[minutes] 2020-06-11 Project Review on W3C website redesign: Choosing a front end framework

Hi all,

The minutes of the meeting to "Choosing a front end framework” are available (and in text below):
 https://www.w3.org/2020/06/11-redesign-minutes.html


> On 10 Jun 2020, at 17:40 , xueyuan <xueyuan@w3.org> wrote:
> 
> Dear all,
> 
> This is a reminder about the upcoming Project Review on W3C website redesign: Choosing a front end framework.
> 
> Reference material can been seen at:
> https://w3c.studio24.net/docs/choosing-front-end-framework/
> 
> Presenter: Coralie Mercier
> Date:  Thursday, 11 June 2020
> Time:  1300 UTC (60 minutes)
> […]
> 
> On 6/5/20 4:39 AM, Coralie Mercier wrote:
>> Dear team, Léonie, fantasai,
>> 
>> Choosing a front end framework is probably one of the key milestones in the W3C Website (phase 1) redesign project, and I would like to make sure that we get the appropriate review to help us make the best decision.
>> 
>> Studio 24 (who as an agency would prefer to work with Apollo but understand if the need to use a standard tool such as Bootstrap is required and will work with this if it's the better option for W3C) have reviewed the questions around HTML/CSS frameworks and have written up the pros and cons of Bootstrap and Apollo.
>> 
>> I’m seeking wide review and feedback until next Thursday, June 11. (cf. mail below, archived in <https://lists.w3.org/Archives/Public/public-website-redesign/2020Jun/0001.html> and W3C blog post <https://www.w3.org/blog/2020/06/w3c-website-redesign-choosing-a-front-end-framework/>.)
>> Team should share feedback on the public mailing list but can do so on the team-website-redesign mailing list if they prefer.
>> 
>> In addition, I would like to schedule the following week a project review. This takes place Thursdays at 1300 UTC for 60 minutes.
>> 
>> The project review would be open to the team, Léonie and fantasai if they are willing and available (as well as any other key participating folks from the community —please let me know who you think I should invite), with guest Simon Jones and Nicola Saunders of Studio 24. The objective of the review is to present the choices, assess any feedback received in the past 7 days, address questions and try to converge on an informed decision.
>> 
>> Coralie


Text version:
=============

            Project review: Choosing a front-end framework
                           11 June 2020

  [2]IRC log.
     [2] https://www.w3.org/2020/06/11-redesign-irc

Attendees
  Present
         Bert, Coralie, denis, fantasai, francois, Gerald, ivan,
         Jean-Gui, Jeff, Kaz, Léonie, Nicola_Saunders, plh,
         Ralph, Richard, Sam, Simon_Jones, Ted, vivien, weiler,
         wseltzer
  Regrets
         Dom, Xueyuan
  Chair
         koalie
  Scribe
         koalie

Contents
   1. [3]Systeam feedback
   2. [4]Framework? Apollo not being a framework in the same way
      Bootstrap is
   3. [5]New changes in the future
   4. [6]Testing
   5. [7]Doc and community
   6. [8]Maintenance
   7. [9]Licensing
   8. [10]Deadline

Meeting minutes

 Systeam feedback

  <vivien> 

  <vivien> The Systems Team discussed this topic yesterday.

  <vivien> You will find minutes here:

  <vivien> [11]https://www.w3.org/2020/06/
  10-systeam-minutes#item01

    [11] https://www.w3.org/2020/06/10-systeam-minutes#item01

  <vivien> The main points were:

  <vivien> Testing: Apollo v2 is really new and we would be the
  first ones to use it, meaning it has not received a lot of
  testing yet, which ultimately means there will be bugs as for
  any new project being release

  <vivien> Maintenance: What kind of maintenance can we expect
  after the final version of the website is released and more
  generally on the long term (to handle issues reported by our
  users or to use on other pages)

  <vivien> Documentation and community: seems to be a key aspect.
  Bootstrap has a huge community and documentation available,
  even if Apollo is on GitHub it does not seem to have any
  community, so we can't really call this an open source project.

