I thought Dice went bankrupt and out of business once people realized that Dice was just making up fake job listings that offered $100k to work on OSS and be a kernel developer.
Break the tasks up into incredible small chunks and use those to estimate times. It is a lot easier to estimate how long it will take to add a new page to the user interface, times ten pages, plus three functional back end tasks, plus interface to the database over MQseries, than it is to ball park an estimate on doing the whole thing. I would say a day and a half per page of user interface, two days per back end functional task, four days to get my half of the MQseries interface working, and an additional 20% for QA, black box testing, and documentation. Doing that I end up with 25 days coding, 5 days of QA testing and documentation for a total of 30 days if I am doing it myself.
Six weeks.
If you just say 'hey we have a app that needs some new user interfaces and talks to the back end over our network (mumble something about MQseries)... how long you think it will take if you and Dubl-u work on it together?' Heck, Dubl-u is easily as 1337 as I am, and if he is in the room so we can do that hacker subtle wink and nod thing... 'no problem, we can have that done in two weeks.'
Reality : take the estimate I did above (detailed quote), double it, factor 1.2x because of the additional guy, end up with 72 man days (30x2x1.2) or about 7 weeks. More likely, because I personally tend to be conservative on my estimates more like 6 weeks.
Note that I included QA, Testing, and Documentation in my original quote (something that most programmers leave out) so here is how a regular senior guy would have quoted it:
New UI pages : 1 per day New functionality on the back end : 2 days per MQseries integration : 2 days.
Total estimate : 18 days. Depending on the guy he will either round that down to three weeks, or up to four weeks.
If he was just ballparking it without asking for a breakdown, he would have said two weeks.
But that didn't answer your question. How can you reconcile the two...
Rate your team, break them down into Newbies, Ubers that guess low, and Ubers that include documentation and test time in their estimates.
Apply this factor: Newbies : 5x Ubers : 3x Ubers that include dox and QA in their estimate : 2x
Or you can back rev it into the timeline : ask the team 'what can you get done in 9 months, give me a detailed list.' Apply the above factors to determine how long it will take to actually deliver (ie, 18 months at the earliest) or start slashing functionality to get the estimate down to 4 months if you really need to go live in 9 months.
Actually nobody hands a newbie a project a 'one-month' project, all they get are little scraps and pieces, things like 'take this routine and change it from a bubble sort to a selection sort to speed up the bottleneck by 50%.'
A newbie remembers something about having written a bubble sort in college, knows that a selection sort exists because he is being asked to code it, and spouts off : sure thing, no problem, I can do this in about one hour.
One hour.
Reality:
He doesn't have the build environment on his machine yet. I am going to assume he already his machine set up with the development environment. Get the source code from where ever it is, and actually find the module that has the routine he needs to change : 1 hour
Become familiar with the source being used, in the context with which it pertains to him so he understands how it should work when he is done : 15 minutes to an hour depending on how well the code is written and documented.
Go to lunch : one hour Check email when back from lunch, check/. for any replies to his messages, check his stock in SCOX to see if it is still going up : half hour.
Selection sort... what the hell is a selection sort... Google / search MSDN for an example of a selection sort : half hour.
Make a new project, cut and paste the sample code, try to compile it, jack with it until it compiles and runs, then keep jacking with it some more until he understands how to code it as a replacement to the bubble sort routine currently being used in the production code : one hour
Comment out the old bubble sort, code in the new selection sort using the proper variables and do a walk through. 10 minutes.
Spend a half hour trying to get the project to compile (unsuccessfully). : half hour
Get a senior guy's attention, learn that he needs to have the patches to his dev environment that can handle something specific to that company because they hacked it to do what they need : half hour.
Compile it, run it, test it : 5 minutes. Newbies never spend enough time testing it.
Documentation? WTF nobody said anything about documentation. Newbie updates the system documentation with a sentior guy showing him how, explaining why it is important : half hour.
Add in the caffeine breaks and you are looking at a solid 8 hours.
