RISC OS DME Launchpad released

To everyone interested,
I just made the first ALPHA release of !Launchpad available for download for ALL RISC OS platforms.

Please Note: I am doing all this for me because I want to have a more useful, convenient and modernized RISC OS Desktop experience. I am sharing my work because some others have asked for it. If this is not for you, please feel free to move along!

What is this?
If you are one of the few who still has no idea about this project, here is a link to the project presentation:

https://paolozaino.wordpress.com/portfolio/risc-os-desktop-modernisation-project/

And here is a link to the full project initial analysis and documentation (this is a dynamic content and tend to change over time):

https://github.com/RISC-OS-Community/DME-Core/tree/main/doc#readme

If you do not like to read then here there are a set of videos (which, to date, according to youtube have been watched by more people than there are in the RISC OS community XD):

https://www.youtube.com/playlist?list=PLhnROSACvoqSpDO_BSaJAyLZzUFpovpYa

Before you continue reading, make sure you have read the warning sign at the bottom of this post ;)

How to get it and try it out
This is the first public ALPHA code quality release of the first DME component and it’s intended for people who wish to help me test it and make it better. If you do not wish to help with testing it, then please ignore this thread
(and wait for the official stable release), thank you. Also note, there are some features that are disabled at the moment, they will be enabled as we progress with testing it.

Also, given its alpha code quality, please do not install and use it on your main system, thank you.

Latest release: 0.0.2-1 (Alpha)

Instructions to download and install it (per platform):

RISC OS 5

Please use my new PackMan repository following the instructions below:

- On your Packman, if you haven’t done this already, first install !CACertificates (this is vital otherwise !PackMan won’t be able to access repositories that use HTTPS only!)

- If you can’t find !CACertificates on your PackMan sources, here is a link to a zip of it that has been fully tested with multiple !PackMan releases, you can download it with your browser and then unzip it in your !Boot.Resources:

- If you’ve installed !CACertificates on one of the previous steps, then make sure the Filer has seen it. Either reboot your RISC OS system OR go to !Boot.Resources and double click on !CACertificates

- Next step is add the riscoscommunity.org packman repository to your list of repositories, by adding the following source:

- Click on save and then click (with menu mouse button, 2nd mouse button) on !PackMan icon on the Iconbar and then click on Update Lists

- Next open !PackMan main window, on the “Sections” menu select “Desktop”

- You’ll find both the DME base and DMELaunchpad in there (please note all DME components start with DME prefix, so they are easy to locate and they should exist all together)

- Click on “DMELaunchpad” and then on the Install button. It will install all the required dependencies, Launchpad, configure your System for DME and also launch !Launchpad automatically for you. So you shouldn’t do anything else at all :)

- Use !Launchpad as you wish and report problems either here or in Bugs or via email if you so wish.

RISC OS 3.50/3.60/3.7x/4.xy and 6.xy

If you use !Packman with the above releases of RO:

- Use the same instructions for RO 5 (all DME is OS aware and should install fine on every RISC OS system via !Packman, please note, on older !PackMan releases and/or old hardware it may take a bit of extra time before !Launchpad icon will automatically appear on your Iconbar)

If you do NOT use !Packman then:

- Download this pre-configured !Boot image here:
https://riscoscommunity.org/zip/riscos/dme/DMELaunchpad_0.0.2-1.zip

- Unzip it

- Copy it on top of your standard !Boot (or use the !Boot Configure facility to install it on top of your !Boot)

- Reboot your system (Launchpad should be automatically loaded after rebooting)

RISC OS 3.10/3.11

Please Note: DME and Launchpad require the Nested Windowing WIMP Module loaded, so, you must have a UniBoot sequence installed (or equivalent procedure) on your Archie HD or your Archie for MiSTer board, Arculator, Web Based Archie.

I have prepared a special version here in a Floppy Disc image, so you can use it directly with your Archie emulator (ArchiEmu, Arculator etc.), Gotek unit, MiSTer core or Flash it on a real Floppy Disc via !ADFFS.