  <vivien> Deadline: You mention it might take more time for you
  if you were to use Bootstrap, what would be the impact on the
  project timeline?

  <vivien> ]]

  [Vivien summarizes the above]

 Framework? Apollo not being a framework in the same way Bootstrap is

  Simon: we might have confused things a bit by calling Apollo a
  framework
  … while it's our agency's approach to building websites
  … the team has been using that [since 2017] in the work we do
  … Nick is looking as v2
  … this approach lets us work on project more efficiently
  … We open sourced it as it's easy to do

  Nicki: Apollo is really about giving us a means to quickly
  start working on a new project
  … consistent

  <fantasai> My comments posted to [12]https://lists.w3.org/
  Archives/Public/public-website-redesign/2020Jun/0004.html fwiw

    [12] https://lists.w3.org/Archives/Public/public-website-redesign/2020Jun/0004.html

  Nicki: uses vanilla html css

  Jeff: the confusion I have re: framework is that I'm not sure
  the breadth of the statement of what apollo does
  … we build web pages for our website
  … are the people going to use apollo?

  Simon: It's for us to help build the CSS
  … once we're done, you'll have a bunch of html templates and
  css
  … so if people want to use the new design they will have to use
  the templates
  … we'll document that in a design system
  … apollo will end up being a small part of what we'll end up
  building for you

  fantasai: I made a whole bunch of comment
  … that I won't repeat
  … It's important to know that Apollo is like a skeleton with a
  bunch of patterns
  … set of CSS rules and maybe some javascript
  … so you don't have to build common components like tabbed
  navigation, sidebar layouts, etc. from scratch
  … if you write a website with html and css, it will not break
  down

  <Ralph> [13]fantasai's comments to public-website-redesign

    [13] https://www.w3.org/2020/06/11-redesign-irc#T13-10-51

  fantasai: maintenance, if you need to improve, is only needed
  … to make it better
  … so not concerned about the maintenance question

  Nicki: Thanks fantasai for the comments
  … what we're trying to do with apollo isn't very far from html
  and css
  … allows us to add a minimal lay of customization

  Léonie: Nicki just confirmed that apollo is an accelerator
  … so it's a good thing
  … I agree maintenability is an important aspect
  … also, endorsement:
  … as soon as W3C chooses a framework there's an implicit
  endorsement
  … also, leading by example
  … apollo seems to stick to the good basic ideas of HTML and CSS
  … and W3C could take this stance
  … lastly,
  … accessibility is important
  … I looked at apollo

  <fantasai> Bert, in some ways yes, but in some ways "fix things
  that are broken" is a large part of what's considered
  "maintenance", and HTML and CSS don't "break" going forward :)

  Léonie: pretty good, with a few things to look at
  … but far more accessible than Bootstrap
  … accessibillity can be gnarly
  … my recommendation is to use apollo.

  Richard: I know next to nothing about these things
  … what I'd like to know is: are these frameworks tools to
  create the templates?
  … or will people like me who contribute pages need to know that
  framework?

  Nicki: apollo is about building components and layouts
  … there needs to be another decision on what CMS to use
  … apollo is about front-end styling

  <weiler> [thank you for the clarification that CMS is in-scope
  and also not ripe]

  Nicki: there should not be much need to change the CSS
  … there will be guidelines on typography and colour

  Richard: Would I, if I want to create a page in the future,
  will have to understand apollo to use the templates that are
  available?

  Vivien: today anyone can start a page from scratch, literally
  … we're trying to move away from that
  … team will lose some freedom in their ability to create web
  pages
  … in the name of a more harmonious website, consistent, and
  with better branding
  … pages will be started from a common ground
  … via a CMS
  … your focus will be on the content, not the page itself

  Richard: I hope that when we put the site together we take i18n
  into account as well as a11y
  … language labels, right to left, etc.

