Bugs and Suggestions

Chris Maunder23-Nov-20 7:30

23-Nov-20 7:30

The entire CodeProject stylesheet is currently being reworked using Bootstrap.

cheers
Chris Maunder

Sergey Alexandrovich Kryukov23-Nov-20 7:57

23-Nov-20 7:57

Sergey A Kryukov

Richard Deeming23-Nov-20 17:42

Richard Deeming

23-Nov-20 17:42

Chris Maunder wrote:
Bootstrap

Version 4 or 5? v5 is currently still in alpha, but sod's law says it will be released the day after your v4 rework goes live, and will have lots of features you really need. Smile | :)

Introduction · Bootstrap v5.0[^]

"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer

Chris Maunder24-Nov-20 5:03

Accept WebP for articles! (Or even SVG..?)

24-Nov-20 5:03

Richard Deeming wrote:
sod's law

I'm keenly aware!

Have you ever watched track cycling where the cyclists play this game of chicken. The person in front always provides a wind-break to the person following, so in a race you actually want to be behind the guy in front (at least until the last metre!). So they ride to the top of the velodrome and go as slow as they can seeing who will break first.

That's how I imagine the Bootstrap team and myself at the moment. I'm picturing launch sometime around 2024.

cheers
Chris Maunder

Sergey Alexandrovich Kryukov22-Nov-20 19:48

22-Nov-20 19:48

Trying to publish my article today, I was surprised by the fact that my .webp image file was rejected by the system.
This kind of action is not just obsolete, it is very unfortunate. WebP is one of the well-established standards for the Web supported by all non-nonsense browsers. It renders JPEG, PNG and GIF morally obsolete, as it it replaces all of them implementing all their features, depending on options: photography, drawings, icons, animations and even video, with dramatic improvement in compression, for both lossy and lossless algorithms.

Side note: it would be good to accept SVG as well, but I can understand that you consider it unsafe due to the possibility of scripting. Even in this aspect, the presence of the <script...> tag could be easily detected and the scripting could be rejected. Vector graphics in articles could seriously improve the overall quality of the site and publications.

Thank you.

—SA

Sergey A Kryukov

Chris Maunder23-Nov-20 6:35

23-Nov-20 6:35

WebP isn't completely ubiquitous yet: Can I use... [^]

SVG: We have partial acceptance and are moving more and more towards using SVG in our system itself. We do have SVG cleaning scripts that mitigate the dangers of SVG.

Now if only everyone used (and was able to use) a no-nonsense browser. We're getting there, but not quite yet. Soon.

cheers
Chris Maunder

Sergey Alexandrovich Kryukov23-Nov-20 6:57

23-Nov-20 6:57

Thank you, Chris.

Your update is reasonable, and I'll wait for the possibility to include SVG in my posts.
Could you notify me when and it it's possible?
I'll have a good use for it.

As to WebP... If you did not consider it, I would seriously ask you to consider.
Here is one thing: it's not important if it is ubiquitous. It is important that it is standardized for the Web browsers, and that the browsers support it well.
It will become ubiquitous if sites and developers accept and use it. This is perfectly safe to do so.

Thank you.

—SA

Sergey A Kryukov

Chris Maunder23-Nov-20 7:29

23-Nov-20 7:29

I've been watching WebP for a while and had actually allowed it briefly a year or two ago but pulled back after noting that many users were still using browsers (eg IE11) that didn't support it.

As I said: soon.

It becomes ubiquitous only when we make the decision to actively exclude certain readers. Those in corporate environments stuck with IE. Those with old tablets who can't afford to upgrade them. Those who are stuck using old versions of browsers due to security policies. At some point we need to simply say "too bad" to those people, and that's never a fun thing to do.

cheers
Chris Maunder

Sergey Alexandrovich Kryukov23-Nov-20 8:06

23-Nov-20 8:06

I have to agree.

IE? Maybe, but... In many US companies I contacted during last decade (not just software), IT discouraged IE and seized to support it.
During first part of 2020, even Edge gave up and embraced Blink+V8.
However, I recently found that Edge is still behind: it does not support very important AV1 (which can be contained in WebM or Matroska). AV1 performance and quality-per-bbs is truly amazing, this is new top standard for Youtube.
Even Mozilla supports it. (Unfortunately, Mozilla also fell behind: I found incomplete support of touch and buggy sound rendering with WebAudio.)

Thank you.

—SA

Sergey A Kryukov

Richard Deeming23-Nov-20 17:40

Richard Deeming

23-Nov-20 17:40

Even Microsoft agree that IE is not a browser[^]. Smile | :)

Beyond that, and Safari being limited to "Big Sur" or later (no, me neither), only KaiOS Browser seems to lack support.