It gets worse when the newbie is estimating in the 1 to 3 days category because he is estimating 16 hours of productive, uninterrupted, caffeine fueled hacking like he would do in college, but those kinds of hack sessions are impossible in a cube farm when everybody is shuffling in over the 8am-10am timeframe, shuffling about at 10:30 for a break, shuffling around at noon and back again at 1 for lunch, and shuffling out the door from 4pm til 6pm... just the traffic, chatter, background noise, friendly 'hey, how you doing?', interoffice email and memos, etc make that totally impossible.
Maybe 8x was a little on the heavy side, but 5x is easily a realistic number.
-Programmers working on large projects estimate large chunks of time. If these people estimate one year, I would at most epect not more than one year and a half.
Tell that to the guys working on Doom III, Duke Nukem Forever, Team Fortress 2. They are pretty senior dev teams with successful projects behind them and all are running easily 2x as long as they had projected.
If a large project estimates a year and ships in 18 months, they had to slash a lot of functionality to get it out the door. You know, like removing the multiplayer aspect of a first person shooter, or removing the cooperative multiplayer aspect originally planned (leaving in deathmatch mode.)
-You may have meant shorter term estimations, and while they are more prone to double, saying that a you should expect a newbie to take a week to get a one day assignment done is completely rediculous.
New programmer fresh out of college: Take his estimate and multiply by 8x. Yes he could get it done in 1 day, assuming he got so cranked up on caffeine his eyes stopped blinking and he worked on that (and nothing else) for 24 hours straight. In the real world a newbie can dedicate about 2 real hours doing a particular task each day, the rest is spent coming up to speed on corporate coding standards and libraries, email, breaks, and not 'in the groove'.
Veteran programmer of average skill, single person project : multiply his estimate by 3x. A third of his day is spent hand-holding the newbie, and another third is spent hand-holding management. The other third is spent programming, but luckily he knows to pad the schedule some (not enough, but some.)
Veteran programmer of uber skill, single person project : multiply his estimate by 2x. This is as good as it gets. A uber veteran programmer knows to leave his email client closed and his door closed so he can stay in the zone. He knows to pad the schedule more than he really thinks he should. And it still takes him twice as long as he expected.
Multiple people working on the same project : increase the timeline by a factor of 1.2 per additional person. If two people ought to be able to do it in 10 days it will take 12. If 11 people (10 additional) ought to be able to do it in 10 days it will take... 1.2^10 = about 6, so 10 x (1.2^10) = roughly 60 days = 12 weeks = 3 months.
-One question is how would they endure a lightning strike (perhaps not as good as an optical medium)?
Given that when you aren't actively moving data onto / off this thing it is supposed to be in your pocket or on your keychain, if it gets hit by lightning then 'data retention' is going to be pretty low on your list of concerns.
That shit isn't funny, and I can pretty much guarantee you weren't there when it happened.
Ok, here's the story. The handlebars can lock using the little lock that is built into the lower triple-clamp, you need to crank the bars over to the left and with the key reach down into the cowling and lock it.
Now just pretending that I had such a motorcycle in college (a 1986 Suzuki GSXR-750) and pretending that I had parked it next to the Sears at the mall, up on the sidewalk fairly close to the brick wall, and pretending that I needed to turn left to drive away from the wall to get back onto the driveway anyways... it is entirely possible (hypothetically speaking) that since the bars were pointing that way anyways a person could put the key in the ignition, start the bike, lean to the left and start to drive away and not remember that the bars were locked and have that futhermocker flip over onto its side in about one second flat, scratching the lower cowl and bruising his ego.
You know, in theory. Pretty damn funny now that I look back at it.
(1) What are the general opinions of programmers on technical writers;
Newbie programmers look down on technical writers with distain, the same distain they have for anybody that writes HTML and calls it programming.
Experienced programmers LOVE technical writers because we hate writing the documentation and are happy to have you do it for them. We will generally bring you cheesecake.
(2) is there someone out there who has first-hand experience in technical writing who can tell me about the work and their experiences;
It requires that you be able to work as a go-between the hardcore techs and the mouth breathing users - so you must be able to relate to and work in terms that either understands, at various parts of your work. Note that there isn't a lot of crossover between the two.
