03:23 < sgupta1> rcurtin: okay I'll have a look.

06:21 < sgupta> rcurtin: I gave read-write permission to all u,g,o and change the ownership of /usr directory to mlpack user. And, it seems to work fine.

13:13 < rcurtin> sgupta: that seems like a little bit of a hack, I think instead you should investigate why boost isntalled with such weird permissions

14:52 < rcurtin> lozhnikov: any chance you can send a link to the DVK-3M you found for sale for $250? I think I would like to use that data point as bargaining fodder :)

14:58 < lozhnikov> rcurtin: I found the advertisement here: http://zx-pk.com/forum/viewtopic.php?f=35&t=6455

15:00 < lozhnikov> here a buyer suggests ~250$ http://zx-pk.com/forum/viewtopic.php?f=35&t=6455&start=20#p73076

15:00 < rcurtin> ah, okay, thank you so much

15:01 < lozhnikov> and the seller accepts the suggestion in the last comment

15:03 < rcurtin> google translate is so much improved compared to ten years ago... full page translate now allows me to understand some of what is going on

15:04 < rcurtin> it used to be so bad, the translations you would get would be nonsense

15:07 < lozhnikov> yeah, they improved the quality significantly

15:07 < lozhnikov> rcurtin: I forgot to say that the starting price was ~450$

15:15 < rcurtin> yeah, I saw that the original price was 26k rubles

15:30 < sgupta> rcurtin: okay I'll look at it.

15:59 < rcurtin> I did a quick calculation, if all the remaining benchmark jobs for shogun time out at 9000 seconds, assuming each only runs once, then the shogun benchmark will finish in about 5.5 days...

16:00 < zoq> hm, I guess if they time out anyway, there is no good reason to wait

16:01 < zoq> I guess, it makes sense to look into the issue and restart the benchmark later

16:03 < rcurtin> would you like to do that? I'm happy to wait a few more days

16:03 < rcurtin> but definitely if we wanted to kill it and look into the failures that is ok too

16:03 < rcurtin> there's no guarantee that the remaining jobs will fail---my estimation was only a worst-case guess (and it's only right if each method runs only once, which I think is the case)

16:07 < zoq> Not sure I have the time to look into the issue today or tomorrow, so I don't mind to wait. On the other side if someone has time to look into it, I think there is no good reason to wait.

16:07 < zoq> I know, you are waiting for a restart.

16:13 < rcurtin> yeah, but no hurry on that

16:13 < rcurtin> 5.5 days is fine with me, if it was a month I would do something about it :)

16:13 < zoq> okay :)