<div dir="ltr">You can splice together conduit pieces with set screw couplers. They're cheap, stable under compression loads, only require a screwdriver to operate, and produce no toxic gases.<div><br></div><div><a href="https://www.homedepot.com/p/Halex-3-4-in-Electrical-Metallic-Tube-EMT-Set-Screw-Coupling-25-Pack-62807B/202288563">https://www.homedepot.com/p/Halex-3-4-in-Electrical-Metallic-Tube-EMT-Set-Screw-Coupling-25-Pack-62807B/202288563</a></div><div><div><br></div><div>-- rec --</div></div></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Fri, Sep 20, 2019 at 8:00 PM Carl Tollander <<a href="mailto:carl@plektyx.com">carl@plektyx.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"><div dir="auto"><div dir="auto"></div>Welding galvanized steel without proper respirators (even outdoors) can kill you. Research this carefully.<div dir="auto"><br></div><div dir="auto">How about some nice thick wall pvc?</div><div dir="auto"><br></div><div dir="auto">Carl</div><div dir="auto"><br><div dir="auto"><div class="gmail_quote" dir="auto"><div dir="ltr" class="gmail_attr">On Fri, Sep 20, 2019, 17:48 Steven A Smith <<a href="mailto:sasmyth@swcp.com" rel="noreferrer" target="_blank">sasmyth@swcp.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">Gary -<br>
<br>
I understand better now...<br>
<br>
I definitely agree that the *most* naive eyeballing methods can be<br>
excruciatingly wasteful.<br>
<br>
I presume that your conduit length requirements are not precise... that<br>
you might be designing them to allow for leaving the window partially<br>
open but otherwise not subject to intrusion or compromise? That seems<br>
to complicate the problem but may pose opportunities. In particular,<br>
*I* might be looking for solutions which leave me with a *minimum* of<br>
leftover conduit by making them longer than their shortest possibles in<br>
some cases. Or looking at it the other way, even if you don't need to<br>
leave the windows open much when "locked" a more complete use of the<br>
material might be obtained by relaxing the length a little without<br>
compromising security (if a given window can only be opened a few inches<br>
for example).<br>
<br>
I will be interested in hearing the results of whatever optimization (or<br>
satisficing) method you use yields.<br>
<br>
- Steve<br>
<br>
PS. regarding guerin's solution, an alternate would be to measure as<br>
suggested, then cut naively until the remaining spaces are larger than<br>
the remaining pieces. Only *then* does one break out the welder and<br>
begin to piece together as-needed. I don't think these are equivalent.<br>
It also occurs to me that *2* pieces of conduit (end to end, unwelded)<br>
in a window channel might be *nearly* as effective as a single piece,<br>
albeit less elegant?<br>
<br>
> Hey Steve. The actual project is nothing elaborate. My house has a<br>
> couple or three dozsen horizontally sliding windows with pretty weak<br>
> locks. Since I've had a couple of break-ins in the past, I decided<br>
> that the easiest way to shore up security for that aspect of the house<br>
> is to just cut short pieces of 3/4 inch conduit to lay horizontally in<br>
> the spaces where the windows slide. When I want to open a window, I<br>
> will just stand its conduit piece up, and when I want to "lock" it<br>
> again, just lay it back horizontally. I asked on FRIAM because instead<br>
> of just eyeballing it and having lots of extra (even potentially<br>
> useful in the future) pieces left over, I'd rather use my (and<br>
> FRIAM's) brain to look at possible ways of optimizing this. Kind of<br>
> fun actually putting my mind to something for a change (retirement can<br>
> be boring if you're not careful).<br>
><br>
> On Fri, Sep 20, 2019 at 5:55 PM Steven A Smith <<a href="mailto:sasmyth@swcp.com" rel="noreferrer noreferrer" target="_blank">sasmyth@swcp.com</a>> wrote:<br>
>> Gary -<br>
>><br>
>> I *patently don't* recommend my method, though it does have some<br>
>> charms. I recently was faced with a similar problem to yours where I<br>
>> needed to cut and install trim around the perimeter of the room (with<br>
>> door openings) I just layed hardwood floor in.<br>
>><br>
>> Rather than go into it in detail (I already did that and realized it was<br>
>> a TL;DR as usual, so cut it) I will just say that I approach these<br>
>> problems as *satisficing* and *constraint* problems rather than<br>
>> *optimization*. Once I had a candidate layout, I simply looked at the<br>
>> results and determined that the *waste* was acceptable. Depending on<br>
>> the circumstances I sometimes prefer to have for example, 2 3' leftovers<br>
>> rather than 1 5' leftover, other times, vice-versa, depending on how I<br>
>> might use said leftovers in some future application (or hedging against<br>
>> a mistake in my measuring/cutting).<br>
>><br>
>> Care to share what your actual conduit/pipe project is?<br>
>><br>
>> - Steve<br>
>><br>
>><br>
>>> Thanks for the links, Peter. I will probably use that software or<br>
>>> similar, to get a quick solution, then look at the MOOCs.<br>
>>><br>
>>> On Fri, Sep 20, 2019 at 2:52 PM Pieter Steenekamp<br>
>>> <<a href="mailto:pieters@randcontrols.co.za" rel="noreferrer noreferrer" target="_blank">pieters@randcontrols.co.za</a>> wrote:<br>
>>>> Two possible approaches are:<br>
