Re: Hotlist window issue
On Sun, Aug 27, 2023 at 11:37:26PM +0100, Harriet Bazley wrote: > I've just noticed that the same thing applies to the global history > window - if I check that in WinEd, the history window in the Templates > file has *neither* 'Allow off screen' nor 'Keep on screen' flags set. Both of these use the "tree" template, whose source can be found here: http://git.netsurf-browser.org/netsurf.git/tree/frontends/riscos/templates/en#n2605 (And similar in the other languages.) B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: User Agent
On Mon, Jun 19, 2023 at 11:55:35AM +0300, jtz...@post.com wrote: > What does "--url_file" do? will NS load all urls listed there? It controls where the URL database is stored, ie history etc. It does not do what you want. We should probably provide a way to define more than one home page in preferences to open multiple windows/tabs, but this may end up being front-end specific rather than in the core. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: User Agent
On Mon, Jun 19, 2023 at 11:21:21AM +0300, jtz...@post.com wrote: > Thanks. > Can the gtk version accept multiple urls on the commandline?. > > The man pages shows "netsurf [options] [url]" (i.e., a single url) & > issuing "netsurf-gtk www.yahoo.com www.google.com" starts with a > single tab (yahoo). That suggests no, then. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: User Agent
On Mon, Jun 19, 2023 at 10:50:52AM +0300, jtz...@post.com wrote: > Thanks Rob for the prompt reply. > > Another question please - are tabs supported?. It depends on the frontend in use. I know the Gtk front end does, and I know the RISC OS one does not. As for the others, I am uncertain. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: User Agent
On Mon, Jun 19, 2023 at 10:17:41AM +0300, Joel T. wrote: > > Hi, > > Can the User Agent string be set, e.g. as in Elinks?. Not without recompiling it, see utils/useragent.c B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: ci.netsurf-browser.org down?
On Tue, Sep 13, 2022 at 08:14:51PM +0100, David Higton wrote: > I was rather surprised to see that netsurf-browser.org and > ci.netsurf-browser.org resolve to completely different IPv4 addresses, > but the latter is unreachable to ping4. The former is hosted by me and Daniel at Pepperfish (along with email and mailing lists). The CI and autobuilder is hosted by Vince in his garage and should not really be considered as an essential service. I'm sure it'll be back soon. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: How to build with framebuffer usage?
On Sun, Jul 10, 2022 at 05:51:41PM +0200, Peter Wiehe wrote: > Hi all, > > how do I build netsurf on Linux with framebuffer usage instead of gtk? > > (I tried already TARGET=framebuffer in the makefile, but that didn't > work.) How precisely did you try? Using env.sh as described in our quickstart guide, this just worked for me first try. What did you type in? What errors did you see? B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Online converters
On Mon, Jan 31, 2022 at 05:35:46PM +, Harriet Bazley wrote: > I haven't used it for a while, but I just tried it on a file from the > Office of National Statistics and so far as I can tell, it worked - > however, the resulting file was so huge (30 pages in !ViewXLS) that > Fireworkz ran out of memory before managing to convert and generate all > 30 Excel sheets, while !ViewXLS only displays the text labels and not > the formula results. A reminder to all: This mailing list is about NetSurf, and is not a general RISC OS support group, and that my advice for opening docx files and similar is to buy a computer that can run LibreOffice. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Wix site images
On Mon, Jan 03, 2022 at 09:28:46AM +1300, Andrew Hodgkinson wrote: > There is almost no universe in which this is better or faster than > just letting the web browser load the original image directly, with > the web site building providing a precise at-in-page-resolution copy > of the image for the page and optionally a click-through to the full > resolution version, if available and if permitted by the person who > put the web site together. This is true, but it's not the reason websites do this. They do this because it drastrically reduces their bandwidth costs and bursts of activity on the httpd, which is important if you're paying AWS's rates. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Netsurf for Palm OS
On Wed, Dec 01, 2021 at 12:59:51PM +, boeing chan wrote: > I was not sure where to direct this email to, but I was wondering if netsurf > is available for Palm OS? Thank you in advance for any reply. Nobody has attempted a port to PalmOS as far as I am aware. I think it might be a tight fit, even on the very last models of Pilot PDA though. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: #5290 crash
On Wed, May 19, 2021 at 09:15:39AM +0200, Paul Sprangers wrote: > Dear all, > > The latest build (#5290 - 2021-05-17) crashes on my system (4té - RISC OS > 5.29) as soon as I open a page. It's very repeatable and I had to > re-install #5289 in order to get to Google again. Do you have the log/backtrace? > I've tried to send a bug report using the 'Bug reporting interface', but > that horribly failed. Howso? B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: webp file handling
On Wed, Feb 24, 2021 at 03:12:51PM +, Tim Hill wrote: > > As discussed elsewhere, it's great to see that .webp files are rendered > by NetSurf but files dropped are ignored if they have the correct (A66) > RISC OS filetype. > > (So I haven't added WebP files to mimemap yet and nor should you!) Try #5263. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Rubbish
On Mon, Jun 22, 2020 at 07:47:10AM +0100, Dave wrote: > > Seems to be getting a lot of rubbish advertising spam on this lists ATM. > ? Sorry for the noise - see my previous posting. We've put a hack in to disable web posting from the archival website while we wait for Mailman to sort out a proper fix. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Recent list spam
Apologies for the recent flurry of spam to the list. This is due to an oversight in the new mailing list software we are using that makes it easy to post from the web interface: there's no way to turn it off it make it hard for bots to sign up. We have an open issue. We would prefer it if people did not report these as spam to their ISP, because they "come" from us and it may mean you end up not receiving legitimate postings from our server. We're trying to get it fixed as soon as possible without just entirely turning off the archival interface. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: appimage and flatpak
On Thu, Jun 04, 2020 at 12:19:59AM -, danielrobidou...@yahoo.ca wrote: > Ok, I've seen netsurf on flathub.org, thats great. > Note: I cant edit my post? No, this is a mailing list :) B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Non SSE Build
On Tue, May 19, 2020 at 11:53:44AM +, David Feugey wrote: > Yes, I tried and I get some illegal instruction error. > > I did have the same kind of problem with Fireworkz, and it was a SSE problem. > But perhaps it's a MMX issue too. I don't know. > > That's why I would like to test with some other build options, and check it > on some old hardware. I believe if you have a Linux box to hand, building any of our ports is reasonably straightforward as we provide pre-built toolchains and instructions on how to drive it all. (We cross-compile everything from Linux because life's too short to build software on RISC OS and Windows natively.) Modifying the Makefiles to pass something like -march=i586 should target CPUs with no MMX oe SSE support. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: Non SSE Build
On Tue, May 19, 2020 at 01:32:28PM +0200, David Feugey wrote: > Hi all. > > NetSurf is very cool. And perfect on old computers. > > Unfortunately, under Windows, it's (probably) compiled with SSE > optimisations. So it's not possible to use it with embedded devices, as > my eBox under XP (it will be probably the same with new dual core eBox > under 7). Have you tried it? Guessing from the filename of the toolchain download, the toolchain targets "686", which is P6/Pentium Pro which pre-dates the Pentium 3 that introduced SSE. B. ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: forms -- the "required" attribute
On Tue, Mar 10, 2020 at 04:23:30PM +, Jim Nagel wrote: > I'm setting up an HTML and tried to use the "required" attribute on > one input field. (Actually my form consists of only that single field.) > > Unfortunately "required" has no effect in Netsurf (#5033 on Ro 5.25). The > user can get away with just clicking Submit without entering any data. > > In the Netsurf documentation, the section on web standards says "Forms: > nearly done". I presume this attribute is one of the undone-yet! Indeed, however: given that you absolutely must check that the required field has been provided at the server end, this only serves to prevent a round trip and server-side error message rendering. This is important because, well, not all HTTP clients will enforce the required attribute, and people could inject bad data by just using wget or curl. In short, you should be checking at the server anyway. B.
