When I was young, we used RCS and we liked it! As the state of the art changes, so do the requirements to stay at the top.
Yes, but it is also important to use the right tool for the job. In this case, it sounds to me like rcs might be that right tool. No central server or repository needed, just a little wrapper or plugin for whatever editor that puts the file into the archive (stored in the same dir with the file) after every edit.
I'm also rather unhappy at SVN this week because it managed to get itself in a horribly confused state in which it told me I needed to run svn cleanup to fix some locks, but running svn cleanup just got me an error message saying that I needed to run svn cleanup to fix the locks. I ended up having to delete and do a clean checkout, and was not impressed.
BTDT, but if I blew away my 10G sandbox I wasn't going to get anything done for a long time. Turned out to be fairly easy to fix. First an svnclean and then svn cleanup. Careful attention to the error messages led me to the problem. Can't remember if it was missing tmp dirs or what. I've had sandbox problems enough times to know svn is far from perfect, but my sandbox has been much more solid since I uninstalled tortoise and just use the plain vanilla cmdline client (from the apache package). Now windows uses more of my linux finger memory, and less problems with data integrity.
Personally I do not recommend products that do not support Linux/FreeBSD. Because I use those operating systems? That's part of it yes, but mainly because Linux tends to expose crappy products. Look at the board in question here: "Badly written table". I have yet to see a product where they cut corners in ONE place only.
That is exactly right. And if they would cut corners and screw up in something as easy to test and fix as a flash BIOS, what are they going to do when faced with a hard problem?
The last socket 939 systems I built (myth tv) had foxconn boards because they were cheap in a closeout bundle from newegg. It was the first time I had used foxconn and they seem fine, but this...
Foxconn refusing to fix because of linux totally negates my formerly positive impressions of the company.
...as all applications on a typical users system have been vetted by a distribution team, and code is only going into those applications with the approval of project managers, the chance of a linux user winding up with a dangerous application on their system is diminishingly small. There's more to it than that. The officially installed software isn't the problem. It isn't just "a linux user" with a problem. It is every linux system whether that means a big server or system with one user account OR any system with any hole in anything running on that system. Anyone who was given or is stealing user-level access to a system could put a program onto that system and become root.
Power draw may fluctuate depending on usage, but expect a relatively constant power requirement. Yes, of course. Because we all know there is no difference between max and min power consumption on a processor or a complete system, even a laptop, so why should there be a difference between max and min consumption on an access point which really is a complete computer system with processor, storage, memory and I/O?
Just because a given access point will use "up to 18watts" does not mean it will always use 18watts. As long as the average power consumption is under what the supply can deliver, all can be well. Several possibilities exist.
Best case is that the supply can deliver short term bursts of power sufficient to meet the demand. This is realistic in most scenarios today, as supplies are typically rated by long-term average power. For example, the 20amp breaker on your typical home circuit will easily supply a few seconds at 30amps. My irrigation pump is rated at 28amps, the breaker at 50amps, and the pump draws over 100amps every time it starts. No problems.
Worst case is the access point requires some local power storage to meet the demand. This might take the form of a supercap or rechargeable battery. Average power usage keeps the battery charged, and peak demands are met from the battery. Only if the battery becomes depleted does the access point have to limit power usage -- perhaps by limiting transmit power or by limiting speed.
the main problem is propertory connectors. there are a grand total of 4 USB connectors. A, B, mini-b, micro-b. But don't forget that each of those connectors come in male and female, making 8 USB connectors.
Now look at the variety of cable types needed...
Remember that while the host is supposed to have female A, some peripheral devices also have female A, making maleA:maleA cables needed. And people need or want extension cables, maleA:femaleA.
So while in theory three cables should be sufficient to connect all peripherals using standard connectors, reality differs.
I sure wouldn't call it suffering. I'd say we are enjoying a huge trade deficit. As another pointed out, we get to trade dollars (which we "print" at very low cost) for tangible stuff. What's not to like?
Suffering comes later, if/when everyone else decides they have enough of our dollars. Then either they won't take any more, or they will send back the ones they have and buy out our entire country.
Twenty years ago I used a Tandy Model 100. Decent keyboard, way too small a display, no moving parts, fairly small and light, and would run a couple of long days on 4xAA batteries. It also had functional applications and a modem built in. Reporters, etc. used them by the thousands. This might actually be a nearly ideal replacement.
It has a LOT more functionality in a reasonable package.
If they place it randomly, this means that on average, they have to download HALF of the file in order to find the watermark.
But that means when the watermark does NOT exist, they have to download and process the entire file. Since most files do not (yet) have the watermark, that task seems untenable so the random location watermark is equally untenable.
