hours programming, max
bigbinc
the most anybody has programmed straight.
(I know this is bad and errors a plenty will occur)
I still think I am lazy especially compared to ER doctors and nurses.
Richard Heathfield
> I am in the zone, its been 8 hours, I have 8 hours more to go, what is
> the most anybody has programmed straight.
I don't know the world record, but my personal worst is roughly 8am to
11-30am (27.5 hours).
--
Richard Heathfield : bin...@eton.powernet.co.uk
"Usenet is a strange place." - Dennis M Ritchie, 29 July 1999.
C FAQ: http://www.eskimo.com/~scs/C-faq/top.html
K&R answers, C books, etc: http://users.powernet.co.uk/eton
Ian Woods
@posting.google.com:
Don't know about each sitting, but there's been a couple of times where
I've done this sequence over a few days:
1) work on code
2) run code
3) go to bed while code runs
4) get up when run is completed
5) go to 1
Far from the healthiest way to live, and I wouldn't advise it.
Ian Woods
gswork
I'd be happier to get 2-3 hours straight sometimes.
IME, you need about 30 mins to get into the right frame of mind, where
your programming 'flows', then about 2 hours is good - followed by a
break, a walk, a chat - something different. There must be a point at
which you get worse and worse, but sometimes you have one of those
days where you keep thinking of new ways, stuff to add - and you feel
compelled to do so. Planning ahead is good.
However, the reverse can happen too - where you may want to sit down
and concentrate but are continuously interrupted by; kids, family in
general, colleagues, scheduled meetings, social activities and so on.
On those days you can never get much beyond that 'starting up' phase
of 30 or so minutes, so you may spend a total of 4 hours at the PC but
actually only get 1-2 hours of programming done, because you keep
being 'short-circuited'.
Anyone else have that happen regularly, or can you just dive straight
back in without a need to 'get into it' again?
Darren Compton
"bigbinc" <big...@hotmail.com> wrote in message
news:d1b33313.03111...@posting.google.com...
I have done a 26 hours stint once but that was not too bad but I once did 60
hours in three days using the four hours off to walk around and clear my
head and get some breakfast, but on the fourth day on the way back to work I
walked into a lamp post and decided that I was too tired so I went home and
slept for about a day and a half.
CBFalconer
> bigbinc wrote:
>
> > I am in the zone, its been 8 hours, I have 8 hours more to go,
> > what is the most anybody has programmed straight.
>
> I don't know the world record, but my personal worst is roughly
> 8am to 11-30am (27.5 hours).
Never got anywhere near that (I tend to fall asleep instead), but
I have been equally foolish while playing bridge. At least the
competition was similarly idiotic, or it would have cost me even
more money.
--
Chuck F (cbfal...@yahoo.com) (cbfal...@worldnet.att.net)
Available for consulting/temporary embedded and systems.
<http://cbfalconer.home.att.net> USE worldnet address!
goose
18 hours (or thereabouts) per day, 7 days a week, for 3 months.
i dont recommend it. unless of course you regard sex, the
ability to communicate verbally and non-fast-food as something
that only happens to other people :-)
also have done the code, compile, run, test, sleep eat cycle
for about three months as well ... thats yet another quick
way to learn more about the phrase "only human" ;-)
goose,
Michael J. Fromberger
big...@hotmail.com (bigbinc) wrote:
My longest single marathon hacking run to date (and probably ever, as I
don't intend to repeat it) was 56 hours. I must say, the code that
resulted is not what I would consider my best work, although it was
impressive for its volume, if nothing else. :)
-M
--
Michael J. Fromberger | Lecturer, Dept. of Computer Science
http://www.dartmouth.edu/~sting/ | Dartmouth College, Hanover, NH, USA
ak
<Michael.J....@Clothing.Dartmouth.EDU> wrote:
>In article <d1b33313.03111...@posting.google.com>,
> big...@hotmail.com (bigbinc) wrote:
>
>> I am in the zone, its been 8 hours, I have 8 hours more to go, what is
>> the most anybody has programmed straight.