(3) what software is used mainly in the process;
Microsoft Word. And some application that converts Word docs into help files.
Not to be dragged too far off topic, but if you honestly believe the outsourcing / migration of jobs to ANYWHERE has anything to do with any aspect of business except dollars, you are sadly deluded. I have had to hang up a zillion times on tech support calls to callers with accents so thick I couldn't understand them.
They are not technological wizards that have backgrounds of 10 years plus diagnosing hardware / network issues and immediately recognize that 8 beeps means your motherboard is complaining about a problem with your video card, know that a SCSI terminator before the new hard drive you just added (instead of on the end of the cable) means that no matter what you do the new drive isn't going to be seen, or know from experience that you are going to need to bump the voltage to 2.2v to run a Celeron 300a at 450MHz, particularly if the first number on the FPO/BATCH number is a 0 (ie, chip is from Costa Rica.)
They are polite script monkeys that are quite happy to read through tech support scripts for $1.50 an hour while punctuating their sentences with 'please', 'thank you', and 'berry good'. That the customers are getting wicked pissed off is of no concern because hey, it is only costing the company $10 a day.
Quite honestly nothing would make me happier than to have the Paki's rain thermo-nuclear bombs on all of the new tech-parks that IBM and Dell just built over there. Every American tech worker is thinking it right now, but few are willing to actually say it.
Outsourcing has nothing to do with the Queens English, or their verbal / written comms skills. It has everything to do with working for 83 cents an hour. Ask the Russian and Chinese folks that are also getting outsource work.
Me and my friends here in India would be loving to help you out of your predicament. My team consists of the finest young plucky software engineers versed in all the latest technologies and will eagerly do this for you for only $10 total. This includes a complete set of documentation and followup, project management and five happy programmers to code your solution using on the best of software development methodologies. All for only $10.
Of course then we have your entire contact database which we can either use for our own personal profits (legal or otherwise) or maybe we just sell it to all your competitors for a nice tidy profit. Then you will be fuxored pretty bad, but hey - at least it only cost you $10.
Look at the other end of the spectrum (simply because 3GHz P3 chips are in short supply) : a 1.4GHz P3 will trounce a 1.4GHz P4 (and both are available.)
Where we get the benefits from each CPU generation is in being able to scale it up so much faster, ie getting 3.2GHz out of a P4 is possible whereas a 3.2GHz P3 is not possible without liquid nitrogen and corporate sponsorship.
There is no substitute for cubic inches, as the saying goes, but larger engines are less efficient.
486 had 1.2 Million transistors. First CPU in the lineup to have an internal Level 1 cache (8k as I recall) Considered the father of the 32 bit computing generation in that lineup.
Zealots : 'in that lineup' means the Intel chips. Yea I know a Sun box probably did it earlier and a Mac box probably did it better - but I can't pull Sun or Mac numbers out of my butt like I can Intel numbers.
The other thing that makes it so appealing to chip makers is that it can withstand much higher temperatures - your regular CPU will fry somewhere between 85 and 100 degrees C, but it is going to take a LOT more than that to fry a diamond CPU.
Granted the motherboard and ZIF socket will probably melt long before the CPU.
Just for the record, that is how I can tell the difference between cubic zirconia rocks and real diamonds, particularly at the 1 carat size. A 1 carat CZ is going to be flawless and a 1 carat diamond is going to be pitted, flawed, and full of inclusions that are visible under a low power glass or even your eye.
It is possible to have a totally perfect 1 carat diamond, but very unlikely. CZ is perfect pretty much every time.
Bah! I had a Commodore 64 in the early 80's and not only was it a Commodore 64 (note : the earliest use of the numbers 64 in a computer name) it also had 65,536 bytes of memory. 65,536 - you see that? Big numbers. Way bigger than anything I have seen in this whole thread. I have seen some 244's, some 1.8's, some 2.4's, some 32's and some 128's, but this machine was a monster with 65,536. Connected to other computers at 300 baud. Not the puny 28.8 or 33.6 or even 56.7 you read about today, or even the dedicated 128 or 256 lines - or even the wickedly overpriced 1.544 T1 lines that everybody dreams about... a hard core 300!