Intel can get 300 million transistors right on something the size of a postage stamp, I expect better from an LCD manufacturer.
It is actually a lot easier to get it all "right on something the size of a postage stamp" because the chance of finding a flaw in the silicon is directly proportional to the area.
Yup. I had one of the original hotmail accounts from before M.S. and it still worked with webdav access even as a free account. That is, it worked until late December sometime when it went away. All of it, e-mail and everything.
When I logged in again just after the new year, it said "a hotmail account has been reserved for you" and I had to accept the new user TOS agreement. No more e-mail, no more webdav, no more of anything that had been familar a month before. New e-mail now comes in to that account, but...
original was for DOS, is still around, and able to do both POP/SMTP
As someone who used the DOS version since 1991 (and I removed it only a few months ago after several years of running it in parallel with WinPMail), I'm quite certain it did neither POP nor SMTP. However you could get gateways from 3rd parties that would gateway the DOS pegasus via POP and SMTP.
I prefer hamster (nntp and imap server). I haven't looked at mercury for a couple of years, but I've used hamster many times to do such conversions and it is trivial to set up for the job.
my wife uses pegasus exclusively, and she won't be happy about switching.
In an interesting coincidence, in December I switched my wife from Pegasus to Thunderbird. She was the last person I support (personally or professionally) who used Pegasus.
I did the switch myself about 2 years ago. I switched my inlaws about 6 months ago. Work switched to outlook (bluck) when they finally decided that e-mail was worth officially supporting.
Pegasus had just withered on the vine. I had been using it since long before the Windows version. Progress was always slow, but the last few years the quirks and bigger and bugs were lasting longer and longer. I was one of those anxiously hoping David could see his way clear to releasing the source code, but now I cannot see that making any difference. Five years ago it would have been great. Two years ago it might have worked. Now? It might help some other project, but I doubt it.
Mounting the partition read only is easily overcome as soon as they have root. Change/etc/fstab, unmount, and remount and the partition is now read-write.
Or maybe by "mount -o remount,rw {partition}" ?
A dedicated syslog server is always a good idea if it fits the budget.
Sure, anything to get the logs off the compromised system is a good idea. Even piping them out to an attached printer, or an RS232 data logger, or WORM drive, or another server, or a workstation, or...
Some areas can get really cheap pre-paid old-tech these days. But without a location, I'll assume you want something that could be just about anyplace in the U.S.
If you need the best national coverage, it will end up costing you at least $8/month to keep alive a plan from. (That's CDMA or TDMA... Tracfone has a newer setup using GSM, but that will cost more money and much worse coverage.) Get a referral from someone before activating, you'll get free minutes and so will they. Starter kit with phone and a few minutes will cost you $20 on up. That's about the cheapest way to get started, but beware that you have to use their phone, and their phones only work with tracfone.
The cheapest national option I've found is if t-mobile has prepaid coverage where you need it (most metro areas, interstate highways, etc). The coverage map on is really good, but do not confuse the prepaid map with the post-paid contract map. Buy a starter kit for about $30-40 (walmart or after rebate, better if you watch slickdeals or fatwallet) or more depending on what phone you want (any t-mobile or unlocked GSM phone with U.S. frequency bands will work if you just buy a prepaid sim on e.g. e-bay), and buy a $100 refill (1000 minutes, use a coupon at e.g. staples or online discount reseller and get it for $80). Those minutes will last for a year, so under $9/month (plus the phone) for 1000 minutes total. It's a HUGE win in year two IFF you don't need minutes. A $10 card will keep your minutes alive for another year, or under $0.10 per month.
P.S. Wife and I have t-mobile prepaid. Several members of my extended family have tracfone. I hate that t-mobile started charging for incoming SMS/MMS messages, but haven't found anything better enough to be worth switching or even to recommend instead.
Try again. You are stuck on section 6b being the only distribution clause in the LGPL. But look at 6a also.
Statically linked is linked at build time and as such it needs the LGPL for the library. This is covered by section 6a of the LPGL.
Dynamically linked means linked at run time, and as such even a GPL'd library is fine as long as you do not include GPL'd header files.
The LGPL is the workaround needed for for statically linked libraries. That is why the FSF created it. At the time, my company and others were asking for it. If we could have done dynamic linking (on DOS, MVS, and certain Unix(tm) platforms) in 1987, we wouldn't have needed the LGPL.
that's why they wrote the LGPL, the license that you CAN dynamically link proprietary code to.
Dynamic linking means it links at runtime. Static linking means it links at build (compile, link, etc) time.
