needed for a production commercial system.
Common elements of the problem group: Lack of test tools. User interface challenges. Weaker feedback path from users to application publisher.
Open source projects that are a mass of time delays:
Drivers for all that hardware that comes with proprietary Windows only software.
Open source projects that need to come into existence:
A TCPIP stack for a new generation transport layer.
A test harness for any entire open source distribution.
There is a fellow running a computer recycling center in Oakland, California who brought to the May 2006 MakeFair in San Mateo a rolling wall with a 9 monitor display. The display was one image partitioned among all 9 displays.
The base of his system was a 3 shelf display rack with casters. It is a common commercial shelving unit.
On the cart he put 3 rows of 3 CRT monitors. Each row of 3 monitors was driven by a salvaged pentium class computer, and each computer had 3 video cards. The video cards pick up an address from the slot they are plugged into, so the video software could send the correct 1/9th part of the image to each display.
The three computers were connected with ethernet cable. The three computers all run linux and X display software. Ethernet is plenty fast enough even on old machines for feeding the displays.
One of the computers was the master. Now what was the software? It was something generic associated with the X display software system but not well publicized. Maybe it was mplayer (which does a number of remarkable video reformatting stunts). See the mplayer documentation files. I remember, he was not using "dual head" video cards (which I have found rather problematic), he just used junkbox video cards and addressed them by slot number. I think X itself has multiple displays in the XF86Config file.
In summary: he had a striking "wall size display". It was portable. It was built using any generic Linux and X. It used recycled computers, video cards and CRT monitors that are available in abundance.
Some of the previous posts surely will get you going on your initial question asking if there are emulation and terminal server pathways for a business to adopt some Linux.
Now, as Monty Python says, for something quite different:
Why not look at re-implementing your basic business tasks in a high level web programming language like Ruby on Rails?
No need to risk millions of dollars, just buy the Dave Thomas Rails book (or something similar) and give yourself three weekends to sketch and try something on your own home Linux box.
Rails leads to an application that runs on a server computer. Your users, initially connect to your application using a web browser. So your business data sits on the server. Your employees run any kind of computer with a Web browser.
So you sketch out on paper what this application will do. The problem will be to understand what specific things are the windows computer only magic stuff. You will have to explore ways to match that functionality or to push data at the magic application and catch the stuff that comes out of said magic application.
So I am basically saying: Try this as an exersize, a low risk and low cost way to determine if at this time you might be able to bypass the whole darn office full of expensive Windows only applications.
I suggest that all the distributions need new level of testing and fixing tools:
What I feel was revealed in the early Ubuntu 6.01 and 6.10 was a number of simple silly application configuration problems.
For instance, I discovered the printing was broken! (fixed by deleting and recreating all the printer drivers).
As noted in earlier posts: the Ubuntu self help wiki pages were completely overloaded with help requests. Every help request was formatted differently, used different terms and had a different subject line. I was fixing stupid stuff I have not dealt with in years. (Spoiled by meticulous time and labor intensive Debian excellence).
The problem: There is not enough manpower or management or bug-fix communication capability in the Ubuntu team to test and validate a distribution with 100 applications running on 100 motherboards with 100 video cards and 100 printer drivers and 20 usb wireless network adapters and 100 ethernet cards. A distribution faces an astronomical number of instances that may reveal another bug.
So here is what we need: A generic application test tool shipped with the distribution and that will run on every computer the distribution is installed on. Something developed like testing built into the Ruby environment. Have a set of simple tests that will verify an application is working OK. Something that will exersize an application, send a report to a server, and retrieve suggestions for changes, and repeat tests.
Whether the "distribution" is Red Hat, or Debian or Ubuntu, the software configuration and quality problem happens because there are an astronomical number of variant combinations of hardware and software. We need a test scheme that runs on every hardware instance. The scheme must test, document, report, work with the local owner, and enable changing software or hardware. On the host end, the test host needs to organize the reports, prioritize the problem, support the programmers and conceivably stage the resolution phase.
------------
The year of the Linux Desktop is happening in India and China with very little publicity.
I hear that large numbers of sites with large numbers of PC's are being rolled out in schools, governments and businesses in China and India. The management, installation and configuration are being done at the lowest possible labor cost by using the remote management and configuration software.
The remote management software is built around a program like webmin, a web administration server. But the granularity of the solution is still hung at administering but not viewing the actual running application.
The remaining problem in these rollouts is despite efforts at having large masses of identical machines, individual machines still develop differences because of specific tasks, added cards, and added hard disks,
It looks to me like the internal contradictions of this "ask slashdot" posting qualify it as a troll.
A real person who installed Ubuntu on their own machine would not mail disks at Ubuntu's expense to friends. A real person with a small amount of experience would realize that the user's investment of time and work and familiarity with the existing computer setup is far more important than the operating system. A real person wouldn't style this kind of intervention as a gift to a friend. - Note, the question appears on Saturday night. This seems to be a slow day for technical news. But this is an intense time for United States election news. There should be plenty of electoral stunts (surely a few suprise emails, no?) including exploitation of all the fraud opportunities in computerized voting.
The upgrade problems point to the need for another kind of testing tool. A tool that will test the complete distribution.
It seems to me that Ubuntu's programmers were willing but bunches of cascading oddball problems exceeded the time and programmer diligence available.
I have been working with Ruby on Rails, which is an environment where huge chunks of function execute with very little detail visible to me as a programmer. It turns out, it is essential to set up tests. You depend on tests to alert you when something you have done has unexpected consequences and breaks stuff elsewhere.
So how would a "distribution testing tool" work? Well, in the end it would exersize each one of your applications and compare the results with a results file. File an error report and download a fix script or fix instructions.
Hmm, lots more to think about here. Is this a wheel needing re-inventing?
Ubuntu has hit the speed and quality limit
on
Ubuntu 6.10 is Out
·
· Score: 1
I think Ubuntu has hit the speed of release vs. quality of packages limit.
The present Ubuntu system of preparing new software releases does not fix enough simple configuration errors before the the little checkbox appears in the Ubuntu software update service saying "Upgrade to 6.06 LTS available now."
