There are a lot of little niggles about the Humax, various bugs (many of which Humax have sorted out), but can anyone remember what it was like when we had VCRs?

Only one tuner, winding and rewinding, fiddling about with tapes, no EPG..........

I'm really chuffed with the Humax, ever since I bought it over 2 years ago; the niggles of missed recordings etc. are not as bad as on VCR days, and every upgrade has made it better and better, the series link stuff is amazing - al we need now is the broadcasters to get their end of it right all the time!

I'd be happy with the latest Humax software version providing there wasn't so many serious bugs (hdd corruptions etc)...if they fix the bugs I'll be happy.

Humax need to fix the lock-ups, anything else is just minor niggles.
A locked-up machine is no more use than a brick.

Personally, I doubt that Humax will ever release a reliable software version.

Frank

What is your definition of reliable?

Mine would be something that runs for say three months solid with no errors, missed recordings or lockups and something that does what it says it should!

There should be no way interactive content should lock a box up under any circumstance.... eg if if doesn't understand the data it should disregard it.

SL & AR are good features, but it is making the Humax unreliable as you don't know if something will record or not.

With scheduled recordings only eg v.10 you knew the box would start at x and finish at x. It was the broadcasters fault if you missed something for not running to time.

Series-Link is working perfectly for me, coping with time changes and extended programs, no failures.

Accurate-Recording will sometimes miss a few seconds from the beginning of a program that has a short introduction period. I've never had an end chopped off.

Technology moves on and is hardly exclusive to Humax.

Of course the Humax is better than a VCR, but then, it isn't a VCR is it! It's a PVR

To come home after two weeks holiday to find that it hasn't locked-up on the first day away and recorded NOTHING.

A consumer electronics product should NEVER lock-up despite its other failings.

Frank

Humax had better hope we're never completely satisfied with a product, else why would we buy the next version. Capitalism is build on consumers always wanting something better. : )

I'm struggling to think of a consumer electronics product that doesn't lock up from time to time. My old Sky+ box locked up just as often as the 9200 - about once every six-months. Frustrating, but inevitable...

I've had a Humax 8000T since the early days, it's still going strong and has never locked up, the adverts causing problems for the 9200 are easily coped with by this older machine, so what have they done to cause the new one to be so easy confused ?!?!?!

I'm struggling to think of an instance when any of my VCRs (and that numbers 10+) have ever locked up. I think that's what's meant by progress.

We had (still have, but never use) two VCRs. The cheap one was £200, the good one nearer £400. And that was a good many years ago. The good one has PDC and screwed up (maybe broadcasters fault) about 10% of recordings. If we REALLY wanted it we left PDC off and padded it an hour or two.
Our first digibox has one tuner and, er, well that's about it. Cost £50 about 3 years ago.

Considering the price and features, and the number of balls it keeps in the air (2 recordings and watch something else or chase play), the 9200 is very good value. It's not perfect, but it has undoubtably had a bigger impact on our tv watching than any other item we've had over the years. (I used to have to set up most of our VCR recordings; now I have to get up early to set 'my' recordings before she gets to it )
It has been generally reliable (but we don't use SL/AR, just auto-padding) - the odd failure is annoying but its benefits far outweigh those slips.

I used to own a hifi that would lock up occasionally. As for it being inevitable..I suppose it is if you consider the target market. It's fast moving with low margins leaving insufficient time for developers. Factor in that the developers are often underpaid and you have another factor lowering the quality.

Modern software development tools make writing software easy. Unfortunately it can lead to sloppy programming. Those programmers that can think beyond the present to the future (required for good code) tend to migrate to well paid jobs.

A chewed up tape is the analogue equivalent of a lock-up in my book - surely everyone sufffered one of those?

It's also easy to forget the number of recordings that were missed on VCR by virtue of the complexity in setting them up. Now before someone has a go at me for using the term complexity, I simply mean that there were far more things that could go wrong in the VCR recording process, which is now reduced to a double click.

Unless, of course, it was just me who regularly recorded the wrong side, mis-programmed the time, ran out of tape, forgot to turn the machine off to make it record etc etc.....

Automatic head cleaning;
VPT video programming by teletext;
Auto switching to LP mode when remaining tape in SP is half the remaining scheduled programme length;
Ability to begin recording without having to be in standby or activate a special timer mode.

Some VCRs I've owned (although typically not those of Japanese origin) satisfied all of the above.

The point being though that all these features were included at time of manufacture and had to be free of bugs as there was no aftermarket upgrade capability.
Maybe (the knowledge) that firmware for DTT receivers is upgradeable by OAD or PC transfer, makes for sloppy programming.