>>
>> (I know this is bad and errors a plenty will occur)
>>
>> I still think I am lazy especially compared to ER doctors and nurses.
>
>My longest single marathon hacking run to date (and probably ever, as I
>don't intend to repeat it) was 56 hours. I must say, the code that
>resulted is not what I would consider my best work, although it was
>impressive for its volume, if nothing else. :)
>
>-M
6 weeks of 18 hours a day 7 days a week - at site by customer
as the management realized that the solution was not at all
what the customer expected. LOL - was extra fun too since
we went there with the intention of installing over the weekend
and then go back the following Monday.
/ak
Programmer Dude
> I'd be happier to get 2-3 hours straight sometimes.
This can be a *severe* problem for me at work. Part of my job
description is providing PC and software support for the department,
and the "help me" interruptions some days result in almost zero
programming work getting done.
I keep telling my boss about the studies that have shown it can
take 20 minutes or more to rebuild the mindset after an interruption,
but they apparently value having an "answer man" more than having
a productive programming. (Sometimes I call in "sick" just to get
some work done!)
> IME, you need about 30 mins to get into the right frame of mind,
> where your programming 'flows',
That's about right by my experience, also. I've used the metaphor
of those "jugglers" you used to see on variety shows that had a
dozen or more plates spinning on slender poles. They'd run around
like mad keeping the plates spinning.
Being interrupted lets all those plates go crashing down. :-(
> then about 2 hours is good...
Once I'm in the zone, 2 hours is about the *minimum*! More typical
for me is 4 to 6 or more. (I'm often amazed at how much time has
passed.)
> There must be a point at which you get worse and worse, but
> sometimes you have one of those days where you keep thinking of
> new ways, stuff to add - and you feel compelled to do so.
So much, I think, depends on your mental state and degree of rest.
When I'm at top of my game, an 18-hour day passes like a breeze.
(And when I'm at the bottom,... might as well do something else!)
> However, the reverse can happen too - where you may want to sit
> down and concentrate but are continuously interrupted by; kids,
> family in general, colleagues, scheduled meetings, social
> activities and so on. On those days you can never get much beyond
> that 'starting up' phase of 30 or so minutes, so you may spend a
> total of 4 hours at the PC but actually only get 1-2 hours of
> programming done, because you keep being 'short-circuited'.
Yep. Exactly. Hate that!
> Anyone else have that happen regularly, or can you just dive
> straight back in without a need to 'get into it' again?
Only if I've been very, very into it for a few days. But for the
most part, seems ya just gotta get all those plates going again.
--
|_ CJSonnack <Ch...@Sonnack.com> _____________| How's my programming? |
|_ http://www.Sonnack.com/ ___________________| Call: 1-800-DEV-NULL |
|_____________________________________________|_______________________|
bigbinc
I have made a complete 130% switch to java and any job I do(hopefully)
in the future will be java based, point being that I can do several
hours of java and not go nuts. It is a plug-play language, I have
almost gone nuts on memory errors with C++, sometimes because of bugs
with someone elses library(what a pain);
Anthony Ventimiglia
> big...@hotmail.com (bigbinc) wrote in message news:<d1b33313.03111...@posting.google.com>...
>
> I'd be happier to get 2-3 hours straight sometimes.
>
> IME, you need about 30 mins to get into the right frame of mind, where
> your programming 'flows', then about 2 hours is good - followed by a
> break, a walk, a chat - something different. There must be a point at
> which you get worse and worse, but sometimes you have one of those
> days where you keep thinking of new ways, stuff to add - and you feel
> compelled to do so. Planning ahead is good.
>
> However, the reverse can happen too - where you may want to sit down
> and concentrate but are continuously interrupted by; kids, family in
> general, colleagues, scheduled meetings, social activities and so on.
> On those days you can never get much beyond that 'starting up' phase
> of 30 or so minutes, so you may spend a total of 4 hours at the PC but
> actually only get 1-2 hours of programming done, because you keep
> being 'short-circuited'.
>
> Anyone else have that happen regularly, or can you just dive straight
> back in without a need to 'get into it' again?