My story is I upgraded from "Breezy Badger 5.10" to the "6.06 LTS" sometime in September.Right off the bat I lost three days of job hunting and programming time piddling around fixing 4 package configuration errors. xorg-xserver required a package roll back, printing required reinstalling the printers, Firefox no longer started Adobe Acrobat on pdf files, and GoogleEarth required an obscure edit to xorg.conf.
I seriously wouldn't wish such a string of aggravatons on a beginner. I absolutely would not install 6.06 LTS as a paid technician or consultant.
How did 6.06 LTS get offered to the public? My guess is a whole bunch of package testers were on vacation.
There has been a steady evolution in Linux distributions and I feel the Ubuntu team has hit a wall: The amount of volunteer manpower isn't enough to get all the packages in a Ubuntu Distribution tweaked before the published release date.
In the short term, my next upgrade is going to be a full separate install. I'd much rather set aside 10 Gigs of disk space ( that is $15 of disk space ya know ) than have my work and projects drop dead for 3 days. In the case of GoogleEarth, fixing one stupid number in xorg.conf took 3 darn weeks. Every single day held up a programming project. Yeoww.
In the long term, Linux distributions need something elegant and useful to test each package in place on beta tester machines with a clear connection back to the package maintainer. The slow release cycle of Debian reflects that quality from hand tuning takes a lot of time.
The departure of this debian developer could be an occurance of organizational disruption, brought on by a guerella organizational disrupter.
Several recent marketing books offer "guerella marketing" as the way to promote products with less conventional advertising.
There is a corresponding discipline, "guerella organizational disruption" used in power politics and power business as a less than ethical set of methods to diminish the oponnent. Provocateurs are surmised to exist and not often acknowledged.
Debian is capable of being disrupted by deliberately fomented discord. Debian has a wide open organizational structure, a relatively small number of contributors and a determined style to accept a wide range of people in their community.
My brief contact with #debian on IRC was an unforgettable piece of brisk rudeness. (I went elsewhere and stay away from debian after that kind of humiliation).
Suppose what happened to me was deliberate guerella dismarketing. Suppose persons on the debian developer channels are actually practicing guerella organizational disruption?
How would you discover or map out the individual who consistently and plausibly practices organizational disruption?
I can see a way to start: Get an archive, rate every message, group rated messages by sender, sort to find the person who sends the consistently most disruptive messages.
I think it would be very interesting to interview one of these disruptive people and find out how they reached their state of mind. Is the person in some sense sponsored or involved with a competing operating system vendor? That would be interesting to find out.
Gimp book shows the most common tasks
on
Beginning GIMP
·
· Score: 1
Akana Peck's Gimp book is mainly a how-to-do it guide for the most common graphics editing tasks.
As the other comments indicate, Gimp has an awkward interface and for the present, the best way to use Gimp is to buy a how-to-do it guide.
I intend to buy the Akana Peck Gimp book the next time I have a graphics editing project.
I attended a Linux user group talk by Akana Peck. Besides being a writer, she is active in the refinement of Gimp.
My experience with Gimp is I get lost in the sea of choices. I don't really have a sense of "order of work" in Gimp.
The user interface problem is so chronic, so classic. It is a real opportunity to do what open source does best: build on what has been done and innovate. Move ahead toward what really ought to come into existence.
What I think Gimp needs is:
1. Gimp should get a top layer added. A user interface menu system built to guide and record. Record what menu texts, and what the user expects, what the user does (right step, trials, and errors). And send this menu dialogue for analysis and improvement. Uhh, as Marshall Kirk McKusick said regarding filesystems: another layer of abstraction.
2. A way to write backup files automatically with menu step names stuffed into the filename so you can backup to a named "step".
3. For the forest of variations situations like logo shading where there are dozens of choices, have a way to run 8, 16 or 32 trials with a single run choice such as "show me thumbnails of my data file processed by every logo routine in Gimp". And then a favorites checkbox to bubble the user's favorites to the top.
4. A process of analyzing user interface menu reports to improve the menuing system. Eleminate unnecessary steps, automate switching to special type of cursor, post messages to a dialogue box etc.
5. Design the user interface menu log files to be capable of becoming custom semi-automatic batch execution files. For example, I want to automate the process of subtracting images of sand patterns for my sand pattern investigations. Once I do it, I'd like to automate the process.
It seems to me that the usability problems of Gimp are a classic dilemma. Gimp can do many graphics processes. It is hard for a beginner to mentally organize an order of work and sequence of tasks. The problem at hand is to design a User Interface. One approach is to parallel Photoshop. How about doing even better: Collect information about user expectations, actons and errors. Use that information to converge on an interface that statistically results in the fewest errors.
cwcowell here is doing some interesting thinking. I would like very much to reflect on it from a slightly different educational perspective.
The last philosophy I studied before I was don-ragged out of St. John's College was Wittgenstein.
Roughly Wittgenstein said: We know that philosophy can be done. Similarly, in program design, we have modeling languages and we can arrange actions or semantic outputs that are as close as we please to the design. ----- From Donald MacKay's Information, Mechanism and Meaning we have a very provocative proposal that very simple chunks of language can have a much more meaning due to the situation or context in which a message is sent. The same expansion of meaning appearing the output of a program would be a bug. ------ One could imagine a Turing Machine that could be programmed to make semanitc outputs imitating my writing here. We could imagine that a really big program could eventually converge on making outputs that looked like any given philosopher. ------ So one of the interesting problems is, even if a philosopher and a computer program use the same logic tables, and the same tables of premises and conclusions and so on.... They still do not look alike. And they are not doing the same thing. The philosopher is still doing an inquiry and the computer is still executing turing machine steps.
I did a server cluster monitoring job for an internet company for 13 months. I was fired and I hear by gossip that 1/2 of my job was offshored about a month after I left.
Offshoring is just a part of the business strategy of pushing for lowest overall labor cost and lowest future labor liabilities.
I found the firing quite painful because I knew the place was a job "windmill" and I was earnestly doing my best.
My firing was a case of if the job hadn't been offshored, I still would have been fired.
