<html xmlns:v="urn:schemas-microsoft-com:vml" xmlns:o="urn:schemas-microsoft-com:office:office" xmlns:w="urn:schemas-microsoft-com:office:word" xmlns:m="http://schemas.microsoft.com/office/2004/12/omml" xmlns="http://www.w3.org/TR/REC-html40">
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8">
<meta name="Generator" content="Microsoft Word 15 (filtered medium)">
<style><!--
/* Font Definitions */
@font-face
{font-family:"Cambria Math";
panose-1:2 4 5 3 5 4 6 3 2 4;}
@font-face
{font-family:Calibri;
panose-1:2 15 5 2 2 2 4 3 2 4;}
/* Style Definitions */
p.MsoNormal, li.MsoNormal, div.MsoNormal
{margin:0cm;
margin-bottom:.0001pt;
font-size:12.0pt;
font-family:"Times New Roman","serif";}
a:link, span.MsoHyperlink
{mso-style-priority:99;
color:blue;
text-decoration:underline;}
a:visited, span.MsoHyperlinkFollowed
{mso-style-priority:99;
color:purple;
text-decoration:underline;}
p.Code, li.Code, div.Code
{mso-style-name:Code;
mso-style-link:"Code Char";
margin-top:0cm;
margin-right:0cm;
margin-bottom:0cm;
margin-left:36.0pt;
margin-bottom:.0001pt;
font-size:10.0pt;
font-family:"Courier New","serif";}
span.CodeChar
{mso-style-name:"Code Char";
mso-style-link:Code;
font-family:"Courier New","serif";}
span.EmailStyle19
{mso-style-type:personal;
font-family:"Calibri","sans-serif";
color:windowtext;}
span.EmailStyle20
{mso-style-type:personal-compose;
font-family:"Calibri","sans-serif";
color:windowtext;}
.MsoChpDefault
{mso-style-type:export-only;
font-family:"Calibri","sans-serif";
mso-fareast-language:EN-US;}
.MsoPapDefault
{mso-style-type:export-only;
margin-top:6.0pt;
margin-right:0cm;
margin-bottom:6.0pt;
margin-left:0cm;}
@page WordSection1
{size:612.0pt 792.0pt;
margin:72.0pt 72.0pt 72.0pt 72.0pt;}
div.WordSection1
{page:WordSection1;}
--></style><!--[if gte mso 9]><xml>
<o:shapedefaults v:ext="edit" spidmax="1026" />
</xml><![endif]--><!--[if gte mso 9]><xml>
<o:shapelayout v:ext="edit">
<o:idmap v:ext="edit" data="1" />
</o:shapelayout></xml><![endif]-->
</head>
<body lang="EN-GB" link="blue" vlink="purple">
<div class="WordSection1">
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US">Iavor, Adam, Eric<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US">I’m happy with the general direction of the plugins stuff, so I’m mostly going to leave it to you guys to plough ahead; but I am happy to respond to questions.<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US"><o:p> </o:p></span></p>
<p class="MsoNormal" style="margin-left:36.0pt"> * I still think it would be better to provide an escape hatch to the<br>
TcS, not merely the TcM, alongside the nice TcPluginM wrapper. Notably,<br>
Simon's new TcFlatten.unflatten needs TcS...<o:p></o:p></p>
<p class="MsoNormal" style="margin-left:36.0pt"><o:p> </o:p></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US">It think the only reason for this is that ‘unflatten’ needs to set evidence bindings, which in turn requires access to tcs_ev_binds. I think that everything else
is in TcM. So I suppose you could carry around the EvBindsVar if you really didn’t want TcS. (And I can see why you wouldn’t; TcS has a lot of stuff you don’t need.)<o:p></o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US">Simon<o:p></o:p></span></p>
<p class="MsoNormal" style="margin-left:36.0pt"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US"><o:p> </o:p></span></p>
<p class="MsoNormal"><span style="font-family:"Calibri","sans-serif";mso-fareast-language:EN-US"><o:p> </o:p></span></p>
<div style="border:none;border-left:solid blue 1.5pt;padding:0cm 0cm 0cm 4.0pt">
<div>
<div style="border:none;border-top:solid #E1E1E1 1.0pt;padding:3.0pt 0cm 0cm 0cm">
<p class="MsoNormal"><b><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif"">From:</span></b><span lang="EN-US" style="font-size:11.0pt;font-family:"Calibri","sans-serif""> Iavor Diatchki [mailto:iavor.diatchki@gmail.com]
<br>
<b>Sent:</b> 10 November 2014 19:15<br>
<b>To:</b> Adam Gundry<br>
<b>Cc:</b> ghc-devs@haskell.org; Simon Peyton Jones<br>
<b>Subject:</b> Re: Typechecker plugins: request for review and another workflow question<o:p></o:p></span></p>
</div>
</div>
<p class="MsoNormal"><o:p> </o:p></p>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Hi,<o:p></o:p></p>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
On Mon, Nov 10, 2014 at 1:31 AM, Adam Gundry <<a href="mailto:adam@well-typed.com" target="_blank">adam@well-typed.com</a>> wrote:<o:p></o:p></p>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<br>
On the subject of that "nearly", I'm interested to learn whether you<br>
have a suggestion to deal with unflattening, because the interface still<br>
works with flat constraints only. Simon's changes should make it more<br>
practical to unflatten inside the plugin, but it would be far easier (at<br>
least for my purposes) if it was simply given unflattened constraints. I<br>
