Apps

Poll: Developers, what workplace change do you want to see in 2011?

If you could an aspect of your development process, would it be a tool upgrade, different tools, a backend system change, or something else? Let us know.

Not many of the developers I know are 100% happy with the technology they use in their workplace. Some developers see the world around them adapting new tools, and they are stuck with the old versions. Other folks wish they were using entirely different languages, tools, and so on. I would not mind seeing a change in my environment's backend systems; TFS is way too heavy for our needs, and I am really enjoying my time with Mercurial on personal projects.

I know a lot of developers who feel like their processes need a major revision. If you could change one aspect of how you do your development, what would it be?

J.Ja

About

Justin James is the Lead Architect for Conigent.

37 comments
Kim SJ
Kim SJ

More and better screens is an instant productivity win, without having to rebuild your environment. Every developer should have a minimum of two 1920 x 1080 monitors, imho.

alanhoward
alanhoward

I'd like to see me back in the workspace.

navyhatch
navyhatch

How about a more reliable grammar checker..."If you could an aspect of your development process,..." taken from the email presenting this article. Although as a FED I vote for upgrading and acquiring the tools needed to get the job done at the least cost to the tax payer.

Tony Hopkinson
Tony Hopkinson

Also for management to be agile as opposed to just paying it lip service by using 'agile' methodologies Want to cross every i and dot every t before work commences, and everything is must have priority one, but we do have a burn down chart. :(

john.a.wills
john.a.wills

I want to get onto a non-UNIX mainframe, with persistent line numbers in the code, cross-reference etc. automatic in the compilation listing, etc.

jhoward
jhoward

I want to get away from running our custom backend services on Windows. Besides the inherent down time of Windows Update requiring reboots every month the licensing eats into our budget meaning our dev environments are often a bottleneck. I will agree however that UNIX vs. Linux gives us headaches as well but more because most of us are more familiar with Linux and the similarities become more of a hindrance than if they were completely different.

oldbaritone
oldbaritone

Mainframe? Line Numbers? XREF? Just charge up your flux capacitor, hop in the DeLorean, and go back to 1960. You'll be thrilled. COBOL. TSO. JCL. Hollerith cards. Green bar. Have a blast!

Sterling chip Camden
Sterling chip Camden

... the behavior of the compiler you're using?

john.a.wills
john.a.wills

UNIX does not have fixed line numbers for text files. Therefore any line numbers the compiler might give are pretty useless for turn-around editing. Therefore the compilers are written without such features. This does not have to be so: there are build-ons for UNIX which allow fixed numbers, but they are expensive and my employer has not thought to buy them. Some of my colleagues have never even known decent systems.

Sterling chip Camden
Sterling chip Camden

... I don;t use listings. I use the compiler error messages to tell me what line number to jump to. If I add code ahead of a line that I want to get to later, then of course I have to add an offset. So?

john.a.wills
john.a.wills

you must have the line numbers all wrong as soon as you insert or delete a line higher up, so all your editing planning goes haywire. My compilers (SQR and PL/SQL) do not give listings with even the right file line number, let alone something I can use for a big editing run.

Sterling chip Camden
Sterling chip Camden

The line numbers reported by the C compiler, Synergy/DE, Ruby, Haskell, and Perl all seem to line up with vim +n for me.

Justin James
Justin James

I saw this problem all the time ages ago... it still isn't fixed? There is zero correlation between what the text editor considers a "line" and what the compiler calls a "line", finding bugs is a real needle/haystack affair. Indeed, I was having this issue last I worked in Perl, the interpreter's idea of "line 52" was not my text editor's idea of "line 52". J.Ja

Kam Guerra
Kam Guerra

I'd like to see an employer pay for the PC/Laptop to do the job. I'd like to see an employer pay for the software to do the job. Show me a company that does those two things, and I'll show you a company that has no job vacancies and a line of people a mile long wanted to get in.

Sterling chip Camden
Sterling chip Camden

... when CFOs complain about the high costs of maintaining employees, and then they refuse to invest a few hundred dollars each in the equipment that would help them to be truly worth the rest of their investment.

gechurch
gechurch

Add me as an echo to this comment. I do consultant work for a dozen or so different companies, and a couple of them amaze me with the PCs they expect their employees to use. I have replaced two Pentium III's in the past two months, which are extreme examples. I have clients with quite a few P4 2.2-3.0Ghz machines too. This makes so little sense. The cost of employees is generally by far the biggest cost for a lot of companies. Wasting employee time like this is near criminal. Add on top of that the additional stress and frustration it creates for the users... there's nothing worse than being on a tight deadline and waiting minutes for something that would take 5 seconds on a decent PC.

