|
01 Apr 18 - 03:00 AM (#3914417) Subject: Tech: FireFox 59 - anyone notice this bug? From: Mr Red my FireFox 59 (portable) is having trouble with my own sites on pages using FRAMEs. FF 54 & 34, Chrome 40 & 53.0.2785.116 and IE11 all handle the same pages cleanly. On my PCs anyway. (Win7) If anyone wants to try on the latest FireFox for me? it loads either the Main or Upper FRAMEs but not both. Post here with coments. TIA. http://stroudceilidhs.co.uk http://dance.mister.red clicking on links in the Upper FRAME can fill the Main FRAME - depending on if they are intended to do that like in "dance" clicking refresh or F5 will usually resolve the problem, which is hardly satisfactory. I may have to resort to yet more JavaScript to automate the refresh - not that straightforward because it has to reload, and it then has to remember it has, when a reload wipes everything! And a FRAME's own code cannot decide it hasn't loaded if it hasn't loaded! No answers on the FireFox (Portable) Forum - yet! It's a jungle out there. |
|
01 Apr 18 - 04:05 AM (#3914431) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo Seems to work OK in Fx 59.0.2 on Android. I'll try a desktop later if nobody else has. Maybe something to do with the way Fx portable does cacheing - but I'm guessing. |
|
01 Apr 18 - 04:09 AM (#3914433) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: GUEST,Peter Laban seems fine to me on 59.0.02 (64 bit) on PC |
|
01 Apr 18 - 04:46 AM (#3914441) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red looking at it - probably a "portable" thingy. But the Chrome versions are "portable". But then FF portable has had trouble debugging my JavaScript since v50 ish, and I rely on FF34 for that (or Chrome). The later portable versions must have a timer running, because any page without a timer does not cause trouble. |
|
01 Apr 18 - 06:39 AM (#3914464) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: GUEST,Mick Pearce (MCP) Works fine on 59.0.2 (64bit) (Mozilla Firefox for Ubuntu canonical-1.0) on Ubuntu 17.10 pc. Mick |
|
01 Apr 18 - 10:45 AM (#3914501) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo Clicking around the dance site was a bit taken aback by a screen which said (in capitals - some of which are bold): WARNING!! THIS SITE CONTAINS ADULT MATERIALS OR MATERIALS THAT MAY BE CONSIDERED OFFENSIVE IN SOME COMMUNITIES. YOU MAY NOT ENTER THIS SITE IF YOU ARE EASILY SHOCKED OR OFFENDED OR IF THE STANDARDS OF YOUR COMMUNITY DO NOT ALLOW FOR THE VIEWING OF ADULT EROTIC MATERIALSscreenshot (link will expire). What do you get up to in these ceilidhs? Or is it an April Fool? |
|
01 Apr 18 - 11:22 AM (#3914508) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red dunno, maybe the webring has expired and been taken-over by miscreants. I will have to seek out the links and whip them till they submit! <noscript>Naughty Webring, hiding behind the - <NOSCRIPT> </noscript> But as GBS said "perpendicular expression of horizontal desire" |
|
02 Apr 18 - 12:56 PM (#3914729) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Newport Boy Frames are a recipe for trouble. I use one webpage which displays in Firefox correctly on my laptop but not on the desktop. On the desktop, the only browser that shows it correctly is Opera, although Pale Moon is almost correct. The frame statement is deprecated in HTML5 and the position can only get worse as browsers update. The clearest brief summary I found is here Phil |
|
03 Apr 18 - 04:07 AM (#3914826) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo Frames may be obsolete but they still seem to be well-supported: MDN. One problem with frames is that search engines sometimes dump you in a frame, with no navigation or indication that the page is incomplete. For example I got here via a search on part of that page. (Admittedly it was not the first hit.) It also changes the URL at the top. |
|
03 Apr 18 - 05:23 PM (#3915008) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red Goggle have long indicated their dislike of frames. The alternative is layers and they are not so easy to design with. You can do more with them, and do more harm too. I use them as pop-ups and drop-down menus anyway. I am considering JavaScript to solve the immediate problem. Not that my Folk site is responsive, but it takes more thought to achieve in layers. I have constructed a site that is responsive, mainly because it can stand alone or sit inside a responsive Wordpress site. But, and here's the catch: to place it in a Wordpress page it is in an IFRAME. To do it in a layer would require more knowledge than I possess about Wordpress. the stand alone page is ask.stroudvoices Loss of Goggle SEO is not high on my priorities, the lexicon alone is 7800 words. If Goggle want to decode it............ tell them it is a JavaScript array. As for dumping you in a frame - JavaScript could sort that - but fer now - priorities! Anyways, I am very grateful for all the insights (and incites). Thanx folks. |
|
04 Apr 18 - 04:09 AM (#3915050) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: treewind Frames are horrid for all sorts of reasons. Of course they are widely supported by browsers because they have been around for a long time and many sites that use them, but that's not a reason to use them in any new site. I used to do frames for consistent navigation menus across all pages, but I stopped that 10 years ago when I discovered PHP. Responsiveness is easy with PHP and CSS (especially CSS3). I've never used layers or Javascript. |
|
04 Apr 18 - 05:01 AM (#3915058) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red PHP - fine if you want to learn yet another language. And spend 10 years using it. For everyone else there is Wordpress, Joomla (yuk) Droopl (?) etc which are supposed to be easy, but, trust me, are yet another learning curve with limitations. And if JavaScript was so infra-dig, how come a bog-standard Fakebook page (sans apps) with Chat switched off - has 180+ JavaScript files listed? One reason is that JS gives your PC the task of dealing with the time consuming stuff wot would load their servers. They have a lot of servers. But a lot of subscribers. I considered using PHP for my audio website but the first hurdle was having to construct an area for every person currently playing an audio. And how many people do you cater for? Similar hurdles with server-side JS. Answer is distributed processing. That answer is client-side. Only JS will do it on a website. That is why Fakebook could grow so fast. There is clever and there is smart. They ain't the same. |