Re: Google and #spf
On Mon, Sep 23, 2019 at 03:27:59PM +0100, Jim Nagel wrote: > Sure makes one wish for that "Javasccript on/off" button on the Netsurf > button bar that we've oft wished for! Please remember that there is no RISC OS front end maintainer for NetSurf anymore - nobody is working on the RISC OS front end beyond making it build with core changes. If anyone wants to volunteer for this please step forward! Some people have expressed interest in the past but have either vanished or discovered they did not have the spare time they thought. B.
Re: Sargasso RSS feed reader
On Thu, Jun 27, 2019 at 12:12:10AM +0100, John Williams wrote: > > James, please look at thread: James Bursa has not been involved in NetSurf for a very long time, and as far as I can tell, apart from his @netsurf-browser.org task alias, is no longer subscribed to this list. (For non-RISC OS people, this is related to an earlier piece of software James wrote before NetSurf on RISC OS.) B.
Re: new release and nested
On Wed, Aug 15, 2018 at 12:43:03PM -0500, Cág wrote: > Michael Drake wrote: > > > Our current layout engine doesn't support dynamic changes to the page. > > The layout engine needs a rewrite to handle this, which is a big job. > > Will this big job make the browser big too? And if so, should it be > planned then? Almost certainly bigger, but we won't know how much bigger until it's done. B.
Re: Please try the latest build
On Wed, Aug 15, 2018 at 09:19:02PM +0100, Chris Newman wrote: > > I got one of these but my mail was posted anyway. (see above). I'm confused > (normal state). > > "You are not allowed to post to this mailing list, and your message has > been automatically rejected. If you think that your messages are > being rejected in error, contact the mailing list owner at > netsurf-dev-ow...@netsurf-browser.org." Look carefully at the To: field in the email you sent: you also send it to our developer's list, which I assume you are not subscribed to. B.
Re: NetSurf web sites
On Thu, Jul 19, 2018 at 08:38:52PM +0100, cj wrote: > In article <921e311a57.davem...@my.inbox.com>, >David Higton wrote: > > Also I notice that the !Fetch_NS app no longer works, complaining of > > an incorrectly formatted filename at line 24. Probably not your > > problem, but I struggle to see how it works and thus how the error > > occurs. Help would be appreciated! > > You have to have a recent version of wget that understands https and > change the command options when using wget. There was a recent thread > in comp.sys.acorn.apps where this arose. I suggested using the option > > --no-check-certificate > > which works around the certificate check. Please note that this defeats the whole point of us switching to HTTPS in the first place. A better solution would be to provide wget and whatnot with the certificates to check. NetSurf itself ships a set of certificates, but I'm not sure if the format is compatible. Check with who ported wget. B.
Re: linking to root of website, not root of drive
On Fri, Feb 23, 2018 at 02:18:18PM +, Jim Nagel wrote: > - The distant server (probably running on Linux) does not care about > RiscOS filetypes and Netsurf correctly displays pages from there when > my !2018.css has RiscOS filetype as text That's because it will be using the file extension to decide what content type to tell the browser, where WebJames uses the RISC OS metadata. > - The local development server (Webjames) does a successful GET of > !2018.css whatever its RiscOS filetype, but Netsurf does not display > pages correctly when !2018.css is Ro-filetyped as text, and *does* > display them correctly when !2018.css Ro-filetype is changed to CSS. See above. B.
Re: mailto:
On Tue, Feb 20, 2018 at 09:02:57AM +, Richard Torrens (lists) wrote: > > This system appears to defeat spammers - most of the email addresses that > have been spammed seem to have been harvested from computers by malware. We're getting dangerously off-topic here, but they mostly harvest them from: - usenet and mailing lists such as this (they are so bountiful, the harvesters simply special-case the mitigations that mailing list archival websites use to obscure addresses) - brute-forcing people's email passwords, not only to send spam through but to harvest addresses from your INBOX and address book. - Guesswork. Attempting to send an email is cheap, so when you get a list of current domains, you simply try all the common stuff, like "lists" or "tim". Doing it on your website with mailto: links isn't worth it these days, spidering the web like that is comparitively expensive compared to the above techniques. B.
Re: linking to root of website, not root of drive
On Mon, Feb 19, 2018 at 03:46:41PM +, Jim Nagel wrote: > - Netsurf displays the pages served up via Webjames perfectly, > except that it is NOT obeying the stylesheet (which is /2018.css ) > even though the Webjames log records a GET success for the CSS file > just as it did for the logo and paper and favicon. > - Yet when Netsurf fetches the identical page from my ISP's server, > it DOES obey the CSS. Complete guess: is your CSS source file set to the CSS type, or plain text type? I'm not sure how NetSurf reacts to CSS files served with an appropriate Content-Type header. B.
Re: Failure to initialise Unicode font library
On Thu, Dec 07, 2017 at 11:12:07AM +, John Williams wrote: > Any suggestions as to what to look at next? Have you added any fonts recently? B.
Re: URL lengths
On Thu, Nov 30, 2017 at 02:32:29PM +, Tim Hill wrote: > > I just can't believe that a URL can contain 33% more letters than the > King James Bible's 3,116,480. > > "four gigabytes per URL" It's probably much greater than that on modern systems, too! B.
Re: URL lengths
On Mon, Nov 27, 2017 at 06:08:46PM +, John Williams wrote: > An genuine real-life number would be better! >From memory, they are dynamically allocated as needed and can be of arbitrary length. This is a lot more efficient than using statically-sized buffers for the worst common case, especially in a piece of software that might be storing information on tens of thousands of URLs (Everything in your history, the links to all the resources on every page you have open, etc). My advice is firstly to either measure and then allocate, or read in the data and reallocate as needed dynamically. Secondly, use a different language; BBC Basic really isn't good for this sort of thing :) (If using a modern interpreted language like Lua, Python, or even Perl, asking this question would never have occurred to you.) B.
Re: problem with Mailman or is it just me?
On Tue, Nov 07, 2017 at 11:50:18PM +, Richard Porter wrote: > On 7 Nov 2017 Jim Nagel wrote: > > > I wonder if a couple of others could try that link, please: try > > subscribing to the Netsurf list from a different valid address of your > > own. Do you get the same message as I did? What ISP are YOU using? > > I'm getting "You must supply a valid email address" whether I supply a > valid forwarder address, an alternative mailbox on my PlusNet account > or a PlusNet alias for my normal address. We've had reports of something similar for another mailing list we (Pepperfish) host. I'll take a look. Could somebody divulge an address that is incorrectly being identified as invalid, and which browser and version they were using? B.
Re: test2
On Sat, Nov 04, 2017 at 04:00:46PM +, Jim Nagel wrote: > I'm getting reports by private email that they received my earlier > "test" message in the Netsurf list. > > But here, I still have received nothing beyond the two messages I > listed in that earlier "test" posting. > > Peculiar. It's not that I am receiving NO messages from the list, > because those two odd ones DID arrive after Oct26. tl;dr: Your mail server is not routable from ours. You're subscribed via nets...@abbeypress.co.uk. root@outmail0:~# host -tmx abbeypress.co.uk abbeypress.co.uk mail is handled by 0 abbeypress.co.uk. (Incidentally, using an MX priority of zero is ill-advised, as there are some people still using qmail, and qmail uses an MX priority of zero to signal it can speak non-standard SMTP to it.) root@outmail0:~# host abbeypress.co.uk abbeypress.co.uk has address 193.35.57.245 root@outmail0:~# telnet abbeypress.co.uk smtp Trying 193.35.57.245... telnet: Unable to connect to remote host: Connection timed out The route seems to go dead inside somewhere colled "Veloxserv" in Wolverhampton. (Also sent to Jim directly.) B.
