Default domain on VRPC
NetSurf is working well on VirtualRPC-SA except for one annoying aspect. DNS lookups fail unless the full server URL is used. So http://wiki/ says not found whilst http://wiki.ant.co.uk/ is fine. Now, VRPC of course uses the WindowsXP network stack and that has ant.co.uk defined as a default domain and other RISC OS browsers such as Fresco and Oregano both work OK in the same VRPC environment. So what is NetSurf missing? Presumably its getting the first DNS lookup failure and then giving up instead of trying to add the default domain? Bug, or feature request? Ta. Mike
Re: Multitasking (not)
In message <[EMAIL PROTECTED]> you wrote: > On 21 Oct 2007, Jess Hampshire <[EMAIL PROTECTED]> wrote: >> In message <[EMAIL PROTECTED]> >> Tony Moore <[EMAIL PROTECTED]> wrote: >> > On 21 Oct 2007, "David J. Ruck" <[EMAIL PROTECTED]> wrote: >> > > On 20 Oct 2007 Jess Hampshire <[EMAIL PROTECTED]> >> > > wrote: >> > > >> > > > I have noticed that sometimes Netsurf singletasks. >> > > >> > > Well thats really going to help developers track down the >> > > circumstances. >> > >> > It's been discussed before. Single-tasking occurs whenever objects >> > are fetched from a secondary url - two timers appear in the status >> > bar and NetSurf single-tasks until the second timer completes, or >> > times-out. I thought that a bug report had been filed, but I cannot >> > find it for now. >> >> I hadn't pinned down the circumstances, I hadn't thought of it as a >> bug as such, but if it is, it'll be nice when it's fixed. > > I've found the bug report now, filed on 1 June 2007 by Harriet Bazley > http://sourceforge.net/tracker/index.php?func=detail&aid=1729711&group > _id=51719&atid=464312 > and still assigned to 'nobody', unfortunately. > > Tony I don't know if its realated, or not, but yesterday I could not quit NetSurf. The iconbar menu quit failed and then also quit from the task manager failed. It wasn't busy so Alt-Break didn't do anything. Other than that it seemed to be working fine. This is the first time I've noticed the problem and I'm not sure how it got into this state (it had been running for a long time). Anyone else noticed a failure to quit? Looks like I should raise a bug report but not knowing how it got into this state won't be very helpful. Mike
Status display sometimes too small
When the pointer is moved over links the URI is displayed in the status icon but some links are incredibly long. Would it not be better to display the URI (and/or descriptive text) in a pop-up bubble? An alternative improvement might be gained by moving the horizontal scroll bar onto a separate row and allowing the URI/status icon to wrap text and expand to show it all (or give the user the ability to extend its size). Another alternative would be to put "Link" on the "NetSurf" menu so that when Menu button is clicked over a link it can be saved. Fresco does this. Mike
Re: netsurf buttonbar
In message <[EMAIL PROTECTED]> you wrote: > In a dim and distant universe <[EMAIL PROTECTED]>, >Richard Porter <[EMAIL PROTECTED]> enlightened us thusly: >> Just one minor point though - when you click on a link that throws up a >> save dialogue, the button bar is attached to the dialogue box. I don't >> think this was intended. > > Also, another minor point is that when you click on a button and NetSurf > /doesn't/ have the caret, odd things happen to other applications. I > suspect this is because the button bar is simply passing certain keypresses > onto the wimp (for any other application to act on). It might be better to > detect whether or not NetSurf has the caret or not (not sure how difficult > this is though, but should be possible by reading NetSurf's task handle and > then reading the active window and comparing window handles. I wrote (and use every day) an app that does what is required. Its called KlikMacro and what it does is to inject key presses (and/or mouse clicks) to the app with focus. It automatically switches the active macro menu for the app that has focus (which of course includes NetSurf). That way, the KlikMacro menu should always be correct when NS has focus. I do other tricks with apps that have multiple windows (e.g. NS) so that it can work out which window has the caret and do some kind of function based on the window title. Many other tricks can be done. KlikMacro is almost entirely written in Basic and has my entire Basic library (quite extensive) available to it so if you have any idea how the wimp works you can quickly program it to do exactly what you want (either in the macro for each button on the menus, or more complex things can be achieved by writing a new function in the Basic library and calling it from a button macro. KM also does much more than I've described here and it has routines for doing the sort of thing Paul describes. If anyone wants to try KlikMacro please email me and I'll send you a copy. [I must get around to putting it on a web site - one day] Mike
Lock-up starting NetSurf from Messenger link
