chaesloc's profile - activity

Where can one file a feature request?

And shouldn't that place be linked with a big red flashing banner from the Sage homepage?

"How would the inclusion of the module be better than say including the small groups database" - the two are not incompatible. The module uses a negligible amount of disk space, is loaded manually, does not break anything, and is trivial to package (i.e. copy its folder to gap folder). So I see no downsides to including it. The upside is that if one normally uses the database, but stumbles upon a group not in it, there will be a workaround. So I guess my main reason is "for completeness".

You program is minimizing y above those lines. The problem is not unbounded only because of the additional constraints 0 < p < 1 given by the top 2 rows of G (=see below).

Using the notation in the article you linked, linear_program(c,G,h) minimizes c * x given G * x < h. You have c=[0,1] and x=[p,y], so you are minimizing y. You want c=[0,-1].

You have G * x = [-p,p,-p-y,-3p-y] < h, which gives y>-h[2]-p, y>-h[3]-3p. You want G*x=[-p,p,p+y,3p+y] and h=[0,1,1,2].

So you want c=[0,-1]; G=[[-1,0],[1,0],[1,1],[3,1]]; h=[0,1,1,2]. As for M, I have no clue what it is.

linear_program(vector([0,-1]),matrix([[-1,0],[1,0],[1,1],[3,1]]),vector([0,1,1,2]))['x'][0]
-7.08356432755e-09

I could not find any Sage function that constructs all groups of a certain (not necessarily small) order - is there one?

There is one for GAP, but it is in the grpconst module, which is not included by default in Sage. Do you find it reasonable to request that it is included by default? After all, the Sage installation instructions recommend installing the GAP small groups database, but that database was created with the module in question.

@Evgeny, it means that the user can decide whether pressing 'enter' (or 'return') posts the comment or has its usual functionality of creating a new line. And on that note, I just noticed that even if I paste a new line character, it works in the editing field, but does not get displayed in the actual comment.
s also do not display properly. Presumably this is to keep comments short, but there should be some way to make a long answer to an answer (rather than to the question).

Hmm, it hadn't occurred to me that feature requests can be classified as bugs. But the point is, many people will google for 'sagemath feature request' (as I did), find no hits, and give up. Of course, if you would rather have feedback only from people who use Sage a lot, the current setup is perfect.

That is precisely what I had in mind. I'd imagine that it isn't too difficult for people like you to paste such requests in a file and submit them together through the appropriate channel, say, once a month. And so far I have the impression that you read a very significant proportion of the posts here. Anyway, thanks! :)

TBH, I can't be bothered for such a small thing. And since most people will not be bothered either, this (and similar 'issues') will likely remain unfixed for some time. The only solution I see is to make bug reporting easy - like a quick post here. I am not trying to criticise anyone or anything, just offering the view of someone much less involved than you.

The canonical place is to go to the Trac server and do it there.

Hints:

• Make sure to do a good search for things first!
• Now you will have to email someone to get an account there; we were getting too much spam.

The Sage support list and Sage devel list are also good places.

As to the banner, we figure people should first ask places like this first, in case it turns out we do have the feature in question :) But we love those requests - and bug reports.

Thanks; I have "marked this answer as favourite", but what I actually wanted was to "accept" it - a tooltip bug? Incidentally, pressing 'enter' submits the comment, while adding a new line seems more convenient.