Re: Form submission timeouts
On Fri, Aug 11, 2017 at 05:16:09PM +0100, Dave Higton wrote: > > If anyone reading this thread can shed any light on MTU issues, > how to diagnose them and how to fix them, it would be very > helpful. Something that has bitten be before is where automatic path MTU discovery does not work (and I suspect it doesn't on RISC OS), and your ISP does not support end-to-end 1500 byte MTUs (many do not). The solution is to set your whole LAN's MTU to 1400 or similar, although setting it only on the RISC OS machines may be sufficient (and not needed on VirtualAcorn because that proxies Windows's IP stack which is full-featured and conforming) assuming the problem is only sending large packets, not receiving them. B.
Re: Browser ID/User Agent
On Tue, Aug 01, 2017 at 09:40:57PM +0100, John Williams wrote: > > I've been doing some stuff recently with browser ID strings, and find that > it is quite complicated! > > There is stuff called Mozilla compatibility, platform, platform details, > specific enhancements etcetera! All very difficult to untangle! > > But NetSurf (RISC OS) resolutely reports itself as: > > NetSurf/3.7 (RISC OS) > > which isn't terribly helpful/compatible with the likes of, for instance: > > Mozilla/5.0 (Linux; Android 6.0.1; SAMSUNG SM-T550 Build/MMB29M) > AppleWebKit/537.36 (KHTML, like Gecko) SamsungBrowser/4.0 > Chrome/44.0.2403.133 Safari/537.36 > > Observations, please! http://git.netsurf-browser.org/netsurf.git/tree/Makefile.defaults#n79 All the accreted nonsense in the UA string you site from your phone is because websites used UA sniffing to decide what to serve, and then web browsers needed to lie in order to be compatible. It was an endless cycle. B.
Pepperfish downtime last night
Hi, Summary: We had some downtime, and we're working on a fix. Services are currently up but may be slower than usual. Expect moments of downtime throughout the day. Last night, at approximately 0330 GMT, one of Pepperfish's virtual machines suffered a kernel issue related to handling of interrupts, causing it to become unresponsive. Sadly this happened exactly as a database update was occuring, causing some corrupted configuration files to be generated and pushed to other servers. The upshot was that DNS broke on our main server and on our secondaries. This had a knock-on effect of also preventing mail being delivered to us. A second issue relating to very poor IO performance on one of our virtual machine hosts has compounded the problem and slowed a correct fix. Given the extremely high uptime of the host (Over 1,200 days), and the lack of anything it any of its logs that might suggest why IO performance has become disappointing, we plan on rebooting it (and thus all the virtual machines it hosts) at some point today. This will mean the following services will be down: - ssh access to 'platypus' - Web server - IMAP server - Mailing list delivery server - One incoming mail server - One DNS server Secondary servers hosted elsewhere should receive and store any email that is delivered while this happens; nothing should be lost. Apologies for the disruption. B.
Re: Netsurf inconsistent behaviour between RISC OS 5.23 and 6.20
On Tue, Jan 24, 2017 at 12:42:08AM +, Brian Jordan wrote: > The home page of www.clubmans.org.uk has a news area in the lower part. > It is intended that this is a scrolling and in all mainstream > browsers I have tried this is the case [1]. it is also the case in > NetSurf #3966 running under RISC OS 5.23 on my Raspberry Pi. However, > using the same browser to view the same page on this emulator with RISC > OS Select 6.20 the overflows. The most likely explanation is, as > ever, that this is a local issue and that I have missed the obvious but > to help define the problem I wonder if some of you would take a look and > let me know what you see and the OS version you are using? If there is an > inconsistency relating to OS versions I will file a report to the NetSurf > developers, if not I shall have to dig a lot deeper here. My first and totally uninformed guess is the different FontManagers and fonts in use. Could you perhaps create a but on the tracker, and attach screenshots of both and the config files for both? B.
Re: !Cache
On Fri, Sep 09, 2016 at 01:37:42PM +0100, Richard Torrens (lists) wrote: > But I have replaced !Cache.Caches.Default.Netsurf with a zip archive > called Netsurf. This appeasrs to work fine, currently it's 492K - with 75 > directories in it. Replacing the directory with an image filing system using a file format that is not designed for random access will render the cache totally and utterly pointless unless you're on 3600 baud internet. Perhaps not even then. B.
Re: Segmentation fault started with no apparent cause
On Sun, Aug 28, 2016 at 09:48:43PM +0100, Andrew Pinder wrote: > My thought is that one conceivable explanation could be the file had > got too big. It was over 42KB in size. It seems to have shrunk down > to 39KB. Both these are trivially tiny. B.
Re: Segmentation fault started with no apparent cause
On Thu, Aug 25, 2016 at 10:21:52PM +0100, Andrew Pinder wrote: > There appear to be 106 URLs in there. I'm not going to test each one > to find where the problem is. Assuming there is nothing personally identifying or secret in your URL file, attaching it to the bug would be helpful. B.
Re: memory-gobbling on Virtual Acorn
On Sat, Aug 20, 2016 at 10:20:40AM +0100, Peter Young wrote: > On 19 Aug 2016 Jim Nagel wrote: > > > I'm using Virtual RiscPC on my ol' Windows XP laptop while away from > > base. Various wrinkles of this setup are still unfamiliar to me. > > Just a silly point, maybe. Why use NetSurf when you have much more > versatile browsers in Windows? NetSurf is excellent as far as it goes, > but can't compete with Chrome or Firefox on the dark side. I very strongly recommend never connecting Windows XP to the internet. B.
Re: Netsurf refresh
On Wed, Apr 13, 2016 at 08:35:01PM +0100, Jim Nagel wrote: > Vince M Hudd wrote on 13 Apr: > > Ctrl-F5 - which may be unfamiliar, but is standard on all the browsers > > I use (including NetSurf) - is the keystroke I use to reload pages > > when working on websites because it *does* fetch a new version of the > > page, rather than just reload it from a cache. > > Interactive help for the item on the Netsurf buttonbar does say > "Reload". Does it do the same thing as Ctrl-F5 or not? Vince's email is pretty clear. B.
Re: Netsurf Cache
On Wed, Apr 13, 2016 at 12:06:04PM +0100, David Pitt wrote: > > How about this from a Titanium :- > > (75142.82) content/llcache.c:3402 llcache_finalise: Backing store wrote > 2590212 bytes in 1394 ms (average 1858114 bytes/second) 1.8MB/sec to an SSD is nothing to write home about. I'd expect at least 20MB/sec from a modern OS on a spinning disc. B.
Re: multi-core processors
On Sat, Mar 26, 2016 at 06:14:29AM +, Brian wrote: > > NetSurf is single-threaded, by design. It won't take direct advantange > > of multiple cores. It'll still be faster on those systems which support > > multiple CPUs, because the OS can run on one while NetSurf runs on > > another. > > It does beg the question. What happens when NetSurf runs in VRPC on a Win > 7 machine with a multi-core processor? Does VRPC take advantage of > multiple cores? No idea, but NetSurf can do nothing to help in that respect. B.