I figure the company has a built in program of firing employees to keep average hourly wages down. The job was rigged with performance standards and written warnings. Every month a couple of people left or quit. Out of 140, there were one or two hourly employees with 5 years of employment. The comapany aggressively hired: it ran 17 position ads by my count. I figure about 4 to 5 people quit or are fired every month.
Looking around, I see a few people who are hanging onto steady gigs, a few who are superbly qualified and have a steady gig even though it is burning them out.
And then a big hunk of us are "Employed until you cost too much." The company calculates when the cost of hiring and training a new employee or the cost of getting the offshore team working is less than giving you a small raise and contributing to your 401K savings plan.
That is the day they fire you. If the company uses this replacement cost method of staff management, every hourly employee will eventually be fired. how about this as a lemma: The only way not to be fired is to become a manager.
Beyond computers, beyond any technical field you might study, there is a meaningful universe that is worth exploring when you go to college. This is the liberal arts.
The technical fields are endless fractals of detail and mental intrigue. The technical fields are like spokes. What is that hub? The hub I call the liberal arts.
Don't sacrifice yourself completely to the "job security altar". Don't allow your college to be simply 4 years of trade school.
Don't leave college without reading Plato and discussing it, proving the pythagorean theorem, studying enough history to understand what fragile and important stuff it is, dissect an animal, do some physics and chemistry, understand and practice rhetoric, learn to play a Bach Invention, learn the awesomely difficult process of seeing and drawing a human figure.
Note that a bunch of mostly technically well educated young men were persuaded to fly suicide airplanes on September 11th 2001. These technically well educated men were unable to recognize or resist forces being applied to their minds by others.
So the liberal arts, which are taught in college, are the skills of literacy and citizenship. Valuable thing to include in your college.
UserLinux came closer to being a simple, successful business desktop than anything before it.
Ubuntu has taken over on my desktop because of better USB integration. Ubuntu handles my USB scanner, printer and camera and UserLinux doesn't.
UserLinux made the extraordinary Debian software and package environment accessible without the without the inadvertent and uncontrolled negative Gurella marketing presence that has undermined the mainline Debian distribution.
I'll tell you a UserLinux story:
Back in the days when Red Hat stopped selling a $50 box I started looking for another Distro.
I tried Knoppix and an interesting thing happened when I mistakenly visited the Debian IRC chat seeking assistance. I was treated with gruff and rude dismissal.
What I think was going on was somebody was engaging in Gurella Dis-marketing. Whoever this was, it was someone deliberately making sure that anybody exploring Debian got a good bitter mothful of rejection. The people in the IRC chat were hybrid child-professional assholes. People who projected a veneer of competency, and had nothing to say except "go away".
So after that, UserLinux looked like a really nice bunch of people with a reasonable tolerance for my interests.
UserLinux has a picture of a folded paper airplane reflecting that it was a careful selection of the best of breed applications from the huge Debian package universe. Unlike Debian it didn't make you "Figure this out if you want to install this software"
The target client for UserLinux was a "business desktop". The charm of the distribution was it installed like gangbusters and you could add anything you wanted from Debian.
I joined the UserLinux project and I contributed a help file. For UserLinux I wrote a help file covering tasks like dual boot setup and Java installation.
So I'll say thanks Bruce Perens and also Linux Format British edition is an excellent Linux publication (sold at Borders Books I think).
Epidemic: patents and immoral withholding.
on
A Flu Pandemic?
·
· Score: 1
This potential epidemic could lead to moral condemnation of the West for withholding lifesaving medicines and instruments.
The vulnerability of healthy young people to cytokine storm death is the same age group who I presume are recruited by extremist Islamic groups for suicide bombers.
The article is a flashing alarm light that will trigger a thousand entrepreneurial projects. Everything from fancy face masks to genuine medical developments.
One thing we need now is a 2 second test for the h5n1 virus. Something like a tagged antibody and an analyzer driven by an led laser operating on a 1 micrometer nasal droplet. (Make a proton magnetometer with some 1/4" neodymium magnets?)
Another thing we need is a medicine for dealing with the cytokine storm secondary cause of death.
There is quite likely to be a whole spectrum of instrumentation and medicine created in response to this potential epidemic.
Now, as these instruments and medicines are developed, many of these creations will fall under the patent and copyright system.
The problem is the patent and copyright system under a guise of law makes it easy to price lifesaving medicines and instruments instantly out of the reach of low income people.
Note, there is already a problem with patented Aids medications and copyrighted medical information being over priced for some AFrican countries.
So, the United States (and other countries) might get in the position of having the technology for saving lives patented, copyrighted and offered for sale, The price will typically be all the market will bear, as seen by diners eating breakfast on the 86th floor of the World Trade Center. Uhh whoops, on the second floor of some conference room in Palo Alto.
A global epidemic may provide a basis for the most intense finding of moral and religious condemnation against all of Western society that adheres to the patent and copyright system as it is without any component of mercy and local reasonableness in price and availability of patented and copyright materials.
Or else, the patent and copyright system should be immediately ammended to allow ideas and innovations to push out and propagte freely.
Following older slashdot postings about Dr. Kay's Squeak language I decided not to download it because it is not GPL'ed.
I hope Guido van Rossum and Larry Wall eagerly absorb Alan Kay's ideas (especially "write a language in itself") and move the ideas out into the accessible world of GPL'd free documentation, and programs.
The non-GPL publication universe is a get rich quick lime pit.
First, if you are migrating to Linux, check out this IBM "redbook". Very helpful for a business:
http://www.redbooks.ibm.com/abstracts/sg246380.h tm l ---------- Among the various distributions that you might use, you can also add UserLinux, see
http://userlinux.com
---------- Userlinux is a subset of the Debian distribution.
The applications within UserLinux are specifically selected to provide a useful Linux business desktop.
The team assembling UserLinux includes Bruce Perens and a global list of administrators and users focusing on assembling a useful business desktop.
I participate in the group. I have used UserLinux for several months. UserLinux has the elegance and power of Debian with a smaller, less redundant set of applications.
Staffers in the Bush administration have "floated" the idea of a Value Added Tax for the United States.
A VAT tax is the kind of simplification that would allow a radical simplification of the income tax.