If NetSurf is already running, selecting a link in an incoming message in Messenger works and NetSurf opens the web page. However, if NetSurf is not running (but Alias$URIOpen_... and Alias$URLOpen_... are set to run it), and I click on a link in a Messenger message there is a pause, then the border of a dialogue window is drawn (with no content) and the machine locks up. The mouse pointer still moves but Alt Break etc. can't break out of the lock-up. If I start NetSurf from other applications, or my own Basic functions that look at Alias$URLOpen_... and do a Wimp_StartTask, it works OK. So it seems that whatever Messenger is doing to fire up NetSurf causes this problem. I thought I would ask if any other NetSurf users have seen the problem (then I'll probably report it to RComp). I'm running NS r4259 with Messenger 5.12 on VirtualRPC-SE -- Mike Hobbs
microsoft-generated html - 'parsing document failed'
I frequently get sent html that has been generated by MS Word, Excel or Access and when NetSurf tries to load it the error "parsing document failed" occurs (using r5369 Sep 18 and earlier versions). OK, I know stuff generated by Microsoft software is likely to be a nightmare for anything other than IE but the issues are: * surely NS could make a better attempt to display something? * what does the error really mean? what exactly failed? * could NS have an option "re-load but ignore styles"? Looking at the html its fairly typical Microsoft html format - excessive use of CSS to define masses of irrelevant stuff like loads of fonts and styles that aren't even used and a sprinkling of unnecessary spans and other useless tags as well as MS-specific definitions inside html comments. NS could safely ignore almost all of this. The most recent document I've been sent is actually a fairly trivial table with 25 rows and 12 columns and an (oversize) email signature (its a bug report table from JIRA). It displays OK in Oregano2 (excepting some of the styles). Even Fresco v2.25 (circa 2000) displays it without the styles. I can't release the document as it stands but if anyone wants a copy I'll edit out the sensitive stuff. Its over 200K, of which about 90% is style definitions and references and over 50% of that is superfluous (typical MS bloat). Mike (speaking personally, not for ANT, but the SMTP server won't let me send mail as if from my demon account)
Re: NetSurf 2.0 + Royal Mail Address Finder
In message <5053ad248brh.li...@phone.coop> Russell Hafter - Lists wrote: [snip] > When entering info into the form here, while the cursor > moves along the space, no characters appear to be entered. [snip] I've seen this intermittently on many builds as far back as I can remember but after system restart its always been OK. Does anyone have any idea what is actually going on when this happens? It doesn't seem to happen with any other application so its got to be a NetSurf issue. Any ideas about how the problem might be isolated? I just wonder if its similar to a DrawPlus problem I've had for years where the menu widths expand to full screen width, then if I shut down various other applications it will sometimes recover. It would suggest applications trampling on memory that doesn't belong to them. Could it be the same with this NetSurf issue? -- Mike Hobbs
NS2.1 Unicode font library could not be initialized
I just tried 2.1 and can't get past the first hurdle... "Unicode font library could not be initialized..." I'm using it on Virtual RPC-SA (RO 4.02) which has been running various versions of NetSurf successfully up to 2.0 and I just rolled back to 2.0 and its fine. The only resource/module that was updated was Tinct and the new version is working fine with NS 2.0. I've tried both new and old !Unicode - no difference. -- Mike Hobbs
Re: Speed of loading NetSurf
In message <28366c6250.roger...@rogerarm.freeuk.com> Roger wrote: > Loading/Running NetSurf (on an Iyonix at least) seems to take 30 > seconds whereas running Oregano2 takes less that 3 seconds. [I am not > including the time that NetSurf, when first run, uses looking at all > the fonts]. [snip] Actually, I would like to suggest that the time taken for font scanning is now an issue since NS 2.1. It now takes several minutes to scan fonts on my home machine. Yes, I know its only on the first time NS is run, but I can't help thinking that this is all wasted time. After all, NS isn't going use all these fonts. Its only likely to need the standard set of ROM fonts. Maybe font scanning could be filtered? And also, once the fonts have been scanned where is the data cached? Is it wasting space somewhere by retaining font data for fonts that will probably never be used? Regarding non-font-related load time I don't seem to have a problem on Virtual RPC. Its certainly not taking 30 seconds. Mike
Re: NS2.1 Unicode font library could not be initialized
