<div dir="ltr">markdown and the web math thing came up in the greater context of the conference if memory serves.<div><br></div><div>Yes, a server version of TeX makes sense for a lot of projects.</div><div><br></div><div>William</div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Jul 19, 2019 at 10:00 AM Mike Marchywka <<a href="mailto:marchywka@hotmail.com">marchywka@hotmail.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Fri, Jul 19, 2019 at 09:04:07AM -0400, William Adams wrote:<br>
> As a person who remembers TeXview.app with great fondness I'll note that he developed TeXview.app for NeXTstep which used<br>
> this IPC method<br>
> [<a href="http://ftp.yzu.edu.tw/CTAN/support/hypertex/hypertex/%5Dhttp://ftp.yzu.edu.tw/CTAN/support/hypertex/hypertex/" rel="noreferrer" target="_blank">http://ftp.yzu.edu.tw/CTAN/support/hypertex/hypertex/]http://ftp.yzu.edu.tw/CTAN/support/hypertex/hypertex/</a><br>
> and see:<br>
> [<a href="http://www.tug.org/TUGboat/tb31-2/tb98panel.pdf%5Dhttp://www.tug.org/TUGboat/tb31-2/tb98panel.pdf" rel="noreferrer" target="_blank">http://www.tug.org/TUGboat/tb31-2/tb98panel.pdf]http://www.tug.org/TUGboat/tb31-2/tb98panel.pdf</a><br>
> William<br>
<br>
Thanks, searching for "Tom" and reading his comments those questions still seem relevant although<br>
do you know if "markdown" or whatever the web math thingy is somehow related to this session?<br>
<br>
I played with a server version of "R" and a custom made java server and always wondered if there<br>
is any use for a server version of tex. That might sound dumb as both can be quite slow but the server<br>
I mentioned made heavy use of caching ( it was for custom ads generated in response to<br>
a specific web page irequest and whatever is known about the viewer ). <br>
Similarly with math or even dynamic web content a lot of it is reused anyway and better<br>
caching could make even slow things practical. In my case however if you want to send a tex like<br>
source code to a browser and have it generate dvi or pdf locally client side then caching would not help.<br>
Although apparently from the commentary here a lot of markdown just generates requests for <br>
rendering by a third party and that is quite slow ( beyond roundtrip web latency )<br>
but not sure why if caching is cheap.<br>
<br>
In any case incremental dvi does seem to have been explored before. <br>
<br>
<br>
<br>
> <br>
> On Fri, Jul 19, 2019 at 8:55 AM Mike Marchywka <[mailto:<a href="mailto:marchywka@hotmail.com" target="_blank">marchywka@hotmail.com</a>]<a href="mailto:marchywka@hotmail.com" target="_blank">marchywka@hotmail.com</a>> wrote:<br>
> <br>
> I started looking at tex and got to texmfmp.c, finding these comments,<br>
> /* IPC for TeX. By Tom Rokicki for the NeXT; it makes TeX ship out the<br>
> DVI file in a pipe to TeXView so that the output can be displayed<br>
> incrementally. Shamim Mohamed adapted it for Web2c. */<br>
> #if defined (TeX) && defined (IPC)<br>
> On a few moments of search, I found AUCTEX which has a WYSIWYG<br>
> like feature. This may be an ideal starting point although<br>
> apparently it is all emacs lisp LOL. In any case if tex has<br>
> an IPC output option that may be a big step closer to what I<br>
> was after. Curious if anyone can comment on this socket output<br>
> or incremental dvi or auctex as it may relate.<br>
> Thanks.<br>
-- <br>
<br>
mike marchywka<br>
306 charles cox<br>
canton GA 30115<br>
USA, Earth <br>
<a href="mailto:marchywka@hotmail.com" target="_blank">marchywka@hotmail.com</a><br>
404-788-1216<br>
ORCID: 0000-0001-9237-455X<br>
<br>
</blockquote></div>