Undo files

A coworker of mine just lost a couple of files he had been working on.

Apparently, in the IDE we use, if you copy and paste files through the explorer-like tree that shows the project's working directoy, and then inocently press "Ctrl+Z" while your mouse is hovering the tree, the IDE will erase those files.

Never mind if you had been furiously typing and doing some heavy work on those "copied" files for the past few minutes..

RobIII

So, what's the name of that IDE so we can avoid it?

tgape

@Zecc said:

A coworker of mine just lost a couple of files he had been working on.

Apparently, in the IDE we use, if you copy and paste files through the explorer-like tree that shows the project's working directoy, and then inocently press "Ctrl+Z" while your mouse is hovering the tree, the IDE will erase those files.

Never mind if you had been furiously typing and doing some heavy work on those "copied" files for the past few minutes..

I recall an IDE like that... except :s/minutes/hours, days, or weeks/. The only things that would protect those files from "Ctrl+Z" was exiting and restarting the IDE, or performing more actions in the explorer-like tree. (Once, I went on vacation for two weeks, shortly after having pasted a file. I left the app running, with the screen locked. About a day after I returned, having performed no file-level changes, I hit Ctrl+Z after having bumped the mouse without noticing it. Bam!)

I cannot remember the name of the IDE. However, IIRC, it was a proprietary, highly specialized program. As a general rule, one wouldn't accidentally wander into it without a change of job. At least, I'd assume such a gratuitous bug would show up in anyone's evaluation testing.

@RobIII said:

So, what's the name of that IDE so we can avoid it?

Zecc isn't sayingd.

Aaah.. aah... Shhtudio !!

Oh my, looks like I caught a cold.

Daid

Reminds me of the painful IAR embedded systems AVR IDE bug. In version 3 you could have the same file open twice, which resulted in "change is saved, changed is not saved"? actions when you pressed save all, or compile.
Version 4 release not long after that, which fixed the bug. Still, countless hours of work lost because of it.

Kemp

@tgape said:

(Once, I went on vacation for two weeks, shortly after having pasted a file. I left the app running, with the screen locked. About a day after I returned, having performed no file-level changes, I hit Ctrl+Z after having bumped the mouse without noticing it. Bam!)

You left your PC on for two weeks with no one using it? Glad I don't pay your power bill. Though those greenhouse gasses aren't going to help much...

@Kemp said:

@tgape said:
(Once, I went on vacation for two weeks, shortly after having pasted a file. I left the app running, with the screen locked. About a day after I returned, having performed no file-level changes, I hit Ctrl+Z after having bumped the mouse without noticing it. Bam!)

You left your PC on for two weeks with no one using it? Glad I don't pay your power bill. Though those greenhouse gasses aren't going to help much...

how much does it cost to run a computer for 2 weeks in sleep mode? Computer use 1-20 watts in sleep mode. Lets say his is bad and uses 50 watts. 50 watt/hours * 24 hours/day * 7 days/week * 2 weeks = 16800 watt hours is about 17 kilowat hours. The average cost of electricity here in the US is about 7-8 cents/kilowatt/hour. Lets say he lives in one of the most expensive states, like California where it is 12 So the total cost of running the computer that two weeks is 17 * 12 = $2.04.

Compared to the cost to the company of shutting down the computer. When tgape gets to work he starts up his computer. Lets say it takes 5 minutes to start up and be ready to go. Lets say tgape makes on the low end of programmers at about $50,000/year. That means the company's expense is probably about $80,000-$100,000/year. Lets say it is low ($75,000). Lets further say tgape has 13 holiday days, 2 weeks of vacation and 5 personal days. That means he works 52*5-(13+2+5) = 240 day/year. That means the company expense per day is $312.50. A work day is 8 hours (480 minutes). That means his time is worth $0.65/minute to the company. 5 minutes at 65 cents is $3.25. So he saved the company over a dollar using extremely conservative numbers.

I'm glad I don't pay your paycheck.

NSCoder

@tster said:

@Kemp said:
@tgape said:
(Once, I went on vacation for two weeks, shortly after having pasted a file. I left the app running, with the screen locked. About a day after I returned, having performed no file-level changes, I hit Ctrl+Z after having bumped the mouse without noticing it. Bam!)

