My sentiment here is that all of this has supported every latent opinion that I had about the project when it was started. Years of maintaining templates, CSS and Javascript on English Wikipedia had taught me not to mess with fonts unless it was very targeted (preferring a specific font for a language fragment or class). It also taught me that anything like this is factors worse on non-English wiki's. I think Erwin was sharing those concerns (probably more vocal than me) for the exact same reason. We might not have been able to show the evidence beforehand in a nicely written document, but we knew this was a very risqué operation. The beta testing had sort of subdued my concerns a bit, but that turned out to be a false sense of security.
With all the good that has happened on the web over the past 10 years, there are simply a few areas that are hopelessly stuck in that past, and those are mostly the areas that are very much bound to (or used to be bound to) OS functionality. These include for instance: Audio/Video, accessibility support and fonts !!!
There are very good reasons for why all these areas are in such a bad shape, history, long support cycles, patents etc. Every time we have tried to make changes in these areas, we seem to underestimate: 1. just how far in the past many of the browser+OS configurations are 2. how much worse 3rd party software can make the situation 3. how inadequate the existing Free/Open alternatives are (especially with regard to global scope/big audiences/high performance) 4. how 'religious' our community is with regard to it's Free/Open foundations.
So for me, the question is not how can we apply pretty serif fonts to headers, the question is what can we do short term and long term to make that happen. Short term: * Accept that the current solution is not working * Rely on Operating System to make the best choice it can, because we cannot do better (return to status quo) * Accept that maybe it might just not be possible right now * Gather statistics on cleartype font rendering (just like we look at tofu). * See if there are ways to make the target group to which the font change is applied narrower/stricter/better defined.
Longer term: * Work with font communities to build true usable, enjoyable Free fonts with global reach. * Work with web/browser development communities to make web fonts more efficient * Work with OS vendors to make sure they are committed to fixing problems in/with fonts, delivering open fonts
It seems to me that those are the areas where we should focus our attention, simply because that will allow us to get to nicely designed pages, without running into those problems mentioned above.
DJ
On Tue, Apr 8, 2014 at 8:14 PM, Erwin Dokter erwin@darcoury.nl wrote:
On 08-04-2014 19:29, Jared Zimmerman wrote:
I don't really have the energy to keep having this conversation, I appreciate that everyone has taken the time to weigh in on this whatever you opinion is on the matter.
I am sorry you feel that way. But I have to make one thing clear:
This is not an aestethic issue. This is a *technical* one.
I do appriciate all the work the designers have done and I do believe the new typography is in essense very well thought out.
However, its *technical* implementation is severy flawed.
It has caused many non-latin projects to force them to override the global font stack in their local CSS to reset it to sans-serif, because parts of their site have become unreadable or illegible. That makes this a *breaking change*, and as such, *must* be reverted.
I do not accept that there will be a "few" readers left with issues; our primary goal is legibility. And if legibility is damaged on a world-wide encyclopedia, how can you even *think* about defending a breaking change?
Regards,
Erwin Dokter
Wikitech-l mailing list Wikitech-l@lists.wikimedia.org https://lists.wikimedia.org/mailman/listinfo/wikitech-l