June news in brief | Drobe.co.uk archives

June news in brief

Published: 19th Jun 2007, 23:52:14 | Permalink | Printable

Software free and non-free and more

Things haven't been active round here for a number of reasons, but first, a catch-up of news you may have missed.

Birthday
The Acorn Archimedes celebrates its 20th birthday this week. The launch of the ARM2-powered machine was announced on June 16 1987 in a national newspaper and released to the public soon after. The humble A305 had 512KB of memory, Arthur in ROM, and sold for 900 quid. The anniversary is also marked by the unveiling of a new RISC OS search engine, riscpedia.net. The website run by Rebecca Shalfield only allows one free search a day unless you pay for more. Searching for 'Castle' returned zero results before locking us out for 24 hours.

Emulation
Free and open source RiscPC emulator RPCEmu is up to version 0.7. Author Tom Walker claims his software is now up to VirtualRiscPC-grade speeds thanks to its dynamic recompiler. It also now supports CD-ROM drives through its .iso file handling, can handle hard disc images up to 32GB in size, boasts a new DOS port and also compiles for big-endian machines - such as PowerPC Apple Macs. RPCEmu's sister emulator Arculator has lost its FDI support for "licensing reasons", mysteriously. Meanwhile, VirtualAcorn have updated some RISC OS software to allow it to run on VirtualRiscPC and RISC OS 6.

Modules
Rik Griffin has published a new version of games C library Popcorn, which now features support for 32bpp screens and sprites, alpha-blending, Iyonix hardware acceleration and other tweaks and improvements. His precision timer module HALTimer is also online as is his XScale-accelerated memory copying module. Rik said the Iyonix's IOP321 can throw around 300MB of data a second with this acceleration.

IDE
APDL have produced a new version of their driver software for the Arcin and Blitz IDE cards. With the updated software, Flash-programmable Arcin devices will be compatible with RISC OS 6. The drive icon height bug is also fixed for RISC OS 4 and above. The new version costs about ten quid. For more details, see here.

Freeware
Democracy tool MPdata, which displays information about your member of parliament, can now report issues and problems to a local authority for you. Maths accelerator module SuperFPEm has been fixed to not crash Moredesk and an Iyonix-compatible version is now available. BBC Micro game Dominion, which was developed in 2004, is now online as a free download - a version for emulators and a RISC OS build are available.

Popular NFS client Sunfish is up to version 2, and can now act as a full filing system as well as an image filing system. The user interface has been improved to make creating connections easier. New utility StickyPad allows users to pop handy post-it notes on their desktops.

Jobs
A full-time ICT technician capable of looking after 40 RISC OS 4.39 RiscPCs is needed at Knightsfield school. All their teachers have PC laptops with VirtualRiscPC on them and data projectors in every classroom. Deputy head Sharon Pointeer said: "We have increased the hours to full-time - 30 hours per week, term time plus two weeks (40 weeks) in the hope that we might attract someone with the right knowledge and skills. This will give the person more scope for development work such as production of software guidance for staff and making of resources, both paper based using ICT and also hopefully computer based materials." The school is also after a Head of Care and after-school activities coordinator.

No jobs
R-Comp's technical support has been on hold this month after the company's Andrew Rawnsley was laid low. A spokesman for the software publisher said: "Due to medical problems with neck and arms and on doctor's advice, R-Comp's Andrew Rawnsley will be reducing typing-time for the next few weeks. This means that email technical support and some telephone support will not be available for the next few weeks. We hope that normal service will be resumed progressively during July. Sales are unaffected by this."

And finally
The NutShells website has upgraded its back-end software to Drupal 5 and moved to more reliable hardware. It is hoped this will allow developers to log in again and update their profiles to point to their updated software.

Speaking of websites, you may have noticed that drobe.co.uk has been running as a minimal service of late. This is due to drobe towers moving house, having no Internet connection, and its regular contributors being tied up with social lives and day-jobs. We'll strive to improve things, but in the meantime, anyone who wants to help out should step forward and get in touch. Otherwise, we might have to switch drobe to an NTK or popbitch style publication.

Links

News? Comments?

