Raspberry Pi 2 Model B

Actually I know where there's a Pygame program that *is* in production and runs 24/7. I'm informed it's been running for about 3 years :)

You were doing well up to here.

Can't say I've had any serious issues myself, and the changes seem to make sense. Also the well advertised upgrade path started with V 2.5 I think.

I had a Python program running in 2.7 for a while, about 1100 lines. I decided to update it to Python 3 when I moved it from a home server to a VPS and added a web front end. It took about 2hrs to port it. 2to3 did most of the work and the hard bits were the bits where I written the code like the C programmer I am.

I think the payoff is in smoother Unicode handling and support for larger character sets and non-English and non-European languages. I haven't converted yet and got bitten when I tried to label an input text box with

everybody's; bug reports came in and it had to change back to the inelegant "uS".

Mel.

"A Real Programmer can write FORTRAN in any language" ...and anybody who has been around for a while has seen it done far too many times.

No, there's worse - much worse. Try FORTH or (shudder) APL.

How about LISP (Lots of Irritating Single Parentheses)?

Indeed, I've always thought it was versatile and productive (once you figure out what the compiler's trying to tell you, or should have told you).

On Tue, 24 Feb 2015 21:08:05 +0000 (UTC), Martin Gregorie declaimed the following:

I used to be a master at VMS F77... But there is no way I could manage to write FORTRAN in PERL (heck, the few times I've had to help someone with PERL, I started by first writing the problem concept in Python -- then mapping the Python to the index of the Camel book)

On Tue, 24 Feb 2015 21:09:57 +0000 (UTC), Martin Gregorie declaimed the following:

What's so bad about (using a transliteration)

{if you number a deck of cards sequentially, that just dealt four 5-card hands}[from a range of 52, generate 20 random values with no duplicates, reshape into a 4x5 array]

On Tue, 24 Feb 2015 16:14:06 +0000, Rob Morley declaimed the following:

And I've experienced one such "programmer"... this (unnamed) person basically looked for code snippets others had written and combined them to create a program... Or, in the specific case, four programs.

Problem: the programs were performing GPIB control of four separate hardware boxes. EACH program started out by opening the GPIB device, which issued a reset of the bus. As a result, each program was wiping out the parameters that had been sent to a prior box by a different program.

Yes, each program worked.

NO, the system of programs did not work!

We delivered to the end user site, and found the system failed... This person went back home on schedule... I had to get an extension to my gate pass, change airlines, etc. while debugging -> rewriting this persons four programs into a single program that did one GPIB reset and then commanded all devices in sequence.

My one and only time programming GPIB.

Rexx is even more readable than Python. I've got Regina Rexx on all my RPis.

"With great power comes great responsibility."

I'd expect at least a line or two of comment per APL token, which kind of negates the notational terseness--unless the notation supports a useful algebra.

manage

You know what I mean: bet you could write Perl as though you were writing FORTRAN, i.e. leave out any code structures or loop constructs that weren't supported in FORTRAN 4.

I've seen COBOL written by somebody who didn't know that, in the absence of conditionals or GOTOs, you fall through from one paragraph to the next, e.g. he code looked like this:

.... ADD LINE-ITEM TO INVOICE-TOTAL. MULTIPLY C BY DISCOUNT-RATE GIVING DISCOUNT. SUBTRACT DISCOUNT FROM INVOICE_TOTAL. GO TO NEXT-PARA. NEXT-PARA. ......

It was said that Ken Iverson (who designed and implemented APL) once wrote a text editor in 25 lines of APL. It was also said that when he needed to modify the editor 6 months later he couldn't do it.

On 02/24/15, Tony van der Hoff pondered and said... Tv> This could be descending into a religious war, but IMO, Perl is the most Tv> annoying, obfuscated, non-intuitive, convoluted, write-only language Tv> ever dreamt up. Larry wall is a linguist, not a computer scientist, and Tv> it shows. Tv> Tv> Of all the interpreted languages, give me Python any day.

Hi all,

I laughed when I read your comments Tony :)

Thanks everyone for your lively debate regarding the relative merits of Python vs Perl. I have also been interested to read of the Python 2.x vs 3.x debate. My general take Python is favoured by a number in the group posting here with a heavier lean towards 2.x versions.

I'll likely lean that way as well for now and may start asking the odd newbie question of you (if that's ok?) as I start my programming adventures.

Sorry not to have been more active in recent days, life seemed to get busy.. you know how it is.

Best, Paul