Richard Nixon floated the idea of a VAT tax for the United States.
VAT is used in Europe. VAT is where each seller collects the VAT tax on her sales and deducts the VAT paid buying stuff for sale. Governments like it because it tends to be "self collecting". It gets revenue from the formerly "underground" economy. It can get revenue from trans national corporations that manipulate "transfer payments" to shift profits offshore.
As a previous poster has noted, the tax system needs an XML data structure and an XML description of the algorithms for the computation of tax.
I've daydreamed like this: Divide the task into an algorithm part and a sample data part. A python root module would emulate the "1040" form. All the child forms would be separate Python Modules corresponding to each schedule or tax form. The XML tagged test data would be included with each package of modules. The package of algorithms plus data would be GPG signed.
This scheme allows free and professional tax applications to co-exist. It enables modular development. It creates credible test data that can be used to cross validate alternate tax packages. It enables trusted open source.
My compromise for fastest, cheapest and best tax filing, given the absence of open source software is:
First I get the paperwork together. All year long I ensure each deduction is written on a sheet of paper and put in a tax box.
That is 90% of the work.
Next I fill out the federal and state returns in pencil by hand.
Finally type in the numbers from the paper forms to TaxActOnline.com.
The really important check point is: Ensure that the online tax bill agrees to the penny with the paper form.
Then resume seeking a better and more peaceful truth beyond these morally vacant warmakers.
For immediate support, build a library. Buy one book at a time as you need it.
(I think the Rute book and an anthology of the Linux Documentation Project are good cornerstone books.)
Balance your library with appropriate college studies. Whatever aspect of computers you work with, there is much worthwhile material to master and develop at length under the leadership of teachers.
Training does not have lasting value. I took a Red Hat CNE course at my own expense and failed the exam. That training has never had a net positive value ever in a job interview. I simply blew a week trying to absorb what I really would have rather learned in an academic setting where thinking and understanding are valued. The exam was figuring out some stupid dorm tricks and doing a whole bunch of editing real fast.
So the situation is patents fill up the graphics software landscape. The following tiers of software are kept as business secrets.
As I note from the HP patent cross licensing email a month ago, the big players cross license their software patents.
Looking at the patent problem some more: The problem is in the relatively uncontrolled enforcement and royalty collection practices.
What do we have in the way of institutions and social practices that model patent licensing?
Well I can think of Wartime. The government ignores patents.
And video tape copying, where a copying is allowed.
And there is Fair use. I don't know that there is any fair use of a patent.
To focus on part of what Bruce Perens says: "...users can be sued..."
Seems to me that "...users can be sued..." is the crux of why the Microsoft patent blitz is so depressing and menacing. ------ Here is my short list of anti software patent responses:
1. Seek legislative redress.
a.I am looking for a politicial action strategy . b.I need slogans and "3 second sound bites" to accompany donations.Legislators need handles to grip to freedom issues involved. 2. Model a reform act on the Automotive Right to Repair Act. It has a great appeal to the public good in its' text. 3. Write a patent claim engine to systematically generate and publish every conceivable software patent claim. Patent the engine and license it so everything it creates is GNU licensed. Have reviewers reseed the engine with key phrases so the engine eventually carves out a substantial software space that is free. Furnish copies of the engine so specialist developers in particular areas can can carve up free specialty areas.
Slashdot Mirror
Part of the problem here is we can't tell the review writer and his point of view from the book author.
A second problem is, the reviewer didn't address Who is this book addressed to? This reviewer appears to be caught up in the book, and the review is largely a retelling of what the book asserts.
A third problem is the review doesn't match the quality usually associated with the publsher: O'Reilly. O'Reilly books are generally in the Stewart Brand "access to tools" cultural-philosophical penumbra. This year incidentially is the 35th anniversary of the 1972 Whole Earth Catalog.
The reviewer hasn't distinguished "What is the question?". Is the book about computers, education, culture, philosophy, existential anxiety or the emptiness of the unexamined life?
I have been recently thinking back to the Energy Crisis of 1972 and 1979 and asking "What have we learned?" "What did we change about our society as a result of those crises?" "What myths have we wrapped ourselves in as a society in the last 28 years?". The choice of questions drastically conditions the audience and assertions.
Yes, my Stylus Scan 2500 cost me $40 used.
This was a quality business scanner & printer for it's time.
I disassembled one ink cartridge to understand how the ink feed works. (Hacksaw around seam in top cover plate, wash pieces in sink ). Now that I have seen how the cartridge is arranged, I use bulk refill ink in a bottle with a stainless steel needle tip. I leave the cartridge in the printer, lift up a piece of tape, squeeze in 20 ml, run clean heads three times.
The flatbed scanner is pretty good. Scans of 2 1/4" b&w negatives and documents are decent.
Well I'd suggest your problem is not choosing between two different brands of conventional computers.
.15 per KWH for carbon based generation to the price of solar electric generation, which is currently perhaps $.40 to $1.20 per KWH, When electricity is priced as if it were solar electricity, low energy computers quickly become the lower total lifetime cost device.
The answer to "what kind of new computer?" is "It must be low energy consumption now."
The first point I would suggest is: Now is the time to begin aggressively moving to low energy consumption computing for general purpose office machines.
The conventional mass market machinery that I see is not dramatically better than my 700 mhz 1998 Athlon. My computer burns 176 watts ( about 90 watts for the CRT monitor, and 90 watts for the cpu with 2 disk drives, My machine also wastes 4 minutes starting ).
The typical $800 business box with a CPU running at 2% load for 99% of every 8 to 10 hour workday is way too much energy consumption for typical office work.
One way of arguing for low energy computers is look at the total cost of the electricity used over the lifetime of the computer. Then reprice the electricity from $
The problem is to deliver a way of doing the work of your organization with dramatically lower carbon burden. If you set a goal of replacing individual machines with machines that use 1/3 to 1/5 the energy of present office machines you will be moving towards a low energy business profile that will look good for 5 to 7 years.
Energy reductions in the 20% to 33% range mean you need cutting edge low energy systems. "Cutting edge" means as consumers we need quality comparative reviews. Some low energy systems may not be right, other configurations may do what is needed elegantly.
