On 11/04/16 19:54, Felix Miata wrote:

Markus Slopianka composed on 2016-04-11 18:55 (UTC+0200): ...

...

konqueror - Two browsers are overkill. ... Not at all. I typically have at least 5 different ones open at once.

...

Konqueror is also behind in web ... Clearly. However, Konq is the ONLY remaining browser with the capability, (when the DE's logical display density does actually match the screen's physical pixel density), to render objects at specified physical sizes. In all newer browsers[1], physical sizes pc, cm, in, pt, mm and others are all arbitrarily rendered at [il]logical sizes[2]. Rather than 12pt equaling 16px at a display density of 96, 12pt is rendered at 12px, 56.25% of the specificed physical size. At 200DPI, 12pt in other browsers comes out to as little as[3] a mere 12.96% of intended physical size.

[1] In Gecko browsers, physical sizes can be produced, but this requires styling with Gecko's proprietary mozmm unit rather than traditional physical units, such as pt or mm. Physical sizes remain physical sizes in other browsers only for printing.

[2] https://www.w3.org/TR/css3-values/#absolute-lengths cf. prior version: https://www.w3.org/TR/REC-CSS2/syndata.html#length-units

[3] A device pixel and a CSS pixel are at least in theory not necessarily equivalent. A device pixel is a physical size fixed by the display device. A CSS pixel is an angular size that varies very considerably in practice.

Whilst I can only speak for the 4.x version under openSUSE 13.1, Konqueror has surprised me on a few recent occasions. It's the only other browser than Firefox I currently have installed, and so in those circumstances where Firefox throws a wobbly at something I test under Konqueror. It's astoundingly fast. Granted, it's not bogged down with the bookmarks and cache I have in Firefox, but I don't get that sort of page load speed even on a new FF install. And it renders a lot of websites surprisingly well to boot. I'd add that Konqueror isn't just a web browser but a file browser, and whilst Dolphin will suffice for most users, some still prefer Konqueror for some of the additional bells and whistles it provides for local file management. gumb -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

On Monday 11 April 2016 13:54:23 Felix Miata wrote:

Markus Slopianka composed on 2016-04-11 18:55 (UTC+0200): ...

...

konqueror - Two browsers are overkill.

... Not at all. I typically have at least 5 different ones open at once.

So? My proposal was about the set of default applications. Two applications of the same type is not a sane default. My proposal was not about removing anything from the repos. -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

On Monday 11 April 2016 14:56:12 Felix Miata wrote:

Two is sane default.

No. Seriously, it is not.

No browser works on 100% of web pages. A second browser is a standard troubleshooting reference

Web developers can still install it (not that anybody actually cares for an outdated rendering engine). For normal users it is just confusing. It's usability 101. Btw: Are you an actual openSUSE KDE maintainer? -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

Markus Slopianka composed on 2016-04-11 21:38 (UTC+0200):

Felix Miata wrote:

...

Two is sane default.

No. Seriously, it is not.

How short have you been an internet user?

...

No browser works on 100% of web pages. A second browser is a standard troubleshooting reference

Web developers can still install it (not that anybody actually cares for an outdated rendering engine). For normal users it is just confusing. It's usability 101.

Installed by default browser count isn't about developers. Web developers need at least one variant of every major browser installed. That so many don't is a big reason why some much of the web is incompetent or otherwise difficult to impossible for many to use. Normal users are the ones vexed when encountering web pages that don't work. Try browser bar if browser foo doesn't work is the first thing various forum supporters for decades have suggested to anyone complaining web page baz does not work.

Btw: Are you an actual openSUSE KDE maintainer?

This might matter why? I'm a tester with hundreds of openSUSE (plus other distros, but more openSUSE than all others combined) installations, someone who on real hardware only for going on two decades has been spotting and reporting bugs that maintainers and developers miss. Back to $SUBJECT, until upstream KDE drops Konq from its default package selection, in the opinion of many including myself, it belongs in every KDE-providing distro's default KDE selection, whether or not any other browser is installed by default along with KDE. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

Richard Brown composed on 2016-04-11 22:23 (UTC+0200):

I've been an internet user for over 20 years. 1 browser is all you need as a default in the year 2016.

It may well be all many people need, but developers as a group apparently have little idea what ordinary people with less than average vision need. The Internet, like FOSS DEs, are overwhelmingly designed by good vision people for good vision people. Things too often break when trying to adapt all the tininess up to usable sizing. Think about it. How good is the incentive for people with vision limitations to engage in an occupation highly dependent on a sense that they don't have and on which most of every day's work depends? It's poor at best, which explains why developers as a group see better than average, and consequently have limited ability to empathize with and produce results that benefit more than inhibit those less blessed.

As a professional QA tester, I'd love to understand more behind your process. Not because I think I can learn anything from it, but because based on the amount of invalid, incorrect, or downright bizarre issues I've seen you report on these lists lately I suspect your installations, while numerous, have no relation to reality.