Kim SJ
Kim SJ

It takes me at least a month to get a new PC the way I need it to operate at the same efficiency as my old PC. So you don't want to do that too often.

Justin James
Justin James

I see it time and time again. An employee making $70k plus benefits, but being give the same $300 low end PC to do... I don't know... DB work which requires massive data files, or graphics work, or development work... or whatever. Or tools like Resharper... it doesn't save me a LOT of time, but it helps me do things with a lot more accuracy, which is worth the money. It is incredibly short sighted for companies to treat these small expenses as a big deal. Fact is, a new PC costs 2 day's worth of overhead for an employee. An ergonomic workstation with all the bells and whistles costs a month's pay, and lasts for a decade. It's penny wise, pound foolish. J.Ja

AnsuGisalas
AnsuGisalas

is all the fluckheads who speed ahead, and then sit there, in front of the lights, having to slowly accelerate back to speed when it changes. It means I have to slow down and wait for them, just because they can't rub two braincells together ;) If they'd take it slow and deliberately, timing their arrival at the lights, we'd all make it through faster.

ian
ian

When I lived in Germany in the early 70's (Dortmund 73-75) the main autobahn round the town had speed indicators ahead of the traffic lights. The speed indicators gave a suggested speed ( eg. 45kph). If you stayed with that speed, you sailed through every light - no probs. That was not the allowed road speed (60kph) but the best speed to hit the lights just right. I remember they also had cameras on the lights. I ran a light once (same time frame) and received a photo in the mail showing the rear of my car including the licence plate, the light on red, time and speed.The photo cost me dm200. Good, efficient technology 40 years ago. Now, living in the US, I stay pretty much around the posted limit. It's good to watch all the traffic speeding off. I catch up with them at the next lights where they've been waiting five minutes for the lights to change.Increased fuel consumption on start, wearing of brakes at slow down, increased stress - for what? I look far enough ahead at the lights and gauge my speed accordingly. I rarely need to stop at lights. and it p**sses the heck out of the people behind me till they realise they don't need to stop either.

oldbaritone
oldbaritone

Since I usually run about 7 mph over the limit, I want to be sure the gauge is correct. Simple method: use a stopwatch on a road with distance markers. 3600/(number of seconds to travel distance) = speed in (distance/hr) Use the same formula for mph or km/h.

HGunter
HGunter

The reason speedometers almost never show the correct speed is that they measure the rotation of a wheel, and produce a number based on the assumed diameter, hence circumference of the wheel. The tyre wears away continuously, so the diameter keeps shrinking and the indicated "speed" rises with the tyre wear. Add in the inherent error factor of the device, and you're probably somewhere within 5 to 10% of the correct speed. This is what makes the availability of accurately-measured distance markers such a good idea - you can calibrate the odometer to determine the error ratio, and use that to calculate your speed, or you can measure the time between markers - at 60 kph you do 1 km per minute, at 30 mph you do a mile in 2 minutes, and so on. But the speedo needs to be interpreted.

Sterling chip Camden
Sterling chip Camden

... if I had $30, I could get home -- but it had to be sans hotels. Now it's more a matter of pride.

AnsuGisalas
AnsuGisalas

to remove a defense of "my speedometer said I was under the limit", as the real speed will always be lower than the speedometer reads. I think it has to do with the measure uncertainty, they compensate to ensure it's at least never showing too little. But apparently that's just over here, then...

AnsuGisalas
AnsuGisalas

that roadside hotels must be really awful to make it worth it... Of course, getting home to family a day sooner, after a work trip can make sense, but in my case the "better late than never" comes to mind. I'd not be able to drive for that long, unless I drive in my sleep, and over here roads aren't straight enough for that... bedrock, you see. :D

Sterling chip Camden
Sterling chip Camden

That's about all I can do these days. I used to drive 24 straight when I was in college. I once did 30 straight when I got lost on my way home trying to take a shortcut. The US is a big country. Actually, I can still do 14 hours in one go -- about once a year.