Right now, low energy office computing is practical. For instance there is the Wyse terminal server client product that I think burns around 35 watts with a display. Lashups like a gumstix, an LCD, a keyboard, and a mouse are almost office computers in the 5 to 20 watt range.
Patents screwed up the first 20 years of airplane development (Wrigh),
patents delayed the development of safe mass production cars (Durea, hydraulic brakes, ignition systems),
patents delayed the widespread availability of hetrodyne radio receivers in the 30's,
patents delayed the widespread availability of copying machines (Xerox),
patents are currently blocking millions of people from cheap drugs for Aids.
The patent game is tightly connected to the merchantile theories of "what do high labor cost highly educated countries create and sell in order to buy what low labor cost countries make? They make "intellectual property", stuff protected by patent an copyright.
So that is the mercantilist clinch that prevents the United States from relaxing patent law. Neither of the major politicial parties can vote against "Keeping America competitive in world trade."
What we need is an economic analysis of the "monopoly costs to the world of unrestrained patent law" and a comparison with the global benefit of "automatic licensing of intellectual property at a sufficiently low user cost that patents enhance the quality and value of all production of the entire world."
The benefits of patents and patent revenue are the dream drug of a capitalist "Ahh, if I have enough patents I will not have any competitive pressures." Xerox rode a bunch of patents for 20 years.
But the reality of patents is the small guy is almost always screwed. The middle sized guys are gobbled up, and the big players play long, slow expensive games tying up entire industries with higher prices and inferior ideas.
It looks to me like the most successful open source software projects are basic tools that stick fairly close to the subject of computer science.
An interesting group of open source projects are in what seems to me to be a "stalemate". Stalemate means there is a proprietary program holding a large audience, and the open source solution is not gaining user share. I wish that if we think hard about specific problems with open source software we can open a gateway to more applications becoming the leaders in their subject area.
Development stalemate programs: More application users and fewer programmers. Often have no test framework to prove accuracy. Stalemate programs often have user interface problems, not programming problems.
I feel the Linux Documentation Project is a particularly painful case of "stalemate". After "Linux Undercover" I think the LDP tightened their copyright license. The result is I never see the LDP reprinted in bookstores, Here is an interesting problem; how could the licensing of LDP material be improved to make it the place of choice for authors, editors and program creators to document their application and it's usage?
---------- A short list of open source projects by "success level"
Open source projects that seem to be "successful".
the kernel, GNU utilities, Perl, Python, Ruby, Rails, the LAMP stack,Template Toolkit, MySQL
Common elements: Basic tools, high quality grammar and syntax, test suites.
Open source projects that are really great but are in some kind of stalemate:
Blender and Gimp - high centered for weak user metaphor and vocabulary
GnuCash, extraordinary for being ignored by CPA's who hold their noses and say "Quickbuhx"
Various project management software, need a test suite to match outputs against the MS product.
LinuxDocumentationProject hasn't solved problem of getting into bookstores with dated reference quality editing and indexing.
The open source alternative to various Computer Assisted Drawing CAD programs.
The open source alternative to Mathematica.
The open source alternative to circuit simulators like Spice.
Generic open source systems for common businesses: Restaurant point of sale, grocery store, wholesale business. The problem here seems to be that the free reference object never receives back the refinements needed for a production commercial system.
Common elements of the problem group: Lack of test tools. User interface challenges. Weaker feedback path from users to application publisher.
Open source projects that are a mass of time delays:
Drivers for all that hardware that comes with proprietary Windows only software.
Open source projects that need to come into existence:
A TCPIP stack for a new generation transport layer.
A test harness for any entire open source distribution.
There is a fellow running a computer recycling center in Oakland, California who brought to the May 2006 MakeFair in San Mateo a rolling wall with a 9 monitor display. The display was one image partitioned among all 9 displays.
The base of his system was a 3 shelf display rack with casters. It is a common commercial shelving unit.
On the cart he put 3 rows of 3 CRT monitors. Each row of 3 monitors was driven by a salvaged pentium class computer, and each computer had 3 video cards. The video cards pick up an address from the slot they are plugged into, so the video software could send the correct 1/9th part of the image to each display.
The three computers were connected with ethernet cable. The three computers all run linux and X display software. Ethernet is plenty fast enough even on old machines for feeding the displays.
One of the computers was the master. Now what was the software? It was something generic associated with the X display software system but not well publicized. Maybe it was mplayer (which does a number of remarkable video reformatting stunts). See the mplayer documentation files. I remember, he was not using "dual head" video cards (which I have found rather problematic), he just used junkbox video cards and addressed them by slot number. I think X itself has multiple displays in the XF86Config file.
In summary: he had a striking "wall size display". It was portable. It was built using any generic Linux and X. It used recycled computers, video cards and CRT monitors that are available in abundance.
Some of the previous posts surely will get you going on your initial question asking if there are emulation and terminal server pathways for a business to adopt some Linux.
Now, as Monty Python says, for something quite different:
Why not look at re-implementing your basic business tasks in a high level web programming language like Ruby on Rails?
No need to risk millions of dollars, just buy the Dave Thomas Rails book (or something similar) and give yourself three weekends to sketch and try something on your own home Linux box.
Rails leads to an application that runs on a server computer. Your users, initially connect to your application using a web browser. So your business data sits on the server. Your employees run any kind of computer with a Web browser.
So you sketch out on paper what this application will do. The problem will be to understand what specific things are the windows computer only magic stuff. You will have to explore ways to match that functionality or to push data at the magic application and catch the stuff that comes out of said magic application.
So I am basically saying: Try this as an exersize, a low risk and low cost way to determine if at this time you might be able to bypass the whole darn office full of expensive Windows only applications.
I suggest that all the distributions need new level of testing and fixing tools:
What I feel was revealed in the early Ubuntu 6.01 and 6.10 was a number of simple silly application configuration problems.
For instance, I discovered the printing was broken! (fixed by deleting and recreating all the printer drivers).
As noted in earlier posts: the Ubuntu self help wiki pages were completely overloaded with help requests. Every help request was formatted differently, used different terms and had a different subject line. I was fixing stupid stuff I have not dealt with in years. (Spoiled by meticulous time and labor intensive Debian excellence).