In message <1246791238.32517.99.ca...@duiker> you wrote: > On Sat, 2009-07-04 at 13:57 +0100, Chris Manning wrote: >> In message <4d8668feccea4ec15c9a84c30e157ad0de1f2...@localhost> >> Mike Hobbs wrote: >> [snipped] > Please email me the full log file and your !Fonts directory, zipped. > > > John. Thanks to John-Mark my problem was fixed (for a while) by making sure I had the latest FontManager. However, I unexpectedly had the problem return and the reason was that I had changed some fonts by synchronizing between my laptop and desktop machines. I mention this here because, as I think others have also pointed out, the mechanism for font scanning probably needs reviewing. It seems that seemingly minor changes to a machine's font library can at best cause NetSurf to spend ages scanning fonts before it starts, or at worst it can crash because the RUfl_cache says it has fonts which have now gone. And all this trouble seems to be over fonts which NetSurf is never likely to use. Sometimes I switch the contents of my font library to avoid having an excessive number of fonts but to load some special fonts for a graphic design job. This can play havoc with NetSurf, yet I'm sure many people do this kind of font swapping. Mike
align=center appears to be broken
I've just noticed that our own web page no longer centers. The page is written by TechWriter and uses "ALIGN=CENTER" in several places (e.g. on H1 and DIV elements). I'm not sure which build first broke this but it was working in the 2.1 release and wasn't on r8933 and its still broken as of r9045. -- Mike Hobbs
Re: Google Summer of Code Roundup
In message <20090915140951.35722...@trite.i.flarn.net.i.flarn.net> Rob Kendrick wrote: > On Tue, 15 Sep 2009 14:06:02 +0100 > Andrew Weston wrote: >> On Tue, Sep 15, 2009 at 12:57 PM, John-Mark Bell >> wrote: >> >>> Indirectly. The RISC OS frontend's lack of maintenance is actively >>> blocking several of these changes being merged. The longer this >>> situation persists, the more likely I am to unilaterally remove >>> support for the RISC OS frontend, as it is holding back >>> improvements for other platforms (and NetSurf in general). >>> has there been much response yet to your call for help? >>> >>> Several people have expressed interest. None of this has translated >>> into code contributions, however. >> >> Have they applied conditions or are they not able to do it? > There are more than two possible causes, and neither of these are it. > B. Well, I don't know how many people are using NetSurf on each platform but I ONLY use NetSurf on RISC OS because its the most useful and as far as I know the only maintained browser for RISC OS. Withdrawing support for RISC OS, to me, is crazy. There are already far more capable browsers for other platforms so I never use NetSurf on anything other than RISC OS. I can't see my stance changing at least until there is javascript support in NetSurf. I'd love to offer support but I don't have appropriate skills or the time. Mike
Re: Google Summer of Code Roundup
In message <800994.16379...@web111312.mail.gq1.yahoo.com>
WP Blatchley wrote:
> Hi all,
> Well despite Rob Kenricks's probably highly accurate description of
> how development for RISC OS is ("hateful"!), I am interested, and have
> expressed that interest on the developers' mailing list. And I'm not
> the only one.
> I've had quick responses to the (so far few) questions I've asked, and
> I believe that the existing NetSurf developers will continue to bend
> over backwards to help anyone with the time and inclination to
> maintain the RISC OS front end. They just haven't got the time or the
> inclination to do it themselves. I'm sure everyone here can understand
> that.
> Right now my time is very limited, just like most other people's.
> However, I am looking over the sources in the little free time I do
> have, and I'm starting to get a feel for things. I fully expect to be
> able to make a positive contribution to the project in time, but I
> can't just pick up such a large software project and start committing
> code without a pretty serious investment of time. So patience is
> required before I, or anyone else I should think, can get up to speed
> with the project and fill in the gaps. I would therefore understand if
> the core developers wanted to drop RISC OS support and merge their
> improvements for now to prevent this stall in development affecting
> the other platforms for any longer than it has already.
> That doesn't mean I or somebody else can't be working on getting the
> RISC OS front end back into shape at a later date, does it? It would
> just mean RISC OS autobuilds had to stop for a period, I think.
> Correct me if I'm wrong!
> WPB
I'm in exactly the same position as WPB. In time I hope to be able
to contribute but currently I'm not "a developer", I'm a software
test engineer and although I write lots of code I've only ever
"developed" RISC OS applications using Basic (because its excellent
for achieving quick results in doing real work). I'll be trying
to improve my C skills (which hopefully will also help my day job).
As with seemingly everyone else, time is in short supply, but
hopefully between us we can do enough maintenance to keep the RISC
OS front end going.
I might be sticking my neck out here but I have a feeling that many
of the regular users of RISC OS (and thus frequent users of NetSurf)
are actually old phogies (like me!) and perhaps feel they are past
learning new skills. In fact somebody said that earlier in the thread.