You left your PC on for two weeks with no one using it? Glad I don't pay your power bill. Though those greenhouse gasses aren't going to help much...

how much does it cost to run a computer for 2 weeks in sleep mode? Computer use 1-20 watts in sleep mode. Lets say his is bad and uses 50 watts. 50 watt/hours * 24 hours/day * 7 days/week * 2 weeks = 16800 watt hours is about 17 kilowat hours. The average cost of electricity here in the US is about 7-8 cents/kilowatt/hour. Lets say he lives in one of the most expensive states, like California where it is 12 So the total cost of running the computer that two weeks is 17 * 12 = $2.04.
Compared to the cost to the company of shutting down the computer. When tgape gets to work he starts up his computer. Lets say it takes 5 minutes to start up and be ready to go. Lets say tgape makes on the low end of programmers at about $50,000/year. That means the company's expense is probably about $80,000-$100,000/year. Lets say it is low ($75,000). Lets further say tgape has 13 holiday days, 2 weeks of vacation and 5 personal days. That means he works 52*5-(13+2+5) = 240 day/year. That means the company expense per day is $312.50. A work day is 8 hours (480 minutes). That means his time is worth $0.65/minute to the company. 5 minutes at 65 cents is $3.25. So he saved the company over a dollar using extremely conservative numbers.
I'm glad I don't pay your paycheck.

Now let's say that he lost a day of work due to leaving the computer on and then pressing ctrl+Z. That works out at about $312.50.

tgape

@Kemp said:

You left your PC on for two weeks with no one using it?

Fundamental assumption error.

The software I was using didn't run on a PC (Intel or otherwise) at that time. I said nothing about whether other people used said computer while I was gone. (In fact, it was in use - but that's an entirely separate WTF which has been covered well here in prior peoples' posts.)

tgape

@tster said:

how much does it cost to run a computer for 2 weeks in sleep mode?

... @tster said:

Compared to the cost to the company of shutting down the computer.

...

Interesting line of thought. Let's flesh out some numbers, there... Computer wasn't in sleep mode, due to some fool's decision to create a tangled mess of NFS shares. I think it used something like 80 watts normally (this was some time ago, and all.) This was Texas, so that puts the first price at something like $2.45.

I was a contractor. My bill rate was $130/hr¹, boot time was 1m². That readily works out to $2.17. So if I could have turned off the computer while I was gone, I would've saved the company some money.

Of course, it would've saved the company even more money for me to not have shut down if I hadn't saved the file multiple times between copying the file and undoing the file. (In fact, I'd actually saved it right before the undo. I'd then ran a test, which failed, so I undid the change. So the cost was restarting the app³ and manually undoing the change.

¹ Of course, I saw little of this money. Something on the order of 16%.

² Assuming, of course, that all of the proper other systems were up. Add 5m for each NFS share which was not available. Since this was Christmas, that could'a been a big number.

³ Re-opening a file took this PoS 3-4 times longer than exiting, restarting, and opening an entire project (many files).

electricity in texas is 9.6 cents/Kwh

belgariontheking

@tster said:

electricity in texas is 9.6 cents/Kwh

Yes of course. Texas is such a small state that if you know the price in one area is like the price across the state. Good one.

Also, Let me not show in any way how that affects any of the calculations. I'll just spout some bullshit and it'll prove my point. Eveyone'll believe that!

pjt33

I've had an old version of Eclipse do that to me, although I think it was still possible to recover the file with its history feature.

morbiuswilters

@tgape said:

Interesting line of thought. Let's flesh out some numbers, there... Computer wasn't in sleep mode, due to some fool's decision to create a tangled mess of NFS shares. I think it used something like 80 watts normally (this was some time ago, and all.) This was Texas, so that puts the first price at something like $2.45.

I was a contractor. My bill rate was $130/hr¹, boot time was 1m². That readily works out to $2.17. So if I could have turned off the computer while I was gone, I would've saved the company some money.