The problem: There is not enough manpower or management or bug-fix communication capability in the Ubuntu team to test and validate a distribution with 100 applications running on 100 motherboards with 100 video cards and 100 printer drivers and 20 usb wireless network adapters and 100 ethernet cards. A distribution faces an astronomical number of instances that may reveal another bug.
So here is what we need: A generic application test tool shipped with the distribution and that will run on every computer the distribution is installed on. Something developed like testing built into the Ruby environment. Have a set of simple tests that will verify an application is working OK. Something that will exersize an application, send a report to a server, and retrieve suggestions for changes, and repeat tests.
Whether the "distribution" is Red Hat, or Debian or Ubuntu, the software configuration and quality problem happens because there are an astronomical number of variant combinations of hardware and software. We need a test scheme that runs on every hardware instance. The scheme must test, document, report, work with the local owner, and enable changing software or hardware. On the host end, the test host needs to organize the reports, prioritize the problem, support the programmers and conceivably stage the resolution phase.
------------
The year of the Linux Desktop is happening in India and China with very little publicity.
I hear that large numbers of sites with large numbers of PC's are being rolled out in schools, governments and businesses in China and India. The management, installation and configuration are being done at the lowest possible labor cost by using the remote management and configuration software.
The remote management software is built around a program like webmin, a web administration server. But the granularity of the solution is still hung at administering but not viewing the actual running application.
The remaining problem in these rollouts is despite efforts at having large masses of identical machines, individual machines still develop differences because of specific tasks, added cards, and added hard disks,
It looks to me like the internal contradictions of this "ask slashdot" posting qualify it as a troll.
A real person who installed Ubuntu on their own machine would not mail disks at Ubuntu's expense to friends. A real person with a small amount of experience would realize that the user's investment of time and work and familiarity with the existing computer setup is far more important than the operating system. A real person wouldn't style this kind of intervention as a gift to a friend.
-
Note, the question appears on Saturday night. This seems to be a slow day for technical news. But this is an intense time for United States election news. There should be plenty of electoral stunts (surely a few suprise emails, no?) including exploitation of all the fraud opportunities in computerized voting.
The upgrade problems point to the need for another kind of testing tool. A tool that will test the complete distribution.
It seems to me that Ubuntu's programmers were willing but bunches of cascading oddball problems exceeded the time and programmer diligence available.
I have been working with Ruby on Rails, which is an environment where huge chunks of function execute with very little detail visible to me as a programmer. It turns out, it is essential to set up tests. You depend on tests to alert you when something you have done has unexpected consequences and breaks stuff elsewhere.
So how would a "distribution testing tool" work? Well, in the end it would exersize each one of your applications and compare the results with a results file. File an error report and download a fix script or fix instructions.
Hmm, lots more to think about here. Is this a wheel needing re-inventing?
I think Ubuntu has hit the speed of release vs. quality of packages limit.
The present Ubuntu system of preparing new software releases does not fix enough simple configuration errors before the the little checkbox appears in the Ubuntu software update service saying "Upgrade to 6.06 LTS available now."
My story is I upgraded from "Breezy Badger 5.10" to the "6.06 LTS" sometime in September.Right off the bat I lost three days of job hunting and programming time piddling around fixing 4 package configuration errors. xorg-xserver required a package roll back, printing required reinstalling the printers, Firefox no longer started Adobe Acrobat on pdf files, and GoogleEarth required an obscure edit to xorg.conf.
I seriously wouldn't wish such a string of aggravatons on a beginner. I absolutely would not install 6.06 LTS as a paid technician or consultant.
How did 6.06 LTS get offered to the public? My guess is a whole bunch of package testers were on vacation.
There has been a steady evolution in Linux distributions and I feel the Ubuntu team has hit a wall: The amount of volunteer manpower isn't enough to get all the packages in a Ubuntu Distribution tweaked before the published release date.
In the short term, my next upgrade is going to be a full separate install. I'd much rather set aside 10 Gigs of disk space ( that is $15 of disk space ya know ) than have my work and projects drop dead for 3 days. In the case of GoogleEarth, fixing one stupid number in xorg.conf took 3 darn weeks. Every single day held up a programming project. Yeoww.
In the long term, Linux distributions need something elegant and useful to test each package in place on beta tester machines with a clear connection back to the package maintainer. The slow release cycle of Debian reflects that quality from hand tuning takes a lot of time.
The departure of this debian developer could be an occurance of organizational disruption, brought on by a guerella organizational disrupter.
Several recent marketing books offer "guerella marketing" as the way to promote products with less conventional advertising.
There is a corresponding discipline, "guerella organizational disruption" used in power politics and power business as a less than ethical set of methods to diminish the oponnent. Provocateurs are surmised to exist and not often acknowledged.
Debian is capable of being disrupted by deliberately fomented discord. Debian has a wide open organizational structure, a relatively small number of contributors and a determined style to accept a wide range of people in their community.
My brief contact with #debian on IRC was an unforgettable piece of brisk rudeness. (I went elsewhere and stay away from debian after that kind of humiliation).
Suppose what happened to me was deliberate guerella dismarketing. Suppose persons on the debian developer channels are actually practicing guerella organizational disruption?
How would you discover or map out the individual who consistently and plausibly practices organizational disruption?
I can see a way to start: Get an archive, rate every message, group rated messages by sender, sort to find the person who sends the consistently most disruptive messages.
I think it would be very interesting to interview one of these disruptive people and find out how they reached their state of mind. Is the person in some sense sponsored or involved with a competing operating system vendor? That would be interesting to find out.
Akana Peck's Gimp book is mainly a how-to-do it guide for the most common graphics editing tasks.
As the other comments indicate, Gimp has an awkward interface and for the present, the best way to use Gimp is to buy a how-to-do it guide.
I intend to buy the Akana Peck Gimp book the next time I have a graphics editing project.
I attended a Linux user group talk by Akana Peck. Besides being a writer, she is active in the refinement of Gimp.
My experience with Gimp is I get lost in the sea of choices. I don't really have a sense of "order of work" in Gimp.