Also held the record for most machines ever sold, at the time anyways. The biggest, and the mostest. Commodore 64. Accept no substitutes.
-Just refresh a few times... that's what worked for me...
Oh man, sounds just like the chicks in accounting at the last company I worked for doing IS/IT work. I was working on a machine under the desk when I heard this beautiful conversation:
G1 : I hit the print button but it isn't printing. G2 : Sometimes they get lost on the way to the printer. G1 : What do I do? G2 : Print it again. Print it five times. Just keep hitting the print key until one of them finds its way to this printer.
Meanwhile I am under the table just losing it, totally unable to maintain composure. Somewhere in that office, I am guessing, there was a printer working doubletime just spewing copy after copy of that accounting report.
Actually the nuc subs do have pumps to run the water that cools the reactors. This is why the electric subs can in theory actually be quieter than the new nuclear subs, although the sound-muffling on the nuc's is very, very good.
As cool as it would be, the 688 doesn't just vent the waste heat by running ocean water directly over the reactors.
-The CPU architectures themselves have actually gotten more efficient.
I have to take issues with this (blowing my opportunity to use mod points in the process...)
Correct me if I am wrong, but I am not aware of any apps that will run faster on a 1.4GHz PIV than on a 1.4GHz PIII. In fact, I believe the P3 would trounce (go faster than) the P4 at the same speed (1.4GHz)
The benefit we get from the P4 only becomes evident when we run it at 3GHz or what have you, something fairly impossible from the P3 in a stock machine.
I seem to recall the same thing happening with the P2 chips running at 233MHz vs the original Pentium at that same speed, but the original Pentium at 66MHz simply smoking (going faster than) a 486/DX2-66.
Get him a new email address, this time make it longer than 10 characters. Something like firstname_lastname_nz@yourhost.com (where nz are just two random letters.) I have a hotmail account like this, recently created, that has received exactly zero spams (then again I was very careful about who I give it out to and I didn't use it to sign up for stuff.)
Do you really (really) want to put this one up to a democratic (secret ballot, like November 4th) vote? With binding results, 50.5% or greater (majority) wins?
Heck, I think in California with the way their Proposition thing works you could probably get it on the ballot.
You have this guy's home address. Why not just go kill him, his wife and all his children, burn down his house and light his dog on fire? That will send a pretty strong message.
I bet he starts listing associates if you tie him to a tree and light him on fire.
You can get a lot more accomplished with a kind word and a gun than you can with a kind word alone.
That was a few years ago - I have matured since then. A little.
Heck, there was a day when I only had one computer at home and it ran at 1MHz, had 65536 bytes of memory (of which about 38,000 were available to program with) and talked to the world at 300 baud (about 30 characters per second) at a cost of $6 per hour. And I was -thankful-. Computer performance (and connection speeds) are all relative, you are working at levels I only dreamed of years ago.
Of course right now I am playing catch-up, have a 1.544Mb/s dedicated pipe to the 'net shared behind a firewall / router to two Dell Laptops with wifi connectivity, a HP laptop, two Dell servers, and two hand build machines, with total CPU speeds in the 7.5GHz range, memory totalling about 3Gigs combined.
I have been known to change residences simply because the new place had a better Internet connection - but don't tell my fiance.
Slashdot Mirror
I thought Dice went bankrupt and out of business once people realized that Dice was just making up fake job listings that offered $100k to work on OSS and be a kernel developer.
Oh yea, they did.
-How can you reconcile the two?
:
... how long you think it will take if you and Dubl-u work on it together?' Heck, Dubl-u is easily as 1337 as I am, and if he is in the room so we can do that hacker subtle wink and nod thing ... 'no problem, we can have that done in two weeks.'
:
...