OpenQA isn't human. I do things openQA cannot. What happens here happens precisely because of human interaction with software and hardware and not by robots doing exactly the same things unerrantly without deviation. It's heavily, probably mostly, ad hoc, for what I like to think is good reason. http://www.quotationspage.com/quote/26032.html isn't limited to human action. If one human can cause a problem, so can others. Better to find them before product release than later. It's why openQA exists, and a component of why I do what I do. That I find anything broken proves that, good as it is and so clearly better than it used to be, openQA still cannot do everything.

A prime rule of testing is ensuring that the tests you do reflect actual valid use cases of real human beings actually using your software in the real world to do real things.

There's a substantial element of judgment in determining in advance what real users would do, what constitutes a "valid" use case. Humans as a group have a keen ability to do things designers never anticipated even with something as simple as a brass key or a kabob stick. So many more deviations are possible with complex, feature rich PC software and the operating systems on which they depend.

From the sound of things you have an amazing amount of hardware, a huge amount of installations, and a commendable dedication to testing many different scenarios on that hardware.

But if those scenarios are all so artificial that it is irrelevant to the majority of openSUSE users,

The majority of users is not what matters here. "The majority" is not everything it's cracked up to be. It's an overlooked minority operating under handicap or otherwise unable to sail on unassisted I care more about. Besides, where's the fun in doing the same thing as everyone else? Is something fun because so many are doing it? Maybe it's more fun, or at least feels more productive, when it's something nobody else does.

and that those many installations are so untrackable that you cannot easily inform developers how to reproduce the issues you find, I fear you are just wasting your time.

I rarely file bugs unless I can provide a reliable reproduction scenario. A lot of time may well be wasted, but "wasted time" is a natural part of problem solving territory. I like to think each problem I find that results in some sort of fix means at least one less user who ultimately will not need to be told he will have to wait six months for a new release to see a solution to his problem. It may be expensive, but it's not without value.

And I'd much rather see you spend the huge amount of time you are willing to spend on the openSUSE Project productively. We have an

Look at: https://bugzilla.opensuse.org/query.cgi?columnlist=changeddate%2Crep_platform%2Cbug_severity%2Cbug_status%2Cresolution%2Cproduct%2Ccomponent%2Cvotes%2Cshort_desc&email1=mrmazda%40earthlink.net&emailreporter1=1&emailtype1=exact&query_format=advanced&known_name=all Do you see a FIXED to total ratio that equates to average or worse QA time required? I see quite the opposite. I suspect you'll be hard pressed to find better except of "reporters" who only file when they already have or expect to provide a fix themselves.

automated robot for testing, maybe you'd like to write tests for

I've tried writing "for hire" about that which I have limited functional knowledge, e.g. a software manual. It didn't work in the past. It isn't something I expect would ever work, even if it interested me, which it doesn't.

openQA so we can cover some of the real world scenarios that you think we are currently missing right now, so you can retire all that hardware and put it to some real beneficial use for some open source projects?

I've a plenty big enough backlog of non-FOSS things to do. New avenues to deter the unfinished or tempt off the main course I don't need or want. Most of what people here see of what I do is mostly a side effect, with the tail too often wagging the dog. My bigger interest is the harder one to stay focused on and committed to, generically speaking, A11Y/U7Y issues not limited to an openSUSE context. e.g.: https://bugs.kde.org/show_bug.cgi?id=353224 (no bold, so hard to see) https://bugs.kde.org/show_bug.cgi?id=346122 (hard to see what cannot fit) https://bugs.kde.org/show_bug.cgi?id=305161 (again, too small to see)

There is no such thing as a 'default package selection' from KDE

As has been made very clear in recent discussions with usptream, the KDE 'stack' is not a stack, but a loosely coupled collection of components which distributions can choose to adopt any, all, or none, from.

Whatever upstream chooses to call that, it's what I meant, not to mention topical. cf. $SUBJECT. :-) -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

Markus Slopianka composed on 2016-04-12 15:45 (UTC+0200):

Felix Miata wrote:

...

The majority of users is not what matters here.

Actually, it does. That's what a default package selection is for. You clearly have much to learn about basics of usability.

Ironic that you think so. My use of "here" was not in intended in an in-thread sense. "Here" was meant in an A11Y/U7Y, for the benefit of a physically challenged significant minority of users, motivation for involvement in FOSS, sense. -- "The wise are known for their understanding, and pleasant words are persuasive." Proverbs 16:21 (New Living Translation) Team OS/2 ** Reg. Linux User #211409 ** a11y rocks! Felix Miata *** http://fm.no-ip.com/ -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org

On Monday 11 April 2016 23:29:14 Markus Slopianka wrote:

...

...

noto-emoji-fonts & noto-sans-symbols-fonts - Emoji and Dingbats are the norm nowadays. They shouldn't show up as placeholder boxes.

Those are worthwhile additions. Can you file a bug so this doesn't get lost? OK. Under which component should I file it?

Again: Which component? That said, I was recently made aware of https://github.com/eosrei/EmojiOne-Color-Font I'm planning to package it in the coming days. Benefit over Noto Emoji: The SVG-in-OpenType approach works fine with all Gecko applications, i.e. the default browser. If that'll work fine: Any chance you'd help me get that font package into the main repos? -- To unsubscribe, e-mail: opensuse-kde+unsubscribe@opensuse.org To contact the owner, e-mail: opensuse-kde+owner@opensuse.org