Or it's just something they use as a cop out to drop the whole sharade. I.e they know they wouldn't win anyway but instead of just dropping the case and loose face they played the "if only we could tell you" card.
To be honest I have actually never ever received a "thanks" email as a reply to any notification that I have sent out so I'm questioning the validity of ACs claim that this actually happens, and if so that it happens in enough scale to justify a no-reply return address.
And a company the size of Microsoft probably have some e-mail address that goes to their first line of support somewhere in india. So why not use that address instead of the no-reply? It's not like the "Read the manual" questions won't be asked just because you use a no-reply in a notification mail.
If you notify your customers about something then you can rest assured that there are some of them that do not fully understand your notification or will have further questions. So they will mail and/or call you anyway via other means which is also why most service announcements that I have seen also contains a "for further questions send a mail to xxx@xxx.com", so the main question is why the reply address then is a no-reply instead of the xxx@xxx.com so that those people could just reply to the mail sent.
Which is why I always use our tech support address as the reply address instead of a no-reply when sending out notifications. But then if you happen to be a competitor in my field then please by all means continue with your no-reply practice.
Most companies probably fit your category. I see many "noreply@company.com" as the return address from service announcements from companies that we have an established relationship with (otherwise we would never get their service announcements in the first place).
Then there are the other kind that is practically worse, I'm the recipient on several mail delivery lists for mutual funds (yes ever changing excel sheets sent via mail are the #1 distribution method of mutual fund companies...) and it seams that it's standard there to have these delivery lists be open so as soon as some other (of the thousands of connected companies) writes a reply in order to question the distributor it's bounced to all the members, at which point some one else replies with "please don't send this to me" which bounces yet another round, and so it goes.
I hear you but you basically need something like core memory for that and unfortunately that is way too slow for today's processors. But yes a very small battery backup in the PSU that would allow the system to shut-down orderly on a power outage would go a long way, the problem here is that with current battery technology the lifetime of such a small battery pack is quite limited so it would need to be switched after a few years, of course not that a big deal really.
There is really no reason why you shouldn't be able to write in a document, experience a power outage, boot the computer, load the document and be able to continue right where you left of, including full undo support back to the first edit to the document.
I can see why you where initially confused since I kind of worded it badly, but I never meant it to be a power outage at the commit, if it had I would never expected a full recovery:)
No, no one in this thread talked about a power outage during a transaction until you brought it up with your previous post. That a transaction will be lost if it happens during a power outage is a no brainer so I don't even know why you keep arguing this, or you like to kick in open doors?
Something tells me that you read only parts of other peoples posts before you post yourself, because "my example" with rename was not my solution, it was my description of how i.e word processors does it today and was part of my explanation on why it does not work (remember here that your claim was that how they currently work is just fine and all a user had to do was to change the settings of said application) so why you reply back with "it does not work" when that was exactly my point is just strange.
Instead of reading your books you should perhaps take some time to read the actual posts that you are replying to, that should lessen your confusion greatly.
The only mention of a power outage during a commit was this particular reply from you but somehow you think that I should read a book about it? Does this book contain information on how to read post that are not yet posted?
And btw great work to just ignore the bigger picture, didn't I mention that I talked about all softwares and not just word processors in the very text you quoted? Autosave in a word processor does not save full undo/redo history, autosave in a word processor does not save in commit fashion but in a two step dance (save as tmp, rename to real filename)
No, the very last commit might be currupted. So in 99% (and probably more) of cases you should be able to continue even after a power outage. However losing just the last changes which should, for a word processor, be just a few sentences or characters maximum, is a far cry from todays systems where you loose everything since the last time you pressed save. Now some word processors does an autosave every 5 or 10 minutes but that is way to seldom (and I'm talking about all and every computer program and not just word processors) and it also AFAIK only save the text as it is and not the full undo/redo history among other things.
No it's not impossible, even back in the 80:s it was possible to store files such that only the very last commit could be corrupted in a power outage. Yes HD/SSD:s does not always tell the truth about fsync but it still does not make it impossible to avoid the major data loss that you see with todays systems.
