ARMbo(o)k news update

Not wanting to pour oil into fire, but…

Having just sent out another newsletter to those who have expressed interest in the ARMbook (or, thanks to an accidental mistype earlier in the year, ARMbok, as it is known to its friends), I figured it was time to brief the ROOL forums on where we’re at.

erm… no, didn’t get anything. I expressed interest in the ARMbook very soon, very publicly and (I’m afraid) a little too vocally. I even think I was amongst the first to express interest. The answer at the time was something along the lines of ‘well, sorry, there’s no place to sign up right now, have patience’.

Which I had. But then I had to have my thyroid out at the beginning of summer and was pretty much out of order for a few weeks, so I didn’t have the time and energy to poll the forum for news regarding the ARMbook, otherwise I even might have noticed.

Anyway, I just wanted to chime in and say that for folks like me – who don’t really have an overview which blog post to check and who’s doing what on which page in the RISC OS universe – it seems a little complicated keeping up to date with the going-ons regarding all things ARMbok ;-) Of course it takes time away from development, but a little webpage, maybe even with a functioning RSS feed, would be really something…

Hi Stephan! If you’ve not received anything, please get in touch with me via email. I can make sure you’re on the lists and resend some recent info. I’ve sent out roughly half a dozen newsletters since last Autumn.

In an era of GDPR, we can only send stuff if formally requested, and as I say, don’t like to fill ROOL forums with stuff unnecessarily, which is why I waited a long time before posting here. The PDF form on our computer website (www.armini.co.uk) is a way of gathering that interest in a way that we can easily organise.

Also, if I’m being honest, I don’t want to “publically” show progress reports (I’m pretty candid in my emails, and talk about things “warts and all”) to all on sundry. My preference is to share progress with those are most interested and eager, which will generally be the ones who get in touch. This is a personal thing, but it is similar to why I don’t post on newsgroups. I prefer to give more candid, open reports to those who are keen and want info (like you) – so please do get in touch. It may even be that we sent info and it got stuck in a spamtrap somewhere.

I feel this way because sometimes I feel like I’m feeding the “trolls” and those who are never going to be customers anyway (this is why I stopped reading most of the c.s.a newsgroups). I’d much rather deal with a smaller group who are keen on the product. Not necessarily good business, but it’s why I do RISC OS in the first place – small group of keen folks, rather than large group of ambivalent folks.

In an era of GDPR, we can only send stuff if formally requested

So the logical consequence in this day and age of GPDR (or DSGVO as we call it over here) is – don’t send stuff, put stuff onto your website and let people read it there. It makes it easier for everyone.

Q.E.D.

Whilst someone kindly mocked-up a new website for us, it didn’t degrade well in Netsurf, so for RISC OS users, it’d have been a problem (Wordpress is not RISC OS’ friend).

For updating/editing/adding content, no – but I haven’t had anybody complaining about problems with RISCOSitory in NetSurf. It’s all down to the WordPress ‘theme’ in use, and (in some cases) the settings within that theme.

To be fair, I do complain to myself about the way it now looks in general, because there are aspects of the current theme I don’t like, but going through different themes and testing them in multiple browsers on multiple platforms can be quite time consuming. But that’s the key: Finding the right theme.

(There’s also the option of designing one from scratch – or by adapting an existing one – and this is something I may yet look into at some point.)

The bottom line, though, is that whatever you do, you absolutely do need to do something about the R-Comp website(s). Sorry.

The bottom line, though, is that whatever you do, you absolutely do need to do something about the R-Comp website(s). Sorry.

I have to agree with Vince. I find the R-Comp web site difficult to navigate because the headings are so obscure. Click on ‘ARMX6’ and you get a page titled ‘ARMini’ with only a glancing reference to ARMX6 and no link (currently disabled) to register interest. The two columns on the home page say ‘R-Comp Software’ and ‘R-Comp Interactive’ and my first assumption is that this is similar to CJE Micros where there are two companies, one registered for VAT and one not. It takes some research to work out that R-Comp does the software and R-Comp Interactive does the hardware. Headings that said ‘software’ and ‘hardware’ would be much simpler. It is also difficult to search for a particular product even if you know exactly what you are looking for – for example CMOS widget – as there is no search facility.