Previous: Super fast RPCEmu released
Next: Archive ed Paul quits magazine

Discussion

Viewing threaded comments | View comments unthreaded, listed by date | Skip to the end

I'm trying to work out what purpose this Riscpedia serves. Any idea? Limiting searches per day to one is pretty useless - I searched for "NetSurf" and got nothing of use.

rjek on 19/6/07 11:55PM
[ Reply | Permalink | Report ]

The "Powered by the Python-Based Serviette Web Server on a RISC OS Computer" bit at the bottom is enough to scare me. Running a website off a RISC OS box? Insane!

takkaria on 20/6/07 1:22AM
[ Reply | Permalink | Report ]

Getting RPCemu compiled on a Linux box - especially with the dynarec is an "interesting" experience.

Jaffa on 20/6/07 8:58AM
[ Reply | Permalink | Report ]

It usually takes me a few attempts with search engines to get the results I want (Spelling errors, not using enough keywords) so one search a day makes Riscpedia a dead duck. Well unless i pay, or use a free engine like google, Yahoo, AOL, Ask.com, Windows live, etc etc.

Im sorry, but Riscpedia quacks no more.

Fuzzy on 20/6/07 9:23AM
[ Reply | Permalink | Report ]

Searching on Riscpedia for 'Drobe' got four results: two comments from myriscos.co.uk, a link to the Acornarcade front page, and one two-year-old Drobe article. So the site isn't empty, it just isn't very complete

Takkaria: I don't see anything wrong with running a website off a RISC OS box provided you're not expecting heavy traffic. Nice low power-consumption, and quite enough memory and I/O bandwidth to pass text entries to the current RISC OS market. I wouldn't want to run slashdot or NTK on one, but that's different

chrisj on 20/6/07 10:50AM
[ Reply | Permalink | Report ]

A ROS search engine is an interesting idea, although (even if it was free) how much it would be used is questionable IMHO. Its a shame that the author has not contacted the community to contribute articles and build it up.

polas on 20/6/07 11:09AM
[ Reply | Permalink | Report ]

I thought I'd seen it all with the "one free search for random punters" offer, but then my trip to 1995 was fully realised when I saw the Web server software on sale for the princely sum of £25 a pop. Still, the Web applications on RISC OS scene is probably crowded with takers for this kind of thing, anyway.

guestx on 20/6/07 12:02PM
[ Reply | Permalink | Report ]

Charging 25 quid for a web server that has this gem in its manual is laughable: "It is our intention that Serviette will support version 1.1 of the Hypertext Transfer Protocol (HTTP)." There are others: "VirtualAcorn is not and never will be supported." "Serviette supports the GET method on both RISC OS 5.11 and RISC OS 6.06. Serviette supports the POST method on RISC OS 4.39 only. RISC OS 6.06 has yet to be tested."

I don't see anything there that you can't do with WebJames and have better performance. And the web search thingy is so dire to be useless, certainly not as good as a Google search with "RISC OS" in the query.

It's a shame she's appeared to make no effort to talk to anybody about it for input - things might have been better.

rjek on 20/6/07 12:42PM
[ Reply | Permalink | Report ]

One search a day!!!

You don't know how lucky you all are! I can't even get one search from here. NetSurf just returns an "Empty document" error.

pv on 20/6/07 12:50PM
[ Reply | Permalink | Report ]

"Running a website off a RISC OS box? Insane!"

Could not agree more. Only question is which constraint memory or Blocking I/O are you going to hit first. Risc OS web serving is just NOT going to scale.

If its low volume then its a failure, If it attracts high volume its going fail.

Jwoody on 20/6/07 12:50PM
[ Reply | Permalink | Report ]

What I wouldn't mind knowing is whether the Web server is just a welded shut version of the one you get in the Python libraries.

guestx on 20/6/07 12:55PM
[ Reply | Permalink | Report ]

Loading the search page seems very slow to me - I appreciate that there are many factors affecting this sort of thing, but do people think that the fact that it is running from (possibly) a RPC or similar class machine make a big impact?

polas on 20/6/07 1:20PM
[ Reply | Permalink | Report ]

Well I just tried riscpedia.net and managed 5 searches before it told me I could only do one search.

VirtualAcorn on 20/6/07 2:40PM
[ Reply | Permalink | Report ]

“It's a shame she's appeared to make no effort to talk to anybody about it for input - things might have been better.”

Except for [link] of course.

