[NTG-context] lua-widow-control module error in LMTX

[Eduardo Bohoyo]

Erratum: When I wrote script, I meant hyphen.

El 26/4/22 a las 11:45, Eduardo Bohoyo escribió:
> Hi:
>
> No line such as "Widow/orphan detected. Attempting to delete".
>
> I see interleaved new groups with the same line always repeating a 
> warning message throughout the whole file. In short, there are 613 new 
> lines with the message "luatex warning > tex: left parfill skip is gone".
> But I didn't give it any importance, because I interpreted that they 
> could be inherent to the module.
>
> I can see "modules         > 'lua-widow-control' is loaded".
>
> But, luckily, I can also see this:
>
> open source     > level 2, order 4, name 
> '/opt/luametatex/texmf-modules/tex/context/third/lua-widow-control/t-lua-widow-control.mkxl'
> resolvers       > lua > loading file 
> '/opt/luametatex/texmf-modules/tex/luatex/lua-widow-control/lua-widow-control.lua' 
> succeeded
> close source    > level 2, order 4, name 
> '/opt/luametatex/texmf-modules/tex/context/third/lua-widow-control/t-lua-widow-control.mkxl'
> module          > lua-widow-control > Already enabled
>
>
> On the other hand, the distribution and size of my paragraphs take 
> great care that their "design" optimises the module's goodness, 
> except, of course, for the first pages of each of the nine dummy text 
> chapters (they start at a third of a page).
> But it is very curious that, even so, only the two orphan lines I 
> mentioned, and only two of the five widows relate to two or three of 
> those "supposedly problematic" first chapter pages (as I said, nine in 
> total). And the case of the broken scripts is even stranger: only two 
> of the six breakages concern a couple of those early chapter pages.
>
> I will go over the lua-widow-control.pdf document once more in case 
> there is a tiny detail I am missing, but I think, if I keep going at 
> this pace, I will end up learning it by heart.
>
> Well, joking aside, thanks again for your advice, Max.
>
> Edu.
>
> El 26/4/22 a las 4:42, Max Chernoff escribió:
>> On 2022-04-25 6:51 p.m., Eduardo Bohoyo wrote:
>>> First things first.  I want to acknowledge and thank you for the 
>>> tough mission that surely involves maintaining this module for the 
>>> benefit of the TeX community and, most especially, for LMTX in 
>>> particular, due to the very reasons you have just explained.
>>
>> Well thanks :) These days I write most of my documents in LMTX, so 
>> the LMTX
>> support is pretty self-serving -- I'm admittedly surprised that there's
>> another lwc + LMTX user.
>>
>>> Regarding your remarks, you are right in your assumptions: I didn't 
>>> have the required lua file installed in its corresponding folder. 
>>> That's what I was missing, and, logically, what made my compilation 
>>> crash.
>>
>> Yeah, it's a pretty easy mistake to make. For Knuth TeX, shared files go
>> in texmf/tex/generic and Plain-specific files go in texmf/tex/plain, but
>> with LuaTeX there's just texmf/tex/luatex so it gets a little confusing.
>>
>>> Now I finally get the pdf. But unfortunately, this "new" pdf is the 
>>> same with the module uncommented as when I had it commented. There 
>>> is no difference at all. And you're right: even context --make 
>>> doesn't solve the problem. To give you an idea, although my dummy 
>>> document has 78 pages, only 55 can really be said to be dummy text 
>>> that can benefit from the module. Well, only in those 55 pages I 
>>> have 2 orphans, 5 widows and 6 broken hyphens.
>>
>> So the first step here is to check the log file. If you see lines like
>>
>>     module          > lua-widow-control > Widow/orphan detected. 
>> Attempting to remove.
>>     module          > lua-widow-control > Widow/Orphan NOT removed on 
>> page X.
>>
>> then that means that lwc found a widow/orphan, but gave up. This usually
>> only happens if the page has only really short paragraphs, but it can 
>> also
>> happen if there aren't any paragraphs that both start and finish on 
>> the page.
>> I've got some neat graphs for this (see the upcoming TUGboat issue), 
>> but with
>> default settings this should happen for much less than 10% of potential
>> widows/orphans, so it seems unlikely that this is happening for every 
>> page.
>>
>> If this actually is the issue, then you can try raising the 
>> "emergencystretch"
>> value in "\setuplwc", but that's probably going to give terrible 
>> results. The real
>> solution is to rewrite something, but that should usually be pretty 
>> rare.
>>
>> ---
>>
>> If you see lines like
>>
>>     module          > lua-widow-control > Widow/orphan detected. 
>> Attempting to remove.
>>     module          > lua-widow-control > Widow/orphan successfully 
>> removed at paragraph X on page Y.
>>
>> but the widows/orphans weren't actually removed, then something 
>> really weird
>> is going on and definitely means that there's a bug in lwc. Rerun the 
>> document with
>>
>>     \setuplwc[debug=true]
>>
>> immediately after "\usemodule[lua-widow-control]" and either reply 
>> with the log
>> or post a new issue on the lwc GitHub.
>>
>> ---
>>
>> If you only see
>>
>>     modules         > 'lua-widow-control' is loaded
>>
>> but not any other lines mentioning lwc, then something weird is going 
>> on, and
>> lwc is loaded but not enabled. If this is the case, same as last 
>> time: enable
>> debug mode and reply with the log.
>>
>> ---
>>
>> If you don't see
>>
>>     modules         > 'lua-widow-control' is loaded
>>
>> at all, then you're not actually loading lwc. This probably means 
>> that ConTeXt
>> can't find the file, there should hopefully be clues somewhere in the 
>> log file.
>>
>>> Thanks again for your work and yours observations.
>>
>> No problem; glad to help.
>>
>> -- Max