When I worked at York (yorku.ca) both Computer Science and Science used labs of X-terminals for teaching. The administration costs were low, and the results were very good.
For upper-year CS students who needed to run compute-intensive apps, we still used workstations, so they'd only hurt themselves when they write a runaway cpu-hog (;-))
Another institution we knew, in Holland, used X-terms and the Win4Lin terminal server product to provide Windows apps to their Sunray clients,
at a very low cost.
If you do go the Xterm/Sunray route, you will need a service machine in the lab to stick CDs into, preferably on the desk of the lab monitor. If you gut your lab PCs and run LTSP and RDesktop, you can probably skip that step.
Because MS tried so very hard to embrace and extand Java, it's reasonable to expect that
the Sun lawyers are going to be very reluctant to change anything, for fear of
starting the war back up.
Sun
engineers, on the other hand, arguably agree
with your desire to move to free reference implemntation, and in the short term,
to fewer restrictions on the JRE.
stjobe wrote:Secundo: They don't want their software to interoperate with anything else than their software.
Interestingly, Microsoft initially wrote public RFCs about the protocol,
licenced it to AT&T, sponsored the plugfest
initialy, and, if I reccollect correctly, flew
Andrew to the first one or two.
It's interesting they have backed
so far from what they once proposed as
"Common Internet File Service" (CIFS).
Arguably, their tactic was to embrace
the IBM-/Microsoft-developed protocol,
now it's to extend and deny.
I'm not sure that's a good thing to be doing in front
of a court of law (:-)).
Interesting, but it doesn't go back far enough!
Back in the dawn of time, a colleague showed me
the mail option in ftp (!), before sending me
off to write GCOS Internet Mail in my choice
of B or C (;-))
If you have to shut down your good OS to run one
application on the bad OS, how long will it take
for you to slip and use some other bad-os application?
Like the beloved-to-crackers windows email client!
Once upon a time, my director had a financial
planning application that was supposed to make
his life easier... but it was a bear to use.
A study some years later showed that the people
who used the financial planner the most had the worst financial performance!
We figured it was because it was taking up the time
they should be spending on all the other kinds
of planning, not to mention the rest of their work.
A really boring machine on my desk,
or an x-terminal (ok, sunray),
well maintainewd, and with a
copy of the root password in
a sealed envelope pinned to the
sysadmin's desk.
If i can just ignore the machine
and use it to access my lab mchines,
then I'm happy. If I can't, and
have to make emergency repairs,
I can always rip open the envelope.
This, by the way, is how I ran
a department at Siemens, with
considerable success.
This system is far below the standards
I had to meet for an orange-book system fifteen
years ago, when B1 was about the minimum standard
you needed to keep confidential data confidential
and be able to have an audit trail that tracked access.
It's a lock on a glass door. Worse,
it's one with an alarm that dispatches a trigger-happy security guard, who will shoot
anyone, including the owner, if they happen to be
around when he arrives.
No, the repository won't let you have
the document because, although you can
be authenticated, you're trying to move
the document to an uncontrolled
electronix device
at a non-TEMPEST-shielded location.
You've acheived autheniation,
and arguably need-to-know, but
you've failed mandatory access
control, trusted path, labelling
and covert channel prevention.
You have
nothing like the security of the repository,
so you don't get the document.
This is somewhat similar to the model libraries
(have to) use, for privacy reasons.
The library's circulation system tracks the loan of
copy 42 of book A to user davecb, until such
time as either
it is returned, or
the user pays for its replacement.
After that, they are required by the laws of most
countries/states to delete the information, and
by all countries to report one circulation transaction
completed. The library then gets a grant based on the
number of transactions, etc.
Net result is that they retain only statistical information,
save during the period the book's "out". Usually
to weeks to 30 days.
Some countries attempt to discover what books
individuals are reading: those countries easily
discover what's still on loan by serving a
subpoena, but have to keep re-serving it again and again
to track usage over time.
For important problems we're trying to solve, my employer
uses the same approach as the navy: handoffs.
As the morning-shift engineer in Boston approaches
the end of his day, he calls the afternoon-shift
engineer in San Francisco, breifs him up to date on
progress he's made, and sends him his test harness.
When the San Franciscan gets ready to quit, he hands off
to his collegue in Bangladore, and when the torpedoman
gets ready to go to bed, he hands off to Grenoble.
My Boston collagues eventualy wakes up, wanders
in to work, and gets breifed by the chap from Grenoble.
Net result? For problms where the approach is inherently
a single-threaded investigation, we get 24 hours of
work a day, less a few hours of discussion.
For tasks that have multiple threads of development going
on at the same time, and therefor require coordination rather
than handoffs, this doesn't help much.
Perhaps learning to go months without
sleep, like Asok in Dilbert, is the best approach here (;-))
Google said While your data is automatically deleted from our servers, you can use the Clear my Files from Google button to manually remove all your files from Google Desktop servers. Note that if these files haven't yet been copied to your other computers, clicking this button will prevent you from finding them when you search from your other computers. The files will, of course, still be searchable from their computer of origin.
This implies that the files only need to be on a google
server for as long as it takes to build the index. It could
delete the source even before it copies the index to the
user's other machines.
Indeed, they probably do, and in the RIM case
make it much more likely that there is a
contract that can at least be deemed to make
U.S. patents appropriate.
--dave
PS: thanks for that suggestion, it makes some
parts of the RIM dispute much clearer!
You're assuming, as does China, etc, that conecting
a program in Uzbekistan from Patentistan makes the
Uzbek program subject to Patanti law.
This is one of the questions that the discussion raises:
is it true? If so, why? And if not, why not?