:
Good question. Hmmm - recent experiences
Break the tasks up into incredible small chunks and use those to estimate times. It is a lot easier to estimate how long it will take to add a new page to the user interface, times ten pages, plus three functional back end tasks, plus interface to the database over MQseries, than it is to ball park an estimate on doing the whole thing. I would say a day and a half per page of user interface, two days per back end functional task, four days to get my half of the MQseries interface working, and an additional 20% for QA, black box testing, and documentation. Doing that I end up with 25 days coding, 5 days of QA testing and documentation for a total of 30 days if I am doing it myself.
Six weeks.
If you just say 'hey we have a app that needs some new user interfaces and talks to the back end over our network (mumble something about MQseries)
Reality : take the estimate I did above (detailed quote), double it, factor 1.2x because of the additional guy, end up with 72 man days (30x2x1.2) or about 7 weeks. More likely, because I personally tend to be conservative on my estimates more like 6 weeks.
Note that I included QA, Testing, and Documentation in my original quote (something that most programmers leave out) so here is how a regular senior guy would have quoted it
New UI pages : 1 per day
New functionality on the back end : 2 days per
MQseries integration : 2 days.
Total estimate : 18 days. Depending on the guy he will either round that down to three weeks, or up to four weeks.
If he was just ballparking it without asking for a breakdown, he would have said two weeks.
But that didn't answer your question. How can you reconcile the two
Rate your team, break them down into Newbies, Ubers that guess low, and Ubers that include documentation and test time in their estimates.
Apply this factor
Newbies : 5x
Ubers : 3x
Ubers that include dox and QA in their estimate : 2x
Or you can back rev it into the timeline : ask the team 'what can you get done in 9 months, give me a detailed list.' Apply the above factors to determine how long it will take to actually deliver (ie, 18 months at the earliest) or start slashing functionality to get the estimate down to 4 months if you really need to go live in 9 months.
That works.
Actually nobody hands a newbie a project a 'one-month' project, all they get are little scraps and pieces, things like 'take this routine and change it from a bubble sort to a selection sort to speed up the bottleneck by 50%.'
:
/. for any replies to his messages, check his stock in SCOX to see if it is still going up : half hour.
... what the hell is a selection sort ... Google / search MSDN for an example of a selection sort : half hour.
... just the traffic, chatter, background noise, friendly 'hey, how you doing?', interoffice email and memos, etc make that totally impossible.
A newbie remembers something about having written a bubble sort in college, knows that a selection sort exists because he is being asked to code it, and spouts off : sure thing, no problem, I can do this in about one hour.
One hour.
Reality
He doesn't have the build environment on his machine yet. I am going to assume he already his machine set up with the development environment. Get the source code from where ever it is, and actually find the module that has the routine he needs to change : 1 hour
Become familiar with the source being used, in the context with which it pertains to him so he understands how it should work when he is done : 15 minutes to an hour depending on how well the code is written and documented.
Go to lunch : one hour
Check email when back from lunch, check
Selection sort
Make a new project, cut and paste the sample code, try to compile it, jack with it until it compiles and runs, then keep jacking with it some more until he understands how to code it as a replacement to the bubble sort routine currently being used in the production code : one hour
Comment out the old bubble sort, code in the new selection sort using the proper variables and do a walk through. 10 minutes.
Spend a half hour trying to get the project to compile (unsuccessfully). : half hour
Get a senior guy's attention, learn that he needs to have the patches to his dev environment that can handle something specific to that company because they hacked it to do what they need : half hour.
Compile it, run it, test it : 5 minutes. Newbies never spend enough time testing it.
Documentation? WTF nobody said anything about documentation. Newbie updates the system documentation with a sentior guy showing him how, explaining why it is important : half hour.
Add in the caffeine breaks and you are looking at a solid 8 hours.
It gets worse when the newbie is estimating in the 1 to 3 days category because he is estimating 16 hours of productive, uninterrupted, caffeine fueled hacking like he would do in college, but those kinds of hack sessions are impossible in a cube farm when everybody is shuffling in over the 8am-10am timeframe, shuffling about at 10:30 for a break, shuffling around at noon and back again at 1 for lunch, and shuffling out the door from 4pm til 6pm
Maybe 8x was a little on the heavy side, but 5x is easily a realistic number.