I've had good results using ImageProcessor[^] in an ASP.NET application to dynamically serve up WebP or JPEG images depending on the request headers. Not sure whether that would be an option for you, given the amount of traffic involved?

If you want an image format with really limited support, try AVIF[^]. Big Grin | :-D

"These people looked deep within my soul and assigned me a number based on the order in which I joined."
- Homer

Chris Maunder24-Nov-20 5:06

Reputation counters missing again

24-Nov-20 5:06

Anyone with an old iPad would be locked out.

Dynamically serving is certainly an option, but from a business point of view I'm not sure what I'd gain at a day to day level right now.

cheers
Chris Maunder

DELETE MY ACCOUNT!

User 840608222-Nov-20 13:14

User 8406082

22-Nov-20 13:14

DELETE MY ACCOUNT!

modified 21-Dec-20 21:02pm.

Re: DELETE MY ACCOUNT!

Greg Utas22-Nov-20 14:17

Greg Utas

22-Nov-20 14:17

At the upper right of this page, click on your user name when you're logged in.
This brings up your profile. On the menu at the left, click on "My Settings".
Click on the "Privacy" tab on the right of the light yellow shaded bar.
Check the box beside "Forget me Forever", and then click on "Save My Settings" at the bottom.

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

Nelek21-Nov-20 22:25

Nelek

21-Nov-20 22:25

The reputation counters are missing again in a weird constellation.

Who's who list right now:
Member 14999520 and Member 14999519 (newest members) without counters... OK, tiny delay explained by Matthew

Member 14999518 to Member 14999502 (following members) with counters... OK, looks fine

Member 14999501 (Last of the page 1) to Member 14999495 (Middle of page 2) without counters... Not matching expected behaviour

Member 14999494 (next one) with counters

Member 14999493 (next one) to Member 14999479 (page 3) without counters again...

The only thing that would explain it for me is that all users without counters (no matter which position of this list) have 0 reputation points.

And that was not like that before...

M.D.V.

If something has a solution... Why do we have to worry about?. If it has no solution... For what reason do we have to worry about?
Help me to understand what I'm saying, and I'll explain it better to you
Rating helpful answers is nice, but saying thanks can be even nicer.

Sergey Alexandrovich Kryukov22-Nov-20 20:13

22-Nov-20 20:13

Hi Nelek,

how are you?

As I can see, they are fixed, maybe thanks to your report.

—SA

Sergey A Kryukov

Nelek22-Nov-20 23:06

Nelek

22-Nov-20 23:06

I'm fine, thanks. And you?

This issue is not a big deal, I know, but is a bit nasty at the same time. It works good for some days / weeks, and then it pops up again for a while.

And I am one of the few who sees it due to the SAOTD. But I suppose that is a bit low in the priority list (if nothing important is connected).

M.D.V.

Sergey Alexandrovich Kryukov23-Nov-20 4:22

23-Nov-20 4:22

Sure.
I'm all right, thank you.

—SA

Sergey A Kryukov

Greg Utas23-Nov-20 0:24

Greg Utas

23-Nov-20 0:24

Hi Sergey,

I noticed your CP blog and was going to reply to this[^], but it's read-only, so I'll just say it here...

I agree that easy to fix is one of the important characterizations of a good architecture, although saying so doesn't tell you how to design a good one. Along with that goes easy to evolve. Overly customizable is a drawback unless users of the architecture can't wait for it to evolve, which would usually be true of a closed-source, third-party framework, but not one developed in-house.

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

Sergey Alexandrovich Kryukov23-Nov-20 4:18

23-Nov-20 4:18

Hi Greg,

Thank you for paying attention for this and sharing your thought.

You are absolutely right. And yes, my saying doesn't tell anything about how to design the architecture, as I did not mean to say that. I would be surprised to see someone's aphorism saying something like that. Also, I believe that the formulation of a problem is the most important thing.

It's interesting that you mentioned excessive customization, even though it has nothing to do with the characteristic I suggested, and it can often work badly against the criterion. I can understand if this is because this is another problem bothering you. Indeed, I used to figure out that some customization was designed as if the goal of it wasn't the function of the system, but simply the inability of the authors to develop proper industrial design: we cannot make good design decisions, maybe the user can... Smile | :)

Thank you.

—SA

Sergey A Kryukov

Greg Utas23-Nov-20 4:39

Greg Utas

23-Nov-20 4:39

Your post mentioned excessive customization as a drawback, which is why I commented on it.

Thanks,
Greg

Robust Services Core | Software Techniques for Lemmings | Articles
The fox knows many things, but the hedgehog knows one big thing.

Sergey Alexandrovich Kryukov23-Nov-20 4:51