Re: multi-core processors
On Fri, Mar 25, 2016 at 04:58:05PM +, Jim Nagel wrote: > Some of the new Arm hardware running RiscOS (e.g. the ArmX6 and > Titanium) has a multi-core processor. So far, of course, RiscOS lacks > the ability to make use of the extra potential. > > Question occurs to me: what's the situation with the other platforms > on which Netsurf runs -- Amiga, Atari, etc? Does multi-core ability > exist in hardware there? > > Do the Netsurf developers have ideas for how Netsurf could take > advantage of the extra processing power when the OSes allow it? NetSurf is single-threaded, by design. It won't take direct advantange of multiple cores. It'll still be faster on those systems which support multiple CPUs, because the OS can run on one while NetSurf runs on another. B.
Re: Very slow download
On Tue, Feb 16, 2016 at 11:23:09AM +, Brian wrote: > In article <20160216093631.gv7...@platypus.pepperfish.net>, Rob Kendrick > wrote: > > On Tue, Feb 16, 2016 at 09:23:45AM +, Brian wrote: > > > #3410 downloaded this morning, unusually, very slowly. Any reason, > > > perhaps? > > > 2016-02-16 09:36:12 (3.64 MB/s) - NetSurf-gcc--3410.zip saved > > [3842257/3842257] > > > Works for me. > > #3413 (104kB/s) which works, sort of. This kind of suggests the issue is at your end, or the ISP you use. B.
Re: Very slow download
On Tue, Feb 16, 2016 at 09:23:45AM +, Brian wrote: > #3410 downloaded this morning, unusually, very slowly. Any reason, perhaps? 2016-02-16 09:36:12 (3.64 MB/s) - ‘NetSurf-gcc--3410.zip’ saved [3842257/3842257] Works for me. B.
Re: Log in to YORK Libraries
On Mon, Feb 08, 2016 at 03:27:40PM +, Bryn Evans wrote: > Go to https://exploreyork.org.uk (which seems to take about 8sec > to load a largish Welcome page) and try to get the 'Log In' > or the 'My Account' (top right) to produce a 'Log In' window. > > There should be a USER entry writable and PIN one below but I never > get that far. Log In It requires both JavaScript and on-the-fly modification of the DOM. So no, not yet. B.
Re: Bug reporting
On Mon, Feb 01, 2016 at 05:42:24PM +, lists wrote: > In article <20160201152524.gp32...@platypus.pepperfish.net>, >Rob Kendrick wrote: > > > Yes, once Vince had identified and fixed the fault with Mantis sending > > some emails, I went and clicked the button to make it send it to you > > again. > > I'm now being bombarded with them > > My two attempts to create logins, plus my re-set password attempt, as well > as yours! Ha! Have you successfully logged in, now? B.
Re: Bug tracker
On Mon, Feb 01, 2016 at 01:32:51PM +, Vincent Sanders wrote: > There has been an issue with the bug system sending emails which has > now been fixed. Given the number of bounce emails I'm now getting, I see that quite a few people have apparent difficulty typing their own email address accurately :) If anybody is still missing expected password reset or sign up emails, please poke me or Vince to press the magic button for you to send it again. B.
Re: Bug reporting
On Mon, Feb 01, 2016 at 03:21:51PM +, lists wrote: > In article <20160201072351.gk32...@platypus.pepperfish.net>, >Rob Kendrick wrote: > > I have asked Mantis to send password reset emails for the more recent > > one. I'll leave the older one in case you happen to know the password > > for it already. > > I have now received a confirmatory email so can properly log in and report > bugs. This was timed 13.34. > > I note Vincent's comment on the bug tracker timed 13.32. Yes, once Vince had identified and fixed the fault with Mantis sending some emails, I went and clicked the button to make it send it to you again. B.
Re: Bug reporting
On Mon, Feb 01, 2016 at 12:18:41AM +, lists wrote:
> In article <437fcf4a55.r...@user.minijem.plus.com>,
>Richard Porter wrote:
> > On 31 Jan 2016 lists wrote:
>
> > > There is "My view|View issues|Change log|Roadmap"
>
> > > NOWHERE ON THAT PAGE IS A BOX THAT SAYS REPORT BUG HERE!
>
> > There should be a link "Report Issue" after "View issues".
>
> I'm not seeing that here - NetSurf #3312
You need to be logged in to report bugs. You have two accounts, one
created on 2013-12-09 12:19 ('stuartwinsor') and one create 2016-01-26
08:49 ('stuart'). I have no idea why you might not have received the
email, perhaps Orpheus classified it as spam or something.
I have asked Mantis to send password reset emails for the more recent
one. I'll leave the older one in case you happen to know the password
for it already.
B.
Re: Older Modules in NetSurfs System upgrader.
On Sun, Nov 08, 2015 at 05:22:18PM +, Ron Briscoe wrote: > > Could, would the team upgrade the NetSurf !System upgrader please? The versions included are the versions we depend on. The !System merge tool won't overwrite newer versions with older. B.
Re: Weird link
On Thu, Oct 22, 2015 at 01:14:04PM +0100, Richard Torrens (lists) wrote: > http://www.oldstylesiamese.co.uk/kitten-list/adverts/male-female-old-style-kittens/ > > There is a link in the html which goes: > > Website > href="http://www.arpegesiamesecats.co.uk";>www.arpegesiamesecats.co.uk This domain does not even exist. B.
Re: #2988 hanging
On Tue, Oct 13, 2015 at 04:39:32AM -0800, Dave Higton wrote: > On Tue, 13 Oct 2015 13:09:27 +0100 George Greenfield wrote: > > > I must confess to having used Ctrl-Break from time to time, when faced > > with a frozen desktop, without evident ill-effects. Is it considered > > worse than powering-off at the switch, which is the only other option > > AFAIK? > > Probably, although I'm not really qualified to answer. You can assume the following simplification: Alt-Break allows you to force a stuck task to call Wimp_CloseDown, essentially causing it to quit. Note that this forced quit isn't clean: any open files will remain so, and any memory allocated via dynamic areas or the RMA will be leaked. Ctrl-Break essentially calls OS_Reset, which causes a software reset that for all but the very most obscure situations is functionally identical to pressing the hardware reset button. The danger with Ctrl-Break and the hardware reset button is that FileCore might be trying to do something at the time (or be struck trying), leaving the file system in an unclean state, requiring the use of something like Disknight to fix. Summary: Always try Alt-Break first (and the a clean shutdown straight after), and only Ctrl-Break/Hardware reset button as a last resort. B.
Re:
On Fri, Aug 28, 2015 at 04:52:04AM -0800, Dave Higton wrote: > On Fri, 28 Aug 2015 12:50:00 +0100 Richard Torrens wrote: > > > Doing a list of genealogical notes, I wanted to space them out so there > > was a blank line between each record. So I replaced each by . > > > > This does not work in Netsurf - but does in Firefox and Chrome > > > > Is this a bug? > > > > The page in question is > > www.torrens.org.uk/Genealogy/Torrens/DataScot/sasines.html > > opens a paragraph. You need to close it again afterwards with . > > Paragraphs are rendered with blank lines between them. Alternatively, add padding using CSS or use . B.
Re: Fail 2890 json
On Fri, Aug 14, 2015 at 09:31:36AM +0100, David Pitt wrote: > Daniel Silverstone, on 14 Aug, wrote: > > > On Thu, Aug 13, 2015 at 22:56:41 +0100, Gerald Dodson wrote: > > > Has any one else had this version fail? > > > > It is distinctly likely that current CI builds will cause issues. We're > > undergoing a *huge* JS transition and forcing JS on while we do it. If > > you encounter issues, turn JS off after loading your browser (note it will > > turn on again next time the browser loads). > > Is this a RISC OS thing? I have just built nsgtk and that looks OK (briefly > tested caveat etc..) with JavaScript enabled. One of the issues we are currently trying to track down appears to be a bug in UnixLib, the library that provides a POSIX API for RISC OS. Your real UNIX system's C library probably doesn't have the same bug so the issue would not present. B.
