This thread is used for opinion and usage polls around Wolfram Mathematica. The poll questions are added as answers to this question. The choices for each specific poll can be found as comments below them. To participate simply up-vote the comments which apply for you. If there is no suitable ans...
@m_goldberg I copied the community poll from the TeX SE site. There, they have a separate meta-post for discussion things about the poll itself, if this was your question.
@m_goldberg Since you already participated and edited in the poll post, I thought I gave you the original reference so that you can look how they do it over at tex.se.
@halirutan. My question: What is the right way to address an issue if I should have one with a poll? For example, in your poll on code distribution, I can not up vote any of the choices , because I have written one large package. Could you remove "small" from you 2nd choice.
@m_goldberg Hehe.. Indeed, I didn't expect that someone who has written a large package doesn't write packages regularly. I cannot edit the comment any longer, so you are free to decide whether you consider yourself a pro package writer or inexperienced.
@m_goldberg This specific poll was more to get a feeling whether or not packages are used often for self-developed code.
@m_goldberg To give a general comment about the whole post itself: I'm really interested in what the average Mathematica user is. This lets one decide whether or not you should put energy in certain projects.
@m_goldberg The IDEA plugin is one example. Another one is that e.g. under Linux the web-cam support is not working. If one decides to implement something to fix this, it is probably a good idea to know large the group of Linux users is.
Ok, started importing all traffic logs at about 30 min/hour of data in order of most recent not-imported log file. 56,600+ total files, but it should already be worth playing with after a day's worth of data is loaded. Off for a walk and dinner.
I must say I personally kind of like how MMA handles conflicts between variables. Sure, this may be strange
Function[var, var + #] &[var][2] /. var -> 1000 Function[var, var + #] &[var2][2] /. var2 -> 1000
But at least there aren't too many rules involved in predicting what will happen. Also there is the choice between With and ReplaceAll. I'm curious what problems you encountered @LeonidShifrin
@JacobAkkerboom What it shows is that you can't pass pure functions with named arguments and expect it to work correctly in general. And that is a pretty big deal, because when you create closures and pass them around, they can be used pretty far from the place they were created, both in time and space, so to speak.
@LeonidShifrin thank you for the excellent example. It still took me quite some time to figure out exactly what was going on. Maybe I should test my understanding by trying to emulate the rules that monitor this, that would be a nice exercise.
@JacobAkkerboom Since you are interested in this topic, here is another example of a different, but somewhat related scoping issue, where rules also break lexical bindings. Ironically, it only took me 4.5 years to find some decent explanation for this particular issue.
@JacobAkkerboom All this stuff has to do with the disruptive nature of rules, and "over-transparent" nature of Mathematica expressions w.r.t. pattern-matching.
@LeonidShifrin Does this mean that in principle you have to make some core functionality of the system as black box which is fixed and does not follow the same rules as the rest?
Is it me or there was a 200 bounty on that question from ViViD that has been removed while the answer is perfect? How do bounties work exactly? Can they be taken back as wish?
@Öskå Correct. VividD posted the bounty after Simon answered, so it did not qualify. Had he (or anyone else) posted after the bounty and gotten more than 2 votes, then they would've been eligible for at least 50% of the bounty