That is very similar to my thinking, I definitley have a small time
frame where it takes me to get flowing, that time depends on what it
is I'm working on, my motivation, and how fresh it is in my mind, So
if it's something I've not had to look at for a few months, sometimes
it may take an hour to get going, but if it was something I've been
working on for the past couple weeks, it's pretty short start time.
Usually I find my marathon sessions happen when I'm being very
productive, If I have ideas in my mind I usually can't sleep, I've had
many times in my life where I lay in bed for ten minutes thinking
about something, then get up and work until dawn on whatever it was
(this applies to more than just programming form me).
The point you make about the interruptions is interesting as
well. There are some days when you just can't get anything done for
what ever reason, and that can be frustrating, but I also have times
when I am away from my computer for a reason and I am still quite
productive. I might spend a day travelling or a weekend away and have
a problem on my mind, with no way to work on it, often I'll think
about it the whole time and find an answer so when I return to
working, I sit right down and do it off the top of my head.
--
(incf *yankees-world-series-losses*)
David
Hi Chris,
I was going to stay out of this thread but I'll add my bit.
I've done the long term coding cycles to rush things through.
At a former company we (most of the team leads) enjoyed staying
through the night and refactoring much of the code. We'd get
dinner brought in and stay into the wee hours or sometimes be
there the next day. The impetus was usually that someone had
a great idea and we agreed to implement it. These were usually
big rewrites to all modules so we'd make an effort to pull
together and make the changes. I've done 36 hours of this but
only while its fun and profitable. I quit before a mistake will
be made.
The answer I wanted to give back a few days ago when this
thread started was something like 3-4 months. Coding is just
part of the effort, so why judge yourself on just that. Sure
I've had 1000 lines in a day when its rather straight forward.
For a long time now I only code (write it on the computer)
if its already designed, written, debugged, and tested in my
head.
I'll be the first to admit that not everyone can or should
do this, and there are times when I know that coding isn't
the best task for the day. I'll focus on what I can do.
Several people have commented on being in the zone. My
solution from almost day one as a paid programmer was to
set aside time for certain tasks. I'd come in early and
write code that I'd designed the previous day or night.
This "in the zone" time was deliberate and only lasted a
couple hours. I had a real office with a door back then.
When others came in and the phone and email started
I'd shift to support roles. If the code that was completed
earlier was shippable, I'd give it to the customers that
needed it. My task at the time was for an in house graphics
product. We had a six month release cycle, but I considered
that only for everyone. If a user (high end or low end)
had a suggestion or need, I'd think about it and eventually
implement it. When it was ready, I'd ship it to just them.
If they liked it, it would make the major release. If we
needed to rework the ideas, so be it. I also had the rule
that such requests weren't changes to an existing behavior.
That needed more formal review. But simple tasks that make
things easier for the user were important. We had a couple
people who tried to automate everything. We'd add lots of
bells and whistles for just them. The great thing was,
these people would call up with an idea and sometimes it
would be coded and ready to ship before I got off the phone
with them. Usally it was just a day or two turn around.
I still kept up with the expected workload. I keep a
couple hours to do docs, designs, and other paperwork.
People need to know what you've given them.
So, I'd advise those of you who can, to plan your days
so that you can find the zone more often. If you can,
support the customer better. Make time to get out of the
cube and visit with the customer. Watch them make mistakes
or do things you hadn't thought of. A customer support role
can be good for this too.
Have fun and enjoy your work,
David
bigbinc
I also think programming for fun and programming for a job, to meet a
deadline are two different things. For fun is a lot easier to go at
all night, trying to meet a deadline is nothing but sweaty palms and
biting of the nails and the question(should I change careers?)
Corey Murtagh
<snip>
> I also think programming for fun and programming for a job, to meet a
> deadline are two different things. For fun is a lot easier to go at
> all night, trying to meet a deadline is nothing but sweaty palms and
> biting of the nails and the question(should I change careers?)
I've always had the same reaction to them, for the most part. The only
real difference is that when I'm doing something for a job, I can't just
put it aside when I get bored with it. But even when it's boring, I
still have no problem throwing an all-nighter at it to get it done.