The LGPL was originally created for STATICALLY linked libraries. Distribute your executable, plus the obj files for your executable which would be needed to relink with a new library, plus the sources for the LGPL'd library, and then you were in compliance with the LGPL. (This was the agreement reached in 1988-89 between the lawyers for the FSF and those for the company I worked for at the time.)
Contrast that with the GPL which would have required the sources (not just obj files) for your app.
Of course, Stallman and the FSF have become a lot more militant in the past nearly 20 years. I suspect it would be harder to reach an equivalent agreement today.
Despite a common misconception of the opposite, the speed of electric fields through copper is faster than light through fiber.
(~0.95 c for copper, ~0.65 c for fiber)
Wow. You are saying that a signal propagates thru copper over 46% faster than thru fiber. That's incredible!
Even more incredible given the normal figure for propagation thru copper ranges from 55% to 85% of c, with 75% being a normal approximation. This results in a more reasonable propagation advantage of 17% to copper instead of your 46%.
But to be fair, propagation speed isn't everything. If it were, we'd have scrapped all the copper and be exclusively using free-air LASER or even wifi.
- There's less need for signal boosters and repeaters - It's much easier to splice copper than fiber
Both of those are usually quoted as an advantage for fiber.
Granted, it is always a pain to splice fiber, but if you are talking a line for 10gb/s data it is no less of a pain to splice copper. And the difficulty in splicing fiber also makes it hard to secretly tap the line.
And re. signal boosters, etc... just compare running a high data-rate copper vs fiber cable over a few km's long link, think trans-oceanic.
There are loads of animals that these things get tested on before the jump is made to human testing. If animals die or are poorly affected, humans should not be tested. So there should be some idea about what the drug is supposed to do, and the given results make me wonder if the regular, long process was followed..
You should wonder. What with various "animal rights" organizations that imply (if not state outright) that animals are equal to humans, animal testing is being reduced.
Hmm, here's an idea... Step 1, minimal animal testing. Step 2, test on humans who claim to be equal to animals. Step 3, test on other humans.
Slashdot Mirror
When I was young, we used RCS and we liked it! As the state of the art changes, so do the requirements to stay at the top.
Yes, but it is also important to use the right tool for the job. In this case, it sounds to me like rcs might be that right tool. No central server or repository needed, just a little wrapper or plugin for whatever editor that puts the file into the archive (stored in the same dir with the file) after every edit.
I'm also rather unhappy at SVN this week because it managed to get itself in a horribly confused state in which it told me I needed to run svn cleanup to fix some locks, but running svn cleanup just got me an error message saying that I needed to run svn cleanup to fix the locks. I ended up having to delete and do a clean checkout, and was not impressed.
BTDT, but if I blew away my 10G sandbox I wasn't going to get anything done for a long time. Turned out to be fairly easy to fix. First an svnclean and then svn cleanup. Careful attention to the error messages led me to the problem. Can't remember if it was missing tmp dirs or what. I've had sandbox problems enough times to know svn is far from perfect, but my sandbox has been much more solid since I uninstalled tortoise and just use the plain vanilla cmdline client (from the apache package). Now windows uses more of my linux finger memory, and less problems with data integrity.
sdb
Personally I do not recommend products that do not support Linux/FreeBSD. Because I use those operating systems? That's part of it yes, but mainly because Linux tends to expose crappy products. Look at the board in question here: "Badly written table". I have yet to see a product where they cut corners in ONE place only.
That is exactly right. And if they would cut corners and screw up in something as easy to test and fix as a flash BIOS, what are they going to do when faced with a hard problem?
The last socket 939 systems I built (myth tv) had foxconn boards because they were cheap in a closeout bundle from newegg. It was the first time I had used foxconn and they seem fine, but this...
Foxconn refusing to fix because of linux totally negates my formerly positive impressions of the company.
sdb
That link, at 1:30pm EDT, for at least for windows en-us, has last weeks' RC build, NOT the 3.0 release.
Dropped the ball, they did.
...as all applications on a typical users system have been vetted by a distribution team, and code is only going into those applications with the approval of project managers, the chance of a linux user winding up with a dangerous application on their system is diminishingly small. There's more to it than that. The officially installed software isn't the problem. It isn't just "a linux user" with a problem. It is every linux system whether that means a big server or system with one user account OR any system with any hole in anything running on that system. Anyone who was given or is stealing user-level access to a system could put a program onto that system and become root.sdb
sdb
Just because a given access point will use "up to 18watts" does not mean it will always use 18watts. As long as the average power consumption is under what the supply can deliver, all can be well. Several possibilities exist.