The user interface problem is so chronic, so classic. It is a real opportunity to do what open source does best: build on what has been done and innovate. Move ahead toward what really ought to come into existence.
What I think Gimp needs is:
1. Gimp should get a top layer added. A user interface menu system built to guide and record. Record what menu texts, and what the user expects, what the user does (right step, trials, and errors). And send this menu dialogue for analysis and improvement. Uhh, as Marshall Kirk McKusick said regarding filesystems: another layer of abstraction.
2. A way to write backup files automatically with menu step names stuffed into the filename so you can backup to a named "step".
3. For the forest of variations situations like logo shading where there are dozens of choices, have a way to run 8, 16 or 32 trials with a single run choice such as "show me thumbnails of my data file processed by every logo routine in Gimp". And then a favorites checkbox to bubble the user's favorites to the top.
4. A process of analyzing user interface menu reports to improve the menuing system. Eleminate unnecessary steps, automate switching to special type of cursor, post messages to a dialogue box etc.
5. Design the user interface menu log files to be capable of becoming custom semi-automatic batch execution files. For example, I want to automate the process of subtracting images of sand patterns for my sand pattern investigations. Once I do it, I'd like to automate the process.
It seems to me that the usability problems of Gimp are a classic dilemma. Gimp can do many graphics processes. It is hard for a beginner to mentally organize an order of work and sequence of tasks. The problem at hand is to design a User Interface. One approach is to parallel Photoshop. How about doing even better: Collect information about user expectations, actons and errors. Use that information to converge on an interface that statistically results in the fewest errors.
cwcowell here is doing some interesting thinking. I would like very much to reflect on it from a slightly different educational perspective.
The last philosophy I studied before I was don-ragged out of St. John's College was Wittgenstein.
Roughly Wittgenstein said: We know that philosophy can be done. Similarly, in program design, we have modeling languages and we can arrange actions or semantic outputs that are as close as we please to the design.
-----
From Donald MacKay's Information, Mechanism and Meaning we have a very provocative proposal that very simple chunks of language can have a much more meaning due to the situation or context in which a message is sent. The same expansion of meaning appearing the output of a program would be a bug.
------
One could imagine a Turing Machine that could be programmed to make semanitc outputs imitating my writing here. We could imagine that a really big program could eventually converge on making outputs that looked like any given philosopher.
------
So one of the interesting problems is, even if a philosopher and a computer program use the same logic tables, and the same tables of premises and conclusions and so on.... They still do not look alike. And they are not doing the same thing. The philosopher is still doing an inquiry and the computer is still executing turing machine steps.
Fired in March, offshored in May:
I did a server cluster monitoring job for an internet company for 13 months. I was fired and I hear by gossip that 1/2 of my job was offshored about a month after I left.
Offshoring is just a part of the business strategy of pushing for lowest overall labor cost and lowest future labor liabilities.
I found the firing quite painful because I knew the place was a job "windmill" and I was earnestly doing my best.
My firing was a case of if the job hadn't been offshored, I still would have been fired.
I figure the company has a built in program of firing employees to keep average hourly wages down. The job was rigged with performance standards and written warnings. Every month a couple of people left or quit. Out of 140, there were one or two hourly employees with 5 years of employment. The comapany aggressively hired: it ran 17 position ads by my count. I figure about 4 to 5 people quit or are fired every month.
Looking around, I see a few people who are hanging onto steady gigs, a few who are superbly qualified and have a steady gig even though it is burning them out.
And then a big hunk of us are "Employed until you cost too much." The company calculates when the cost of hiring and training a new employee or the cost of getting the offshore team working is less than giving you a small raise and contributing to your 401K savings plan.
That is the day they fire you. If the company uses this replacement cost method of staff management, every hourly employee will eventually be fired. how about this as a lemma: The only way not to be fired is to become a manager.
Beyond computers, beyond any technical field you might study, there is a meaningful universe that is worth exploring when you go to college. This is the liberal arts.
The technical fields are endless fractals of detail and mental intrigue. The technical fields are like spokes. What is that hub? The hub I call the liberal arts.
Don't sacrifice yourself completely to the "job security altar". Don't allow your college to be simply 4 years of trade school.
Don't leave college without reading Plato and discussing it, proving the pythagorean theorem, studying enough history to understand what fragile and important stuff it is, dissect an animal, do some physics and chemistry, understand and practice rhetoric, learn to play a Bach Invention, learn the awesomely difficult process of seeing and drawing a human figure.
Note that a bunch of mostly technically well educated young men were persuaded to fly suicide airplanes on September 11th 2001. These technically well educated men were unable to recognize or resist forces being applied to their minds by others.
So the liberal arts, which are taught in college, are the skills of literacy and citizenship. Valuable thing to include in your college.
UserLinux came closer to being a simple, successful business desktop than anything before it.
Ubuntu has taken over on my desktop because of better USB integration. Ubuntu handles my USB scanner, printer and camera and UserLinux doesn't.
UserLinux made the extraordinary Debian software and package environment accessible without the without the inadvertent and uncontrolled negative Gurella marketing presence that has undermined the mainline Debian distribution.
I'll tell you a UserLinux story:
Back in the days when Red Hat stopped selling a $50 box I started looking for another Distro.
I tried Knoppix and an interesting thing happened when I mistakenly visited the Debian IRC chat seeking assistance. I was treated with gruff and rude dismissal.
What I think was going on was somebody was engaging in Gurella Dis-marketing. Whoever this was, it was someone deliberately making sure that anybody exploring Debian got a good bitter mothful of rejection. The people in the IRC chat were hybrid child-professional assholes. People who projected a veneer of competency, and had nothing to say except "go away".
So after that, UserLinux looked like a really nice bunch of people with a reasonable tolerance for my interests.
UserLinux has a picture of a folded paper airplane reflecting that it was a careful selection of the best of breed applications from the huge Debian package universe. Unlike Debian it didn't make you "Figure this out if you want to install this software"
The target client for UserLinux was a "business desktop". The charm of the distribution was it installed like gangbusters and you could add anything you wanted from Debian.