Different strokes for different folks, I guess.
--
Corey Murtagh
The Electric Monk
"Quidquid latine dictum sit, altum viditur!"
Programmer Dude
> For a long time now I only code (write it on the computer)
> if its already designed, written, debugged, and tested in my
> head.
Just for another point of view (and something of a subject change),
we had a discussion here recently about a different way of working,
of "design through coding". The idea is that experienced programmers
can sit down and start roughing out an idea in working code and
incrementally refine it.
This can be dangerous if you are unwilling to throw away all that
work should you decide you've gone down a wrong path. But this
isn't much different than throwing away your notes or mental ideas
in a similar situation.
The advantage is that, if you get it right, you're ahead of the game.
There's also the fact that ideas in your head, or on paper, may look
good, but not turn out so good when implemented.
This rather goes against the 90% design/10% coding ideal, but it
can work well (IME, anyway).
Some languages facilitate this more than others. Lisp, e.g., is a
**great** language for this sort of "doodling". Something like Ada
probably isn't.
Back on subject....
> Several people have commented on being in the zone. [...]
> This "in the zone" time was deliberate and only lasted a
> couple hours. I had a real office with a door back then.
[sigh] That would be so nice. Yes, I agree, one can set up ones
work situation to provide "zone time". I tend to work 9-6, while
most of my co-workers (I'm the only programming in the department)
work earlier "shifts" (we have flex time). My zone time starts
about 4 pm.
Or, as I mentioned, sometimes when I call in "sick". (-:
ak
<Ch...@Sonnack.com> wrote:
>David wrote:
>
>> For a long time now I only code (write it on the computer)
>> if its already designed, written, debugged, and tested in my
>> head.
>
>Just for another point of view (and something of a subject change),
>we had a discussion here recently about a different way of working,
>of "design through coding". The idea is that experienced programmers
>can sit down and start roughing out an idea in working code and
>incrementally refine it.
>
>This can be dangerous if you are unwilling to throw away all that
>work should you decide you've gone down a wrong path. But this
>isn't much different than throwing away your notes or mental ideas
>in a similar situation.
>
>The advantage is that, if you get it right, you're ahead of the game.
>There's also the fact that ideas in your head, or on paper, may look
>good, but not turn out so good when implemented.
>
>This rather goes against the 90% design/10% coding ideal, but it
>can work well (IME, anyway).
>
>Some languages facilitate this more than others. Lisp, e.g., is a
>**great** language for this sort of "doodling". Something like Ada
>probably isn't.
I think it depends a bit on the project, if one is moving in
well-known waters vs using new technology/new 3rd party libraries. In
the latter it is a must to tinker a bit and see if it works as the
docs state plus it may be a bit faster than reading 500 pages of
user-unfriendly manuals. I reckon that one should at any rate throw
away the code after figuring out how it works, well partly at least -
building on the attempts could create rather difficult to read - hard
to maintain programs.
/ak
Michael Littlejohn
I am a Mohawk College student, so I have spent DAYS programming without
sleep. At the end of last semester, I spent approximately two and a half
days programming in C++ without a wink. The only thing keeping me from
passing out were doughnuts and espresso.
BTW, the program I was making calculated statistics for the New York subway
system such as average direction, rotation, G force, etc.
"Ian Woods" <news...@wuggyNOCAPS.org> wrote in message
news:Xns943619D2EC29...@217.32.252.50...
Darrell Grainger
> I am in the zone, its been 8 hours, I have 8 hours more to go, what is
> the most anybody has programmed straight.
When I was young and foolish I would routinely program for 2 days
straight. I was a hacker. I was self-taught. I was in highschool and had
the summer to do what I wanted. I had a mother who would bring me meals
and make me eat.
As an adult who has a job and a life, I'd say the maximum I program for is
6 hours straight. On average I program for 3 or 4 hours straight. I find
taking a break every once in a while improves the quality of my work. My
employer prefers a decent amount of good code rather than a huge amount of
bad code. Most importantly, I work on a team and code I write might need
to be maintained for over a decade.