I'm not saying that the firmware in a PVR is not far mor complicated an application than the microcode in a VCR, but surely the tools available to the programmers now are far easier to work with. I can only imagine that those tools also are inadequately programmed, meaning the PVR code behaves unpredictably.

Well I think my Tivo locked up once in the last 3 years (and incidentally supported Series Link and padding simultaneously).

You have never answered the question I asked some time ago about what a Tivio cost to buy and in subscriptions?

I think I bought 5 Tivos in all. One cost me £200 the other 4 cost £100 (there was a big clear out from the Electrical retailers when it was apparent it was a marketing flop). I bought a lifetime subscription on 2 of them for £200. Having used these for 7 years I reckon the cost in subs was about £2.50 per month.

The trouble is that the tools are geared toward making writing code quick and easy. Programmers are pushed toward sloppy thinking and leaving everything to the OS or runtime. Memory leaks and the occasional crash are regarded as unimportant as long as the product will survive a show-and-tell to marketing.

Management (especially in consumer goods) pushes toward an early release and won't give the developers the extra time needed to think things through. A significant number of developers just don't want to think things through either.

That's why I distrust Open Source code. There's some good stuff out there but most of it is written with one task in mind. The team rush through the coding to get to the 'cool' bit. Once it's sort-of working they lose interest and rush off to the next cool thing.

I suppose that's true of life in general. So many people seem unable or unwilling to put the effort in to plan ahead and don't want to see things through to completion.

But looking on the bright side...

My 9200 hasn't crashed yet and it's been used daily for 3 months since I bought it (during the day by my kids and during the evening by me and my better half). It sometimes misses a few seconds from the start of C4 programmes but it's never (so far!) missed a scheduled recording. It's not corrupted the recordings list and browsing the menus/EPG is speedy and responsive.

All in all I think the 9200 hardware is excellent (still on V1 h/w for all the new TBX boxes!) and the latest .21 firmware version has obviously been seen through to completion by a development team that has put in a lot of effort and has looked ahead

So it's not all bad

In my line of work, I have some embedded software programming to do. Being in a small business also means that it is often just down to me to ensure it is bug free. And I always have my boss breathing down my neck to find out why it's taking so long. However, being small means that what minor bugs there are (or at least, were) can be fixed when they are found, and released into new production batches, rather than wait 6 months, and consolidate all fixed bugs into one new piece of software. More effort is made to make sure it is bug free because it's expensive returning goods from around the world, and that thankfully hasn't happened yet.

Now, in contrast, our internal sales processing software is buggy as hell, as I get to spend an hour a week on improving it, and it often gets used by others in the office in the meantime with the warning of part of it not working properly! I don't care that much, because no one else outside the office has to use it!

So I do agree that programmers can be sloppy, but it also depends on how they are managed, and what the incentives are to fix bugs before they are shipped. If you know it is easy to reprogram in the field, then less importance is given to making it right in the first place.

I guess another important thing is developing with the intention of further development. Humax have included lots of extra functionality in the box, but if there wasn't a solid base to start with, this can be prone to making things worse where knock-on effects can cause serious problems! Even to the fact that people who are having problems with their box thinking it's hardware may have inadvertently stumbled on a rare unknown path that has somehow disrupted the normal flow of events in the software!

I spent sixteen years working on a suite of data recovery/forensic investigation tools. I nursed them from DOS, through Win16 then to Win32. It didn't help that when issues arose in the d/r side they had to be fixed ASAP and we never had proper project planning or supervision.

Despite that the code was easy to maintain and had few bugs. Me and guy working on it just had to learn how to fly by the seat of our pants /safely/.

The biggest trick was never skimping on identifier names (which produces near-self documenting code) and declaring functions like there was no tomorrow. The latter isn't really an option for embedded devices with limited stacks but it does mean that you end up developing your own bespoke language for the project at hand.

Oh and of course developing data recovery tools teaches you never to trust any data you got from outside your own process. In fact being slightly suspicious of data that your own code generated is sensible

Some people will never be happy. We have had our 9200 for just over a year and are happy with it.
Since the introduction of "Digital" hardware and software programming we have been generally dissatisfied with the performance of electronics equipment.
Analogue gear did a job quickly and efficiently and when it went wrong it was out with the soldering iron and replacement parts to restore the performance.
The operation of digital stuff generally seems to depend on the weather, moods,time of the month or whatever...possibly even the capabilities of the programmers (within their allowed time constraints). So, by now we should have got used to the poor performance of "modern"electronic hardware and the "you buy it first and then we will fix it" approach, this being one step up from the "register and pay for it and we will deliver it when we have made some" approach. (Rant over)
That said I think that the 9200 is one of the best bits of digital kit we have bought, it actually does the job I bought it for ( admittedly within constraints)....so thank you Humax