Re: Do Netsurf's errors get logged? (Syslog, Wimplog)
On Wed, Aug 12, 2015 at 12:53:57PM +0100, cj wrote: > > *help wimplog says: "The Wimplog module logs errors sent to > > Wimp_ReportError (ie, shown as error boxes on the screen) to a file > > using !Syslog." > > Wimplog intercepts the standard Wimp_ReportError wimp error messages. > If Netsurf uses its own internal error reporting, then Wimplog will > not see it. Correct. Wimp_ReportError is evil and NetSurf does not use it unless the event is cataclysmic. B.
Re: Unable to locate logical address space.
On Thu, Jun 25, 2015 at 03:02:41PM +0100, Peter Young wrote: > NetSurf development #2811. > > Since I got the ARMX6 I occasionally and unpredictably get the message > "Unable to locate logical address space" when I try to run NetSurf. > Today this happened shortly after a restart, and up till now I have > only used NetFetch and Messenger Pro. FWIW the Next slot is 1024K. > According to Organizer's icon bar organ I have 1.825M free memory, so > it can't be that NetSurf is running out of memory. Can anyone comment > on this, please, and should I submit a bug report? I do not believe this error is from NetSurf, or UnixLib, or the GCC stubs. At least, a grep in the source code suggests not. I would suggest following up with ARMX6's technical support. B.
Re: More disc cache improvements
On Tue, May 05, 2015 at 11:48:46AM +0100, cj wrote: > In article <20150505103028.gg19...@kyllikki.org>, >Vincent Sanders wrote: > > This change should be beneficial to RISC OS users as filecore is > > (apparently) dreadful at this kind of usage pattern. > > I wonder if this is because RISC OS files are 'defragmented' all the > time - I think I am correct in saying that when a file becomes too > big for the contiguous space it currently occupies, it is copied into > a larger block of free space. This means that as the file grows, it > may be regularly being copied somewhere else on the disc, and with > files that may be hundreds of MB in size, this will be slow - very > much so when the drive is an SD card. Ish. FileCore has supported fragmented files for a long time (ISTR E implements this). It was only earlier versions that didn't support fragmented files and thus had to rearrange on file growth. B.
Re: Hot List
On Thu, Apr 30, 2015 at 04:33:28PM +0100, Brian wrote: > In article <54b9784cbfli...@torrens.org.uk>, Richard Torrens (lists) > wrote: > > Could you change the saving of the hot list? > > > At present it only saves when NS is quit - a pain if you have just made > > changes and NS quits1 > > > It either needs a Menu Item: Save Hot List or it should save > > automatically when changed. > > Reading this thread with some dismay. However, would it be practicable to > sponsor support of the the RISC OS side of NetSurf under the ROOL > sponsorship scheme? I should be glad to contribute if such a scheme were > to be implemented. I don't think offering a small amount of cash is going to make much difference if there is no interested, skilled, or bored enough developer to do the work. B.
Re: Hot List
On Thu, Apr 30, 2015 at 10:27:06AM +0100, John Rickman Iyonix wrote: > someone? wrote: > > > It takes less than an hour to install the OS and get a NetSurf build > > going. > > Is it feasible to do this using an RasPi? (Pi 2 Raspian) Yes. However you will not be able to use our pre-built cross-compiling toolchain, which is built for AMD64 Linux. Which means you'll have to build your own. And on an ARM that's going to take a while. Sadly our build process is optimised to use the fastest cheapest hardware available; we've never considered that anybody would be developing from anything else. B.
Re: Hot List
On Wed, Apr 29, 2015 at 10:36:35PM +0100, Nick Roberts wrote: > > I spent some time writing a HOWTO document outlining how to set up a > > cross-compiling environment using a virtual machine (so it can be done > > on Windows machines as well as on Linux). Sadly few people seem to have > > taken advantage of it. > > Could you point me at it? I may (emphasize _may_, not _will_) be able > to take on some new development work in a few weeks. > > I've hesitated to volunteer for NetSurf in the past because Real Life > has been getting in the way of my hobby developments for the last few > years, but that may be going to change. If you already have a Linux system to hand, then this documents everything you need to know: http://git.netsurf-browser.org/netsurf.git/tree/Docs/QUICK-START B.
Re: Hot List
On Wed, Apr 29, 2015 at 09:02:11PM +0100, Andrew Pinder wrote: > In message <54bbf0d883ch...@chris-johnson.org.uk> > on 29 Apr 2015 cj wrote: > > > One question I have mused over for some time is what is the fraction > > of NetSurf users that are RISC OS users and what fraction use other > > platforms. > > Data on the number of downloads per month for each platform would give > some kind of answer to that question. A very poor answer: Linux and the BSDs for example often ship their own copies: we don't provide binaries for those platforms on our website. > BTW, I was at the Wakefield Show and was expecting to donate to the > NetSurf project there as in previous years. Then I realised that none > of the developers were there :-( Sadly it was not economically viable. I seem to recall that last year we only just managed to break even (given stall prices and travel expenses), as well as the fact that we had basically nothing new to show for RISC OS users beyond core improvements, all of which everybody who uses NetSurf on RISC OS already knew about via this list and release announcements. B.
Re: Hot List
On Wed, Apr 29, 2015 at 02:42:43PM +0100, cj wrote: > In article <54bbe96175li...@torrens.org.uk>, >Richard Torrens (lists) wrote: > > write a request to comp.sys.acorn.programmer to see if you can > > interest anyone there in joining the team. There may well be > > someone who is capable, could be interested, but does not subscribe > > to this list. > > There has been talk in various forums about the problem. > > The first thing that gets stressed is that in order to compile > NetSurf (essential in order to test even the smallest change) you > need to be using GCC and the autobuilder under linux. That is an > immediate and very major obstacle to several people who have > expressed some interest. I spent some time writing a HOWTO document outlining how to set up a cross-compiling environment using a virtual machine (so it can be done on Windows machines as well as on Linux). Sadly few people seem to have taken advantage of it. It takes less than an hour to install the OS and get a NetSurf build going. B.
Re: Hot List
On Wed, Apr 29, 2015 at 10:52:19AM +0100, Harriet Bazley wrote: > I'm not sure it's very realistic to expect a random C coder out of the > kindness of his heart to download the source to a program of the complexity > of a web browser, work out how one specific piece of it currently functions, > and alter that part without breaking anything else... but I suppose that's > how open source software is intended to work... NetSurf is well compartmentalised. I suspect anybody with any experience of RISC OS C programming would find the task pretty simple. It's not as if you need to touch or look at the CSS selection code or box layout code. B.
Re: Which font does NetSurf use?
On Sun, Apr 05, 2015 at 09:10:41PM +0100, Gavin Wraith wrote: > When NetSurf renders the character pi (π) how does it > choose which font to use? If memory serves, it uses RUfl (RISC OS Unicode Font Library), which depending on if the host OS supports Unicode font rendering (ie, not ROL's OS) will either select the glyph in the current font if it has it, or will switch to a different codepage to render it. How it chooses the font is based on a heuristic involving font weight and style (sans/serif, bold, oblique, approximate size metrics, etc). If there are multiple possibilities, my bet is it picks the first it finds. I don't think anybody has touched this code in a long time, but here it is: http://git.netsurf-browser.org/librufl.git/tree/src B.
Re: Updated disc cache
On Fri, Apr 03, 2015 at 03:13:17PM +0100, David Pitt wrote: > I also think NetSurf's performance is severely hampered by the slow > processors available to RISC OS. No, the CPUs are perfectly adequately fast. A Raspberry Pi can do many megabytes a second when running Linux. RISC OS's IO layer and file system stack is simply too old-fashioned and simplistic (even by 1980s standards). B.