I joined the UserLinux project and I contributed a help file. For UserLinux I wrote a help file covering tasks like dual boot setup and Java installation.
So I'll say thanks Bruce Perens and also Linux Format British edition is an excellent Linux publication (sold at Borders Books I think).
The vulnerability of healthy young people to cytokine storm death is the same age group who I presume are recruited by extremist Islamic groups for suicide bombers.
The article is a flashing alarm light that will trigger a thousand entrepreneurial projects. Everything from fancy face masks to genuine medical developments.
One thing we need now is a 2 second test for the h5n1 virus. Something like a tagged antibody and an analyzer driven by an led laser operating on a 1 micrometer nasal droplet. (Make a proton magnetometer with some 1/4" neodymium magnets?)
Another thing we need is a medicine for dealing with the cytokine storm secondary cause of death.
There is quite likely to be a whole spectrum of instrumentation and medicine created in response to this potential epidemic.
Now, as these instruments and medicines are developed, many of these creations will fall under the patent and copyright system.
The problem is the patent and copyright system under a guise of law makes it easy to price lifesaving medicines and instruments instantly out of the reach of low income people.
Note, there is already a problem with patented Aids medications and copyrighted medical information being over priced for some AFrican countries.
So, the United States (and other countries) might get in the position of having the technology for saving lives patented, copyrighted and offered for sale, The price will typically be all the market will bear, as seen by diners eating breakfast on the 86th floor of the World Trade Center. Uhh whoops, on the second floor of some conference room in Palo Alto.
A global epidemic may provide a basis for the most intense finding of moral and religious condemnation against all of Western society that adheres to the patent and copyright system as it is without any component of mercy and local reasonableness in price and availability of patented and copyright materials.
Or else, the patent and copyright system should be immediately ammended to allow ideas and innovations to push out and propagte freely.
Following older slashdot postings about Dr. Kay's Squeak language I decided not to download it because it is not GPL'ed.
I hope Guido van Rossum and Larry Wall eagerly absorb Alan Kay's ideas (especially "write a language in itself") and move the ideas out into the accessible world of GPL'd free documentation, and programs.
The non-GPL publication universe is a get rich quick lime pit.
First, if you are migrating to Linux, check out this IBM "redbook". Very helpful for a business:
h tm l
http://www.redbooks.ibm.com/abstracts/sg246380.
----------
Among the various distributions that you might use, you can also add UserLinux, see
http://userlinux.com
----------
Userlinux is a subset of the Debian distribution.
The applications within UserLinux are specifically selected to provide a useful Linux business desktop.
The team assembling UserLinux includes Bruce Perens and a global list of administrators and users focusing on assembling a useful business desktop.
I participate in the group. I have used UserLinux for several months. UserLinux has the elegance and power of Debian with a smaller, less redundant set of applications.
A VAT tax is the kind of simplification that would allow a radical simplification of the income tax.
Richard Nixon floated the idea of a VAT tax for the United States.
VAT is used in Europe. VAT is where each seller collects the VAT tax on her sales and deducts the VAT paid buying stuff for sale. Governments like it because it tends to be "self collecting". It gets revenue from the formerly "underground" economy. It can get revenue from trans national corporations that manipulate "transfer payments" to shift profits offshore.
As a previous poster has noted, the tax system needs an XML data structure and an XML description of the algorithms for the computation of tax.
I've daydreamed like this: Divide the task into an algorithm part and a sample data part. A python root module would emulate the "1040" form. All the child forms would be separate Python Modules corresponding to each schedule or tax form. The XML tagged test data would be included with each package of modules. The package of algorithms plus data would be GPG signed.
This scheme allows free and professional tax applications to co-exist. It enables modular development. It creates credible test data that can be used to cross validate alternate tax packages. It enables trusted open source.
My compromise for fastest, cheapest and best tax filing, given the absence of open source software is: First I get the paperwork together. All year long I ensure each deduction is written on a sheet of paper and put in a tax box. That is 90% of the work. Next I fill out the federal and state returns in pencil by hand. Finally type in the numbers from the paper forms to TaxActOnline.com. The really important check point is: Ensure that the online tax bill agrees to the penny with the paper form. Then resume seeking a better and more peaceful truth beyond these morally vacant warmakers.
I recommend a short view + long view approach:
For immediate support, build a library. Buy one book at a time as you need it.
(I think the Rute book and an anthology of the Linux Documentation Project are good cornerstone books.)
Balance your library with appropriate college studies. Whatever aspect of computers you work with, there is much worthwhile material to master and develop at length under the leadership of teachers.
Training does not have lasting value.
I took a Red Hat CNE course at my own expense and failed the exam. That training has never had a net positive value ever in a job interview. I simply blew a week trying to absorb what I really would have rather learned in an academic setting where thinking and understanding are valued. The exam was figuring out some stupid dorm tricks and doing a whole bunch of editing real fast.
So the situation is patents fill up the graphics software landscape. The following tiers of software are kept as business secrets.
As I note from the HP patent cross licensing email a month ago, the big players cross license their software patents.
Looking at the patent problem some more: The problem is in the relatively uncontrolled enforcement and royalty collection practices.
What do we have in the way of institutions and social practices that model patent licensing?
Well I can think of Wartime. The government ignores patents.
And video tape copying, where a copying is allowed.
And there is Fair use. I don't know that there is any fair use of a patent.
To focus on part of what Bruce Perens says: "...users can be sued..."
Seems to me that "...users can be sued..." is the crux of why the Microsoft patent blitz is so depressing and menacing.
------
Here is my short list of anti software patent responses:
1. Seek legislative redress.
a.I am looking for a politicial action strategy . b.I need slogans and "3 second sound bites" to accompany donations.Legislators need handles to grip to freedom issues involved.
2. Model a reform act on the Automotive Right to Repair Act. It has a great appeal to the public good in its' text.
3. Write a patent claim engine to systematically generate and publish every conceivable software patent claim. Patent the engine and license it so everything it creates is GNU licensed. Have reviewers reseed the engine with key phrases so the engine eventually carves out a substantial software space that is free. Furnish copies of the engine so specialist developers in particular areas can can carve up free specialty areas.