-Programmers working on large projects estimate large chunks of time. If these people estimate one year, I would at most epect not more than one year and a half.
Tell that to the guys working on Doom III, Duke Nukem Forever, Team Fortress 2. They are pretty senior dev teams with successful projects behind them and all are running easily 2x as long as they had projected.
If a large project estimates a year and ships in 18 months, they had to slash a lot of functionality to get it out the door. You know, like removing the multiplayer aspect of a first person shooter, or removing the cooperative multiplayer aspect originally planned (leaving in deathmatch mode.)
-You may have meant shorter term estimations, and while they are more prone to double, saying that a you should expect a newbie to take a week to get a one day assignment done is completely rediculous.
Differe
Time estimage guidelines:
... 1.2^10 = about 6, so 10 x (1.2^10) = roughly 60 days = 12 weeks = 3 months.
New programmer fresh out of college: Take his estimate and multiply by 8x. Yes he could get it done in 1 day, assuming he got so cranked up on caffeine his eyes stopped blinking and he worked on that (and nothing else) for 24 hours straight. In the real world a newbie can dedicate about 2 real hours doing a particular task each day, the rest is spent coming up to speed on corporate coding standards and libraries, email, breaks, and not 'in the groove'.
Veteran programmer of average skill, single person project : multiply his estimate by 3x. A third of his day is spent hand-holding the newbie, and another third is spent hand-holding management. The other third is spent programming, but luckily he knows to pad the schedule some (not enough, but some.)
Veteran programmer of uber skill, single person project : multiply his estimate by 2x. This is as good as it gets. A uber veteran programmer knows to leave his email client closed and his door closed so he can stay in the zone. He knows to pad the schedule more than he really thinks he should. And it still takes him twice as long as he expected.
Multiple people working on the same project : increase the timeline by a factor of 1.2 per additional person. If two people ought to be able to do it in 10 days it will take 12. If 11 people (10 additional) ought to be able to do it in 10 days it will take
-One question is how would they endure a lightning strike (perhaps not as good as an optical medium)?
Given that when you aren't actively moving data onto / off this thing it is supposed to be in your pocket or on your keychain, if it gets hit by lightning then 'data retention' is going to be pretty low on your list of concerns.
That shit isn't funny, and I can pretty much guarantee you weren't there when it happened.
... it is entirely possible (hypothetically speaking) that since the bars were pointing that way anyways a person could put the key in the ignition, start the bike, lean to the left and start to drive away and not remember that the bars were locked and have that futhermocker flip over onto its side in about one second flat, scratching the lower cowl and bruising his ego.
Ok, here's the story. The handlebars can lock using the little lock that is built into the lower triple-clamp, you need to crank the bars over to the left and with the key reach down into the cowling and lock it.
Now just pretending that I had such a motorcycle in college (a 1986 Suzuki GSXR-750) and pretending that I had parked it next to the Sears at the mall, up on the sidewalk fairly close to the brick wall, and pretending that I needed to turn left to drive away from the wall to get back onto the driveway anyways
You know, in theory.
Pretty damn funny now that I look back at it.
(1) What are the general opinions of programmers on technical writers;
Newbie programmers look down on technical writers with distain, the same distain they have for anybody that writes HTML and calls it programming.
Experienced programmers LOVE technical writers because we hate writing the documentation and are happy to have you do it for them. We will generally bring you cheesecake.
(2) is there someone out there who has first-hand experience in technical writing who can tell me about the work and their experiences;
It requires that you be able to work as a go-between the hardcore techs and the mouth breathing users - so you must be able to relate to and work in terms that either understands, at various parts of your work. Note that there isn't a lot of crossover between the two.
(3) what software is used mainly in the process;
Microsoft Word. And some application that converts Word docs into help files.
(4) what seems to be the average pay?
No clue, sorry.
Dude there is a reason you are not doing documentation, and 'because you are a programmer' isn't it.