Re: Updated disc cache
On Fri, Apr 03, 2015 at 01:30:14PM +0100, nets...@avisoft.f9.co.uk wrote: > In article <54ae82a927ch...@chris-johnson.org.uk>, >cj wrote: > > I can see why RISC OS gets indigestion with the cache. Have just > > deleted the cache on the Iyonix, and there were over 21,000 > > directories and over 19,000 files. > > Yes, there seem to be lots of directories - many empty. The non-empty > ones often only contain directories. And the ones containing files rarely > have more than 1 file in! > > Strikes me that there are too many directory levels. Insetad of having > !Cache.Caches.Default.NetSurf.m.A.3.G.A3GKETD > it may be better to have > !Cache.Caches.Default.NetSurf.mA3G.A3GKETD > > I suspect much of the delay for small files is due to checking, creating, > and traversing directories! The depth was chosen so it would work on poor-quality file systems that only allow a handful of entries in a directory, such as FileCore :) B.
Re: JavaScript
On Sun, Mar 29, 2015 at 05:53:19PM +0200, David Feugey wrote: > Sorry. The mailing list seems not to like messages from Gmail. Let's try > again. It seemed to arrive just fine... B.
Re: AJAX
On Thu, Mar 19, 2015 at 10:10:44AM +, Brian wrote: > I receive a monthly newsletter. The latest cannot be read. NS says 'This > browser is not AJAX enabled'. What's that all about, please? Is there a > way round it so that the newsletter can be read? AJAX -> Asynchronous Javascript and XML. It is a term used to describe applications that run in the browser and use XML to communicate to the server. Why anybody would consider using this for a newsletter is beyond me. Regardless, anything AJAX won't work in NetSurf yet. > Should I complain to the sender, perhaps? They are a large company. You could always ask them to send a news letter as, you know, text. In an email. Companies seem to forget you can do that. B.
Re: Fwd: netsurf and gtk+ build error
On Thu, Mar 19, 2015 at 12:30:17PM +0400, ny wrote: > Vincent, I don't understand what the option "Q=", but I ran the > command as you pointed out, and that's what I got: > > CFLAGS="-Wno-error" make Q= > > I put the output file on the hosting, as attaching files is > prohibited: http://dropmefiles.com/O598L That link doesn't work for me. Just put the text in the body of your email. Or use a pastebin. B.
Re: #2654
On Wed, Mar 18, 2015 at 12:15:07PM +, David Pitt wrote: > > In RISC OS, if a directory is copied, its original datestamp is loat, > > however if a !Runimage is present then the application directory shows the > > datestamp of the !Runimage, thus preserving the datestamp of the original > > full-save. > > Is this with VRPC? A quick experiment shows that this is not the case in RPCEmu in a RAM Disc (ie, FileCore), but is the case with HostFS. (RISC OS 5.) B.
Re: #2654
On Wed, Mar 18, 2015 at 10:25:41AM +, David Pitt wrote: > The underlying issue is how the host treats directory moves, real RISC OS or > for exampler SDFS on the Pi, maintains the original date stamp on a move. > > HostFS on VRPC Mac, RPCEmu Mac and VRPC Windows 7 on a directory move assign > the current time at the move. Then this does sound like a bug or omission in HostFS, not NetSurf. B.
Re: #2654
On Wed, Mar 18, 2015 at 09:15:30AM +, David Pitt wrote: > > And it works. Tested on Mac VRPC which does not preserve the directory > timestamp on a move. Ah, is this a bug in VRPC? This would explain a lot. (I've been experimenting in RPCEmu trying to work out wtf this bug is about; directory time stamps work as I'd expect them to there, regardless of any !RunImage or lack of one.) B.
Re: Javascript version
On Thu, Mar 12, 2015 at 08:22:48PM +, Peter Slegg wrote: > >> I guess this would be relevant only for the "json" version of Netsurf > >> (or would simply do nothing in the "jsoff" version). > > > >There is no jsoff version anymore. > > > >B. > > There isn't a Javascript version for the Atari build yet. > It would be nice to have the option but I don't know what the > performance would be like. I was talking about the RISC OS autobuilds and releases. I can't see a JavaScript release having less performance, just a (much) larger binary. B.
Re: wee feature request for 3.3 -- a Javascript on/off button
On Wed, Mar 11, 2015 at 03:42:37PM +, Jim Nagel wrote: > Rob Kendrick wrote on 10 Mar: > > > On Tue, Mar 10, 2015 at 01:06:49PM +, Jim Nagel wrote: > >> It would be handy if the Netsurf buttonbar had an item that would: > >> (a) toggle Javascript on or off > >> (b) indicate whether it's on or off. > > > I don't see any way that this is going into 3.3; we are currently > > prepping a release, and this is not a small feature to implement at all. > > For now, then, how about just a hotkey that would open the same window > that you get from iconbar "Choices > Content"? Maybe Ctrl-J or F10. No. It is too late for 3.3. Please file a wishlist/feature request on the bug tracker. > I guess this would be relevant only for the "json" version of Netsurf > (or would simply do nothing in the "jsoff" version). There is no jsoff version anymore. B.
Re: wee feature request for 3.3 -- a Javascript on/off button
On Tue, Mar 10, 2015 at 01:06:49PM +, Jim Nagel wrote: > It would be handy if the Netsurf buttonbar had an item that would: > (a) toggle Javascript on or off > (b) indicate whether it's on or off. I don't see any way that this is going into 3.3; we are currently prepping a release, and this is not a small feature to implement at all. Please feel free however to add it as a wishlist item to the bug tracker. B.
Re: A curious case of attachment filetyping]]
On Wed, Dec 03, 2014 at 01:30:58PM -, Learning Partners wrote: > > > On Wed, 3 Dec 2014 Dave Higton wrote: > > snip > > > > I'm confused. "Attachments" normally refers to email, which would > > appear to have nothing to do with Netsurf. > > > > Can you please give more detail about what you're doing and how > > you're doing it? > > > I am using Netsurf to access Orpheus webmail for emails. Attaching files > typed as Word and PDF last Friday was fine, they went as applications. What do you mean, "went as applications" ? One hopes they went as files. Can Orpheus not tell you what file type is being uploaded? B.
Re: !Cache management
On Mon, Nov 24, 2014 at 01:15:26PM +, Jim Nagel wrote:
> A typical item within this folder is filetyped as text but contains
> gobbledygook.
"Text" is the default file type for files created using UnixLib. Given
these files are not meant to be consumed by humans, I don't see a
problem there.
(The actual gobbledydook is the cached data and headers.)
> What is supposed to control expiry? Can't find anything about expiry
> within the !Cache application itself.
!Cache is not NetSurf-specific. The control settings are in each
application that uses it.
>Ah, Netsurf choices--cache. Disc cache size is set at 1024M (which
> I presume is meant as a max), and expiry at 28 days. Those figures
> seem to be defaults, because I have never altered them.
>
> Don't recall any discussion or explanation when Netsurf adopted !Cache
> -- when was that? Maybe it happened while I was away and I missed it.
It happened when Vince implemented disc caching :) This was some time
ago, perhaps even before our last stable release. If that's the case,
the last release announcement would have included a mention. If it
isn't, then it'll be mentioned via version control for our test builds.
> !Cache, according to its helpfile, was written by Adam Richardson
> ("Snowstone") in 2007. The current version 1.13 on his website has a
> runfile dated 2007-june-14. The runfile of my active copy in Boot
> Resources (which presumably came with a recent-ish version of Netsurf
> inside its usual boot-update file) is dated 2014-09-16, but !Sidediff
> shows it is identical to the 2007 runfile.
