Dev Cl #1717
In article <53d83f69e8bbai...@argonet.co.uk>, Brian wrote: > In article <53d83d6e6ebrian.jord...@btinternet.com>, Brian Jordan > wrote: > > In article <53d83be5c4bbai...@argonet.co.uk>, Brian > > wrote: > > [Snip] > > > Who'd have thought it?? Certainly not me. I'm intuitive but not that > > > intuitive. I assumed that # was as per American practice and is part > > > of any/all part codes. > > I think the logic is that the "#" is in the caption alongside the box > > in much the same way as we often see "£" alongside boxes when a sum of > > money needs to be entered. Several of us seem to have fallen foul of > > this and had to rewrite our reports, I reccommend writing reports in a > > text editor and saving them as insurance and not putting the "#" in > > the version number. > OK, Brian, but I'm not into guessology. # is not a British convention > anyway so far as I am aware, for some 80ish years, it's a hindrance! Why > put it there when it serves no useful purpose, I ask? Perhaps if anyone is interested to see just how confusing the use of the # is have a look at Wikipedia - Number sign. In American practice I've often seen it added onto British nomenclature to, possibly, mean item number, part number, mark number, or whatever, without adding any useful intelligent information. Having handled many 100,000's of British manufacturing and development drawings, spares and component data sheets, as well as specifications, I don't recall ever seeing the use of the # sign. If I recall correctly, parts columns were invariably headed Part No. Without defining what # means it follows that it is meaningless. Who was it who said, 'if you would converse with me, you must first define your terms'?
Re: Cursor doesn' follow typing
On Thu, Feb 13, 2014 at 10:35:13PM +0100, pal...@inwind.it wrote: > Hi, > I am a newcomer to the list and have a problem with netsurf which I am unable > to solve. > I'm trying to use netsurf on a i386 Debian Jessie box. There is no such thing. Do you mean Debian Testing? Sid? What version of NetSurf are you using? > If I start netsurf with the default homepage www.google.com and type a string > of 7 "a"s in the search field, > the cursor in the field moves slower than the right end of the string. After 7 > "a" the cursor is between the 6th and the 6th "a" instead of being at the > right > of the 7th "a" as it should be. It seems that the routine that moves the > cursor > underestimates the space occupied by the string by a factor of 6/7. > Do you have any suggestion to solve the problem? A patch :) More seriously, is this the GTK front end, or the framebuffer? If the later, which renderer? SDL? Frambuffer? X? B.
Cursor doesn' follow typing
Hi, I am a newcomer to the list and have a problem with netsurf which I am unable to solve. I'm trying to use netsurf on a i386 Debian Jessie box. If I start netsurf with the default homepage www.google.com and type a string of 7 "a"s in the search field, the cursor in the field moves slower than the right end of the string. After 7 "a" the cursor is between the 6th and the 6th "a" instead of being at the right of the 7th "a" as it should be. It seems that the routine that moves the cursor underestimates the space occupied by the string by a factor of 6/7. Do you have any suggestion to solve the problem? Best, Palama
url field
I noticed a couple of oddities that may just be the Atari version. 1. I opened an incomplete url and got a "bad url" error. So I closed the Netsurf window and tried to open another. Instead of an empty window I got the bad url error again. I had to exit and restart Netsurf. 2. I tried to append some characters onto the url field but instead of the extra characters being added the url was replaced with the pasted item. Using version 1718 but I know these have been around for a while. Regards, Peter
Re: dates of !Boot and !System within download
In article <53d954c84abrian.jord...@btinternet.com>, Brian Jordan wrote: > I work on the assumption that if a merge-boot or merge-system is required > that NetSurf will let me know with an appropriate message. Me too. I only bother with the other stuff in the download if !NetSurf complains. -- Stuart Winsor
Re: dates of !Boot and !System within download
In message Jim Nagel wrote: > Brian Jordan wrote on 13 Feb: > >> In article <9b5f51d953@abbeypress.net>, >>Jim Nagel wrote: >>> Suggestion to the Netsurf team: > >>> Please, could you include an extra "Readme-date" file in the download >>> zipfile that would simply state in plain text the date of the last >>> actual change and words to the effect that "If you have updated !Boot >>> and !System since this date, there is no need to do it this time." > >>> It would save the chore of going to every machine and repepetititively >>> performing the merge-boot and merge-system rigamarole when it isn't >>> necessary. (The rest of the job of updating the !Netsurf application >>> on all stations can be done over the network.) Thanks. > > >> I work on the assumption that if a merge-boot or merge-system is required >> that NetSurf will let me know with an appropriate message. Otherwise I >> just let NetSurf get on with it. Am I under vigilant or are you over >> vigilant? It seems to me that the NetSurf team have far bigger fish than >> this to fry. > > Maybe I am being overpunctilious, but I'm just following the > instructions in the existing&unchanging Readme that comes as part of > the download. > >>From what you say, I can see that Netsurf might give a warning about a > module being out of date. But the Unicode stuff is not modules: > would Netsurf warn if any Unicode files are missing or outdated? > > I am in the 'ALWAYS update !Boot and !System' camp. I often notice, when doing so, that the hourglass/percentage twitches briefly into life, which suggests that something is being overwritten, i.e., altered, presumably necessarily. -- george greenfield