- Download the ADF Floppy image here:
https://riscoscommunity.org/zip/riscos/dme/DMELaunchpad_0.0.2-1ADF.zip

- Unzip it

- Copy the ADF file where your emulator can use it or on your MiSTer board in the subdirectory for the Archie Core software

Note about the sources:

I am still finishing my build automation on GitHub, so sources should be available in few weeks from now, if I do not get too busy with work during the week!

The whole DME software suite is distributed under CDDL v1.1 license

Thanks :)

W A R N I N G

This project contains a lot of modernisation ideas in RISC OS Desktop design that may cause disconfort and even emotional reactions to “die hard RISC OS conservatives”, therefore they are strongly advised to ignore the project and run away from it ;)

- first install !CACertificates (this is vital otherwise !PackMan won’t be able to access repositories that use HTTPS only!)

Does it use the default system certificates or its own? (if the latter, why?!).

The default certificates can be updated easily using https://davehigton.me.uk/Progs/UpdCaCert.htm which will fetch them from… here someplace.

here is a link to a zip of it that has been fully tested with multiple

Oh yes, because pointing to a link for security certificates on a non-authoritative website is great practice.
I get that you want to make it easier for people, but one shouldn’t encourage bad habits.

Does it use the default system certificates or its own? (if the latter, why?!).

Can you actually put a subject in that question please? You mean !PackMan? If you mean PackMan, then no it doesn’t work without !CACertificates.

The default certificates can be updated easily using https://davehigton.me.uk/Progs/UpdCaCert.htm which will fetch them from… here someplace.

Has ROOL made this the default method yet? If so, can we please have an announcement and have the method added to the ROOL repository and not one of our personal websites? I strongly believe that everything that becomes standard procedure should be hosted on ROOL’s RiscPkg repositories.

Oh yes, because pointing to a link for security certificates on a non-authoritative website is great practice.
I get that you want to make it easier for people, but one shouldn’t encourage bad habits.

Rick, this is a comment for ROOL, not me. I agree that RISC OS should have a standardised way to download official certificates, but that is something that ROOL has to fix, not me.

Also please let’s not start now a completely off topic discussion on this, so let’s open a specific thread on the matter, thanks! :)

It’s good that this is available, but could you consider maybe saying what it is and why we want it? :)

Just pulled the package info file on my phone to read.

Basically – it is intended to replace the Apps pseudo filing system.

@ Chris

but could you consider maybe saying what it is and why we want it? :)

Done, the main post has been updated with all the project documents and the vide playlist that track some of the progresses, so you can also see Launchpad (which is the first of the DME components made available) works and looks like.

Also added a warning for the “die hard RO conservatives”, hopefully they will read it before they get a too strong emotional reaction from this entire project XD

I think those sorts of things tend to be called “Trigger warning”. ☺

One final thought:

Please Note: I am doing all this for me because

That’s how most useful software gets created.

Initial bug report,
two minor issues have been found on using certain default parameters correctly, so I have issued a new build (-3 for the PackMan users, while for everyone else is tonight date 31st of Jan 2023).

Links above are already updated, so just re-download if you so wish.

In case the upgrade fails on PackMan, just uninstall the previous release, and try again, there was an issue with the previous release packaging that can cause some problems durign this upgrade, it has been fixed on 0.0.1-3, but PackMan will try to remove previous package with the already stored triggers and not with the new one (this is normal, but can cause torubles in this case).

Thanks.

I just tried this for the first time. I went into Sources, added the new URL (https://riscoscommunity.org/packages/rpkg/arm32/betatesting), clicked Save, and got “The requested URL returned error: 503 Service Unavailable”.

I just tried this for the first time. I went into Sources, added the new URL

Chris, which repositories do you ahve enabled?

Mine is working fine, so maybe others are havign troubles tonight?

[edit]
I have checked for you, tonight it seems that ROOL’s raspberry-pi-programs-armv5 is having problems, you can temporarly disable it and retry.

In case you wish to also try without mine, just disable it and check :)
[/edit]

It looks like you’re right. Apparently when you click Save it enumerates through all of them, so the “requested URL” is not necessarily the one you just entered!

I came across this Packman issue yesterday for another reason, but it would be useful if it reported the URL it was trying to access at the time.