realise that this would require the plugin loop to be pushed further<br>
out, however, which has other difficulties.<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Not sure what to do about this. With the current setup, I think either way, the plugin would have to do some extract work. Perhaps we should run the plugins on the unflattened constraints, and leave to the plugins to manually temporarily "flatten" terms
from external theories? For example, if the type-nat plugin saw `2 * F a ~ 4`, it could temporarily work with `2 * x ~ 4`, and then when it figures out that `x ~ 2`, it could emit `F a ~ 2` (non-canonical), which would go around again, and hopefully get fully
simplified.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:12.0pt;margin-left:0cm">
<br>
A few other issues, of lesser importance:<br>
<br>
* I still think it would be better to provide an escape hatch to the<br>
TcS, not merely the TcM, alongside the nice TcPluginM wrapper. Notably,<br>
Simon's new TcFlatten.unflatten needs TcS...<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
I don't mind that but, if I recall correctly, to do this without more recursive modules, we had to split `TCSMonad` in two parts, one with the types, and one with other stuff. Eric and I did this once, but we didn't commit it, because it seemed like an orthogonal
change.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:12.0pt;margin-left:0cm">
* Is there a way for my plugin to "solve" a given constraint (e.g. to<br>
discard the uninformative "a * 1 ~ a")?<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Yes, you'd say something like: `TcPluginOK [(evidence, "a * 1 ~ a")] []`<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
The first field of `TcPluginOK` are things that are solved, the second one is new work.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:12.0pt;margin-left:0cm">
* It is unfortunately easy to create infinite loops by writing plugins<br>
that emit wanteds but make no useful progress. Perhaps there should be a<br>
limit on the number of times round the loop (like SubGoalDepth but for<br>
all constraints)?<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Indeed, if a plugin keeps generating new work, we could go in a loop, so maybe a limit of some sort is useful. However, note that if the plugin generates things that are already in the inert set, GHC should notice this and filter them, so we won't keep going.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
* Perhaps runTcPlugin should skip invoking the plugin if there are no<br>
wanteds?<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Oh, there is an important detail here that needs documentation! GHC will call the plugin twice: once to improve the givens, and once to solve the wanteds. The way to distinguish the two is exactly by the presence of the wanteds.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Why might you want to improve the givens? Suppose you had something like `x * 2 ~ 4` as a given: then you'd really want the plugin to generate another given: `x ~ 2`, as this is likely to help the rest of the constraint solving process.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<br>
* The use of ctev_evar in runTcPlugin is partial, and fails with a<br>
nasty error if the plugin returns a non-wanted in the solved constraints<br>
list. Would it be worth catching this error and issuing a sensible<br>
message that chastises the plugin author appropriately?<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Aha, good idea. Bad plugin! :-)<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:12.0pt;margin-left:0cm">
* Finally, I presume the comment on runTcPlugin that "The plugin is<br>
provided only with CTyEq and CFunEq constraints" is simply outdated and<br>
should be removed?<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Yep, thanks!<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
<blockquote style="border:none;border-left:solid #CCCCCC 1.0pt;padding:0cm 0cm 0cm 6.0pt;margin-left:4.8pt;margin-right:0cm">
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Apologies for the deluge of questions - please take them as evidence of<br>
my eagerness to use this feature!<o:p></o:p></p>
</blockquote>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
Thanks for your feedback! Also, if you feel like doing some hacking please do so---I am quite busy at the moment so I don't have a ton of time to work on it, so any help you be most appreciated. I know Eric is also quite keen on helping out so we can just
coordinate over e-mail.<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
-Iavor<o:p></o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p> </o:p></p>
</div>
<div>
<p class="MsoNormal" style="mso-margin-top-alt:6.0pt;margin-right:0cm;margin-bottom:6.0pt;margin-left:0cm">
<o:p></o:p></p>
</div>
</div>
</div>
</div>
</div>
</div>
</body>
</html>