On Sun, 24 Jan 2010 22:52:40 +0000, Bob Eager wibbled:
While we're on the topic, I've been wondering what people would recommend if I wanted to get into radio links?
I don't much care if it's 433MHz, 868MHz or something else.
I'd like:
a) Very cheap and easy to interface (ideally without having to manage horrible timing sensitive programming, but could live with this if I had too) b) Range - about 20m through masonry, 40m free air. c) Bandwidth - few 10's to 100s kilobits/sec will be fine.
Zigbee looks interesting but may be a bit heavy with all the protocol layers. I'm happy to have something that lets me bang raw frames out and worry about simple ACK/retries and frame integrity myself.
I guess I'm after a good layer 1 solution, but I wouldn't object to something that comes with layer 2 built in providing it's not silly and restrictive (eg 16 addresses).
Any pointers? I see *many* modules available in all sorts of flavours, so it's a bit of "wood and trees" problem for me...
On Sun, 24 Jan 2010 23:31:01 +0000, geoff wibbled:
Ah. Thanks for the examples. That answers my musings...
You think each manufacturer would put in a bit of R&D and come up with one chipset and stick with it and possibly even one core design to all their boards (surely only the IO layers and software vary?).
But I can clearly see I give them too much credit...
I like you :)
and "bollocks".
(Don't read that wrong - I'm not into Rocky Mountain Oysters)
I fired that student. He wanted to do it all in perl. After 6 months he had achieved nothing. He had been co authoring an online science fiction story.
For example Glowworm have 5 different pcbs which are, to all intents and purposes, identical (OK, two have some neons on), all with different designs and part numbers
Almost every boiler in existence has a fan which is different to every other (chaffoteaux did achieve it for a time and the Glowworm ultimate series are similar from the 30 to the 80 (then the 100 and 120 are different)
Worcester and Vaillant both have quite complicated pcbs which many of them look identical to various other ones and have similar shape and interfaces - they could easily have been built for one pcb fits all and the differences performed in s/w, ... But ... no, a whole range of slightly different pcbs
On Mon, 25 Jan 2010 00:36:43 +0000, The Natural Philosopher wibbled:
Seems to be the curse of Britain. Engineering isn't respected, more so it seems to be actively disrespected. Companies are run by people with MBAs and not clue as to what the core of the company is about, more interested in puffing themselves and their mates up than actually makeing the company service better.
I'd love to work for myself... Trouble is I have no clue what to do. I'm a sysadmin, which doesn't help, though I do develop in a limited sphere (specialist network servers mostly)
Sysadmin work is more satisfying when you are permanant - make the stuff work, keep it working, make it better, less labour intensive so you can make something else work better. Hit and run contracting doesn't do it for me.
Knocking out widgets for mobile phones that 10000 people are happy to pay a quid for seems to be the way to do it these days.
Yes .. avoid 2.4 Ghz, too crowed by half these days. 433 or 868 but be aware that 433 has its problems with large signals from "ham" use and sometimes TETRA..
Yes, when's the last time you heard anyone proudly say my son or daughter is studying engineering at Oxbridge?.
Nope far, far, better accountancy or law or medicine 'tho the last one perhaps is almost eng in a way;)...
Right..
If you don't know what you want to do and you haven't dun it off your own bat then don't do it .. stay working for someone else its not all fun running your own outfit;!..
And a lot of the day to day headaches... Like finding/smarming/keeping customers, and relieving them of their money in a timely fashion. And boring crap like PAYE, NI and pensions.
I never thought for a moment being self employed would be easy. I did a spell of contracting once as a Ltd company and the paperwork was tedious
- and accountants robbed me of all the savings I had calculated by being Ltd status. I can handle tax, but fiddling^H^H^H^H^Hoptimising company expenditure and doing written accounts are just voodoo to me.
If I did - I would favour writing and flogging nifty software. No manufacturing/funding/supply issues and relatively small capital costs bar time. It's easy to sell (automatic download, licensing and money extraction website) - that just leaves the marketing. And depending on the complexity/importance of the software, the support issues. That's why I think some people would rather knock out widget phone apps because no- one who buys something for a few quid is likely to kick up a stink if it breaks. OTOH, software costing 100's-1000's is a different game.
The other option is an internet service of some sort. That keeps the grubby software in my hands (user compatability problems much lessened) and is probably quite suitable for a sysadmin who likes to keep servers ticking smoothly. Again - what? It's hard to find an angle, yet people keep popping up with small and random things that in hindsight seem quite obvious, so I conclude the pool of Good Ideas (TM) is not yet depleted.
A while back, I wanted to do a meta search engine specialising in house sales because I was frustrated about how crap and irregular most agents sites are and the fact you had to go to lots of different sites. That's more or less moot now because Rightmove + Property Bee firefox addon has achieved 80% of my idea - but it goes to show the idea was sound in principle.
I guess two routes to success go like:
1) Have popular idea;
2) Implement fast before some other sod does;
3) Make it and keep it the defacto standard.
or
1) Have very niche idea;
2+3 less important as less interested competition...
Well, it parses, because # in that context is a comment delimiter.
This however is a valid and functional perl expression:
$#_
(highest index of the default array @_)
But as you have thrown down the gauntlet:
1) It's mature;
2) Larry doesn't break crap randomly between releases - there are well advised deprecation policies;
3) It is stable at execution time
4) It is way more code efficient than C for string processing and a host of other common case scenarios, whilst retaining the flavour of C.
5) It has virtually one single place to go for extra modules (CPAN)
6) It's portable in a way C could only dream of.
7) It runs on practically anything from ARM upwards
8) It is relatively efficient for a script.
9) It's not python
10) It's not Ruby
(just joking re 9+10 - good for flamewars though...)
But just because you *can* write one line of obfuscated perl, or even perl Haiku, doesn't mean you *should*. Even the regexps, which are notoriously hard to read, can be written multiline, indented and commented which helps a lot.
HomeOwnersHub website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.