IvanDobski on 20/6/07 3:26PM
[ Reply | Permalink | Report ]

polas: How about: [link]

adamr on 20/6/07 3:51PM
[ Reply | Permalink | Report ]

From the Google link above we find this statement;-

"My Acorn RiscPC is now up and running again as my web server ([link]) with Serviette as the web server software."

From the address link we find this statement:-

"You are connected to 84.92.157.78 running RISC OS 5.11"

How has she managed to do this? ;-)

ajb on 20/6/07 4:53PM
[ Reply | Permalink | Report ]

Im sure CJEMicros run their website off a Risc PC...

OliverB on 21/6/07 10:43AM
[ Reply | Permalink | Report ]

We have, of course, had a RISC OS search for over 5 years, namely acornsearch.com. Sadly that domain has gone to a cybersquatter, but the search is integrated at the top of this very page, as 'RISC OS Search -> Websites'

FWIW Rebecca seems to be aiming for a Yahoo style directory search, but without ability to browse the directory.

caliston2 on 21/6/07 10:57AM
[ Reply | Permalink | Report ]

Quite frankly I am a bit upset, how almost everyone here seems to immediately wash down this Riscpedia/Serviette project. It is still in its early stages, and it is not often that we attract new software developers.From what she wrote on usenet (follow adamr's link) it sounds like she is trying to develop some kind of distributed webserver. This sounds very interesting to me. Imagine a server rack full of a9homes, with only a few powered up to manage the rack and answer initial request, powering up the rest as the load increases. Energy consumption would be very low compared to traditional webservers, eleminiating thermal problems at the same time. The whole thing would be very scalable as well. The name suggests some kind of online-encyclopedia like wikipedia.

@Jwoody: Not all websites are targetted at large audiences, so not all websites with a low volume are a failure. So far we know very little about this new site's purpose.

JGZimmerle on 21/6/07 2:50PM
[ Reply | Permalink | Report ]

I to was disappointed at how negative all the comments have been to Riscpedia and the Serviette project and also surprised at the fact that so few people had been aware of it.

I remember reading in February's edition of Archive a 4 page article by the creator of this initiative explaining their aims and requesting feedback, I also think I read another version of this article somewhere else possibly in Eureka? so she has not been quite as secretive as you may think.

However this does seem to be a project that will need a lot of cooperation from developers and people running RISC OS related websites.

Vic

vshears on 21/6/07 4:00PM
[ Reply | Permalink | Report ]

As far as Riscpedia is concerned, I don't think that it is a bad idea as such, however to me in the current incarnation it seems that charging people to search it is not particularly useful. IMHO it would be a lot more beneficial (and available as a showcase of the power of her webserver) if the author opened it up and invited people to browse and submit articles freely so that it could be built up into a powerful resource by the community.

It is nice to see new development and I wish author good luck with her project

polas on 21/6/07 4:43PM
[ Reply | Permalink | Report ]

I would question the legality of charging for access to something that (badly)re-parses a results page from altavista. Not tried the 'riscpedia' search yet, I've used up my '1' search....

piemmm on 22/6/07 7:00PM
[ Reply | Permalink | Report ]

Blimey!, the poo has really hit the whirly thing; I never realised there were so many hornet's nests in the RISC OS Community.

I would like to clarify the following points made about Serviette and Riscpedia: I felt very uncomfortable running a RISC OS web site on a server that wasn't a RISC OS computer but saying I'm insane for doing so really helps the future of RISC OS; Had I opted for a different OS, Serviette for definite would not now exist! My web server is now an Iyonix and not one of my RiscPCs. To ease the load on my Iyonix should it be required, I have a load balancing gadget plus an A9 and two RiscPCs at my disposal. Serviette is built up from Python's Socket library and does not use Python's supplied HTTP or CGI libraries. Serviette differs from other RISC OS web servers in that it runs Python scripts, handles Python Server Pages and has the potential to call Python objects. As for charging GBP25, is it now a trend that all RISC OS software must be free of charge? The price of Serviette was more for providing support than for the software itself.

My idea for a distributed XML-based RISC OS database seems to have passed many of you by. Riscpedia was meant to be the first web site to use such riscosnet XML files with which to populate itself. I decided to combine the display of an XML record (the XML tags will be hidden in due course) alongside a list of hyperlinks. Last week, only the hyperlinks side was implemented and because you were obviously unaware of my riscosnet idea from February, you picked up the wrong stick and started to beat me with it. The 6000 or so records in XML format have now been added to the site although the XML syntax used is wrong and they are not pretty printed. The search engine options have now been removed as spidering is now working.

