the OP said:
So, I should mention that this script that analyses the logs that I'd like to write would help with the everyday work.
Yes, because it can effectively free up your time.
the OP said:
I should also mention to him that writing this script will make me go deeper into understanding the logs and therefore would enhance my skills.
Yes
th OP said:
Should I mention to him that I see myself as a programmer in the future and therefore I want to have programming part of my job?
A tougher one to answer because it depends on your your Boss's attitude and whether or not your projecting yourself as "willing to learn", "I want to get out of here" or "I would like to advance here".
Thinking out loud:
Remember Who, what, where, when and Why?
Who: You,
What: You have a problem of some sort
Where: You or IT (Work, Work at home?)
When: How long to you think it would take
Why: Why do you need it.
* I'd like to talk to you about an idea I have. Can we schedule some time. I need about x minutes to present it.
* I'd like to improve my efficiency of going through the log files by trying to write some scripts to help me analyze them. I don't foresee this impacting on what I do now, but I would really like to work on this in my spare time. I could, for instance, give you an update at least once every two weeks on my progress.
What if I give you my ultimate list of requirements (what I'd like it to do) by next Friday and what I think is essential. Now, you could present the ultimate list or not, but it's IMPORTANT that you design your program using the ULTIMATE list. Trust me.
Aside: I have always learned that when specifying software and hardware, it's important for me to plan for the Ultimate system. When you do that, it's easier to add stuff rather than to re-write the package. That's how I managed to not have to rewrite a software package that was used for 16 years. The technology required the re-write.
In one case, I anticipated spreadsheets, but they didnt exist at the time. PC's didn't exist at the time, but in my specs, I could have very easily added write comma separated data.
We wrote the program so it could do recipies as a set of @filenames. Somebody, just like you, and in his spare time created a FORTRAN program to CREATE the recipe files.
The program we wrote was a proof of concept PID controller for 7 heaters with recipe controls back in the early 80's. The system was written in FORTRAN on a Real-time operating system. It also required a real-time and a non-real-time task and used a VT100 style terminal. Definately a radical idea for it's time. Removable Hard disks were 10 MB and about the size of a microwave.
We ran out of memory. It was hard fitting this thing in 28K of 16 bit words with disk overlays. We were limited in being able to used thermocouples in groups of 4 and they had to be R,S, J or K. There was no spreadsheet output, no Ethernet (didn;t exist) and files had to be transferred via Kermit. There was no ability to add it.
One of the processes used a "C" thermocouple. While the software was "designed" to handle it, it was unable too. We were also forced by management to use an OS system we were somewhat familiar with rather than use one with better memory management and we were told to write it in FORTRAN and not C. I knew C at the time, but not FORTRAN.
So, we did anticipate the ability to handle multiple thermocouple types, but it probably would have had to be done in hardware. We were tight on the Budget, $25K in parts.
Later, the DC power supplies used became a model to replace the AC supplies and a autotransformer and making a custom panel. So, the DC supply reduced the expense of making a voltmeter, ammeter and variac panel and gave us the ability to measure real power.
Even later, when the temperature controllers were obsolete and had to be replaced, recipie based controllers could do the job. The communocations option was ordered and in the next iteration was used, again someone wanting to improve what they had.
Log the time and temperature data with a laptop rather than by hand and to write a program to program the controllers.
In one case the researchers had to know the power of one of our AC powered heaters. The only solution was an AC 3 phase power monitor and a current transformer to measure 0-40 V and 0-30 A of a single phase output. That device added about $1000 to each heater and $1500 or so was the cost of a DC power supply.
There were a lot of challenges and technology upgrades were giving me fits. It was fun because we had the power supplies made for 208 input, the meters were analog and the current ended up being measured from an internal sense resistor.
Later a replacement had 0-5V out for voltage, 0-5V out for current, was a 1 RU case and had a universal input with digital meters for about the same price. This also created a problem with spares being used for swapping out, but to cut down significantly on rack space the 1 RU supplies HAD to be used as well as the 1/2 Rack, 4 RU height modules.
Power distribution became a headache as well.
I guess, I was trying to tell a story.
At one time I had a system that would do I-V-t measurements in an environmental chamber with programs to help analyze the data. I built in an RPN calculator with units conversion to input cross sectional area. You could type
PI 2 cm 2 ^ * or 100 um 10 mm 100 A * / which would do L/A in the formula L/(cross-sectional area) where L was 100 um, and the cross sectional area was 10 mm x 100 Angstroms. The units would automatically be converted to cm.
---
The KEY is management wants choices and recommendations. You need to make sure management hangs for the bad decision. I recommended that we use an SMU unit for measurements. Boss said no, re-use what we have and the SMU's use feedback and Feedback is bad. It took 15 years, but the latest incarnation uses, you guessed it, SMU's. The old boss retired.
Just like way back, we might have not had the memory limitation had we chosen an unfamiliar operating system, but we probably developed it faster. C probably could have been more portable, but FORTRAN went away. Those decisions were not mine. They were, however, recommendations.
You also have to watch moving targets. You make a decision based on time t, and you have to buy all the capital equipment at time t. At time t+delta t, the entire decision process would have changed and that's the time you needed to use the stuff.
IEEE-488 has pretty much been replaced by USB and Ethernet. In the early days parallel was the way and RS232 was slow. There were even PC based stuff such as EEProm programmers or CNC control that used the parallel port. Will they work on Windows 7? Not likely.