|
Who cares, just code it and let the client worry about the rest 
If you have built castles in the air, your work need not be lost; that is where they should be.
Now put foundations under them.
|
|
|
|
|
Well, since most of what I write is for my own use... 
XAlan Burkhart
|
|
|
|
|
Why is this missing?
TBH our product is more than a tad slow compared to the competitors' but the shop is still running because we are a lot more accurate in our results compared to the ones that deliver results in a jiffy but are far from the expected level of accuracy.
Signature is a waste of time. I'll have one when I've got enough time to create. Well, I had enough a few seconds back.
|
|
|
|
|
I think Accuracy is too specific for the project.
In most projects there is no accuracy involved... the data is either correct or incorrect.
|
|
|
|
|
On the contrary, I would say most project do actually have accuracy involved. You can't assess accuracy only interms of floating point numbers! It could be anything...the way a string is tokenized, giving out all the possible combinations for a route etc. Most things cannot be classed into right or wrong. May be most projects involve using stock libraries, so this thing wouldn't have crossed their minds. But ask the people producing those libraries and accuracy matters very much, at least as much as others. Also, a certain library is used by the people who use it only because it is seen as "correct", but this correctness isn't binaty; rather it is accuracy that I am trying to talk about.
Signature is a waste of time. I'll have one when I've got enough time to create. Well, I had enough a few seconds back.
|
|
|
|
|
Koder wrote: Most things cannot be classed into right or wrong. ... but this correctness isn't binaty;
I would love to see your non-binary unit tests 
I imagine they would never turn a light green to say you did your job right, they would instead just always say "We need more accuracy!!!"
|
|
|
|
|
I think accuracy is assumed. Memory use, responsiveness, disk storage, CPU use, power use and network use are all somewhat subjective. You need to use memory, but how much is too much.
Accuracy on the other hand is a black or white issue. It either works or it doesn't*. If it doesn't work, your right, nothing else matters.
*I recognize there are some areas where the software might be approximating (forecasting) and in those cases accuracy would be a gray area, but we, the 99% don't deal with that
|
|
|
|
|
Accuracy too is highly subjective. Why do you think there are multiple contenders in scientific software market, or for that matter any "result producing softwares"? It is not always to do with the finances to buy a "better" software that people opt for "lesser" software. As I replied above, probably because most project use "tried and tested" libraries, accuracy is assumed.
And no, it is most definitely not black and white. Accuracy isn't binary at all.
The quote was amusing. 
Signature is a waste of time. I'll have one when I've got enough time to create. Well, I had enough a few seconds back.
|
|
|
|
|
If I understood right... you are right, but then we must always consider the expectations.
A software made to do X must do X. If it does Y, it is not the software we want.
So, yes... accuracy is the most important... but it is the real objective of the software... right?
But then, if you have 2 software that completely fulfill your needs... but one is slow, the other uses a lot of memory... which one (considering only those 2) you will chose?
I certainly will chose the faster but memory hungry one. (if it is too memory hungry, than it will be slow too... or even will never do the job).
So...considering the software do what it needs to do (so, it must be accurate in what it does)... it then need to:
1) Be fast/responsive.
2) Use less memory? Less disk space?
I will keep my vote on responsiveness... even if I can partially agree with you.
|
|
|
|
|
we all are developing software for our users/customer. If they are happy, we are happy by any mean.
"Customer satisfication" is the main moto. Which leads to Faster response from software.
so i think, Responsiveness is the main part affected than all.
If the software is very slow in response to customer, then what is the use of disk space, memory use, CPU use.
Rating always..... WELCOME
The only reason people get lost in thought is because it's unfamiliar territory.
|
|
|
|
|
Our software needs to be responsive [microseconds count] and there is a shed load [increasing by around 1.5 wheelbarrows per month] of communications.
The other considerations are only really an issue in relation to TCO for our user base. There's no CPU thrashing [AFAIK] so if capacity is reached we just plug in another board. The same goes for memory, a lot of data has to be kept in memory - see responsiveness - and paging would slow things down too much so all the active data needs to be kept in memory. If we run out, we just add some more; to the best of my knowledge we're a long way off any hardware / OS application limits in this respect.
Panic, Chaos, Destruction. My work here is done.
Drink. Get drunk. Fall over - P O'H
OK, I will win to day or my name isn't Ethel Crudacre! - DD Ethel Crudacre
I cannot live by bread alone. Bacon and ketchup are needed as well. - Trollslayer
Have a bit more patience with newbies. Of course some of them act dumb - they're often *students*, for heaven's sake - Terry Pratchett
|
|
|
|
|
Message Closed
modified 1-Nov-11 16:16pm.
|
|
|
|
|
While we are on the subject of responsiveness... check your gmail acount and respond. 
|
|
|
|
|
I'm developing an internal LOB app, and responsiveness is the #1 issue. All the desktops have multi-core CPUs and >2GB ram and for the most part the application I work on is the only thing running. I feel like by focusing on responsiveness the other issues mostly take care of themselves.
Sending too much stuff across the WAN? First place we'll notice it is in a slower screen loading.
Spinning the CPU? Probably some inefficient code that is causing the app to hang or load data from a worker thread slowly.
Using to much memory? The first thing we'll notice is a drop in responsiveness, then we'll find the reason was memory.
We had a memory leak at one point, but it wasn't reported as a memory use issue, it was reported as "after a few hours it gets really sluggish".
|
|
|
|
|
Yes, it must use memory, it must use the CPU, it must be stored on disk...
What a freakin' stupid survey. 
|
|
|
|
|
My applications are old fashioned. They don't work with the power off 
It's an OO world.
public class Naerling : Lazy<Person>{}
|
|
|
|
|
Power use does not mean controlling on/off
And Dont think a project means something only creating software, a project gets bigger than that when power consumption become a big factor
Thanks got I am not a doctor, cause doctor cant use debugging
|
|
|
|
|
All sorts of equipment other than personal computers make power use a high priority.
The most notable are smartphones, running a GPS radio receiver, Bluetooth, wifi, and not to forget the phone itself. Tack on a spinning CPU, and the battery will be dead in no time.
There are other types of equipment as well, such as airplanes. More power consumption usually means more heat, and convection doesn't work as well at 30000ft than it does at sea level.
|
|
|
|
|
Consider PLCs or OPLCs... they use power and they are programmable... and run a program as well
Alberto Bar-Noy
---------------
“The city’s central computer told you? R2D2, you know better than to trust a strange computer!”
(C3PO)
|
|
|
|
|
Applications can reduce power just by switching stuff off when not in use (Even the CPU). Although this is obviously more for mobile apps.
|
|
|
|
|
Responsiveness is top?! How ironic. My PC is hundreds of times faster than 20 years ago but is it any more responsive?
|
|
|
|
|
Kamran Behzad wrote: My PC is hundreds of times faster than 20 years ago but is it any more responsive?
Today's software is expected to do much more than software from 20 years ago.
|
|
|
|
|
20 years ago... Try running Windows 3.0 on a modern PC and I'm sure it will be responsive.
Steve
|
|
|
|
|
I'm surprised. I thought memory use would be clear #1...
At least, for windows forms...
|
|
|
|
|
It costs around $45 US for 8 GB of memory (name brand DDR3 1333 1.5V) now.
John
modified 31-Oct-11 18:35pm.
|
|
|
|
General 
News 
Suggestion 
Question 
Bug 
Answer 
Joke 
Praise 
Rant 
Admin 
Submit an article or tip