Best case is that the supply can deliver short term bursts of power sufficient to meet the demand. This is realistic in most scenarios today, as supplies are typically rated by long-term average power. For example, the 20amp breaker on your typical home circuit will easily supply a few seconds at 30amps. My irrigation pump is rated at 28amps, the breaker at 50amps, and the pump draws over 100amps every time it starts. No problems.
Worst case is the access point requires some local power storage to meet the demand. This might take the form of a supercap or rechargeable battery. Average power usage keeps the battery charged, and peak demands are met from the battery. Only if the battery becomes depleted does the access point have to limit power usage -- perhaps by limiting transmit power or by limiting speed.
sdb
Now look at the variety of cable types needed...
Remember that while the host is supposed to have female A, some peripheral devices also have female A, making maleA:maleA cables needed. And people need or want extension cables, maleA:femaleA.
So while in theory three cables should be sufficient to connect all peripherals using standard connectors, reality differs.
sdb
I sure wouldn't call it suffering. I'd say we are enjoying a huge trade deficit. As another pointed out, we get to trade dollars (which we "print" at very low cost) for tangible stuff. What's not to like?
Suffering comes later, if/when everyone else decides they have enough of our dollars. Then either they won't take any more, or they will send back the ones they have and buy out our entire country.
sdb
RTFA !
Twenty years ago I used a Tandy Model 100. Decent keyboard, way too small a display, no moving parts, fairly small and light, and would run a couple of long days on 4xAA batteries. It also had functional applications and a modem built in. Reporters, etc. used them by the thousands. This might actually be a nearly ideal replacement.
It has a LOT more functionality in a reasonable package.
Battery life is iffy, but probably adequate.
Display seems OK. Sunlight is probably an issue.
But how is the keyboard, really?
sdb
If they place it randomly, this means that on average, they have to download HALF of the file in order to find the watermark.
But that means when the watermark does NOT exist, they have to download and process the entire file. Since most files do not (yet) have the watermark, that task seems untenable so the random location watermark is equally untenable.
sdb
Intel can get 300 million transistors right on something the size of a postage stamp, I expect better from an LCD manufacturer.
It is actually a lot easier to get it all "right on something the size of a postage stamp" because the chance of finding a flaw in the silicon is directly proportional to the area.
sdb
Yup. I had one of the original hotmail accounts from before M.S. and it still worked with webdav access even as a free account. That is, it worked until late December sometime when it went away. All of it, e-mail and everything.
When I logged in again just after the new year, it said "a hotmail account has been reserved for you" and I had to accept the new user TOS agreement. No more e-mail, no more webdav, no more of anything that had been familar a month before. New e-mail now comes in to that account, but...
sdb
original was for DOS, is still around, and able to do both POP/SMTP
As someone who used the DOS version since 1991 (and I removed it only a few months ago after several years of running it in parallel with WinPMail), I'm quite certain it did neither POP nor SMTP. However you could get gateways from 3rd parties that would gateway the DOS pegasus via POP and SMTP.
sdb
I prefer hamster (nntp and imap server). I haven't looked at mercury for a couple of years, but I've used hamster many times to do such conversions and it is trivial to set up for the job.
sdb
my wife uses pegasus exclusively, and she won't be happy about switching.
In an interesting coincidence, in December I switched my wife from Pegasus to Thunderbird. She was the last person I support (personally or professionally) who used Pegasus.
I did the switch myself about 2 years ago. I switched my inlaws about 6 months ago. Work switched to outlook (bluck) when they finally decided that e-mail was worth officially supporting.
Pegasus had just withered on the vine. I had been using it since long before the Windows version. Progress was always slow, but the last few years the quirks and bigger and bugs were lasting longer and longer. I was one of those anxiously hoping David could see his way clear to releasing the source code, but now I cannot see that making any difference. Five years ago it would have been great. Two years ago it might have worked. Now? It might help some other project, but I doubt it.
R.I.P.
sdb
I just thought of something that would make a great patent. A light switch that only stays on for a specified period of time, say 10 minutes.
You mean like the timer switches that are widely available?
sdb
Mounting the partition read only is easily overcome as soon as they have root. Change /etc/fstab, unmount, and remount and the partition is now read-write.
Or maybe by "mount -o remount,rw {partition}" ?
A dedicated syslog server is always a good idea if it fits the budget.
Sure, anything to get the logs off the compromised system is a good idea. Even piping them out to an attached printer, or an RS232 data logger, or WORM drive, or another server, or a workstation, or...
sdb
Some areas can get really cheap pre-paid old-tech these days. But without a location, I'll assume you want something that could be just about anyplace in the U.S.