  Nicki: We're very aware of the global nature of W3C
  … we're keen to add to apollo something for i18n

  Simon: bidirectional text, for example

  Richard: It would be essential to have support for bidi content

  Simon: Right

  Nicki: we want to support CSS logical properties
  … fallback will be needed

  Richard: great to hear

  <Zakim> Ralph, you wanted to ask to what Simon was referring
  when he said "might document"

  Ralph: Richard just proved my point in IRC: he does know a lot!
  … I had a question for Simon
  … in response to Jeff's question on components apollo provides
  … I heard you say that there are things you *might* document
  … which things might be important to our long-term use of the
  site?

  Simon: slip of the tongue
  … we intend to deliver a design system
  … to explain the templates
  … so that the team can build more web pages

  Simon: it's going to be more than apollo
  … we're aiming to document as much as possible
  … sorry for using "might"

  Ralph: if we want to create components on our own, will there
  be documentation for that?

  Simon: that depends on the CMS question
  … not a hundred percent sure

  Kaz: Many mechanisms for W3C pages
  … some manually edited, some generated, some mirrored from GH
  … we should think about which part should be handled by which
  and how to combine
  … and how to integrate all of them is the key

  Simon: the design system (documentation) is intended to be
  static html
  … we can integrate this in any system you use

  <Zakim> Bert, you wanted to say that "maintenance" and "making
  it better" is not that distinct.

  Bert: back to a point fantasai made
  … maintenance indeed may not be needed
  … but between maintenance and improvement there is a fine line
  … sometimes people need just a little bit more
  … [Bert gives an example of adding ability to have three images
  in a row with captions, by adding a few lines of CSS]
  … in any case, we need to know a bit of the existing
  stylesheets in order to extend them for
  (improvement|maintenance)
  … so it's not just writing one for 20 years

  <Zakim> fantasai, you wanted to note that we have a lot of
  hand-coded pages and that's likely to continue

  fantasai: we do have a lot of hand-coded pages on the site
  … they need to continue to work

  <Ralph> [in Vivien's response to Richard's question I heard
  something else that would be nice to clarify internally. In our
  current environment any Team member is able to post a trivially
  bare HTML file on w3.org, and then it can be styled later. The
  threshold to First Publish is very low. If the CMS requires
  more knowledge before First Publish, that raises the threshold]

  fantasai: we'll continue to add hand-coded pages
  … that's how we experiment and show off
  … whatever we're building will need to accommodate that
  … Bootstrap has a lot of classes
  … apollo has much fewer things
  … will only include things that we need

  <Ralph> [ah! Fantasai's "hand coded pages" comment is a variant
  of my concern]

  fantasai: I second what Léonie said

  <jeff> [Correct, Ralph, and it's not just the Team.]

 New changes in the future

  Nicki: Apollo is a starting point
  … we keep it under review
  … from a client support point of view too
  … we handle that on a case by case basis
  … in which case we may decide to roll back into apollo if it
  fills a gap that's been identified

  <gerald> [/me would expect the team would continue to be able
  to create arbitrary html/css/etc files]

 Testing

  Sam: how easy will features be ready to roll out, e.g.
  webauthn?

  Simon: I read the question
  … I don't think it makes a difference what framework we use
  … it's more of a programming thing
  … e.g. symfony or CMS
  … apollo or bootstrap would not prevent you from doing this

  Coralie: that's Vivien's topic as Apollo v2 is really new and
  we would be the first ones to use it, meaning it has not
  received a lot of testing yet, which ultimately means there
  will be bugs as for any new project being release

  Vivien: V2 is just going out

  Vivien: concerned a bit that we're going to be a guinea pig for
  apollo v2
  … from systeam's point of view this is something we try to
  avoid
  … ability to handle bug reports?
  … during and after the project

  Simon: the current v2 that Nicki is working on is based on v1
  available the last 4 years
  … like any project we'll be testing as we build
  … thoroughly

  Richard: I get a little nervous when people say "we're going to
  do a11y testing" and don't mention i18n"