So the question I ask you all is whether the distributed riscosnet XML idea is a good one and worth continuing with or am I just wasting my time. If no one is prepared to generate a riscosnet XML file of their own, then Riscpedia is reduced to emulating Acorn Search as you say all-be-it with a regular expression-based search facility (which no one will probably learn how to use!). Each riscosnet XML file will be owned by the person who generated it and sites such as Riscpedia will be able to read such files and extract whatever information they require. When and if such files exist, Riscpedia can then discard its current XML record for a particular entry and use yours instead. Without the backing of the RISC OS Community, I might as well give up on the idea now.

The reason for restricting access to the site was to ensure I didn't exceed my bandwidth limits and to encourage a number of small donations to be made to reinburse me for some of the energy I have and will put into the riscosnet XML schema/RISC OS Encyclopedia idea. I accept that there is probably very little money in RISC OS at the moment but the distributed database scheme I have outlined has the potential to attract new users. I certainly wouldn't be encouraged to join the RISC OS Community on the strength of what I read in a public forum.

Please read what I wrote earlier this year and let me know what you would like me to do and what you would be prepared to do as I have never felt more like an outsider than I do now.

So, the moral of this story is to ensure your audience have all the relevant facts, don't release half-baked ideas regardless of how excited you are and don't ask for money as it often offends! Wow, three new things learnt in one week!

Rebecca on 23/6/07 12:08PM
[ Reply | Permalink | Report ]

Rebecca,

Don't you know that Free in RISC OS land is now old hat they also want you to develop a "killer" app and then pay them to use it as well. "Free" is so old hat, so there.

Seriously well done for trying to do something on RISC OS and ARM harware. It amazes me that everyone moans about RISC OS not being able to do this and that and when you provide something your shot at.

I hope that Serviette goes from strength to strength.

bluenose on 23/6/07 12:58PM
[ Reply | Permalink | Report ]

bluenose:

Not offering our help are you? ;-)

Rebecca:

Its fair enough that you want money for your efforts.

Now that you've told everybody it now makes sense.

I'm with JGZimmerle on this one - this is potentially a really cool thing.

Keep your end up :-D

epistaxsis on 23/6/07 1:31PM
[ Reply | Permalink | Report ]

I like the really cool bit about Non RISC OS users only getting one free search in 3 hours.

I can see this starting a whole new trend of IE and Firefox users faking Fresco, WebsterXL & Netsurf just to get a free search.

Well at least with IE they have a bit of a ahead start on faking Fresco as it already crashes for no reason!

bluenose on 23/6/07 1:48PM
[ Reply | Permalink | Report ]

There are probably many like myself who don't fully understand the complexity of what you are doing but are appreciative of your efforts. Many of us are grateful for anyone supporting RISC OS. Congratulations on what you've done and keep at it. Some of us do appreciate it.

rmac on 23/6/07 3:21PM
[ Reply | Permalink | Report ]

I'm with armac, too, Rebecca!

CharlesB on 23/6/07 4:55PM
[ Reply | Permalink | Report ]

Rebecca:

What about promoting python as a clientside programming solution on RISCOS. Would this not play better to its strengths?

markee174 on 23/6/07 6:33PM
[ Reply | Permalink | Report ]

To quite a few:

It is amazing: Rebecca came up with an interesting piece of software for RISC OS and surely one offering features not yet present since otherwise she would not have written Serviette but used something there already. But instead of saying things like "wow", "great" or at least "good" or "interesting" quite a few comments are along the lines "insane", "stupid" and "bad".

Well, everybody is allowed a personal opinion - no problem with that. But on the other hand more often than not lack of functionality is one of the main critisisms for RISC OS. Thus applications and those behind them offering new functionality or enhanced features should be thanked and applauded - be it for trying in the first place and/or the results... and that even if they developed something the person commenting doesn't need for herself or himself. Any new feature or functionality be it added as part of RISC OS or as an application or module or ... is good since it helps RISC OS go forward. And don't forget that quite a few good apps have been the result of some programmer starting to write the app for his or her own need! And thus every try to do so is good as well since without anybody even trying we wouldn't have as many applications around.

