OK, so there's a counter argument that we should also have moved on from using 200MHz or less RiscPCs by now as well, but this is a talking point! ;)

I'm sure Microsoft have financial motivations for ditching OSs ASAP to force (or encourage :) people to upgrade, so let's not go giving ROL or Castle ideas eh? :)

If more people upgraded then there would be more progress to the OS because more programmer time would be afforded. I don't care if you smugly don't like the short obsolescence time, OS developers don't generally work for free.

Simply put, if you're not running the latest version of the OS (Select or ROS5 depending on your hardware), you might as well not use RISC OS at all because you're contributing this much to the platform:

Nothing.

Chris. Just me.

It's up to the individual if they wish to run OS3 but I don't think that programmers should feel any obligation to make their software work on it.

I suspect that one of the reasons that so much recent software does still work on OS3 is that the market is so small that programmers have to try and cater for the largest possible audience.

I don't think ROL could possibly drop support for OS4 for sometime since they're still selling it!
Even once they move over to 4.37 on ROM (or whatever version it will be (4.4 I think)) it would be reasonable to expect support for a year afterwards on OS4.
Cheers!

(Oh except you, you're running RISC OS 4.0, and you're wearing trainers :)

Yeah everyone should upgrade to the latest version... where can we get figures on how many people run which version of RISC OS? Maybe everyone has who's still active users. Can't wait for select on ROM though, faster bootups all round! :)

***retires into the corner, muttering to self ***

I would need to replace !Boot and then reinstate everything which was configured in my old !Boot, and back up and reformat the hard disk and remove !LongFiles. Then I (probably) need to patch some of my software which will stop working.

For many people, it is not worth the time and money to upgrade a perfectly functional computer which is worth two hundred quid at most.

Take NetSurf as an example. It would have been very easy to use the ImageFileRender module, found in Select, to render images in various formats found on web pages. In fact Justin Fletcher implemented a rough and ready implementation of this almost overnight - compare this to the development time in porting open source image libraries, the extra consumption of system resources, and so on. The time saved could have been spent improving CSS support or adding new GUI features.

Select has many benefits for users and programmers (the programmer benefits would benefit users too if the user base was large enough to make development worthwhile). You might think RISC OS 3.7 is good enough for your needs, but you can't make progress by standing still.

Yes, 230UKP is a lot - that's one argument for leaving RISC OS 4 behind. You shouldn't have to upgrade twice to get into a position to use Select, and the ROM release should solve this. It depends on the pricing strategy, of course.

What would be nice is if all new computers shipped with Select by default. I'm including the Iyonix in that statement.

e.g
I just sold some RISC OS 3.7 ROMs for GBP16.50 inc P&P, and I will now probably upgrade part of my computer i.e VRAM using that money (and some other money.). If I hadn't sold RO3.7 I wouldn't have planned to upgrade the VRAM.

where can we get figures on how many people run which version of RISC OS?

> Nothing.

Unless you're buying hardware and software that isn't Iyonix/Select. Or writing code, or helping in all those non programy ways people keep talking about.

What of all the programs that stop working whenever you upgrade?

I think Select on ROM will tempt 3.x users more, and maybe even 4.0x users who don't like softloading.

I wonder if they'll be a VARiscPC-SE "Select in ROM" upgrade too - so you can boot straight from the 6Mb ROM image like RedSquirrel.

It would be rather daft if you couldn't do that already.

The Select features seem more interesting than the ROS4 features, but the combined cost is too high.

New computers need to come with the latest stable version of Select if the userbase is to grow enough to make producing software that takes advantage of Select only features a good idea

I don't think ROL could possibly drop support for OS4 for sometime since they're still selling it!

I think that's the most important point wrt to this article.

Diodesign:

I can't believe there's still people using RISC OS 3.x, let alone RISC OS 4.0x.

Presumably you mean 4.0x let alone 3.x ;-)

if you're not running the latest version of the OS ... you might as well not use RISC OS at all because you're contributing this much to the platform: Nothing.

That statement's a bit much.
But even ignoring the inhenrent wrongness of it, I would imagine those people you refer to would categorise themselves as 'users' and not contributors, then.

Given ROL's failure to produce a 32-bit version or an Acorn-chipset-independent version, I think the simplest option might be for ROL to go into liquidation and Castle to acquire the rights to Select to integrate into RO5, making it an even more compelling upgrade for RISC PC owners. ROL has been failing to deliver anything other than incremental improvements for years now. Castle is at least showing some initiative.

[Sits back and waits for the flames]