> (I know this is bad and errors a plenty will occur)
Glad you know it. Rather than measuring how many hours you can program
straight (which strikes me as a macho thing) why not measure something a
customer would like... how long does it take you to code a project with
zero defects that meets the customer requirements? You'll find this more
challenging and people will be willing to pay you for this skill.
--
Send e-mail to: darrell at cs dot toronto dot edu
Don't send e-mail to vice.pr...@whitehouse.gov
William
news:3FBA57AB...@Sonnack.com...
>
> Just for another point of view (and something of a subject change),
> we had a discussion here recently about a different way of working,
> of "design through coding". The idea is that experienced programmers
> can sit down and start roughing out an idea in working code and
> incrementally refine it.
>
> This can be dangerous if you are unwilling to throw away all that
> work should you decide you've gone down a wrong path. But this
> isn't much different than throwing away your notes or mental ideas
> in a similar situation.
I frequently code as part of the design process, but not usually with
the idea that what I code will be in the final product. What I'm usually
doing is proof of principle work to see if the language, tools, techniques
are applicable to the proposed solution. What I end up with, if it
succeeds, will probably turn up as the basis for the final product (more
rarely it will be the final product) but I don't get hung up on that idea
at the design stage. (It's also something of a mental un-blocker to put an
idea into a tangible form - I use drawing programs a lot in design as
well. I'm visually-oriented.)
-Wm
Corey Murtagh
> On Sun, 16 Nov 2003, bigbinc wrote:
<snip>
>
>>(I know this is bad and errors a plenty will occur)
>
> Glad you know it. Rather than measuring how many hours you can program
> straight (which strikes me as a macho thing) why not measure something a
> customer would like... how long does it take you to code a project with
> zero defects that meets the customer requirements? You'll find this more
> challenging and people will be willing to pay you for this skill.
I don't know about you, but some of my best code has come near the end
of marathon coding sessions. Elegant, defect-free, and strangely enough
very readable. I'm an intuitive type anyway, and this really comes
through later in a session.
Hands up everyone who's ever woken up the day after a marathon, looked
at their code and gone "Whoa, pretty. Did I write that?" :>
Not to say it happens all the time. My defect rate is about normal (for
me) after about 20 hours, and seldom increases if I continue. Once my
throughput starts to drop, I'm about done.
Of course the /real/ indicator that the session is over is waking up to
find keyboard imprints in my forehead :>
Programmer Dude
> Hands up everyone who's ever woken up the day after a marathon,
> looked at their code and gone "Whoa, pretty. Did I write that?" :>
[raises hand]
Of course, I've also had the other feeling, "What in the WORLD was
I thinking?!?!?!"
But it does give one a nice warm fuzzy feeling when one works with
code one wrote ten years ago and finds it reasonably easy to do.
Olathe
Of course it's not healthy. Those gotos are evil.
bigbinc
People always mention the 75% design/90% design and 10% coding and
always say that the hero programmer will suffer burnout and create
buggy code. I think the great systems and the systems that you havent
even heard about yet will require strange, abnormal programming
practices. I pretty sure that Linus of linux didnt plan out his OS
with UML diagrams and 90% design methodology, I bet he was up at 3am
night after night working on his OS of his. I think the freaks will
advance computer technology.
Daniel Lidström
I think a reason for the code to be better after long hours of coding
(it has happened to me too!) might be because the brain has everything
fresh in memory. If I leave my code for a couple of days it takes a
longer time to write the first couple of lines, and I may even make
mistakes I have already done once. But once I get the complete picture
of how my code works, I am better able to write the best code, which
usually is after some hours.
--
Daniel
ak
Lidström <som...@microsoft.com> wrote:
//I think a reason for the code to be better after long hours of coding
//(it has happened to me too!) might be because the brain has everything
//fresh in memory. If I leave my code for a couple of days it takes a
//longer time to write the first couple of lines, and I may even make
//mistakes I have already done once. But once I get the complete picture
//of how my code works, I am better able to write the best code, which
//usually is after some hours.
the longer it takes to resume after a break, the
crappier the code :p
/ak