Re: dates of !Boot and !System within download
Brian Jordan wrote on 13 Feb: > In article <9b5f51d953@abbeypress.net>, >Jim Nagel wrote: >> Suggestion to the Netsurf team: >> Please, could you include an extra "Readme-date" file in the download >> zipfile that would simply state in plain text the date of the last >> actual change and words to the effect that "If you have updated !Boot >> and !System since this date, there is no need to do it this time." >> It would save the chore of going to every machine and repepetititively >> performing the merge-boot and merge-system rigamarole when it isn't >> necessary. (The rest of the job of updating the !Netsurf application >> on all stations can be done over the network.) Thanks. > I work on the assumption that if a merge-boot or merge-system is required > that NetSurf will let me know with an appropriate message. Otherwise I > just let NetSurf get on with it. Am I under vigilant or are you over > vigilant? It seems to me that the NetSurf team have far bigger fish than > this to fry. Maybe I am being overpunctilious, but I'm just following the instructions in the existing&unchanging Readme that comes as part of the download. >From what you say, I can see that Netsurf might give a warning about a module being out of date. But the Unicode stuff is not modules: would Netsurf warn if any Unicode files are missing or outdated? -- Jim Nagelwww.archivemag.co.uk || See you at the SW show? March 1 www.riscos-swshow.co.uk
Re: dates of !Boot and !System within download
On Thu, Feb 13, 2014 at 12:56:17PM +, Brian Jordan wrote: > I work on the assumption that if a merge-boot or merge-system is required > that NetSurf will let me know with an appropriate message. Otherwise I > just let NetSurf get on with it. Am I under vigilant or are you over > vigilant? It seems to me that the NetSurf team have far bigger fish than > this to fry. This has been discussed before, and if memory serves, Brian is correct in everything he says here. B.
Re: dates of !Boot and !System within download
In article <9b5f51d953@abbeypress.net>, Jim Nagel wrote: [Snip] > Suggestion to the Netsurf team: > Please, could you include an extra "Readme-date" file in the download > zipfile that would simply state in plain text the date of the last > actual change and words to the effect that "If you have updated !Boot > and !System since this date, there is no need to do it this time." > It would save the chore of going to every machine and repepetititively > performing the merge-boot and merge-system rigamarole when it isn't > necessary. (The rest of the job of updating the !Netsurf application > on all stations can be done over the network.) Thanks. [Snip] I work on the assumption that if a merge-boot or merge-system is required that NetSurf will let me know with an appropriate message. Otherwise I just let NetSurf get on with it. Am I under vigilant or are you over vigilant? It seems to me that the NetSurf team have far bigger fish than this to fry. -- _ Brian Jordan Virtual RPC-AdjustSA on Windows 8.1 Pro RISC OS 6.20 _ --- This email is free from viruses and malware because avast! Antivirus protection is active. http://www.avast.com
dates of !Boot and !System within download
I have just downloaded the current development version, NetSurf-gcc-json-1718/zip and am in the throes of installing it on several machines here. Datestamps on the !Boot and !System folders and all their contents were lost again sometime before 2014-02-04 (possibly before 2013-04-25, which is the previous time I logged this). It's apparently difficult for the autobuilder(?) to maintain true datestamps on these files. Suggestion to the Netsurf team: Please, could you include an extra "Readme-date" file in the download zipfile that would simply state in plain text the date of the last actual change and words to the effect that "If you have updated !Boot and !System since this date, there is no need to do it this time." It would save the chore of going to every machine and repepetititively performing the merge-boot and merge-system rigamarole when it isn't necessary. (The rest of the job of updating the !Netsurf application on all stations can be done over the network.) Thanks. Here are actual internal dates of modules in today's download. The most recent update was in fact more than a year ago. In the !System.310.Modules directory: CryptRandom 0.13 (01 Jun 2012) © Theo Markettos t...@markettos.org.uk Iconv 0.12 (20 Jan 2013) SharedUnixLibrary 1.10 (10 Apr 2006) (C) UnixLib Developers, 2001-2006 Tinct 0.14 (25 Apr 2009) [fr] In the !System.310.Modules.Network directory: AcornURI 1.04 (20 May 2006) © Christian Ludlam The !Boot in today's download contains !Boot.Resources.!Unicode, which is an application containing 78 files (encodings and other stuff). All of them are stamped 2014-02-04, and unfortunately none of them have internal dates. -- Jim Nagelwww.archivemag.co.uk || See you at the SW show? March 1 www.riscos-swshow.co.uk
Re: Dev Cl #1717 -- bug-tracker peeve
Rob Kendrick wrote on 13 Feb:
> ...back then the same issue happened ("occurs in revision" field not
> accepting an "r") and yet nobody complained!