There is already some jurisprudence in the U.S. that
says no, for the specific case of a web browser, and
for the region in which the court has jurisdiction
over. Other cases and jurisdictions have yet to be heard from.
My customer uses a patented machine
and process in Uzbekistan (aka Canada),
for which I have an Uzbek (Canadian) patent.
Were they downloading and running something
on their machine that requires a U.S. patent,
the question changes.
Which, as you see, brings us back to the
original "foreign web page" discusson,
where the U.S., Germany or these days China claim
jurisdiction over anything viewed by their
citizens, often because the user copied it to
local video memory as part of viewing it.
Slashdot Mirror
For upper-year CS students who needed to run compute-intensive apps, we still used workstations, so they'd only hurt themselves when they write a runaway cpu-hog (;-))
Another institution we knew, in Holland, used X-terms and the Win4Lin terminal server product to provide Windows apps to their Sunray clients, at a very low cost.
If you do go the Xterm/Sunray route, you will need a service machine in the lab to stick CDs into, preferably on the desk of the lab monitor. If you gut your lab PCs and run LTSP and RDesktop, you can probably skip that step.
--dave
Sun engineers, on the other hand, arguably agree with your desire to move to free reference implemntation, and in the short term, to fewer restrictions on the JRE.
--dave
Java, due to MS's efforts to subvert it, is probably the hardest to free up, but this is a good, workmanlike step in the right direction.
--dave
Interestingly, Microsoft initially wrote public RFCs about the protocol, licenced it to AT&T, sponsored the plugfest initialy, and, if I reccollect correctly, flew Andrew to the first one or two.
It's interesting they have backed so far from what they once proposed as "Common Internet File Service" (CIFS).
Arguably, their tactic was to embrace the IBM-/Microsoft-developed protocol, now it's to extend and deny. I'm not sure that's a good thing to be doing in front of a court of law (:-)).
--dave
--dave
Like the beloved-to-crackers windows email client!
--dave
A real filesystem and an MMU are worth a lot!
--dave
A study some years later showed that the people who used the financial planner the most had the worst financial performance! We figured it was because it was taking up the time they should be spending on all the other kinds of planning, not to mention the rest of their work.
--dave
If i can just ignore the machine and use it to access my lab mchines, then I'm happy. If I can't, and have to make emergency repairs, I can always rip open the envelope.
This, by the way, is how I ran a department at Siemens, with considerable success.
--dave
--dave
This system is far below the standards I had to meet for an orange-book system fifteen years ago, when B1 was about the minimum standard you needed to keep confidential data confidential and be able to have an audit trail that tracked access.
It's a lock on a glass door. Worse, it's one with an alarm that dispatches a trigger-happy security guard, who will shoot anyone, including the owner, if they happen to be around when he arrives.
--dave
You've acheived autheniation, and arguably need-to-know, but you've failed mandatory access control, trusted path, labelling and covert channel prevention.
You have nothing like the security of the repository, so you don't get the document.
--dave (former professional paranoid) c-b
The library's circulation system tracks the loan of copy 42 of book A to user davecb, until such time as either
After that, they are required by the laws of most countries/states to delete the information, and by all countries to report one circulation transaction completed. The library then gets a grant based on the number of transactions, etc.
Net result is that they retain only statistical information, save during the period the book's "out". Usually to weeks to 30 days.
Some countries attempt to discover what books individuals are reading: those countries easily discover what's still on loan by serving a subpoena, but have to keep re-serving it again and again to track usage over time.
--dave
It's a commercial product from Siemens, which I used years ago for Sietec's large-scale imaging product.
There is probably a Linux port: We ran it on almost everything in existance (;-))
--dave
As the morning-shift engineer in Boston approaches the end of his day, he calls the afternoon-shift engineer in San Francisco, breifs him up to date on progress he's made, and sends him his test harness.
When the San Franciscan gets ready to quit, he hands off to his collegue in Bangladore, and when the torpedoman gets ready to go to bed, he hands off to Grenoble.
My Boston collagues eventualy wakes up, wanders in to work, and gets breifed by the chap from Grenoble.
Net result? For problms where the approach is inherently a single-threaded investigation, we get 24 hours of work a day, less a few hours of discussion.
For tasks that have multiple threads of development going on at the same time, and therefor require coordination rather than handoffs, this doesn't help much.
Perhaps learning to go months without sleep, like Asok in Dilbert, is the best approach here (;-))
--dave
That means google can delete it as soon as it is delivered to the rest of your other computers.
--dave
This implies that the files only need to be on a google server for as long as it takes to build the index. It could delete the source even before it copies the index to the user's other machines.
If true, expect a fix tomorrow (:-))
--dave
Because we put the drywall up ourselves the hourly rate was a bit higher than the cubicle-mechanics, but the quality was way better.
This was using steel bracketing and standard drywall & drywall screws... The month after we painted them pretty (non-beige) colors.
--dave
--dave
Participants who use two-factor authentication, and supposedly there are a few, only need one token card.
See http://www.projectliberty.org/
--dave
--dave
--dave
--dave
PS: thanks for that suggestion, it makes some parts of the RIM dispute much clearer!
This is one of the questions that the discussion raises: is it true? If so, why? And if not, why not?
There is already some jurisprudence in the U.S. that says no, for the specific case of a web browser, and for the region in which the court has jurisdiction over. Other cases and jurisdictions have yet to be heard from.
--dave
Were they downloading and running something on their machine that requires a U.S. patent, the question changes.
Which, as you see, brings us back to the original "foreign web page" discusson, where the U.S., Germany or these days China claim jurisdiction over anything viewed by their citizens, often because the user copied it to local video memory as part of viewing it.
--dave