I came across this Packman issue yesterday for another reason, but it would be useful if it reported the URL it was trying to access at the time.

Some of the package lists are hosted on riscos.info and that was offline yesterday. I emailed Theo and he said a log file had filled up, now cleared, so those package lists should be back online. AFAIK the lists ROOL hosts (on packages.riscosopen.org) have been available as usual.

Thanks Sprow,
last night I was tired, saw the ROOL link and thought it was a ROOL repository, but you’re right, the link is hosted by ROOL, but the repo was riscos.info

All,
just released a new build (-4) that fix a bug we discovered when drag’n’dropping new objects in the Objects Filer when there were few Objects selected. The Sorting was not considering the selected objects properties. Now it does.

For PackMan users:

- Just run your Packman, click on Update Lists, to receive the update and then click on Upgrade All.
- Also this upgrade process will ensure to save your unsaved data from your running Launchpad (during the upgrade) and migrate them to the new release, close your old Lacunhpad and run the new one with all your data in – PackMan process is getting seriously Posh! :D

For other editions:

- Just reload the package from the link you used previously and after download, unzip and install on top of your old one.

Enjoy!

As Rick says:

Basically  it is intended to replace the Apps pseudo filing system.

Quite so, though I see it as playing two roles:
(1) A test bed for Paolo’s DME infrastructure.
(2) Yet another application shortcuts manager.
I expect only Paolo can judge its usefulness as (1), but any user can check if (2) helps them.
The DME project description doesn’t mention an application shortcuts manager, and Launchpad joins a crowded field, most of which are more discreet and frugal of screen space. So perhaps other demonstrations of DME might be more convincing?
Meanwhile, I note that setting up and maintaining shortcuts to find one’s way into apps or data directories can be avoided by using David Llewellyn-Jones’s DirMenu from his RISC OS page. With DirMenu seen, you simply

AddTinyDir 

*DirMenu Boot:^.Apps

Hi Bernard,
thanks for your feedback.

(2) Yet another application shortcuts manager.

Not quite, Launchpad has a lot of features among which:

- it allows also to create “Icons” (aka objects) that are Macros. These aren’t at all shortcuts, they are psudo code that can be used to, for example, create “buttons” that set signals on the Raspberry Pi GPIO, to turns lights on/off, or change selected scene in OBS (running on a PC) or do Home Automation. In other words you can use it as a “Control Pannel”
You probably have noticed that the “iconing” in Launchpad isn’t the traditional way the Filer does it and that is because there is more infrastructural code that will allows as we proceed with releases of it to have dynamic icons (aka icons that can be redefined either via Acorn Plugin protocol OR simply via signaling) and also replaced by Draw files (in the future) and/or controls (for instance a state icon on/off, or value from 0 to x etc.)
- it allows export of your entire, portion, category of object list as a CSV file (this also for the Macros and launchgroups)
- it is designed to be touch screen friendly
- it is designed to interface (or intercommunicate) with other apps on RO, aka supports WIMP messaging protocol to send and receive messages
- it is designed to support Acorn Plugin protocol to expand its functionalities

But, if you wish to consider it just what you’ve mentioned at your point 2, then please feel free to, however that is only your personal opinion.

The DME project description doesn’t mention an application shortcuts manager

I hope you have read the disclaimer for the project’s presentation, if not I report it here for your convenience:

Disclaimer

This page should be considered an introduction to the project it describes and not 
exhaustive of all the features and components offered by the Project itself.

The Desktop Modernisation is an ongoing effort and so things may change over time.

So perhaps other demonstrations of DME might be more convincing?

You are free to judge it as it is now as well. JFYI, as I have mentioned in my first post here, I am doing this for my personal use and sharing it with all the people that have requested to have access to it. Hence, I feel no need, pressure or desire to prove anything.