|
04 Apr 18 - 07:38 PM (#3915234) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: treewind Point taken about moving the work to the client when there's millions of users hammering on the web site. I know about WP, Joomla and Drupal. Especially Wordpress: I have several web hosting customers using it and it does let them get on with building their own sites without me having to get involved much. I don't understand the issue about playing audio - the HTML5 audio element plays audio and doesn't need PHP or separate areas for each user. But we're drifting way off topic... |
|
05 Apr 18 - 03:46 AM (#3915264) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red Playing audio: selecting a subset of say 110 out of 1800 tracks (ie filtered on the subject of say "canal"). The list of the subset has to be stored. On PHP that requires a dedicated area/array. But how much temporary storage do you allocate for multiple users? How many users to cater for? When to decide they have left? It's a jungle out there. The logical answer is to distribute the job. And that is JavaScript territory. I well remember seeing the code that a streaming service provided for the radio station when it needed changing. PHP larded with JavaScript (or did I mean the other way round). If the professionals need it, I take my cue from them. (limit of 10 streaming users since you ask) And at the end of the day - do all your users use HTML5 browsers? When I started this software I tried to get an acquaintance to look at one of my websites and tried a lot of prompting with carefully crafted questions that didn't involve "What version of IE?". The best analysis came up with IE8. Flash (etc) territory. That was 3 years ago. My software reecognises the possibility. How does PHP do that? I regard the few as valuable. They contribute to my audio, and it is hoped they will keep doing so. It is a collection of old people talking about their life in Stroud. It is Folklore goddamnit. They are old. They are not computer experts. Yes, I do know quite a few luddites, and they are not all Folkies! Their mantra is "I HAVE A COMPUTER!". You can assume that is of any vintage. or disenfranchise them! |
|
08 Apr 18 - 03:53 AM (#3915812) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red I should have added there are two more reasons I stick with JavaScript 1) the results of the search engine (my JavaScript) present in chunks of 5. So there has to be some form of memory - client-side. 2) the identical software is given to two local Gov archive repositories and as documentation they get the software to run in a browser regardless of internet connection. And all links are de-linked, but leave some URLs in text. This is necessary for their net integrity, imagine clueless browser peeps doing a "what if" and endangering their whole archive. I don't know how they offer the set to the public, but a spreadsheet is included to cover options. Back to the OP: I have revamped the HTML of the subdomain which removes one frame level and it has made no difference to the problem. A clickie covers the eventuality. It must be a small number of users that find the same symptoms. FWIW dance.mister.red/dance44.htm uses an IFRAME and doesn't exhibit problems. I have probably done all I can for now. Thanx 4 all ur help, guys. |
|
08 Apr 18 - 11:10 AM (#3915875) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo I notice that http://dance.mister.red gives a warning: An unbalanced tree was written using document.write() causing data from the network to be reparsed. For more information https://developer.mozilla.org/en/Optimizing_Your_Pages_for_Speculative_Parsing 1 dance.mister.red:35... whereas http://dance.mister.red/dance44.htm does not. It's possible that this reparsing is triggering the problem in portable Firefox. You could also check if earlier versions of Firefox gave that warning and see if you can get rid of it by following the advice on the MDN page. Firefox is constantly increasing the degree of speculative preloading of network resources so Fx59 may well be doing this more agressively than hitherto |
|
08 Apr 18 - 01:36 PM (#3915926) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red I do set up the frames in code. But I tried writing a plain old bog-standard HTML and the same symptoms appear. I also tried putting script in the head, the body, and after the body as a proof of concept that would monitor the situation and FF59 didn't show any existence of it so that route is a dead end. I have, in the past, tried code that looked from one frame into the other and same via the holding frame but I think I read somewhere it isn't allowed anyway. Maybe the BT CDN is amplifying speculation. But that is speculation! I will look into the link thanx. Mind you each version of FF I have displays debugging differently. Ho Hum. |
|
08 Apr 18 - 02:46 PM (#3915936) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo I don't see what's unbalanced. Just defining a frameset may implicitly change the base address and force a re-parse - but I'm guessing again. You could set network.http.speculative-parallel-limit to zero and see if it makes a difference. So long as it only affects portable Fx it's not much of a problem. If it's more widespread I would convert the common frames into iframes. |
|
12 Apr 18 - 05:28 PM (#3916919) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: Mr Red I think it is the end of the story. The solution was (not elegant) to put a simple page in the lower FRAME that merely replaced itself immediately. And in the process re-instated the code to allow hash data to select other than the default page. Which can then process the instruction to highlight a date (if data are in the hash) ie http://dance.mister.red/#stroud#20180609 The reason it has taken so long is the number of parameters to perm, and a wish (unfulfilled) to do it right and find the erroneous methodology. It could have been the internet provider (BT) but I can't tell. Clunky does it though. Thanx everyone. |
|
13 Apr 18 - 03:07 AM (#3916970) Subject: RE: Tech: FireFox 59 - anyone notice this bug? From: DaveRo http://dance.mister.red/#stroud#20180609I think that URI is technically invalid - I don't think you can have a fragment containing a #. See RFC3986 3.5 and earlier definition of 'pchar'. It seems to work OK here in mobile Firefox but look out for odd effects with location.hashes. |