Justin James
Justin James

... which I know to be accurate. Over here, police often set up signs that post the speed limit and have a speed radar in them to show you your actual speed (to remind folks to slow down). I see these every few months and always verify my speedometer's reading against the radar readout. The last thing I want is to get a speeding ticket due to an inaccurate speedometer! :) J.Ja

AnsuGisalas
AnsuGisalas

there's the speed we think we're driving, and then there's the speed we're actually driving. Most people think they're driving a lot faster than they really are. Especially over long distances. If we set our "target speed" to 85, we end up driving around 65 anyway, on account of other traffic and traffic lights and other such. A road has a "traffic permeability" specific to it, depending on it's layout and other users, and in the long run it stays pretty even. Often it turns out that trying to go faster simply fails in the long run - average speed stays around the road-specific permeability, or at least stays close to it. Rates of change and rates of flow - it boggles the mind. Must be fun to model traffic... BTW 800 miles, driving 9 hours straight?

AnsuGisalas
AnsuGisalas

Some lights are pulsed - have to be on the same control grid of course. And they have to be worth pulsing. Did you measure your own speed as the real speed, or as what the meter reads? Over here at least, the meter shows about five more than the real speed.

Justin James
Justin James

I've noticed that many roads with lights have a particular "rhythm", where if you make it through a yellow light (or get through a light at the right speed just as it turns green) at some magic speed, and maintain that magic speed perfectly (cruise control), you can often go 10, 15 miles without stopping, so long as there is no traffic, even though the lights may be every mile or two apart. I first discovered this on Route 18 in NJ, between New Brunswick and the exit to 527 (past there, the road changes dramatically and loses the lights every mile thing going on). The "magic speed" can be multiplied with the same results. What's really odd, is that the speed limit is often not this magic speed, at least on the roads I've discovered it on. It's close to it, but not it. On Route 18 (where I first found it), it was 50 MPH (45 MPH road). The difference in total travel time between 50 MPH and 45 was MUCH greater than the extra 5 MPH, because you slipped through the lights. 35 MPH was also faster than 45, because the increments were 15 MPHs, and 35 MPH would have you slide through all the lights, as would 65 MPH. I think the idea was that the road designers figured that at night, everyone would go 5 over, and in traffic, people would be going a bit slow. I've since discovered this on a number of roads in a few different states. I don't know if it is just lazy street light timing (they all turn at the same time, and just happen to be spaced about evenly apart) or on purpose. J.Ja

Sterling chip Camden
Sterling chip Camden

If you're traveling 800 miles (here in the US), the difference between the 65 mph speed limit and the 85 that everyone travels could save you as much as 3 hours, which could mean the difference between making it in a day and having to pay for a hotel. Of course, I'm convinced that US speed limits are determined based on the expectation that everyone exceeds them by 20%.

Justin James
Justin James

It's funny, a few years ago I committed myself to obeying the traffic laws 100%, and I've stuck with it. I consistently arrive at places just as fast as I used to, but everyone is convinced that I don't. In general, people don't understand how think critically about numbers. In this case, the absolutely amount (the cost of the item, the speed of the car) is not the relevant number, it's the second order number (cost of employee, time saved on the drive) which is much more important. Comparing first order numbers only work when comparing the same thing (two employees' salaries, speeds of two cars). :) And this is reason #472 that I lament the decline of math education in the US! "When will we ever need this?" should be answered with "all the time, but you won't realize that you are using it, so shut up and pay attention!" by teachers and parents. J.Ja

AnsuGisalas
AnsuGisalas

That's a bit like speeding on the motorway. They haven't made the calculation for potential savings. I used to go 120 kmph on the motorway over a 10 km stretch, because "I was in a hurry"... Then one day I took a deep breath and calculated to maximum saved time; 120 kmph > travel time 5 mins. 100 kmph > 6 mins. And the amount of extra stress and risks I took on myself for that one minute saved... That's why I think the human subconsciousness isn't good at maths... sure it can do basic stuff, but anything with rates of change is beyond it, as is - apparently - how increasing speed gives diminishing returns.

Sterling chip Camden
Sterling chip Camden

... but I really meant "try"

Justin James
Justin James

The Mercurial experiment has been great for personal use. My next step is going to try to see where I can use IronRuby instead of C#, especially now that version 1.1 has been out for a bit and integrates with VS2010. That's the deal maker for 99.9% of .NET devs. :) J.Ja

Editor's Picks