In my experience the 'young whippersnappers' move on to other
platforms where there is masses of new stuff for them to get their
teeth into and show off their skills. I use RISC OS all day, every
day, to accomplish real work and its the easiest to use and most
responsive and intuitive and time-efficient OS I use. Frankly,
Windows and all the flavours of unix-like OSs keep annoying me by
their over-feature-rich, cosmetically snazzy, but bloody awkward
UIs. All I want is a reasonable browser on RISC OS (and I've been
waiting 20 years for it!) NetSurf is good as far as it goes, but
there is lots to do. As has happened with so much other RISC OS
software, it looks like if its going to progress I'll have to do it
myself (together with a few other willing bods).
Now, ask yourself, how much code could have been written if
contributors had not been raking over the coals in this thread?
Shall we get on with the job now? For my part, I'm just looking
through the sources and over the weekend will attempt to get it
building. If that goes OK I'll look at something on the todo list
that I feel I can do and take it from there.
Finally, I would like to say how grateful I am to the NetSurf
developers for getting it thus far. I'm disappointed that they don't
feel inclined to maintain the RISC OS front end because I'm sure they
could do a better job than me, but, as the creators of NetSurf, they
are entitled to do what they like.
Mike
!Run always open
This has been bugging me for a while but its very minor... !NetSurf.!Run is always open whilst running NS. This means my backup (to propagate new versions to another machine via USB memory) complains. I know the reason is the NS-specific font removal obey commands for when NS exits. I just wondered if there might be a better way of doing this that doesn't leave the file open? I currently edit !Run every time I update NS. -- Mike Hobbs
Re: !Run always open
In message <62c4bd3551.old_coas...@old_coaster.yahoo.co.uk> Tony Moore
wrote:
> On 12 Jul 2010, Mike Hobbs wrote:
>> This has been bugging me for a while but its very minor...
>>
>> !NetSurf.!Run is always open whilst running NS. This means my backup
>> (to propagate new versions to another machine via USB memory)
>> complains.
> Maybe I misunderstand the problem, but why not quit NetSurf before you
> try to backup?
To answer this and other comments and suggestions:
1 The 'backup' tool I use is DirSync and its extremely useful
but it doesn't seem to be able to copy files that are open
(even though RISC OS does allow multiple opens for read)
2 I don't have the luxury of changing the way DirSync works
and the author has stopped maintaining it. Although I
offered to continue to to maintain it, that suggestion fell
on deaf ears.
3 I could close NetSurf before syncing directories but that is
little different from having to dismiss an error box - it
needs manual intervention. NS is running almost 24x7 too.
4 I can (and do) edit the !Run file to allow it to close after
NS starts but its a nuisance when updating NS regularly.
5 Integrity of the backup is not an issue because its just a
NS file that cab be easily recovered from elsewhere, but,
whilst DirSync can ignore directories it can't be told to
ignore files. See 2 above.
6 The suggestion to get the NS binary to remove the fonts on
exit would be most convenient, or simply don't bother to
remove them at all. This is what I'd like implemented.
--
Mike Hobbs
Re: NetSurf RISC OS archive files
In message <519ca0651ft...@netsurf-browser.org> you wrote: > In article <519c9f4858joh...@ukgateway.net>, >John Williams wrote: >> I have noticed recently that the NetSurf download for RISC OS varies >> between zip (&a91) and archive (&ddc) between versions. >> So, is the variation of archive type just an accident of convenience for >> members of the team, and are any other users affected practically by this >> inconsistency? > The NetSurf archives are always zip files. Sounds like something on your > system gives them different RISC OS filetypes when they're downloaded. Do > you use different apps to download NetSurf? > Archive (&ddc) is always used for .zip files here. &a91 was introduced some time ago, presumably to distinguish ZIP from other types of archive file. SparkFS knows how to handle &a91 and &ddc (which may be various formats). It defaults to using &a91 when creating zip files. I find the use of &a91 very annoying because depending on where the files are sent/used the filetype may or may not be recognised and when transferring to HostFS or NFS file systems ",a91" may be appended, making them unrecognizable to applications in Linux or Windows. Depending on whether SparkFS has been Filer_Booted the filetype "Zip" may be either &ddc or &a91 so I guess NetSurf is using the textual filetype rather than the hex when saving zip downloads. I don't have a foolproof solution although at various times I've tried to get SparkFS to default to the old &ddc, but it won't cooperate. Usually, I just manually change the &a91 to &ddc before transferring to a foreign file system or sending via email. In my opinion &a91 was a bad idea. It would be nice to know how to set mimemap and extensions translations so that all archive types are handled seamlessly but I have not succeeded in doing this. -- Mike Hobbs
default form submission by keypress Return
When page content includes something like this: ... ... and the user presses Return, every other browser will send the post request with the various arguments from the input elements (i.e. they default to doing the submit as if the user clicked the Go button), but NS seems to default to actioning the Clear button. If I change the Clear button to then pressing Return sends the post request as if the Go button was clicked (same as other browsers). However, NS gives a "Warning: button can not be activated" on Clear. So it seems that NS can't have forms with more than one submit button. Lack of javascript is a drawback we have to cope with, but this issue also means cgi scripts can't use multiple submit buttons and allow for the usual default action of pressing Return submitting the form with all the arguments. Obviously there are various workarounds but I just wondered if there was any chance that NS could default to actioning the first submit button rather than the last. -- Mike Hobbs
cookie association with specific windows?