My point was just a general one that software needs to anticipate that
different users might input stuff in ways that are obvious with
hindsight.
--
Jim Nagelwww.archivemag.co.uk
|| See you at the SW show? March 1 www.riscos-swshow.co.uk
Re: Dev Cl #1717 -- bug-tracker peeve
On Thu, Feb 13, 2014 at 03:24:19AM -0800, Dave Higton wrote:
>
> I think Netsurf is alone in RISC OS applications in having a version
> number incorporating a hash.
It's not a version number, it is a test build number. And it's odd that
nobody complained about test build numbers beginning with "r" when we
used Subversion instead of git, and how back then the same issue
happened ("occurs in revision" field not accepting an "r") and yet
nobody complained!
B.
Re: Dev Cl #1717 -- bug-tracker peeve
On Thu, 13 Feb 2014 10:59:06 GMT Jim Nagel wrote: > You'd think the software behind the bug-tracker could simply ignore > the "#" if people type it in this box. It would seem natural enough > for a user to type it, since "#" is shown loud&clear in the Netsurf > info box as part of the version number. > > Hits on the same pet peeve as websites that throw hissyfits if you > type a credit-card number exactly as shown on the card with the spaces > that make it human-readable and human-checkable. Ditto phone numbers. Agreed! This does suggest that Netsurf should display its version number in the Info box WITHOUT the hash. The hash doesn't add anything. This thread indicates that it has confused users. I think Netsurf is alone in RISC OS applications in having a version number incorporating a hash. Dave Receive Notifications of Incoming Messages Easily monitor multiple email accounts & access them with a click. Visit http://www.inbox.com/notifier and check it out!
Re: Dev Cl #1717 -- bug-tracker peeve
On 13 Feb 2014 Jim Nagel wrote: > You'd think the software behind the bug-tracker could simply ignore > the "#" if people type it in this box. It would seem natural enough > for a user to type it, since "#" is shown loud&clear in the Netsurf > info box as part of the version number. Well there is a # to the left of the text input box so it's reasonable to assume that you have to enter what comes after the #. > Hits on the same pet peeve as websites that throw hissyfits if you > type a credit-card number exactly as shown on the card with the spaces > that make it human-readable and human-checkable. Ditto phone numbers. I agree although don't have a problem with credit card or phone numbers. I do get annoyed with sites that try to be too clever and then give you an error message if you enter a sort code with hyphens or a date with slashes. -- Richard Porterhttp://www.minijem.plus.com/ Skype: minijem2 mailto:r...@minijem.plus.com I don't want a "user experience" - I just want stuff that works.
Re: Dev Cl #1717 -- bug-tracker peeve
Brian Bailey wrote on 11 Feb: > ... I'm not into guessology. # is not a British convention > anyway so far as I am aware, for some 80ish years, it's a hindrance! Why > put it there when it serves no useful purpose, I ask? You'd think the software behind the bug-tracker could simply ignore the "#" if people type it in this box. It would seem natural enough for a user to type it, since "#" is shown loud&clear in the Netsurf info box as part of the version number. Hits on the same pet peeve as websites that throw hissyfits if you type a credit-card number exactly as shown on the card with the spaces that make it human-readable and human-checkable. Ditto phone numbers. -- Jim Nagelwww.archivemag.co.uk || See you at the SW show? March 1 www.riscos-swshow.co.uk