That said, I blew my chance of moderating this so I could reply.
Not to be dragged too far off topic, but if you honestly believe the outsourcing / migration of jobs to ANYWHERE has anything to do with any aspect of business except dollars, you are sadly deluded. I have had to hang up a zillion times on tech support calls to callers with accents so thick I couldn't understand them.
They are not technological wizards that have backgrounds of 10 years plus diagnosing hardware / network issues and immediately recognize that 8 beeps means your motherboard is complaining about a problem with your video card, know that a SCSI terminator before the new hard drive you just added (instead of on the end of the cable) means that no matter what you do the new drive isn't going to be seen, or know from experience that you are going to need to bump the voltage to 2.2v to run a Celeron 300a at 450MHz, particularly if the first number on the FPO/BATCH number is a 0 (ie, chip is from Costa Rica.)
They are polite script monkeys that are quite happy to read through tech support scripts for $1.50 an hour while punctuating their sentences with 'please', 'thank you', and 'berry good'. That the customers are getting wicked pissed off is of no concern because hey, it is only costing the company $10 a day.
Quite honestly nothing would make me happier than to have the Paki's rain thermo-nuclear bombs on all of the new tech-parks that IBM and Dell just built over there. Every American tech worker is thinking it right now, but few are willing to actually say it.
Outsourcing has nothing to do with the Queens English, or their verbal / written comms skills. It has everything to do with working for 83 cents an hour. Ask the Russian and Chinese folks that are also getting outsource work.
Me and my friends here in India would be loving to help you out of your predicament. My team consists of the finest young plucky software engineers versed in all the latest technologies and will eagerly do this for you for only $10 total. This includes a complete set of documentation and followup, project management and five happy programmers to code your solution using on the best of software development methodologies. All for only $10.
Of course then we have your entire contact database which we can either use for our own personal profits (legal or otherwise) or maybe we just sell it to all your competitors for a nice tidy profit. Then you will be fuxored pretty bad, but hey - at least it only cost you $10.
Caffeine: 3,7-dihydro- 1,3,7-trimethyl- 1H-purine- 2,6-dione; 1,3,7-trimethylxanthine; 1,3,7-trimethyl- 2,6-dioxopurine.
C8H10N4O2; mol wt 194.19. C 49.48%, H 5.19%, N 28.85%, O 16.48%.
Occurs in tea, coffee, mate leaves; also in guarana paste and cola nuts.
-:-
Get some of this in your system and get back to us.
Look at the other end of the spectrum (simply because 3GHz P3 chips are in short supply) : a 1.4GHz P3 will trounce a 1.4GHz P4 (and both are available.)
Where we get the benefits from each CPU generation is in being able to scale it up so much faster, ie getting 3.2GHz out of a P4 is possible whereas a 3.2GHz P3 is not possible without liquid nitrogen and corporate sponsorship.
There is no substitute for cubic inches, as the saying goes, but larger engines are less efficient.
486 had 1.2 Million transistors.
First CPU in the lineup to have an internal Level 1 cache (8k as I recall)
Considered the father of the 32 bit computing generation in that lineup.
Zealots : 'in that lineup' means the Intel chips. Yea I know a Sun box probably did it earlier and a Mac box probably did it better - but I can't pull Sun or Mac numbers out of my butt like I can Intel numbers.
The other thing that makes it so appealing to chip makers is that it can withstand much higher temperatures - your regular CPU will fry somewhere between 85 and 100 degrees C, but it is going to take a LOT more than that to fry a diamond CPU.
Granted the motherboard and ZIF socket will probably melt long before the CPU.
Just for the record, that is how I can tell the difference between cubic zirconia rocks and real diamonds, particularly at the 1 carat size. A 1 carat CZ is going to be flawless and a 1 carat diamond is going to be pitted, flawed, and full of inclusions that are visible under a low power glass or even your eye.
It is possible to have a totally perfect 1 carat diamond, but very unlikely. CZ is perfect pretty much every time.