There are so, so many things wrong from an economic point of view here. For one, your rates for electricity are too high. As you said, this was some time ago but you are using current residential rates. Rates for commercial electricity 10 years ago would be half of that. You also ignore that turning computers on and off wears out components faster, compared to leaving them on all the time.

Even if tster was wrong about the numbers of your particular situation, he is right overall as your situation represents an odd outlier. Kemp's criticism was silly.

However your biggest mistakes show a stunning ignorance of business. An employer does not pay you to turn on a computer. Your billable rate is not what an employer pays you for just sitting around. The employer pays you because it makes them a profit. In this sense, the cost of that 1 minute could potentially be quite a bit more than your billable rate. You also fail to factor in the work involved in shutting down and starting a computer. It may take 1 minute to boot, but if you are anything like me you have dozens of apps and web pages open at any one time and taking the time to save your position in each of these and then resume it later will certainly add up.

Overall, it's difficult to really calculate what the cost of turning on and off a computer is for a company, but to broadly claim that turning off a computer for 2 weeks saves enough money for a company to justify it (as Kemp said) is simply retarded.

@belgariontheking said:

@tster said:
electricity in texas is 9.6 cents/Kwh
Yes of course. Texas is such a small state that if you know the price in one area is like the price across the state. Good one.
Also, Let me not show in any way how that affects any of the calculations. I'll just spout some bullshit and it'll prove my point. Eveyone'll believe that!

what the fuck is wrong with you? I can see that you are just spouting bullshit and not proving anything by the fact that that's exactly what you just did. Electricty prices (residential, which is the highest of all the sectors) in texas average 9.6 cents and hour. I seriously doubt the company was paying 33% more than the average in their state for electricity.

He then makes crazy assumtions:

1. That computers a long time ago used somewhere between 4 and 80 times more electricity in sleep mode than they do currently.

2. That his old computer somehow takes only 1 minute to go from the off state to the ready-to-work state. As morbius already pointed out, I'm not talking about the time it takes for the computer to show the operating system. I'm talking about from off to having the IDE up and regular programs open and running.

3. That his bill rate is the only expense to the company for him working.

BTW, since belgariontheking seem to be unable to multiply numbers and has to have me spell out the entire computation again:

9.6 cents/kwH * 17 kwH = $1.63 which is less than $2.17.

As a follow up, sometimes it is cost efficient to restart your computer because it might make the computer run faster once it has restarted (especially old versions of Windows). But this has NOTHING to do with electricity costs.

Lets put it this way: If you are worried about electricity costs of developer workstations, you are worried about the wrong thing.

morbiuswilters

@tster said:

Lets put it this way: If you are worried about electricity costs of developer workstations, you are worried about the wrong thing.

QFT. That's what this really comes down to. All this jerking off about the cost of electricity is meaningless because it ignores the opportunity cost of lost productivity. One lost minute of productivity a day for every employee can be the difference between life and death for a company in an economy where product cycles are measured in months. Obviously, you can't sit and tabulate every expense like some kind of OCD beancounter; that kind of micro-management is counter-productive and harmful to morale. However, trying to argue that a few dollars per employee per year lost on electricity is comparable to hours of lost productivity is idiotic. Anyone who thought "Oh Jeez, how wasteful of company resources to leave a developer workstation on for two weeks!" should reconsider any ambitions of management or starting a small business because you clearly suck at doing cost-benefit analysis.

Feek

@tster said:

As a follow up, sometimes it is cost efficient to restart your computer because it might make the computer run faster once it has restarted (especially old versions of Windows). But this has NOTHING to do with electricity costs.

Lets put it this way: If you are worried about electricity costs of developer workstations, you are worried about the wrong thing.

I agree. My input was going to be "Who gives a shit?" Do we realize we're arguing over the difference of about a dollar here? Sometimes I think I want to hang out with programmers more to keep in touch with the people in my industry. Then I read threads like this and realize that I'd have to hang out with programmers more. :)

On a side note, why are there so many of us Massholes that frequent this board? It's not like Alex advertises on the T...

Edit: ...and what morbius said, too.

belgariontheking

@tster said:

what the fuck is wrong with you?

I'm drunk. Well I was at the time.