If you need the best national coverage, it will end up costing you at least $8/month to keep alive a plan from. (That's CDMA or TDMA... Tracfone has a newer setup using GSM, but that will cost more money and much worse coverage.) Get a referral from someone before activating, you'll get free minutes and so will they. Starter kit with phone and a few minutes will cost you $20 on up. That's about the cheapest way to get started, but beware that you have to use their phone, and their phones only work with tracfone.
The cheapest national option I've found is if t-mobile has prepaid coverage where you need it (most metro areas, interstate highways, etc). The coverage map on is really good, but do not confuse the prepaid map with the post-paid contract map. Buy a starter kit for about $30-40 (walmart or after rebate, better if you watch slickdeals or fatwallet) or more depending on what phone you want (any t-mobile or unlocked GSM phone with U.S. frequency bands will work if you just buy a prepaid sim on e.g. e-bay), and buy a $100 refill (1000 minutes, use a coupon at e.g. staples or online discount reseller and get it for $80). Those minutes will last for a year, so under $9/month (plus the phone) for 1000 minutes total. It's a HUGE win in year two IFF you don't need minutes. A $10 card will keep your minutes alive for another year, or under $0.10 per month.
Check out the best prepaid plan comparison I've seen.
sdb
P.S. Wife and I have t-mobile prepaid. Several members of my extended family have tracfone. I hate that t-mobile started charging for incoming SMS/MMS messages, but haven't found anything better enough to be worth switching or even to recommend instead.
Try again. You are stuck on section 6b being the only distribution clause in the LGPL. But look at 6a also.
Statically linked is linked at build time and as such it needs the LGPL for the library. This is covered by section 6a of the LPGL.
Dynamically linked means linked at run time, and as such even a GPL'd library is fine as long as you do not include GPL'd header files.
The LGPL is the workaround needed for for statically linked libraries. That is why the FSF created it. At the time, my company and others were asking for it. If we could have done dynamic linking (on DOS, MVS, and certain Unix(tm) platforms) in 1987, we wouldn't have needed the LGPL.
sdb
that's why they wrote the LGPL, the license that you CAN dynamically link proprietary code to.
Dynamic linking means it links at runtime. Static linking means it links at build (compile, link, etc) time.
The LGPL was originally created for STATICALLY linked libraries. Distribute your executable, plus the obj files for your executable which would be needed to relink with a new library, plus the sources for the LGPL'd library, and then you were in compliance with the LGPL. (This was the agreement reached in 1988-89 between the lawyers for the FSF and those for the company I worked for at the time.)
Contrast that with the GPL which would have required the sources (not just obj files) for your app.
Of course, Stallman and the FSF have become a lot more militant in the past nearly 20 years. I suspect it would be harder to reach an equivalent agreement today.
sdb
Despite a common misconception of the opposite, the speed of electric fields through copper is faster than light through fiber.
(~0.95 c for copper, ~0.65 c for fiber)
Wow. You are saying that a signal propagates thru copper over 46% faster than thru fiber. That's incredible!
Even more incredible given the normal figure for propagation thru copper ranges from 55% to 85% of c, with 75% being a normal approximation. This results in a more reasonable propagation advantage of 17% to copper instead of your 46%.
But to be fair, propagation speed isn't everything. If it were, we'd have scrapped all the copper and be exclusively using free-air LASER or even wifi.
- There's less need for signal boosters and repeaters
- It's much easier to splice copper than fiber
Both of those are usually quoted as an advantage for fiber.
Granted, it is always a pain to splice fiber, but if you are talking a line for 10gb/s data it is no less of a pain to splice copper. And the difficulty in splicing fiber also makes it hard to secretly tap the line.
And re. signal boosters, etc... just compare running a high data-rate copper vs fiber cable over a few km's long link, think trans-oceanic.
sdb
BTDT, anyone remember Agenda Computing?
Oh, and while it did make it out of beta (officially if not functionally) the PrismIQ wasn't exactly a bundle of joy either.
sdb
I'm going to keep burning baby-seal-fat lanterns for my illumination. /sarcasm
<skip>sarcasm</skip>
You could make biodiesel out of that baby seal fat...
sdb
There are loads of animals that these things get tested on before the jump is made to human testing. If animals die or are poorly affected, humans should not be tested. So there should be some idea about what the drug is supposed to do, and the given results make me wonder if the regular, long process was followed..
You should wonder. What with various "animal rights" organizations that imply (if not state outright) that animals are equal to humans, animal testing is being reduced.
Hmm, here's an idea... Step 1, minimal animal testing. Step 2, test on humans who claim to be equal to animals. Step 3, test on other humans.
That could work!
sdb