In recent NS builds (past several months) I've noticed a change in behaviour when logging in to our company bug tracking system. At one time I could go through the Login page and from then on I would be able to access the whole bug tracking database through its web UI. At some point that changed so that now, a) if I go through the normal Login process it fails to accept my credentials; b) I have found that by going through an alternative URL that throws up the Login page it results in success; c) even though I've logged in, when I open a new NS window by clicking on a link that should open a bug report (e.g. http://jira/browse/XXX-nnn), it tells me I'm not logged in; d) if I enter the same URL as in c) above into the window that was opened as per b) above, then it displays the page without any trouble. On the face of it, this looks like the cookie-stored login status is associated only with the window that was used for successful login. Is this expected? Do other people see similar behaviour? Should I file a bug report? I'm currently using the 2.7 release on VRPC-SA but this affects earlier releases and development builds. -- Mike Hobbs
Re: cookie association with specific windows?
> On Thu, May 19, 2011 9:41 am, Mike Hobbs wrote: [snip long description of login difficulties] Whatever was causing this appears to be fixed in r12436. Well done and thank you! I was inspired to check out the latest build when I saw Jim Nagel's post "login problem with Netsurf 2.7". Mike
No more Google for us
Well, thats it then... Google requires javascript !! -- Mike Hobbs
Re: NetSurf on Virtual Acorn
In message <533d3d5a8bt...@netsurf-browser.org> Michael Drake wrote: > In article , >Alan & Sally wrote: >> NetSurf 2.9 runs nicely on this set up but I have had no luck with any >> of the later development versions. I have just downloaded 3.0 Dev #833 >> with the same result. > It seems that since NetSurf 2.9 DNS doesn't work when running on > VirtualRPC. It works on real hardware and, I believe, on RPCEmu (can > anyone confirm?). I use recent dev builds on VRPC-SA and VRPC-SE and DNS is fine. The only issue I have is that when I connect to the company VPN (Cisco software) the DNS breaks and nothing will resolve. As soon as I disconnect the VPN it works again. When NS DNS lookups fail, the Windows7 apps like Firefox seem to do DNS lookups OK. Since the resolver addresses are as far as I can see identical in both Windows side and RISC OS I don't know how it breaks. Other networking such as connection to mail server and NFS mounts all continue working and its only DNS that seems to be broken. > None of the NetSurf developers have VirtualRPC so we aren't in a position > to investigate this. Perhaps you could contact the VirtualRPC developers, > if it's still being maintained? If you want to try anything, pass it my way. I'm using VRPC and NS all day every day at home and at work. Mike
NetSurf wont start with certain (required) fonts
I just upgraded MessengerPro (R-Comp) and to be able to display unicode characters certain fonts are required (e.g. Libertine, DeJaVu). However, when these fonts are installed Netsurf fails to start up. To be more precise, its fails when Libertine or DeJaVuSans fonts are installed but its OK with DeJaVu (which only has Serif style). The other fonts have quite a few styles. Could this be the problem? Is there a workaround for this? -- Mike Hobbs ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
Re: NetSurf wont start with certain (required) fonts
In message Michael Drake wrote: > On 11/10/2022 02:53, Mike Hobbs wrote: >> I just upgraded MessengerPro (R-Comp) and to be able to display >> unicode characters certain fonts are required (e.g. Libertine, >> DeJaVu). However, when these fonts are installed Netsurf fails to >> start up. >> >> Is there a workaround for this? > What version of NetSurf are you using? > Please try a recent development build (e.g. 5367): > https://ci.netsurf-browser.org/builds/riscos/?C=M;O=D Aha! That did the trick. I was using 3.10 release. Now 3.11 (5367) Many thanks. -- Mike Hobbs ___ netsurf-users mailing list -- netsurf-users@netsurf-browser.org To unsubscribe send an email to netsurf-users-le...@netsurf-browser.org
