by [anonymous]
1 min read12 comments

-7

Our optimization process sucks.

Every now and then someone makes up a thread concerning the optimization of LW, we have a nice discussion, one or the other idea gets a number of upvotes, sometimes quite a lot, but in the end, nothing happens.

What's more, it is predictable that nothing will happen, which kills the motivation to contribute.

  • The LW staff (those who hold the authority to implement major changes!) needs to commit a considerable amount of time for the optimization of LW. If necessary, hire someone and make his main job LW optimization.
  • The people contributing to optimization threads need to get serious feedback from the staff. It is just frustrating to feel that those who are in charge of implementing changes will not even take the time to think my arguments through.

It is just a tragic waste to put so much time and effort into LW, and about nothing into optimizing it.

New to LessWrong?

New Comment
12 comments, sorted by Click to highlight new comments since:

It is just frustrating to feel that those who are in charge of implementing changes will not even take the time to think my arguments through.

Where is that coming from? You've posted exactly one comment (on this username, anyway) about LW changes, earlier today.

Counter-point: A big thank you to Lucas and other contributors (whose names I don't know) who provide maintenance and improvements free of charge!

You are welcome to contribute enough money to hire a full-time LW coder.

[-][anonymous]40

I had rather something like a crowdfunding approach in mind. ( If someone spends spends much time on LW, shouldn't ze be willing to give some money for its improvement ?! )

But judging from the reactions to this thread, it seems that I did overestimate the number of People who think this way.

[-]Emile140

I don't think that's a very accurate description of the situation.

The situation is that there are very few resources allocated to implementing improvements to the website (that requires money, or time and skill and dedication). Occasionally there are threads where people propose a gazillion improvements and, of course, only a tiny fraction of those improvements has chances of being actually implemented. Whether this counts as a "problem" is debatable, as far as I can tell in any community - online or offline - there is much more talk of "things that someone should do" than actual changes implemented.

If you want to see more improvements, fork out some money! Or, learn how to implement changes, the code is open source after all.

[-][anonymous]00

There is much more talk of "things that someone should do" than actual changes implemented.

Agreed. This happens so much that it gets articles written about it. Example:

http://www.tilde-one.com/articles.php?id=230

A news flash that you may have read elsewhere, yet doesn't seem common enough: everyone has dozens of creative ideas. Everyone is the idea guy. Everyone wants to change the world with their fantasical vision that would probably end up being something with the word "Dragon" in the title.

He's talking about game design and not forum code updates, but I think this is a dilemma that applies to a variety of fields.

  • The LW staff (those who hold the authority to implement major changes!) needs to commit a considerable amount of time for the optimization of LW. If necessary, hire someone and make his main job LW optimization.

Money and time do not grow on trees. All of the LW staff have full-time jobs already; hiring a full-time developer would involve either firing a current employee of FHI/CFAR/MIRI or not hiring an additional employee.

The people contributing to optimization threads need to get serious feedback from the staff. It is just frustrating to feel that those who are in charge of implementing changes will not even take the time to think my arguments through.

There is no one person in charge of implementing changes. In general, feature requests are added to this, and then somebody may or may not write the code. Some of the undone feature requests date back to 2009.

The full issue list and code is available online, if you want to try to optimize your proposal for optimizing optimizing LessWrong.

[-]tim60

In addition to the points already raised, you are completely ignoring the many changes that have resulted from LW users raising ideas and/or concerns. Briefly, off the top of my head, the front page and upper menus were redone for more intuitive use, new comments added to a previously viewed thread now have a pink border, links to the latest open thread/rationality diary were added to the sidebar, and the karma display for posts and comments now allows you to see the percent of positive votes.

Your picture of a terrible, stagnant optimization process is inaccurate.

If you want something done that requires effort, it tends to be a good idea to contribute as much as that effort as possible. "I've got a great idea. Could the staff please implement it for me?" will not be as well received as "I've got this great idea, a patch that implements it, and unit tests. Could the staff please approve/reject it and merge it for me?"

(Not that you shouldn't gather feedback on your ideas first)

[-][anonymous]20

Turn it on it's head: make a guide for how to devalue or perhaps ruin Less Wrong. Then, don't do those things to contribute to incremental optimization.

[-]gjm00

As others have said, the only way this will happen is if LW's users make it happen. (Or if MIRI or some other LW-associated entity suddenly gets a big pile of money and can't think of anything more valuable to do with it than paying someone to improve LW.)

To which end, it would be interesting to know: When "unofficial" people have in the past offered to implement LW improvements, what's the response been? Have they done the work? Has it been integrated into the LW site?

It is possible that some potential contributors are put off by the possibility that they'll do a load of work and then find that no one is willing and able and has enough time to look at their changes and put them into action. Does evidence exist that would enable such people to tell how (un)reasonable that concern is?

Only a few recent changes weren't authored by one of the top 4 contributors. That wouldn't discourage me from starting work on a new feature, but I wouldn't spend more than a day or so without testing the waters (e.g. requesting approval for new unit tests).