Meanwhile, I note that setting up and maintaining shortcuts to find one’s way into apps or data directories can be avoided by using David Llewellyn-Jones’s DirMenu from his RISC OS page. With DirMenu seen, you simply

Thank you very much for providing this example that allow me to show some of the advantages offered by Launchpad, nothing against David’s approach thought (I believe he did a great job dealing with the old Acorn Apps way).

as you mention in your example:

*DirMenu Boot:^.Apps

That implies familiarity with the Obey/Command line, it also imply writing a script that needs to be executed (possibly) at every boot and hopefully that no errors are made by the user doing all this process. Eventually placing the script within !Boot OR adding its execution at boot via !Configure

None of that is required with Launchpad. All you have to do with Launchpad is select the Category you want to add a bunch of Apps into and just drag and drop them on Launchpad main window, that’s all.

It will also take care of checking each of these apps at boot, so icons are always displayed correctly and avoid look up of other apps in the same directory (this is particoularly useful if you have things like !CoreUtils installed, which can mess up aliases on your RISC OS if it’s looked into).

For network based objects (aka remote objects), you can also configure Launchpad NOT to look them up, this can help bosting boot performance, or you can ask it to look them up, Launchpad initialise in Multitasking, so even if that may require a longer time to initialise, it won’t stop or slow down your entire RISC OS boot up sequence.

There are more advantages, but I think this answer is already very long and I am not here to “sell” anything ahahaha :D

Again, thanks for the practical example.

Paolo adds some very interesting detail here, so thank you. But I sense a dilemma. In this forum post he included the objective:

Easy to use for everyone, included who has never seen RISC OS before

which is similar to Cloverleaf’s. Yet above he says:

I am doing this for my personal use

I am not here to “sell” anything

I know his development time is limited but, to widen the appeal of RISC OS by means of a project with the aim:

more “modern” desktop experience

entails not only code that is optional and easily installed (great!) but that is also guided by an element of ergonomic design input from ‘outside’, in addition to alpha testers’ feedback. Am I expecting too much? Probably, since I’m offering only discussion, not code — my apologies! I for one will continue to exercise Launchpad on a Pi 400 to understand it and offer comment.

By the way, are !FirstBoot and !OSUpdates still on the agenda (mentioned in the above referenced forum post)?

Dear Bernard,

But I sense a dilemma. In this forum post he included the objective:

Easy to use for everyone, included who has never seen RISC OS before

which is similar to Cloverleaf’s. Yet above he says:

I am doing this for my personal use

I am not here to “sell” anything

I guess Rick was absolutely spot on when he told me it’s common culture here to complain about everything ahahahah :D

I think you are quoting things, but your quotes seems to lack other discussions that have happened “in between” (and, unfortunately also over different socials/meetings/etc), where, for example, some folks here commented things like:

“but WE don’t need it”

Where the use of ‘we’, I have discovered later, means “my personal opinion, and only my personal opinion, but I am not confident enough to state it that way, so I mask it behind an hypotetical plural to make it sound stronger”.

Hence, I “repuposed” the entire project for my self only. In the end, if it’s not wanted by the famous "we"s, then even better, I don’t need to do all the extra work that is required to distribute things to others ;)

However, after this decision, others (this time the plural is real) asked me to share the project and tools, so, to avoid any more fake "we"s to make more embrarasing comments, I have come to the final idea of:

I am doing this for me and sharing it only with all the people who have asked to have access to it.

Now, that’s brilliant (quote from IT Crowd) I think. The fake "we"s can’t attack it anymore, it’s just for me so the historical “pessimism” can’t be applied either! I may have created the perfect “weapon” against the typical RISC OS grumpy user ;)

I know his development time is limited but, to widen the appeal of RISC OS by means of a project with the aim:

more “modern” desktop experience

entails not only code that is optional and easily installed (great!) but that is also guided by an element of ergonomic design input from ‘outside’, in addition to alpha testers’ feedback.

And feedback, when pragmatic, constructive and avoids getting in the muddy waters of my previous reply is always very welcome. Otherwise I just ignore it.

However, you seems (both by email and on here) to kind of being asking me to support 3rd parties tools that were written to work around the terrible limitations of the original Acorn’s “Apps”. I price those people who have bravely fought the demons in the Acorn’s Apps, they are true heros! I am no hero, so I have avoided those demons all together ;)

But, the fact that I am not interested in supporting those tools, should not stop you or make you engage in some sort of (even if polite) set of judgements to kind of imply anything. You do not need to do that at all, the great news is that there is another advantage offered by Launchpad to actually help you in your endevour to support those utilities, Launchpad is Open Source and you’ll be able to add such support yoursuelf if you so wish.