Because the old tech (i.e the typewrites) worked that way, you typed and your work was 'stored'. His problem was that he thought that new tech (i.e a computer) would be like the old tech but better. Naturally he was wrong, which should be telling about how bad computers actually are.
Of course you could use the SS admin tool to unlock a file, but that required admin rights which you would never get as a plain dev there. So you had to file a unlock request and hope that it would be solved some day in the future. For all the warts with CVS it is nowhere as much trouble as SS. I think that one of the main problems with SS is that it used the windows file share as it's transport protocol so if it was kind of usable for you then perhaps you didn't have to use it over a shared network with other devs?
Actually thinking about it all things with a computer is crazy insane interface wise. I.e why is the power cable so easily to pull out when VGA, Serial and Parallel cables was screwed to the motherboard?
I know of one instance back in the 80:s when a company switched from typewrites to computers and after a person had written some text and closed the word processor and "do you want to save the document?" appeared he clicked "no" because he thought that it wanted to create a copy since logically the text was stored since he had written it. We today might think that he was stupid but I think that he was right and it's we that are stupid. Why wouldn't a modern computer simply store each and every change to a document automatically?!
The same with this guy, why are files deleted and not always under version control. They could still be counted as free storage and it's location could be recycled if the real free storage is insufficient once he does a save of a new file.
There is really no reason why you shouldn't be able to write in a document, experience a power outage, boot the computer, load the document and be able to continue right where you left of, including full undo support back to the first edit to the document.
Slashdot Mirror
Downhill? I thought it went under water.
Yeah, especially on a laptop one would think that this should have been fixed ages ago.
I have a Samsung SmartTV and it's basically 1-2s from start to picture (sound is almost instant).
Or it's just something they use as a cop out to drop the whole sharade. I.e they know they wouldn't win anyway but instead of just dropping the case and loose face they played the "if only we could tell you" card.
To be honest I have actually never ever received a "thanks" email as a reply to any notification that I have sent out so I'm questioning the validity of ACs claim that this actually happens, and if so that it happens in enough scale to justify a no-reply return address.
And a company the size of Microsoft probably have some e-mail address that goes to their first line of support somewhere in india. So why not use that address instead of the no-reply? It's not like the "Read the manual" questions won't be asked just because you use a no-reply in a notification mail.
If you notify your customers about something then you can rest assured that there are some of them that do not fully understand your notification or will have further questions. So they will mail and/or call you anyway via other means which is also why most service announcements that I have seen also contains a "for further questions send a mail to xxx@xxx.com", so the main question is why the reply address then is a no-reply instead of the xxx@xxx.com so that those people could just reply to the mail sent.
Which is why I always use our tech support address as the reply address instead of a no-reply when sending out notifications. But then if you happen to be a competitor in my field then please by all means continue with your no-reply practice.
Most companies probably fit your category. I see many "noreply@company.com" as the return address from service announcements from companies that we have an established relationship with (otherwise we would never get their service announcements in the first place).
Then there are the other kind that is practically worse, I'm the recipient on several mail delivery lists for mutual funds (yes ever changing excel sheets sent via mail are the #1 distribution method of mutual fund companies...) and it seams that it's standard there to have these delivery lists be open so as soon as some other (of the thousands of connected companies) writes a reply in order to question the distributor it's bounced to all the members, at which point some one else replies with "please don't send this to me" which bounces yet another round, and so it goes.
I hear you but you basically need something like core memory for that and unfortunately that is way too slow for today's processors. But yes a very small battery backup in the PSU that would allow the system to shut-down orderly on a power outage would go a long way, the problem here is that with current battery technology the lifetime of such a small battery pack is quite limited so it would need to be switched after a few years, of course not that a big deal really.
Yes, all trucks in Europe must carry a black box that records everything.
Since you linked to the article I demand, !DEMAND!, that some one forks your post!