I think that your original calculation of how much it cost to do A vs B was the very definition of a waste of time in the first place. Any further discussion of it merely sends the thread into a spiral.

Physics_Phil

@Feek said:

On a side note, why are there so many of us Massholes that frequent this board? It's not like Alex advertises on the T...

There seems to be an anomalously high proportion of Aussies here as well, at least amongst the active members.

@belgariontheking said:

@tster said:
what the fuck is wrong with you?
I'm drunk. Well I was at the time.
I think that your original calculation of how much it cost to do A vs B was the very definition of a waste of time in the first place. Any further discussion of it merely sends the thread into a spiral.

That explains it. I have a rule about # drinks and the visiting of thedailywtf.com.

Also, internet forums mostly exist to waste time, so I'm not too concerned about wasting my time proving that someone else is a moron.

Kemp

Wow, you guys really get riled up by random comments =P I know, interwebs, serious business, and all that. Besides, in your rush to prove me wrong you forgot something important. He's not going to come into work and sit there staring at the screen while the computer boots (thus wasting the money you mentioned), he'll be grabbing a drink from the water cooler, or checking his voicemail, or saying hi to someone, or checking the notes people have left scattered around for all the "IMPORTANT" and "URGENT" things that they want doing (he *has* been away two weeks). Either way, I think the time you spent trying to prove me wrong was far more than he would have spent waiting for the computer to finish booting.

Feek

@Kemp said:

Besides, in your rush to prove me wrong you forgot something important. He's not going to come into work and sit there staring at the screen while the computer boots (thus wasting the money you mentioned), he'll be grabbing a drink from the water cooler, or checking his voicemail, or saying hi to someone, or checking the notes people have left scattered around for all the "IMPORTANT" and "URGENT" things that they want doing (he *has* been away two weeks).

In your rush to get the last word in, it you forgot something important: "Who the fuck cares?!?"

belgariontheking

@tster said:

That explains it. I have a rule about # drinks and the visiting of thedailywtf.com.

I do to. I'll explain it in algorithmic form

while (numDrinks < 3)

{ haveADrink();}

gotoWork();

while (numDrinks < 6)

{ haveADrink();}

wasteMyTimeAndOthersOnTDWTF();

Yes, numDrinks is a global variable.

gotoWork is defined to check whether it's time to be at work. It doesn't send me to work at 1am on a Friday night.

doWork() is not defined. neither is doAllTheWork().

also, wasteMyTimeAndOthersOnTDWTF() is named with bad English (should be wasteMyAndOthersTimeOnTDWTF) but it's a legacy enterprise function and would require a lot of changes just so I can have good English in my function names.

@belgariontheking said:

Yes, numDrinks is a global variable.

Your problem might be that you have multiple threads accessing this function at once and you end up calling haveADrink 10 to 12 times before calling the time wasting lagacy function. You might want to add semaphores.

morbiuswilters

@tster said:

@belgariontheking said:
Yes, numDrinks is a global variable.

Your problem might be that you have multiple threads accessing this function at once and you end up calling haveADrink 10 to 12 times before calling the time wasting lagacy function. You might want to add ~~semaphores~~ uppers.

FTFY.

fyjham

@tster said:

Your problem might be that you have multiple threads accessing this function at once and you end up calling haveADrink 10 to 12 times before calling the time wasting lagacy function. You might want to add semaphores.

Not only that but the haveADrink function has been known to corrupt memory, and sometimes when called often enough it leaves your program thinking numDrinks is a much lower number than it really is...

bstorer

@fyjham said:

Not only that but the haveADrink function has been known to corrupt memory, and sometimes when called often enough it leaves your program thinking numDrinks is a much lower number than it really is...

And let us not forget that it also causes unexpected system shutdowns, which lowers access requirements to private members, resulting in an increase in viruses and dialog boxes saying, "Where am I? What did I do last night?"

And now the following prompt appeared:

Yes, No, Cancel

I pressed N.

The Cancel button highlighted in darker grey.

Nothing else happened.

In case you're wondering: 'Yes' means "save again", 'No' means "close without saving", 'Cancel' means "keep it open".