>>>> a) Solve the problem yourself. Use one or a combination of standard algorithms ( eg you mentioned linear programming and greedy algorithms, there are many more of course) and/or your own custom algorithm. If you wish to go this route and want to learn about the subject, I recommend the series of MOOCS by Stanford's Tim Roughgarden <a href="https://www.coursera.org/specializations/algorithms" rel="noreferrer noreferrer noreferrer" target="_blank">https://www.coursera.org/specializations/algorithms</a><br>
>>>> Or, I think yours is probably a knapsack -type problem and the MOOC <a href="https://www.coursera.org/learn/discrete-optimization" rel="noreferrer noreferrer noreferrer" target="_blank">https://www.coursera.org/learn/discrete-optimization</a> covers that relatively well.<br>
>>>> b) But if you just want to get the solution you can use optimization software like <a href="https://www.ibm.com/za-en/products/ilog-cplex-optimization-studio" rel="noreferrer noreferrer noreferrer" target="_blank">https://www.ibm.com/za-en/products/ilog-cplex-optimization-studio</a> (they have a free edition that will be good enough for your application) will solve it for you without you necessarily knowing how the software does it.<br>
>>>><br>
>>>> On Fri, 20 Sep 2019 at 21:00, Gary Schiltz <<a href="mailto:gary@naturesvisualarts.com" rel="noreferrer noreferrer" target="_blank">gary@naturesvisualarts.com</a>> wrote:<br>
>>>>> I'd like advice on possible ways to solve the following problem<br>
>>>>> (plumbers must surely face this all the time). I need to cut a set of<br>
>>>>> metal tubes of varying lengths from standard length (6 meter)<br>
>>>>> galvanized conduit stock. The goal is to find the number of tubes I<br>
>>>>> need to buy, and the order of cuts to produce the minimum amount of<br>
>>>>> leftover, unused tube. I'm interested in what types of solutions<br>
>>>>> people use for similar 1-dimensional problems, e.g. linear<br>
>>>>> programming, greedy algorithms, etc. (I've been Googling). I'm only<br>
>>>>> looking to cut around 15-25 pieces, so my gut feeling is that an<br>
>>>>> exhaustive search of all possible solutions, though probably NP-hard,<br>
>>>>> would be feasible to perform. Working programs, as well as libraries<br>
>>>>> in any language would be a bonus.<br>
>>>>><br>
>>>>> ============================================================<br>
>>>>> FRIAM Applied Complexity Group listserv<br>
>>>>> Meets Fridays 9a-11:30 at cafe at St. John's College<br>
>>>>> to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
>>>>> archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
>>>>> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
>>>> ============================================================<br>
>>>> FRIAM Applied Complexity Group listserv<br>
>>>> Meets Fridays 9a-11:30 at cafe at St. John's College<br>
>>>> to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
>>>> archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
>>>> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
>>> ============================================================<br>
>>> FRIAM Applied Complexity Group listserv<br>
>>> Meets Fridays 9a-11:30 at cafe at St. John's College<br>
>>> to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
>>> archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
>>> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
>><br>
>> ============================================================<br>
>> FRIAM Applied Complexity Group listserv<br>
>> Meets Fridays 9a-11:30 at cafe at St. John's College<br>
>> to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
>> archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
>> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
> ============================================================<br>
> FRIAM Applied Complexity Group listserv<br>
> Meets Fridays 9a-11:30 at cafe at St. John's College<br>
> to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
> archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
> FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
<br>
<br>
============================================================<br>
FRIAM Applied Complexity Group listserv<br>
Meets Fridays 9a-11:30 at cafe at St. John's College<br>
to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer noreferrer noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
archives back to 2003: <a href="http://friam.471366.n2.nabble.com/FRIAM-COMIC" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam.471366.n2.nabble.com/<br>
FRIAM-COMIC</a> <a href="http://friam-comic.blogspot.com/" rel="noreferrer noreferrer noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
</blockquote></div></div></div></div>
============================================================<br>
FRIAM Applied Complexity Group listserv<br>
Meets Fridays 9a-11:30 at cafe at St. John's College<br>
to unsubscribe <a href="http://redfish.com/mailman/listinfo/friam_redfish.com" rel="noreferrer" target="_blank">http://redfish.com/mailman/listinfo/friam_redfish.com</a><br>
archives back to 2003: <a href="http://friam.471366.n2.nabble.com/" rel="noreferrer" target="_blank">http://friam.471366.n2.nabble.com/</a><br>
FRIAM-COMIC <a href="http://friam-comic.blogspot.com/" rel="noreferrer" target="_blank">http://friam-comic.blogspot.com/</a> by Dr. Strangelove<br>
</blockquote></div>