But on the other hand I'm not surprised of the critisism since for quite a few it seems to be more fun to critisize instead of offering help, suggestions or at least encouraging. Just think what you would say if you see a glass of wine, beer, water or whatever filled to 50% level. Do you claim it to be "half empty" or "half full". I hope it is the latter since that is a more positive thinking ... Or if somebody asks if someting can be programmed, do you say "No, that can't be done because ...." or do you say "It can be done, if ..." Both statements (can) state the same information but stated in a differend manner with the latter being a lot more positive. Looking at Serviette and the possible lack of performance of an IYONIX pc for greater load you can either say "Serviette on RISC OS is stupid since RISC OS systems can't handle the load" or "If much load is put on Serviett running on RISC OS then some enhancements like a faster IP stack or threads in RISC OS are needed".

Summing up: Thanks, Rebecca and don't let them strange comments put you off - it's not you outside the community!

hzn on 24/6/07 10:17AM
[ Reply | Permalink | Report ]

The problem is a RISC OS Web server is not very elegant from a technical point of view. To service requests you need to either look up information in memory or disk. If you use memory then compared to a PC with upto 4Gb you are onto a hiding to nothing. If you look up on disk then as soon as your Risc OS program makes a request in C, Python,Basic or what ever, only interrupts can happen whilst the I/O completes. Where as a PC with Windows or Linux can hanfle many at the same time. End result means a PC can handle 10-100 times as many requests as RISC OS box. RISC OS might be okay for low volume traffic but it would be mad to suggest that google upgrade their server farm to RISC OS.

Yes you could front a number of RISC OS server machines, but why bother when you could handle more with a single PC.

Personally I think RISC OS is technically elegant with its GUI, its not technically elegant as a server be it Web or SQL and people should stick to its strengths.

I guess one should not be surprised that as the RISC OS market is getting down to the faithful few and that a fair percentage of the few don't understand the technical limits of their chosen platform, otherwise they would have left the party some time ago.

Jwoody on 24/06/07 1:05PM
[ Reply | Permalink | Report ]

In reply to Jwoody:

"Personally I think RISC OS is technically elegant with its GUI, its not technically elegant as a server be it Web or SQL and people should stick to its strengths."

I agree with the first part but *disagree* with the second!

If we always stick to what RISC OS and the apps for it can do we will never proceed to more features. Why should anybody look into a faster IP stack and/or threading and/or ... unless there is need or use or request for that.

hzn on 24/06/07 5:08PM
[ Reply | Permalink | Report ]

hzn: "Why should anybody look into a faster IP stack and/or threading and/or ... unless there is need or use or request for that."

I think that's what Jwoody is trying to point out. For this to work comfortably, a faster, more powerful machine is strongly advisable. There is no shame in using a PC or Mac to help RO users, provide RO related information or sustain a RO business. Strictly speaking a RO server can work, but if one looks at requirements today and sees where it's going, the demands to both the OS and hardware go far beyond what's realistically possible. Yes, I realise most visitors and requests would probably come from the relatively small amount of RO users, but even then the possibilities are tight.

I tend to agree with Jwoody. People should stick to RO's strengths and build from there. Too few people and too little resources prevent any significant developments, so I believe we should be modest with any expectations, without losing sight of what's realistically possible and achievable.

"I guess one should not be surprised that as the RISC OS market is getting down to the faithful few and that a fair percentage of the few don't understand the technical limits of their chosen platform, otherwise they would have left the party some time ago."

Firstly, the RISC OS market already got down to the faithful few, just like most other little known OS's dating back to the Golden Age. That's not a criticism, merely an observation which should help some people let go of history and comparing today with yesterday. Nonetheless, I do think most users are aware of many technical limits of RISC OS, as evident by the lack of certain application software - so not strictly because the resources are lacking, but because the platform is technically incapable. Again, not a criticism, merely an observation.

While I strive to keep using RISC OS, I won't pretend these limits are non-existent. Rather, it focusses RO's use from a general purpose machine to a specific application computer, where swift, ease and productivity are paramount. I'm sure many other RO users are very happy with their machines, not because of what it cannot do, but the few things it can do extremely well.

hEgelia on 25/06/07 1:55PM
[ Reply | Permalink | Report ]

Changing the topic somewhat (but still on the article) - I am really impressed with RPCEmu and would like to thank and congratulate Tom Walker on his work so far - it is very useful on my Unix machine and works very well.

polas on 25/06/07 3:43PM
[ Reply | Permalink | Report ]

In reply to hEgelia:

I have the impression you agree with Jwoody in that people should strick to RO's strengths - or the. That is exactly what I do not agree with. If we stick to RO's strengths only then that is pretty limitating as for what can be done. Enhancements and innovation tend to come when people try to exploit other paths.