  Simon: We'll have to look at that, it's a very good point

  <Zakim> fantasai, you wanted to contrast with hand-coded CSS

  fantasai: on the i18n side, probably it's going to be easier
  … w3c's website currently uses content negotation
  … the CMS will have to take this into account
  … apache's approach is sophisticated, most CMSes can't do that
  … from the css framework perspective I'm less concerned
  … while we're considering the testing question
  … if Studio 24 had offered to hand-code everything
  … that would be like a similar thing to consider
  … we would also be able to changes things
  … the code behind apollo is not particularly complicated
  … some of the leg work is done for common patterns and that's
  what apollo brings
  … you could argue testing bootstrap @@ vs hand-coded website
  … apollo is closer to that

  Simon: true

 Doc and community

  Vivien: Simon gave answers in that regard already
  … but comparing with Bootstrap, where there's a huge community
  … you can find issues and fixes
  … with apollo, we saw the doc but no existing client issues,
  for example

  Vivien: are you expecing your clients to use GH as a platform
  for issues / feedback?

  Simon: it's not really a community project the same way
  bootstrap is
  … so far we've not had clients put issues to GH
  … they use our support system
  … that could change in the future
  … so far GH is our code repository

  Vivien: One comment Denis made in the systeam call yesterday
  … w3c using apollo may increase the exposure on apollo

  Simon: of course we hope to increase exposure but not
  particularly on apollo

  <fantasai> Wrt maintenance, I hope vivien and others have read
  [14]https://superfriendlydesign.systems/articles/
  should-you-use-bootstrap-or-material-design-for-your-design-sys
  tem/

    [14] https://superfriendlydesign.systems/articles/should-you-use-bootstrap-or-material-design-for-your-design-system/

  Simon: since we don't deploy it for others necessarily to use,
  but it's our tool

  <fantasai> which was linked from the Studio24 doc

 Maintenance

  [Maintenance: What kind of maintenance can we expect after the
  final version of the website is released and more generally on
  the long term (to handle issues reported by our users or to use
  on other pages)]

  Simon: that depends on our contract with the client
  … we have support contracts with some clients
  … if we add things to apollo that you want to use, we'll be
  happy to let you know and advise you

 Licensing

  Simon: any work we do is transferred to the clients
  … apollo is open source
  … built to be reused

  <wseltzer> [sounds good to me]

  <wseltzer> [feel free to ask me if you have license questions]

 Deadline

  Vivien: if we were to use a framework different from apollo
  … S24 mentioned they'd need more time
  … and that might incur delays
  … what would be the impact from your pov?

  [Richard departs]

  Simon: apollo having pre-built things, there would be overhead
  … we would deprioritize other work

  <Ralph> Coralie: our timeline says we need to make a decision
  mid-June

  <Ralph> ... can we get back to you next week?

  <Ralph> Simon: yes

  Bert: I see the apollo is written in sass
  … that would be a new tech for us to use
  … would we have to use sass

  Simon: you don't have to
  … we'd discuss that
  … there are other ways

  <Ralph> [15]https://en.wikipedia.org/wiki/
  Sass_(stylesheet_language)

    [15] https://en.wikipedia.org/wiki/Sass_(stylesheet_language)

  <simonrjones> thanks for everyone's time and comments, a very
  useful call.

  Simon: we'll take your lead

  <Ralph> 

  <Ralph> Sass is a preprocessor scripting language that is
  interpreted or compiled into Cascading Style Sheets (CSS).

  <Ralph> ]]

  fantasai: someone said that we want our stylesheets need to be
  readable and an example
  … I concur
  … using bootstrap would [mot do that]
  … we should turn off css minimization
  … view source

  Simon: we agree

  <Ralph> big +1 to making it trivially easy to 'view source'

  <caribou> [using something like bootstrap might send the wrong
  idea that simple css is not manageable]

  <koalie> I concur with caribou's statement

  <koalie> and generally with what fantasai said (and Bert, and
  others)


   Minutes manually created (not a transcript), formatted by
   [16]scribe.perl version 121 (Mon Jun 8 14:50:45 2020 UTC).

    [16] https://w3c.github.io/scribe2/scribedoc.html


--
Coralie Mercier  -  W3C Marketing & Communications -  https://www.w3.org
mailto:coralie@w3.org +337 810 795 22 https://www.w3.org/People/Coralie/

Received on Monday, 15 June 2020 19:44:23 UTC