Am I expecting too much?

By my limited understanding of your comments, yes, you are expeting too much. Why would someone expect anything from an OS that has delivered very little in so many years, is something that I will probably never understand.

If this can be of help, my way of “swimming” in the RISC OS waters, is simple, I expect nothing at all, whatever comes from this will be an awesome achievement. Why do I see it this way? Because RISC OS has technologically died in 1992. Last year it has made exactly 30 years of “commemorations” and “memories” the good, glorious times when Acorn delivered remarkable things.

All that has happened in the last 30 years is trying to keep “the corpse” of a once very interesting project (and, for certain timing aspects, even a technological miracle for the time), still usable by the nostalgics and enthusiasts. There have been no major breaktrhough in refactoring its vintage architecture, its “hack” of a multi-tasking, its weaknesses. BTW, with this comment I am NOT criticising the work done by all the people that actually tried to do something for RO, I am just comparing the reality with the state of the average OS in 2022 (not even the top ones).

So, why would you expect anything at all? Again, it’s beyond me.

since I’m offering only discussion, not code — my apologies!

No need to apologies. An apology implies that I may have been hurt or felt offended or something on these lines, while I am feeling curious of why you even need to take the path you did, ‘cause again, I expect nothing from RO, and therefore everything everyone does for it, for me it’s an awesome achievement :)

I for one will continue to exercise Launchpad on a Pi 400 to understand it and offer comment.

Thank you very much, I truly appreciate your and everyone else help, maybe, together, we’ll manage to create a set of components that may make RO more useful to who wish to use it within the context of it’s own limitations and quirkiness. For me it’s always a pleasure to turn on one of my RISC OS device and feel once again the little me coding on his favorite OS of the end of the 80s :)

By the way, are !FirstBoot and !OSUpdates still on the agenda (mentioned in the above referenced forum post)?

Yes, everything (and more than what is listed) is still on the agenda, the only limit for me is the time available to code and debug on RISC OS, but , as a glimpse of what’s going on:

The upgrade process we are testing for Launchpad is also part of !OSUpdates.

!FirstBoot’s macro interpreter is completed, so it can basically produce widnows and all required UI items out of a macro script, which is required to generate all “Wizard” like process to configure whatever part of the OS at the first boot (or after a major upgrade). A lot of work now has to focus on receiving the config scripts from DME Components and the mechanism a component can request to repeat the configuration of a portion or the entire OS if and when required.

One of the cool bits of !FirstBoot is that it will allow to configure Territory etc at the first boot.

There is a lot more going on, but most of them are in what I call “modelling state”, in other words, pieces of code that prove a feature can be implemented on RO or not, so it will take a while before they will be available. But, the delivery process kinda give it away that they will all be components and they’ll “appear” in PackMan as they become available. This because some user has requested to have a progressive delivery instead of waiting till the whole thing is completed.

Hope this helps, again thanks for your comments and interest :)

“but WE don’t need it”

Where the use of ‘we’, I have discovered later, means “my personal opinion, and only my personal opinion

My personal opinion, which is of no great importance to anyone but me, is that video games are an interesting screen effect requiring regular input of adjustment factors, but they are not a meaningful way to spend time.
I’m sure the video-games industry will be quaking with fear at my lack of interest – or maybe not.
They’ve survived and thrived in the 40-something years since I built a TV games computer and then handed it to my sisters as I’d done the interesting bit.

I see launchpad as an example use of a developing facility which, unlike cloverleaf code, has a well advertised source repository. People can take from that, and they can contribute to it, no cost and no commitment required.

I guess Rick was absolutely spot on when he told me it’s common culture here to complain about everything

There’s a reason why people on the flip side of the planet call Brits “The Whinging Poms”. ;)

It’s a mentality thing. Compare popular Aussie soaps (Neighbours, Home & Away…) with popular Brit soaps (EastEnders, Corrie…).
One has social issues in a sunny place, maybe with a beach, and bright colourful characters. The other has social issues in a gloomy crapsack world ¹, and characters you wouldn’t ever want to meet in a dark alleyway.