As for the need of a more powerful machine: no. Some time ago I did the odd benchmark with RISC OS systems (Risc PC with 10MBit card, Risc PC with 100MBit card, IYONIX pc classic, Panther, Buffalo Linkstation as NFS, 1GHz Pentium system and 3GHz Pentium system, the last three with 100 MBit Ethernet). The by far fastest file transfer was using FTP from Penitum box to IYONIX pc, both running debian Linux and both machines were still usable during transfer. Throughput was at the limit of what 100MBit ethernet can offer. Whatever other combination and protocol tested at least 30% slower, most even slower. Thus by using more efficient IP code like the debian stack network speed should go up siginificantly on RISC OS with current hardware.

Or just look back a few years where graphics capabilities of e.g. A5000 were dreadful but that was fixed with ColourCard and a similar solution from some other supplier - they simply ignored that the graphics hardware was no strengh. Or take the odd fast IDE interface - again harddisc access was no strength of RISC OS but lucky us, some companies ignored that fact.

But there are cases where fixing should be easier since we're talking about pure software: DOSFS with the 2GB size limit which is more and more of a bad joke. It does really astonish me that even suppliers of an allegedly "cutting edge OS" for RISC OS hardware fail to fix this - but perhaps "cutting edge" does not mean that it is something to be considered very good and modern though I was given to understand just that. Instead of that round buttons and more screen savers (all not the energy saving kind) are developped - a shame!

By now I tend to use a RISC OS and a Windows box side by side and depending on the work at hand use one or the other - some 2-3 years ago this was different since then I was able to use the RISC OS box for quite a bit more of my everyday work...

hzn on 25/06/07 4:21PM
[ Reply | Permalink | Report ]

hzn: "I have the impression you agree with Jwoody in that people should strick to RO's strengths - or the. That is exactly what I do not agree with. If we stick to RO's strengths only then that is pretty limitating as for what can be done. Enhancements and innovation tend to come when people try to exploit other paths."

Yes. I certainly agree with your point of view, except that I do no longer anticipate much scope for "enhancement and innovation" in the market as it is today. That may strike you as cynical or pessimistic, but it seems most entrepreneurs have left and what remains is an enthusiastic minority relying on a few tired and unimaginative companies turning every penny to stay afloat. Castle and ROOL provide a possible future, along with users such as yourself who are willing to invest. The 'ROL camp' are a dead-end in my opinion, even though technologies such as RO Six are fairly impressive.

Like I said before, we should initially stick to RISC OS's strengths, then build from there. These strengths are what set RISC OS apart from the rest and these can act as cornerstone for further, viable, development. In its current condition, RISC OS can't try too much too fast without significant risk.

"As for the need of a more powerful machine: no. Some time ago I did the odd benchmark with RISC OS systems (Risc PC with 10MBit card, RiscPC with 100MBit card, IYONIX pc classic, Panther, Buffalo Linkstation as NFS, 1GHz Pentium system and 3GHz Pentium system, the last three with 100 MBit Ethernet). The by far fastest file transfer was using FTP from Penitum box to IYONIX pc, both running debian Linux and both machines were still usable during transfer. Throughput was at the limit of what 100MBit ethernet can offer. Whatever other combination and protocol tested at least 30% slower, most even slower. Thus by using more efficient IP code like the debian stack network speed should go up siginificantly on RISC OS with current hardware."

Sounds interesting. Well, if someone could integrate this network stack into RISC OS, that would be fantastic. Perhaps this could be done with RISC OS Open / RO5 ? If RISC OS Six ever gets released for the Iyonix, perhaps its network speed could prove interesting as well. Still, for the moment, I wouldn't try running a server on RISC OS.

"Or just look back a few years where graphics capabilities of e.g. A5000 were dreadful but that was fixed with ColourCard and a similar solution from some other supplier - they simply ignored that the graphics hardware was no strengh. Or take the odd fast IDE interface - again harddisc access was no strength of RISC OS but lucky us, some companies ignored that fact."

