For the people who talk on the IRC channel and haven't already done this, what do you think about moving to an #edbrowse on libera as a result of the ongoing troubles?
"Kevin Carhart" <kevin_carhart@fastmail.fm> writes: > For the people who talk on the IRC channel and haven't already done > this, what do you think about moving to an #edbrowse on libera as a > result of the ongoing troubles? Here's the backstory for those who aren't following. This hits kind of close to home, since my girlfriend was Freenode staff and we actually met on Freenode. I dropped off Freenode suddenly, because I didn't want to deal with this stuff. Essentially, the head of staff at Freenode (Christel) sold it in 2017 to a bitcoin millionaire, Andrew Lee. Lee claimed that he would be hands-off. But last month, he made a power play, appointing people to staff positions and trying to have others removed from staff positions. Freenode has always been volunteer-run. The volunteer staff resigned en masse. Here are links to a couple of their resignation letters. https://fuchsnet.ch/freenode-resign-letter.txt https://gist.github.com/aaronmdjones/1a9a93ded5b7d162c3f58bdd66b8f491 I registered #edbrowse on irc.libera.chat. A few of us are in ##edbrowse, but we may as well drop the extra # if we do move to the new network. -- Chris
On Fri, May 21, 2021 at 03:30:52AM -0700, Chris Brannon wrote:
> Essentially, the head of staff at Freenode (Christel) sold it in 2017 to
> a bitcoin millionaire, Andrew Lee. Lee claimed that he would be
> hands-off. But last month, he made a power play, appointing people to
> staff positions and trying to have others removed from staff positions.
> Freenode has always been volunteer-run. The volunteer staff resigned en
> masse. Here are links to a couple of their resignation letters.
>
> https://fuchsnet.ch/freenode-resign-letter.txt
> https://gist.github.com/aaronmdjones/1a9a93ded5b7d162c3f58bdd66b8f491
>
> I registered #edbrowse on irc.libera.chat. A few of us are in
> ##edbrowse, but we may as well drop the extra # if we do move to the new network.
Agreed on the move and the dropping of the extra #. Tbh I think that was only
there because of the freenode channel naming policy and the fact that, at the
time of registration, it was a little debatable how official it was as the
project channel. Since then it's become that so that makes sense. Also, when
I got some more time I was going to propose this move anyway due to
what's happening at Freenode.
Cheers,
Adam.
Paging Silas, do you subscribe to edbrowse-dev now? I researched what Silas is talking about a little bit and here is what I found out. I can't tell if it's significant or just odd. I went to both jdb and the firefox console, and I can confirm that our behavior diverges from FF behavior, but it's also different from what I expected in a strange way In edbrowse, if I load a page and then say document.body.innerHTML, I get stuff Then if I say document.write("hello") and then document.body.innerHTML, nothing changed and it echoes the same string. In the FF console if I run the same things, I was expecting it to append "hello" on the existing innerHTML. But it completely clobbers. And correspondingly, the render completely changes. (of course, only in the browser, where you can mess with the already-downloaded as much as you like.) So now the entire page has had its innerHTML changed to just "hello". So I don't know if ssb22 is saying that an append or a clobber is what should happen, but I think he's pointing out that currently in edbrowse neither happens.
On Wed, Jun 09, 2021 at 10:14:50PM -0700, Kevin Carhart wrote: > I researched what Silas is talking about a little bit and here is what I found out. I can't tell if it's significant or just odd. > > I went to both jdb and the firefox console, and I can confirm that our behavior diverges from FF behavior, but it's also different from what I expected in a strange way > In edbrowse, if I load a page and then say document.body.innerHTML, I get stuff > Then if I say document.write("hello") and then document.body.innerHTML, nothing changed and it echoes the same string. > In the FF console if I run the same things, I was expecting it to append "hello" on the existing innerHTML. But it completely clobbers. And correspondingly, the render completely changes. (of course, only in the browser, where you can mess with the already-downloaded as much as you like.) So now the entire page has had its innerHTML changed to just "hello". > > So I don't know if ssb22 is saying that an append or a clobber is what should happen, but I think he's pointing out that currently in edbrowse neither happens. > I also always assumed append would be the thing to do here, but I checked [1] and it seems that if the document is already loaded when document.write is called it defaults to blasting the document first. [1]: https://developer.mozilla.org/en-US/docs/Web/API/Document/write
So just to be clear: document.write, if already browsing, should be, effectively, document.body.innerHTML = blah The page is completely gone and replaced. I don't think it's a hard change, but it's kinda drastic. Also, two or three document.writes in a row - does each clobber the previous, or do they append and then when the script is done, that is the new page? Karl Dahlke
Appends, so from the FF console if you send document.write(1);document.write(2);document.write(3) document.body.innerHTML returns "123"
On Fri, Jun 11, 2021 at 09:14:45PM -0700, Kevin Carhart wrote:
> Appends, so from the FF console if you send
> document.write(1);document.write(2);document.write(3)
>
> document.body.innerHTML returns "123"
>
As far as I can tell it's like a stream. When loading, the document is
opened for writing and thus document.write() appends. Once loading has
completed (when browsing) the document is closed. Thus, when
document.write() is called on an already closed document it first calls
document.open() which calls document.clear() (like using fopen with "w" as
the mode). Subsequent document.write() calls on the same document then find
the document is already opened and thus append. I'd need to check mdn again
but I'm not sure if there's also a document.close() call as well. All of
this sits below the DOM manipulation calls such that the DOM needs to be
adjusted after document-altering calls I think. I'm not sure how close this
is to what's currently implemented (especially after the last change) in
practical terms (e.g. if anyone widely uses document.open etc).
Cheers,
Adam.