I still don’t understand why there are two separate companies, R-Comp and R-Comp Interactive. Perhaps these are just trading names for each other for some obscure legal purpose.

It used to be said that “the customer is always right” but I know that approach has fallen out of favour. I am not simply throwing brickbats here, I am trying to be constructive. On the plus side, the A9home web site (which still says ‘last updated May 2005’) and the (non-existent) Impression-X web sites are definitely worse.

One purpose of a web site is to present information so that simple queries about factual matters are answered directly rather than forcing the casual enquirer to telephone to ask about it. This sets a hurdle that many might not bother to surmount. It does however allow intelligence about possible customers to be gathered in a ‘person to person’ telephone call with a hurdle to discourage frivolous enquiries so it might be a deliberate and sensible ploy.

I find the R-Comp web site difficult to navigate because the headings are so obscure. [etc].

This and more besides. There’s also the issue that you visit [www.]rcomp.co.uk, and all the links within are to www.arsvcs.demon.co.uk/whatever. This suggests either a weird hosting set up across two locations, or hard-coded links in pages that weren’t updated when the domain was first used – neither of which is a good look.

A quick double check, and I can see it’s the latter: If I call up the website, and look at the source, I see the frame contents are loaded from arsvcs.demon.co.uk, but the links within the pages are relative; clicking on one and then reloading the page with the arsvcs.demon replaced with rcomp in the URL still loads the pages.

That alone is an easy fix, then – just fix the frame links to be relative (assuming there are no other similarly odd links littered throughout the site¹) – but it’s really barely scratching the surface.

It takes some research to work out that R-Comp does the software

You mean does some software.

and R-Comp Interactive does the hardware.

And some software – games and audio titles.

¹ But if so, I know someone who has a piece of software that can fix that quickly and easily. ;)

To throw in another €0,02 – the use of frames (which the rest of the world has all but abandoned ☺) results in the ironic situation where choosing the RISC OS link at the bottom of the main menu loads riscos.com and causes it to appear stuffed entirely into the left bar (Firefox).

I try to write my blog to pass validation ;-) but something I have had to do is add <base target="_top"> at the top of the page to break out of badly handled frames on other people’s sites.

Andrew – you need to either do <a href="..." target="_top"> to open a link out of the frames, or <a href="..." target="_blank"> to open it in a new window/tab. I would suggest that offsite links are opened in a new window/tab so that your content remains in place.
And shouldn’t it be ROOL you point to these days? ;-)

Thanks Rick – I hadn’t noticed that mistake – all the other links act appropriately, just not that one. Am currently suffering from hosting issues as I updated the pages to include ARMbook yesterday at lunchtime, and they STILL haven’t propogated AFAIK. This may be the straw that breaks the camel’s back.

Also, it should probably be RISC OS Developments that I point too these days – it is their property! ;)

Moved to Aldershot (you might want to put the kettle on)

Vince Hudd said:

but I haven’t had anybody complaining about problems with RISCOSitory in NetSurf

Well not exactly RISCOSitory but the SW show site doesn’t show the banners downloads but Firefox does.

So you can’t get to advertise RISCOS if you aren’t using an alternative OS and browser to download the publicity. Catch 22

but the SW show site doesn’t show the banners downloads but Firefox does.

Safari also shows the banners, http://www.riscos-swshow.co.uk.

A look at the source with NetSurf and StrongED shows that that part appears to be commented out.

The preceding paragraph titled “Charity Stand” is commented out but the comment terminator is --!>. Replacing that with --> satisfies NetSurf.

‘Not exactly RISCOSitory’ because it’s not RISCOSitory – at all; I wonder if you thought I was responsible for the site, Doug?

Whoever is responsible will hopefully be reading this forum and see the mistake they’ve made, as pointed out by David.