And that author was me, what I did write was:
There is really no reason why you shouldn't be able to write in a document, experience a power outage, boot the computer, load the document and be able to continue right where you left of, including full undo support back to the first edit to the document.
I can see why you where initially confused since I kind of worded it badly, but I never meant it to be a power outage at the commit, if it had I would never expected a full recovery :)
Hope he doesn't forget to put on the dark blue hanky as well!
No, no one in this thread talked about a power outage during a transaction until you brought it up with your previous post. That a transaction will be lost if it happens during a power outage is a no brainer so I don't even know why you keep arguing this, or you like to kick in open doors?
Something tells me that you read only parts of other peoples posts before you post yourself, because "my example" with rename was not my solution, it was my description of how i.e word processors does it today and was part of my explanation on why it does not work (remember here that your claim was that how they currently work is just fine and all a user had to do was to change the settings of said application) so why you reply back with "it does not work" when that was exactly my point is just strange.
Instead of reading your books you should perhaps take some time to read the actual posts that you are replying to, that should lessen your confusion greatly.
The only mention of a power outage during a commit was this particular reply from you but somehow you think that I should read a book about it? Does this book contain information on how to read post that are not yet posted?
And btw great work to just ignore the bigger picture, didn't I mention that I talked about all softwares and not just word processors in the very text you quoted? Autosave in a word processor does not save full undo/redo history, autosave in a word processor does not save in commit fashion but in a two step dance (save as tmp, rename to real filename)
Hardly, he designed a new bumper. Some one later built a car, used his design for the bumper and used the car in a robbery.
No, the very last commit might be currupted. So in 99% (and probably more) of cases you should be able to continue even after a power outage. However losing just the last changes which should, for a word processor, be just a few sentences or characters maximum, is a far cry from todays systems where you loose everything since the last time you pressed save. Now some word processors does an autosave every 5 or 10 minutes but that is way to seldom (and I'm talking about all and every computer program and not just word processors) and it also AFAIK only save the text as it is and not the full undo/redo history among other things.
Don't think they have:
f.ultra@ubuntu:~$ svn help | grep obliterate
f.ultra@ubuntu:~$
No it's not impossible, even back in the 80:s it was possible to store files such that only the very last commit could be corrupted in a power outage. Yes HD/SSD:s does not always tell the truth about fsync but it still does not make it impossible to avoid the major data loss that you see with todays systems.
Because the old tech (i.e the typewrites) worked that way, you typed and your work was 'stored'. His problem was that he thought that new tech (i.e a computer) would be like the old tech but better. Naturally he was wrong, which should be telling about how bad computers actually are.
You just ignored that I wrote that it was 15 years ago?
Of course, I was just ranting about the complete disaster that is computer interfaces regardless of operating system or vendor.
Of course you could use the SS admin tool to unlock a file, but that required admin rights which you would never get as a plain dev there. So you had to file a unlock request and hope that it would be solved some day in the future. For all the warts with CVS it is nowhere as much trouble as SS. I think that one of the main problems with SS is that it used the windows file share as it's transport protocol so if it was kind of usable for you then perhaps you didn't have to use it over a shared network with other devs?
Actually thinking about it all things with a computer is crazy insane interface wise. I.e why is the power cable so easily to pull out when VGA, Serial and Parallel cables was screwed to the motherboard?
I know of one instance back in the 80:s when a company switched from typewrites to computers and after a person had written some text and closed the word processor and "do you want to save the document?" appeared he clicked "no" because he thought that it wanted to create a copy since logically the text was stored since he had written it. We today might think that he was stupid but I think that he was right and it's we that are stupid. Why wouldn't a modern computer simply store each and every change to a document automatically?!
The same with this guy, why are files deleted and not always under version control. They could still be counted as free storage and it's location could be recycled if the real free storage is insufficient once he does a save of a new file.
There is really no reason why you shouldn't be able to write in a document, experience a power outage, boot the computer, load the document and be able to continue right where you left of, including full undo support back to the first edit to the document.
Except of course that the Winnebago story is just an urban legend. http://www.heraldtribune.com/n...