That’s not to say Brits can’t do comedy – you’ve noticed the IT Crowd, and a number of prize winning American comedies on my streaming setup were originally British, but were remade because… what, they couldn’t cope with the accent? That’s just English being spoken correctly. ;)

so I mask it behind an hypotetical plural to make it sound stronger

👍

I am doing this for me

Best reason to do something, unless there are lots of currency symbols involved.

and sharing it only with all the people who have asked to have access to it.

I’d just drop it on a website. Less hassle for you in the end, and if the Royal We are Not Amused, well, frankly they can eff right off.

Yeah, I guess I’m slowly evolving into a cantankerous old git, or maybe just somebody who’s no longer willing to make allowances to try to please the unpleasable?

The fake "we"s can’t attack it anymore

Given that it’s being done “by you for you”, do you really care if anybody moans?

It’s like all these people who complain about how awful Twitter is these days, while regularly going back to Twitter. It’s dumb.

[note: I pop by Twitter once in a while to see what a certain name familiar around here has said to various inept Tories (or pretenders), but I don’t have an account nor do I plan to ever have one; ditto Facebook, Insta, Tiktok, and the other dozen soul cemeteries with stupid names]

I am no hero, so I have avoided those demons all together ;)

If one wants to tame Apps (sort of mostly), allow me to point you at AppUtils at the bottom of the page: http://kappa.me.uk/download.htm

Directories, organisation, categorisation. It’s a little fiddly to set up (blame how Apps works), but once setup and stuck into your Boot, everything starts up nicely with a properly useful Apps pseudo-filingsystem.

Because RISC OS has technologically died in 1992.

Damn, that’s harsh.

1998, at least!

There have been no major breaktrhough in refactoring its vintage architecture, its “hack” of a multi-tasking, its weaknesses.

Sadly, most of the work in refactoring things didn’t happen with this incarnation of the OS. If you read Gerph’s Rambles, quite a bit was done to try to improve things.

But, for both that version of RISC OS and this one, there’s a limitation on what can actually be done and that limitation is to not break stuff for arbitrary reasons. The market these days is small, the number of apps with errant zero page accesses means that a compatibility page had to be implemented in order to fudge things on high vector builds, because whatever the reason, having a new version of the OS that causes apps (that may never be further updated) to crash is simply unacceptable.

Hmm, is one of the major suppliers of hardware still building custom low vector releases? It’s not necessary with the compatibility page – even I’ve given up on using low vector builds, stuff just works, even the broken stuff. ;)

Anyway… If you compare with some other operating systems, they either play fast an loose with the API, which is much easier with open source as one can just rebuild stuff to get things working, or they’re a total nightmare of interacting pieces of tangled linguine (Windows, I’m looking right at you).
RISC OS has one API and doesn’t support alternatives (such as multiple different versions of CLib, for example). It’s good in that there’s no DLL Hell and things pretty much just work (an update can’t randomly break stuff ²), but on the other hand that immutability of the API makes it hard to enhance and extend.
Consider modern input methods. There’s already a bit of a bodge to get UTF-8 keys into the system, and there’s a sort of a hack for getting pen pressure to use a more responsive device than a mouse. But graphics tablets are old hat these days. It’s a finger poke, with coordinates and poke pressure. Oh, and potentially up to, what is it, five or six that can be tracked? Plus tracking of specific gestures, like the difference between a one finger swipe and a two finger one, or the ubiquitous un-pinch to zoom in.
Where’s the API for this? Where the hell would it even be put in the current API?

So, why would you expect anything at all?

Because everybody wants the moon on a stick these days.