Well more a comment on Netsurf as a native RO browser but the mention of RISCOSitory on the page obviously was the red herring and putting 2 and 2 together I got a big 99 on that for assumptions!

Note, this doesn’t include WiFi – that will be extra as/when we do it as part of the RISC OS Wifi implementation (likely in conjunction with ROOL)

Does that mean WiFi support will now be closed source and won’t be included in the OS? It was my understanding WiFi was being added via a bounty and will be open source?

Do you mean to say the WiFi chipset driver in the ARMbook will be chargeable? Or are you proposing to fund WiFi development and recover the cost?

Clearly I’m misunderstanding something.

The RISC OS wifi would be open source, the update for the ARMbook would be chargeable. I haven’t thought through the mechanics of it yet. Basically, I just wanted to deliver ARMbook at low price, and only charge for Wifi once done (unless it magically happens “for free”).

Bear in mind that we have one of the better track records for open source submission – we bought the OS and open sourced it, i.MX6 is the least “closed source” of the current boards/ports, I think, and John has always been keen for early submission of updates/changes. As such, why would we gimp our own OS by making Wifi closed source. That’d be crazy!

The preceding paragraph titled “Charity Stand” is commented out but the comment terminator is —!>. Replacing that with —> satisfies NetSurf.

So actually NetSurf was displaying the page correctly, and Firefox was in the wrong!

I see the page works as they expected again now anyway.

Firefox was in the wrong

Yes… if you believe they’re both some kind of HTML validator.

On the other hand, if you were expecting them both to be browsers… no.

So actually NetSurf was displaying the page correctly, and Firefox was in the wrong!

Nope.

https://html.spec.whatwg.org/multipage/parsing.html#comment-end-bang-state

A slightly more readable version: https://html.spec.whatwg.org/multipage/syntax.html#syntax-comments

if you believe they’re both some kind of HTML validator.

This is exactly it. Using a number of browsers (with different rendering methods) will show that the comment appears as anticipated but does not show that the content is correct.

There is a specification for each version of markup.

And then there is reality where it’s a twisted perverted mishmash of half understood and generally broken markup. The browser’s job is to try to do something sensible with the input.

When I wrote my little OvHTML (HTML to OvationPro) filter, a basic stuck-in-the-3.2-era parser wasn’t that hard (it’s nothing like today). The hard part was deciding what to do in the special cases. Missing closing tags were common. I came across a div align center closed by a /center tag, and three different ways to spell fuchsia. It quickly became that the “special case” condition was clean markup. :-/
[and now, years later, I discover that the 8859/1 specifier is to be ignored and 1252 assumed because practically nobody ever got that right]

Indeed, on my site, the /frobnicate section was created by a program I tried out (under Windows) that bestowed upon the world some rather horrific markup. People using such software would see “it works in IE” and not the underlying carnage.

A browser does not validate.
A browser attempts to show, as best it can.

--!>

OT. Moved to General.

Any chance some of this could be taken to a thread in General please? The SW show website doesn’t have much to do with ARMboks ;)

Any chance some of this could be taken to a thread in General please?

Done.

--!> moved to General

Whilst someone kindly mocked-up a new website for us, it didn’t degrade well in Netsurf, so for RISC OS users, it’d have been a problem (Wordpress is not RISC OS’ friend).

I have five created websites running on Wordpress, and four of them look just fine on NetSurf and the fifth looks good enough if you don’t know how it would look on a modern browser.

So I have no idea what you did. Just pick the right Wordpress theme and everything is fine with current NetSurf. But, going from the look of your websites, you are probably aiming for full compatibility with Browse, Fresco, Oregano (2), Webster and ArcWeb.

There’s now rather more extensive information about the ARMbo(o)k online at http://www.armini.co.uk/ as promised. I’d like to add more pictures (eg. illustrating Big Mode), but they’re quite time consuming to create, and time is a very precious resource right now. I hope it will answer some of the questions people may have, and explain about the product/project as clearly as possible.

Laptops have been shipping since May, and we’re already on the second big shipment from the Far East, so ARMbooks are very much “a thing”. They will also be present en-masse at the London show.