I can volunteer to skim your notes here to see what qualifies for a doc update and will repost the relevant lines. This system here is not the most comfortable, but search works, and it is a fantastic archive. Just today I looked for the former TOC plugin, found the discussion quickly, from 2009! Greetings to all BoltWire seniors, Martin
Dan schrieb am Freitag, 24. Juli 2020 um 19:47:33 UTC+2: > I think the issue with {+title+} and {+:title+} may be something else. > Seems the edit html page is processing those in the edit box, so when your > changes get saved it stops being dynamic, if you don't catch it. I thought > I had that fixed, but it's a tricky problem. I'll do some more tinkering > and make sure it's working properly. > > I did test your point about having dots info field names (some.info) and > it is not a problem for me using the <<function>> option. But the problem > with the {+page::field+} format seems to be the colons break the pattern. > It was not really designed for that, but it's an easy fix. Just add a "." > to the pattern on engine.php ~line 151. I'll add that to the next release. > Don't see any reason to not make that available. > > You do make a good point about the css. I read this discussions ( > https://stackoverflow.com/questions/8284365/external-css-vs-inline-style-performance-difference > ) > and while there are pros and cons, it seemed to support your conclusion > that small files can be embedded just as easy as doing an external link. It > would make things easier for me too. I love the idea of having a single > skin html file! > > But here's why I'm not sure I want to go to that approach. 1) You can > always embed the entire css file in the header of your html file and you > don't have to link to an external css file. So you can do what you want no > problem. But 2), for those who prefer an external css (I find it easier to > maintain) you need a system more like what we have now. In other words, our > current system allows both approaches, no problem. And 3) To allow the > instant editing of css and even html settings, I need a way to keep a > template with the placeholder codes (for colors, etc), which is separate > from the css actually used (where the colors are already selected). Now > that can be generated dynamically of course, without much trouble, but that > adds to the processing time--as the css has to be regenerated each time and > then downloaded. > > As it is now, we generate the desired final css, it's get cached online, > and then downloaded and cached on the local browser too. So you get time > savings on both ends. I think for simple skins it might be worth doing it > inline. But to maintain all our options, I think we need a more robust > solution. > > Just my thinking here... :) > > Cheers, > Dan > > PS. I love the tip about tagging helpful notes here for inclusion in the > docs. Would it be worth it to create a forum section where these things > could just be cut and pasted into. Then periodically, I (or someone) could > copy those to the appropriate pages. Or for that matter, if there's an > existing page, it could just get posted into the comments so it's available > right away, and the pages could get updated that way... That would be much > easier for me, but maybe too much of a hurdle for others? > > > On Wed, Jul 22, 2020 at 3:14 PM mz <tcc.p...@gmail.com> wrote: > >> You are right and I was wrong. Sorry! >> I did not remember the notation before the colons of the auth files. My >> last edit was a couple of years ago. >> Maybe you put a legend on top of the list to refresh memories. Anyway: *: >> skinname works perfect. >> >> {+title+} gives me always "css" when <title>{+site+} | {+title+}</title> >> in two different html files even after switching pages. This is a minor >> issue when you know you can use {+:title+} without setting any data vars. >> >> Another minor issue in html files: >> It can be my installation, but {+somepage::some-info+} works, {+somepage:: >> some.info+} with dot works not. >> <<function info field=some.info page=somepage>> does not help either as >> long as you keep the dot in the field name . >> Using the function, the page parameter can contain a dot (page=some.page). >> But not with the short notation: {+some.page::some-info+} does not work in >> html files. >> >> Inlining css and javascript is for me the superior solution and I will >> stick with this as long as I can keep the css under 1000 lines (easy). This >> is not expensive for mobile users because it is compressed and therefore >> does not add much to the page weight. I would estimate gzipped inline css >> and js count up to 20kb if its long - the weight equivalent of a single >> thumbnail image. More expensive for mobile are the second and third >> roundtrip to get the css and js files separately. Here are the milliseconds >> going you like to avoid. On desktop the weight is not a problem at all. >> Inlining is much easier to manage on your (BoltWire) side, too. Maybe you >> rethink your decision, going for a design switch? >> >> In the past you gave a lot of docs update nuggets in your answers right >> here. >> I suggest that everyone who gets an answer like this copies the relevant >> part of the text and post these lines with a new common headline like >> "Docs-Update: docs/concepts/skins". >> This would be easy to find. >> >> Greetings, Martin >> >> >> >> >> Dan schrieb am Mittwoch, 22. Juli 2020 um 18:23:45 UTC+2: >> >>> Thanks for the feedback Martin! I'm still trying to get a grip on >>> exactly how to best make these skins work myself! If you could send me the >>> entry you put in your site.auth.skins page I can see what's going on. Try >>> doing something like *: butterfly and see if that doesn't work. In >>> site.config, you use skinDefault, skinMobile, and skinPrint to change the >>> default for those scenarios when no custom skin is set. Otherwise those are >>> ignored. >>> >>> I really like the new site.auth.skins system as I can get very fine >>> tuned in setting skins for different sections of my site. So for example, >>> in my latest skin I have a version of NEO with a side menu and one without. >>> Otherwise identical. So I can define which skin to use on any page or >>> section of my site easily. Seems to be working great. >>> >>> The code.skin.name.css file may have variables in it like {+text-color+} >>> for example which css can't process if you just link to it. It's pulling >>> the css if you link to it correctly, but css has no idea what to do with >>> those. There's also data fields in some of those files. So here's how to >>> edit the css. Go to site.skins, select the installed skin you want to edit, >>> then click edit css. There's also an option to edit skin settings. This >>> edits the code.skin.name.css page, but also automatically updates the >>> style.css page with the correct values. That gives you the best of both >>> worlds--a cachable external css that you can update at any time. >>> >>> Two notes on that last part... You do need to refresh the page to see >>> your changes as the browser can cache it. And if your server caches it, >>> your changes won't be visible either. One work around is to add ?ver=1.1 >>> after the style.css in the link to it in the html file--and then change >>> that number. It's cumbersome but it forces a refresh of the file. Very >>> handy. I'm tempted to automate this process by adding a counter to the css >>> edit page and then dynamically bringing in that by using >>> style.css?ver={+counter+} or something like that. Trivial but time >>> consuming. We'll see if there is demand. >>> >>> I think the {+title+} works. Is it possible you were on a page that >>> ended in css when you saw that? It should change dynamically with each page >>> you visit. >>> >>> The inline code was removed some versions ago. Alas, the docs are not >>> updated for version 7 yet! For now, use the new block markup. It's proving >>> extremely useful and effective. Here are your main options: >>> >>> <<text 'some markup'>> (this is equal to ==some markup==) >>> <<function forward some.page>> (runs some function) >>> <<markup some.page>> (same as above actually) >>> <<code code.embed.whatever>> (inserts code from a code page) >>> <<cache some.page expires-86400>> (same as markup but caches html for >>> performance) >>> >>> You can also add conditionals to those lines to control when they are >>> triggered. These only work in skins, but they give you virtually unlimited >>> options for how the skin runs. If you have edit permissions on BoltWire you >>> are welcome to help update the docs on this point... >>> >>> Cheers, >>> Dan >>> >>> On Wed, Jul 22, 2020 at 10:00 AM mz <tcc.p...@gmail.com> wrote: >>> >>>> Some observations: >>>> Activating a new skin did not work for me even after setting skin.auth >>>> credentials and updating site.config. Setting in config.php works. >>>> >>>> I found that skin.name.css is not used at all, even if the html file >>>> links to this file. style.css which is the third file in the skin folder >>>> and linking to it works. Deleting this file I saw that the default >>>> style.css is used. >>>> >>>> Using {+title+} in the html file will give you the last part of the css >>>> file. {+:title+} works perfect, as well as all other data and info vars I >>>> tried. Even {+someotherpage::test+}. >>>> >>>> The section "inline code" on page docs/concepts/skins suggests that you >>>> can use code in your skin files with == some code ==. This does not work >>>> or >>>> I misunderstood that section. >>>> >>>> All in all I am very pleased. >>>> Greetings, Martin >>>> >>>> >>>> >>>> >>>> -- >>>> You received this message because you are subscribed to the Google >>>> Groups "BoltWire" group. >>>> To unsubscribe from this group and stop receiving emails from it, send >>>> an email to boltwire+u...@googlegroups.com. >>>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/boltwire/18e293a3-eea7-4f5c-8e3b-4ff5991d4793n%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/boltwire/18e293a3-eea7-4f5c-8e3b-4ff5991d4793n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> -- >> You received this message because you are subscribed to the Google Groups >> "BoltWire" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to boltwire+u...@googlegroups.com. >> > To view this discussion on the web visit >> https://groups.google.com/d/msgid/boltwire/c64e34fd-43c9-4070-bd8a-6ef6edcd175dn%40googlegroups.com >> >> <https://groups.google.com/d/msgid/boltwire/c64e34fd-43c9-4070-bd8a-6ef6edcd175dn%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > -- You received this message because you are subscribed to the Google Groups "BoltWire" group. To unsubscribe from this group and stop receiving emails from it, send an email to boltwire+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/boltwire/2f6dd1a0-e12d-473d-80a9-d8cfbbdb0a3fn%40googlegroups.com.