Indeed, but as you say - looking back a few years. Those 'few years' are in fact about 10 to 15 years ago. A whole other market, with active development, many more users and up-to-date hardware and software which could compete reasonably well with PC / Mac offerings. Those times allowed for certain endeavours, but times have changed drastically since then, which is reflected in the market of today.

"But there are cases where fixing should be easier since we're talking about pure software: DOSFS with the 2GB size limit which is more and more of a bad joke. It does really astonish me that even suppliers of an allegedly "cutting edge OS" for RISC OS hardware fail to fix this - but perhaps "cutting edge" does not mean that it is something to be considered very good and modern though I was given to understand just that. Instead of that round buttons and more screen savers (all not the energy saving kind) are developped - a shame!"

Absolutely. The DOSFS limit is ridiculous and should have been resolved years ago. However, I do wonder who you're referring to as suppliers of a "cutting edge OS" ? Certainly not RISCOS Ltd, I hope, but they did develop more significant things than merely 'round buttons' and more screen-savers. It's easier to recognise some of their changes outwardly, but they do not convey the entirety of Select. Still, it's a pity they never addressed the urgent DOSFS limitation.

"By now I tend to use a RISC OS and a Windows box side by side and depending on the work at hand use one or the other - some 2-3 years ago this was different since then I was able to use the RISC OS box for quite a bit more of my everyday work..."

Same here. Indeed, I've come to love the rare, but extremely productive, RISC OS sessions. Though I must admit, the Mac gets switched on more often now for all the general purpose tasks.

hEgelia on 25/06/07 10:14PM
[ Reply | Permalink | Report ]

hEgelia:

First, I think we should distinguish between ordinary users and developers. Your point of view "stick to RISC OS's strengths" is the right one for users, but I think that developers (by that I mean anyone with the technical ability to work on the sources and/or hardware) should make hzn's POV their own.

As for "about 10 to 15 years ago. A whole other market" I fully agree. RISC OS' sources are becoming available and there is more high-quality open source code available on the internet than ever before. Today there is no need for a big user base to finance expensive development work. Anyone with a bit of knowledge (or the motivation to learn it) can download very good, mature code from other operating systems like OpenSolaris or FreeBSD and use it to improve RISC OS. Very little knowledge of software engineering is required for this, as the existing code only has to be adjusted to work with existing RISC OS interfaces.

JGZimmerle on 25/06/07 11:58PM
[ Reply | Permalink | Report ]

Please login before posting a comment. Use the form on the right to do so or create a free account.

Search the archives

Today's featured article

The new apple of my eye
Would you swap your dusty Acorn for a polished Apple computer? Martin Hansen has been checking out the world of Steve Jobs and his range of shiny kit.

15 comments, latest by adh1003 on 6/1/09 1:06PM. Published: 17 Nov 2008

Random article

New RiScript supports CMYK PDF export
Free update is Christmas present for punters

9 comments, latest by hzn on 2/1/08 12:57PM. Published: 27 Dec 2007

Useful links

News and media:
Iconbar • MyRISCOS • ArcSite • RISCOScode • ANS • C.S.A.Announce • Archive • Qercus • RiscWorld • Drag'n'Drop • GAG-News

Top developers:
RISCOS Ltd • RISC OS Open • MW Software • R-Comp • Advantage Six • VirtualAcorn

Dealers:
CJE Micros • APDL • Castle • a4 • X-Ample • Liquid Silicon • Webmonster

Usergroups:
WROCC • RONE • NKACC • IRUG • SASAUG • ROUGOL • RONWUG • MUG • WAUG • GAG • RISCOS.be

Useful:
RISCOS.org.uk • RISCOS.org • RISCOS.info • Filebase • Chris Why's Acorn/RISC OS collection • NetSurf

Non-RISC OS:
The Register • The Inquirer • Apple Insider • BBC News • Sky News • Google News • xkcd • diodesign

© 1999-2009 The Drobe Team. Some rights reserved, click here for more information
Powered by MiniDrobeCMS, based on J4U | Statistics
"We incurred the wrath of Chris Williams, the main journalist at Drobe, who told us to 'Take our filthy spam and shove it where the sun doesn't shine'"
Page generated in 0.1274 seconds.