I expect nothing from RO

Sometimes the lack of * from RISC OS is a benefit. It’s all too easy to get sidetracked with a device that makes it dead easy to marathon dozens of Smarter Every Day videos. Well, cool, I now know exactly how a Venturi works (in beautiful slow motion), but that thing I set out to do? Never got done…

coding on his favorite OS of the end of the 80s :)

Strange thing is, I left RISC OS for a while and did some coding on Visual Basic under W98 and later XP. For all it’s faults, VB (and I mean proper VB, not the new .net stuff) is bloody brilliant. It makes writing software a bit of a doddle.
I have nowhere near the amount of technical knowledge for Windows as is necessary for RISC OS. I don’t even know how to start up an application or how the event loop works. I didn’t need to know.
Imagine how different things might have been if something like that was available for RISC OS, rather than “well you set the fifth bit of the eighth word, then set the first register to point at that block of data, then you call this system interrupt and pray it doesn’t explode”.

But I came back to RISC OS. Why? Because it’s fun.
Windows is functional. It gets stuff done. But the UI (3.1 through to XP, I’ve not bothered going beyond) is a pain in the arse. RISC OS has a UI that looks like somebody actual sat down and thought about how this stuff is supposed to work. Windows? Looks like the result of two different committees that refuse to interact with each other.

some user has requested to have a progressive delivery

You have a willing guinea pig. ;)

¹ These days, indistinguishable from reality.

² It does happen, but it shouldn’t.

@ Steve

I see launchpad as an example use of a developing facility which, unlike cloverleaf code, has a well advertised source repository. People can take from that, and they can contribute to it, no cost and no commitment required.

Haaaa, this, very, very, very much this, thank you! :)

I’d just drop it on a website. Less hassle for you in the end, and if the Royal We are Not Amused, well, frankly they can eff right off.

You are 200% right, but I also read what ROOL’s is doing. It’s easy, it’s on every single page, article and forum post:

“Safeguarding the past, present and future of RISC OS for everyone”

with the everyone in bold, as font and as statement. Something the "We"s constantly forget to read apparently, while they should read it before posting a comment, IMHO.

ROOL’s people may not be moderating, but their desire of being a central point of reference for the entire community is clear, and I am totally ok with that, hence I post my updates on here as well.

Given that it’s being done “by you for you”, do you really care if anybody moans?

I absolutely couldn’t care less LOL.

But Bernard has always been nice and also a contributor of ideas, so definitely deserves an answer in IMO.

Damn, that’s harsh.

1998, at least!

LOL, ok ok I give you the 1998 :)

For me the 1992 is the year for 3 simple reasons: Intel 486DX2 (25 MIPS beast, low power consumption), MS Windows 3.11 (with full 386 mode, so up to 256MB of RAM and the Intel V86 support to run DOS apps in V86 VMs under Windows) and the release of the PCI bus, which made PCs have video cards that could perform better than any Amiga and Archie. all these 3 were released in 1992 while in the same year Acorn released the A3010,A3020 and A4000 so you decide ;)

But, for both that version of RISC OS and this one, there’s a limitation on what can actually be done and that limitation is to not break stuff for arbitrary reasons.

This is true, but that limitation no longer applies as strictly as it did, because legacy Apps could be executed within an emulation layer. New HW is so many times faster than the old platforms that Apps would still run faster than on the old HW.

This thread is not the right place to discuss this, so I do not wish to have a zillion off-topic comments (’cause I am following this thread in case of bug reports on Launchpad), but if we are all interested in this discussion we can open a new thread on the matter.

Because everybody wants the moon on a stick these days.

This is sadly so true… :(

But I came back to RISC OS. Why? Because it’s fun.

Haaa, brilliant (to quote again the IT Crowd lol), and that is the whole point, so thank you! :)

Hi, are there any plans to offer a download without the need to use PackMan.

I do admit that PackMan is a nice tool and good use for many users – but not for everybody. I even usually recommend that users should try it an that it might be good for them.

After some testing my decision is to not use PackMan. I do appreciate that PackMan does force you to use the default folder structure – very important since I do not use that at all (except that the boot application is in the root of the boot disc) since I simply do not like it (and never did since over 30 years ago).

To be able fully use PackMan I’d be forced to reinstall everything it supports since PackMan lacks a scanning feature where I tell it where to look for installed apps to add to the list of my apps (agreed that fact that Acorn managed to define how to put version numbers in Modules but not into applications is a hinderance for this). Why should be do that when my setup is working nicely and I keep 100% control over what is where and what is installed on my system.

So please be so kind to offer a direct download link. Thanks