Time stamps on files are advisory, not gospel. :) The 2014-09-16 is
probably when you merged your first build of NetSurf that included it.
B.
Re: Server error on Genes Reunited (HTTP 503.0)
On Sun, Nov 23, 2014 at 03:35:12PM +, Andrew Pinder wrote: > Hi all > > I received an email from Genes Reunited containing a link. Clicking > on that gives > > HTTP Error 503.0 - Service Unavailable > > It lists as "Most likely causes" > An invalid identity in the application pool could cause this error. > The application pool is no longer running because of configuration > or reaching application failure limits. > The concurrent application request limit was reached. > > It also tells me that "This error occurs when the worker process was > unable to start." > > Accessing the Genes Reunited website from a PC seems to show no > problems so it is possible that NetSurf can't yet handle the link. It > it will be of help I can add it to the bug tracker. This error is from the server. NetSurf won't even have had access to any of the page's HTML or images or anything. The server is probably trying to be clever and is getting confused when something that doesn't announce itself to be Mozilla connects, or similar. Not at lot we can do. B.
Re: !Cache management
On Sun, Nov 16, 2014 at 09:49:06PM +, cj wrote: > In article <20141116145722.gc21...@platypus.pepperfish.net>, >Rob Kendrick wrote: > > Currently, if NetSurf crashes (or your computer does) without > > quitting cleanly, newly-created cached items will be "leaked"; ie > > they will not be stored in the index which is written out on exit. > > Ahhh - that could explain it. My machines normally run 24/7, and the > only time Netsurf is quit is when I move to the most recent version, > which may be a week or more between upgrades. I do get the occasional > crash of Netsurf, so it could be that a week or more of cache > contents is not being indexed. This does sound like it may be the cause then. A solution (ie, detection of an unclean exit and thus a garbage collection and index rebuild etc) is on the to do list. Sadly I think the only solution at the moment however is to empty the whole cache directory manually. B.
Re: RISC OS experimental build with new compiler
On Sun, Nov 16, 2014 at 06:10:17PM +, Fred Bambrough wrote: > In message <20141116101839.ga21...@platypus.pepperfish.net> > Rob Kendrick wrote: > > > Hi, > > > > I have built a copy of RISC OS NetSurf using a much newer version of the > > compiler (GCCSDK's prerelease based on 4.7). If people could give a few > > minutes to playing with it to give us some confidence of it working, > > that'd be great. (Also any comments on performance difference, better > > or worse, would be of interest.) > > Was playing nicely. Downloaded a couple of files from ROOL and checked out > a few web sites. Then it crashed on my closing the window. Would you like > the log? Where to if so? Emailing it to me directly is probably best, rather than opening a bug for something that doesn't really exist on the bug tracker. B.
Re: !Cache management
On Sun, Nov 16, 2014 at 02:33:57PM +, cj wrote: > I checked on the PandaBoard today and lo and behold, the cache > contained more than 8 GB of files. This machine was also configured > for a 1 GB max and 10 days expiry. I have now deleted and disabled > the cache on the PandaRO. > > On all my machines I use recent versions of RISC OS 5.21, and recent > development versions of Netsurf. > > Am I misunderstanding what the cache settings are? Currently, if NetSurf crashes (or your computer does) without quitting cleanly, newly-created cached items will be "leaked"; ie they will not be stored in the index which is written out on exit. We plan to have this situation detected on next start and automatically remove orphan items. At the moment, the only simple way to recover the space taken up by these lost cache items is to delete the contents of the NetSurf folder inside !Cache. B.
Re: RISC OS experimental build with new compiler
On Sun, Nov 16, 2014 at 10:57:43AM +, David Pitt wrote: > In message <20141116101839.ga21...@platypus.pepperfish.net> > Rob Kendrick wrote: > > > Hi, > > > I have built a copy of RISC OS NetSurf using a much newer version of the > > compiler (GCCSDK's prerelease based on 4.7). If people could give a few > > minutes to playing with it to give us some confidence of it working, > > that'd be great. (Also any comments on performance difference, better > > or worse, would be of interest.) > > > http://www.rjek.com/netsurf-gccsdk-4.7.zip > > It has had a few minutes on the Raspberry Pi, it works, it didn't > explode at all. In fact as far as I can see it is indistinguishable > from #2335. Check about:testament to make sure it really is my build you're using :) B.
RISC OS experimental build with new compiler
Hi, I have built a copy of RISC OS NetSurf using a much newer version of the compiler (GCCSDK's prerelease based on 4.7). If people could give a few minutes to playing with it to give us some confidence of it working, that'd be great. (Also any comments on performance difference, better or worse, would be of interest.) http://www.rjek.com/netsurf-gccsdk-4.7.zip B.
Server downtime
Hi, Sadly one of our servers suffered a bit of an oopsy yesterday evening at around 1745 (see URL below), which required us to go to one data centre to pick up a spare server and then go to another to install it. It should all be working now (as of around 2130). For NetSurf, the machine handled mailing lists, website, and git. Please let me know if there are any outstanding problems. http://www.rjek.com/octopus-heatsink.jpeg <-- unleaded solder holding down heat sink on the south bridge failed, probably due to a combination of age, stress, and thermal cycling. B.
Re: "Warning from Netsurf"
On Mon, Sep 29, 2014 at 12:27:16PM +0100, Jim Nagel wrote: > 1. By what process was this message generated? JavaScript. > 2. Is it indeed Netsurf that generates the message? If so, it's > ironic that I'm using a very recent build, Netsurf #2124. No, it's JavaScript. > 3. Why is the error message not recorded by Syslog? It's not a WIMP message box, it's a NetSurf one. B.
Re: Date stamps on resources such as !Cache
On Fri, Sep 26, 2014 at 06:45:25PM +0100, Andrew Pinder wrote: > Fair enough, and thanks for the link and explanation. However, I > still don't understand why the date stamps are updated only > intermittently, Because the resources come as part of our compiler toolchain, and we don't rebuild that very often. (It's rarely changed.) B.
Re: Date stamps on resources such as !Cache
On Fri, Sep 26, 2014 at 12:46:56PM +0100, Jim Nagel wrote: > Rob Kendrick wrote on 26 Sep: > > On Fri, Sep 26, 2014 at 07:20:56AM +0100, Andrew Pinder wrote: > >> I will be happy to install an updated version of !Cache, especially if > >> the problem I have happens to be solved. In the meantime I would > >> appreciate it if the datestamps could be preserved so it doesn't > >> appear to be updated. > > > Sadly this is an unavoidable artifact of the version control system we > > use. Each development build has with it a list of changes that can be > > viewed using the Jenkins interface, however. > > Please explain "Jenkins interface" and how to get to it. This is a > new term on me. Just me? (Mein Unwissendheit entschuldigen, bitte.) It is linked to off http://ci.netsurf-browser.org/ - beware that it is for advanced users only. These are development builds, after all. B.
Re: Date stamps on resources such as !Cache
On Fri, Sep 26, 2014 at 07:20:56AM +0100, Andrew Pinder wrote: > I will be happy to install an updated version of !Cache, especially if > the problem I have happens to be solved. In the meantime I would > appreciate it if the datestamps could be preserved so it doesn't > appear to be updated. Sadly this is an unavoidable artifact of the version control system we use. Each development build has with it a list of changes that can be viewed using the Jenkins interface, however. B.