Bah! I had a Commodore 64 in the early 80's and not only was it a Commodore 64 (note : the earliest use of the numbers 64 in a computer name) it also had 65,536 bytes of memory. 65,536 - you see that? Big numbers. Way bigger than anything I have seen in this whole thread. I have seen some 244's, some 1.8's, some 2.4's, some 32's and some 128's, but this machine was a monster with 65,536. Connected to other computers at 300 baud. Not the puny 28.8 or 33.6 or even 56.7 you read about today, or even the dedicated 128 or 256 lines - or even the wickedly overpriced 1.544 T1 lines that everybody dreams about ... a hard core 300!
Also held the record for most machines ever sold, at the time anyways. The biggest, and the mostest. Commodore 64. Accept no substitutes.
-Just refresh a few times... that's what worked for me...
:
Oh man, sounds just like the chicks in accounting at the last company I worked for doing IS/IT work. I was working on a machine under the desk when I heard this beautiful conversation
G1 : I hit the print button but it isn't printing.
G2 : Sometimes they get lost on the way to the printer.
G1 : What do I do?
G2 : Print it again. Print it five times. Just keep hitting the print key until one of them finds its way to this printer.
Meanwhile I am under the table just losing it, totally unable to maintain composure. Somewhere in that office, I am guessing, there was a printer working doubletime just spewing copy after copy of that accounting report.
Actually the nuc subs do have pumps to run the water that cools the reactors. This is why the electric subs can in theory actually be quieter than the new nuclear subs, although the sound-muffling on the nuc's is very, very good.
As cool as it would be, the 688 doesn't just vent the waste heat by running ocean water directly over the reactors.
-The CPU architectures themselves have actually gotten more efficient.
I have to take issues with this (blowing my opportunity to use mod points in the process...)
Correct me if I am wrong, but I am not aware of any apps that will run faster on a 1.4GHz PIV than on a 1.4GHz PIII. In fact, I believe the P3 would trounce (go faster than) the P4 at the same speed (1.4GHz)
The benefit we get from the P4 only becomes evident when we run it at 3GHz or what have you, something fairly impossible from the P3 in a stock machine.
I seem to recall the same thing happening with the P2 chips running at 233MHz vs the original Pentium at that same speed, but the original Pentium at 66MHz simply smoking (going faster than) a 486/DX2-66.
ROFL - yea, one thing to be sure on : you can count on your inbox having at least one spam message so you know everything is talking correctly.
Get him a new email address, this time make it longer than 10 characters. Something like firstname_lastname_nz@yourhost.com (where nz are just two random letters.) I have a hotmail account like this, recently created, that has received exactly zero spams (then again I was very careful about who I give it out to and I didn't use it to sign up for stuff.)
Do you really (really) want to put this one up to a democratic (secret ballot, like November 4th) vote? With binding results, 50.5% or greater (majority) wins?
Heck, I think in California with the way their Proposition thing works you could probably get it on the ballot.
-They have been known do stupid things like block entire countries.
They only block the evil countries. I mean it isn't like they blackholed Germany or England or Finland...
You have this guy's home address. Why not just go kill him, his wife and all his children, burn down his house and light his dog on fire? That will send a pretty strong message.
I bet he starts listing associates if you tie him to a tree and light him on fire.
You can get a lot more accomplished with a kind word and a gun than you can with a kind word alone.
That was a few years ago - I have matured since then. A little.
Heck, there was a day when I only had one computer at home and it ran at 1MHz, had 65536 bytes of memory (of which about 38,000 were available to program with) and talked to the world at 300 baud (about 30 characters per second) at a cost of $6 per hour. And I was -thankful-. Computer performance (and connection speeds) are all relative, you are working at levels I only dreamed of years ago.
Of course right now I am playing catch-up, have a 1.544Mb/s dedicated pipe to the 'net shared behind a firewall / router to two Dell Laptops with wifi connectivity, a HP laptop, two Dell servers, and two hand build machines, with total CPU speeds in the 7.5GHz range, memory totalling about 3Gigs combined.
I have been known to change residences simply because the new place had a better Internet connection - but don't tell my fiance.