Re: bug about SSL security certificate -- Mantis grrrr
On Thu, Sep 25, 2014 at 05:11:27PM +0100, Jim Nagel wrote: > Tony Moore wrote on 25 Sep: > > 'Reported in CI build #' doesn't make it clear that the '#' should be > > omitted from the answer. If the '#' is included, Mantis posts an error. > > Is "robust" the right word when I say I think software should be > robust enough to anticipate such variations in user input and quietly > take the bits it wants. No, that's not robustness, that's telepathy. No matter how many filters or functions or pattern matches you attempt (each with its own list of maintenance requirements and bugs), somebody on netsurf-user will still end up asking why it refuses "Net Surf Version Three small dot thing 2". B.
Re: bug about SSL security certificate -- Mantis grrrr
On Thu, Sep 25, 2014 at 03:01:14PM +0100, Jim Nagel wrote: > Rob Kendrick wrote on 25 Sep: > > Did you say "Netsurf 3.2" or "3.2" ? > > The former. Is Mantis not robust enough to cope with both? It's robust in the sense that it detected you gave it nonsense rather than inserting inconsistent data into the database. > But upon doublechecking my iconbar, I find I'm wrong. I'm using "3.1 > (Dev Cl #1718)" on this Iyonix. Must be one of my other machines > where I have 3.2 -- will check. When you see Dev CI (not CL) versions, you'll want to mention this number, too. > Info expected by Mantis in some fields was not obvious. > Is the best course then to leave such fields blank and just make sure > the freeform fields include all relevant info? If in doubt, provide more information. Yes, opening the submit button in a new window is probably the best way to go about using it. B.
Re: bug about SSL security certificate -- Mantis grrrr
On Thu, Sep 25, 2014 at 02:38:07PM +0100, Jim Nagel wrote: > Jim Nagel wrote on 25 Sep: > > Tried to log in to Mantis just now to report this bug, but no luck. > > So I'll post the problem here while it's fresh in mind. > > > Sorted out the Mantis login problem, thanks to Vince Sanders. > So I duly logged in and filled in a formal bug report; clicked Submit. > Mantis response (I paraphrase from memory): > Invalid report. Netsurf build number required. > I had said Netsurf 3.2. Did you say "Netsurf 3.2" or "3.2" ? > So I click "go back". > And there is a blank page expecting me to fill in ALL the info again. > Gr. This is a problem with NetSurf, not the bug tracker. Modern browsers annotate history information with form data. We have a plan to do that too. B.
Re: My wish: To have remote LogMeIn as is available in MS-DOS
On Sat, Sep 20, 2014 at 02:56:36PM +0100, Michael Bell wrote: > I am not a very knowledgeable computer user and I rely fairly heavily > on my computer guru, who was a RISCOS enthusiast but who has defected > to MS-Dos for the usual reasons. I like RISCOS for reasons both good > and bad; I dislike the way MS-Dos is open to "help" more plainly > called "parasitism" but maybe any big pot attracts blowflies. I do a > lot of writing and I like the way Select and Adjust allow you to mark > the start, and separately the finish of a section of text. But my guru > wants to shift me over to MS-Dos because it would make his life > easier. > > So, I would like a RISCOS equivalent of LogMeIn I doubt you mean MS-DOS. What does LogMeIn do? Is it somehow related to web browsers and NetSurf? B.
Re: dropzone on a web page
On Thu, Sep 11, 2014 at 03:05:00PM +0100, Jim Nagel wrote: > Jim Nagel wrote on 11 Sep: > ... But I'm curious about the technique involved in registering a > file's pathname by this "dropzone" method. Can Netsurf cope with this > in principle? >Does the "dropzone" technique in itself involve Javascript? The technique used on that site, yes. B.
Re: dropzone on a web page
On Thu, Sep 11, 2014 at 02:25:52PM +0100, Jim Nagel wrote: > Of course, it could be that the calculation is performed by Javascript > so Netsurf couldn't handle the job anyway. That it be. B.
Re: NetSurf 3.2 released
On Mon, Sep 01, 2014 at 09:21:51PM +0100, Tim Powys-Lybbe wrote: > On 1 Sep at 18:42, Michael Drake wrote: > > > The NetSurf developers are happy to announce NetSurf 3.2. This release > > contains many bug fixes and improvements. > > > > Regrettably it still does not work on VRPC on a Mac. For all web-sites > it gives a browser time-out message after 1 (or so) milliseconds. > What can I do now? Sadly, the options are limited to "buy a developer a Mac and a licence to VRPC and hope they can find the problem" and "use RPCEmu instead". B.
Re: Bug reporting
On Wed, Jul 09, 2014 at 03:48:11AM -0800, Dave Higton wrote: > OK, but that's not what it actually says. The options under > "Category" are: > > ABEND > Amiga-specific > Atari-specific > BeOS-specific > Cocoa-sepcific > Framebuffer-specific > GTK-specific > Javascript > Layout > RISC OS-specific > Win32-specific > [All Projects] General > > I, as a user of a single OS, cannot tell whether the problem is > specific to that OS, yet the field is required input from me. Then "General" is the one to go for, at a guess. > I don't want to make a big deal out of it, but it does look to > be a misnomer, and I was just looking for some clarification of > its significance to the developers. I've got it. > > AAMOI: what is the ABEND option about? Abnormal End. This is probably a hang-over from the default configuration of Mantis. B.
Re: BBC Programme Pages
On Tue, Jul 08, 2014 at 07:08:05PM +0100, John Williams wrote: > > I notice that, when I click on a programme page from the Radio 4 Schedule > Page, the main photo is pixellated, whilst other photos on the page are OK. > > Under Ubuntu in Firefox, the photos are rendered correctly. > > I suspect this may be something to do with NetSurf trying to render a > thumbnail instead of the main JPEG - perhaps an EXIF file with embedded > thumbnail? > > However, other browsers get it right, so why can't my preferred browser, > NetSurf under RISC OS, get it right? > > Look at, say, http://www.bbc.co.uk/programmes/b044mc1g - the page for the > programme I've just listened to - to see what I mean. > > Or is there some other explanation? First guess at looking at it in Firefox is that the image is streamed, ie, a low-quality small version is mentioned in the HTML to make page render prompt, and then JavaScript replaces it with a higher-quality one once the page render has completed. B.
Re: Disc cache worth it?
On Wed, Jun 25, 2014 at 01:44:01PM +0100, Steve (plusnet) wrote: > On 2014-06-25 13:36, Rob Kendrick wrote: > >(Getting into the technical distinctions between WIMP and non-WIMP > >tasks > >was something I was trying to avoid getting into on the user's > >mailing > >list :) > > Fair enough, but if you're going to say things like "the whole system > stops while the individual chunks are written" and "Under RISC OS, > file system writes stop /everthing/ until they complete" you can't > expect > me to just sit by and say nothing... Of course not :) But from the user's point of view, everything does stop :) B.
Re: Disc cache worth it?
On Wed, Jun 25, 2014 at 01:32:58PM +0100, Steve (plusnet) wrote: > Secondly, it's not true to say file system operations block everything > in RISC OS - some file systems (notably ADFS) implement "background > transfers" which allow much of RISC OS to continue to operate while the > underlying file operation runs (e.g. via DMA or some such). However, > many > file systems _don't_ implement background transfers - e.g. SDFS - so > these will hog the system for the duration of the low-level operation. But even with background transfers, the Wimp will typically not continue, no? My understanding is that OS_File SWIs that write data only return when the data is written, and thus the application cannot call Wimp_Poll to allow multitasking to continue - and even if you're using UnixLib threads this doesn't help you because threading does not continue over such SWIs. (Getting into the technical distinctions between WIMP and non-WIMP tasks was something I was trying to avoid getting into on the user's mailing list :) B.
