Reader Posts
« previous | TPM CAFÉ READER POSTS HOME | next »
Please Do Not "Recommend" This Post.
I hesitate to write this, for it risks pushing some important posts off the recently-added list. However, I think it is important enough to warrant a small discussion, particularly since I believe it will get no movement from TPM site administrators without significant support from the TPM reader community.
Genghis has recently lamented the overall lack of quality in many of the reader posts. The causes for this are undoubtedly numerous, and the solutions are not immediately evident. However, as I noted in a comment to Genghis' post, one possible idea that might help is capping the number of recommendations allowed per reader over some time frame. Each time a reader would like to recommend a particular post, an e-mail would be sent to his/her valid address with a link requesting confirmation of the recommendation. Maybe 2-10 such recommendations per week.
Such a system is most definitely not immune to abuse, but at a minimum, its implementation might help elicit a bit more thought on the part of genuine TPM readers before recommending any and all posts. The very ease of making recommendations is perhaps the single largest contributing factor to part of the problem that Genghis describes - namely the part relating to sub-standard posts on the Most-Recommended list. Will the larger overall problem of post quality be corrected by this action? Probably not. But perhaps it would help serve as a staring point.
I write this here because I do not see how TPM site administrators will ever find it profitable to spend time and money on implementing such a system unless there is overwhelming calls from the TPM readership to do so.
Wondering if there is some sort of mechanism by which we simple readers could form a petition or just generally describe change requests such as this to TPM administrators. I don't immediately see any kind of "Contact Us" info on the site.
If you liked this post, please do NOT recommend it.
If you hated it, please don't recommend it anyway.











Comments (15)
As I mentioned in the other thread, currently you don't have to be registered to recommend which would interfere with your solution. TPM didn't make this a requirement for a reason, but I don't know why. Perhaps to make the site more inclusive. I wonder how many non-members recommend and what kind of posts they recommend. This is something that should be easy for the staff to figure out. If non-members tend to recommend what the staff consider to be crap posts, then I say drop their recommendation "privileges". That in itself might help a lot.
But even if they didn't do that, they could enforce the limit with cookies (though not email). You can get around cookies but most people either don't know how or couldn't be bothered for the sake or recommending someone extra posts.
May 2, 2008 5:09 PM | Reply | Permalink
Probably, yes.
Forgive me if I overstep the bounds of decorum - I seem to have detected a grain of either software developer or IT administrator sprinkled among your other posts and comments. Assuming this is correct, I hereby delve into the techo-jargon:
I'm not sure ookies are the way to go. Yes, the average computer user might not know how to easily delete them. However, I submit 1) most TPM readers and bloggers in general are above-average users. 2) Many IT departments, proxy-servers, browsers, etc. are set by default to either reject permanent cookies or delete at them end-of-session anyway. 3) The coding effort to place, then read and parse such cookies on the server-side is not too much less than implementing an e-mail verification scheme.
To put some anecdotal metrics on it, I am a mathematician by training, but have been engaged in software development and server maintenance for the better part of 5 years. Was once able to implement such a user-based e-mail verification scheme in just over a day! Surely some hot-shot webmaster could implement something like this in a matter of hours, no? Basically, I guess I'm making the argument that maybe registration should be required for recommendation access. Thoughts?
May 2, 2008 5:20 PM | Reply | Permalink
No need to jump to conclusions. It says so in my profile.
Obviously you wouldn't want to use cookies for a critical function. But what's the worst that can happen here? So someone recommends more than their share of 5 posts. As long as most people don't do it, the purpose is served. Personally, I would rarely be bothered to delete my cookies just so I could recommend someone's posts.
You wouldn't need email. Just make people register and pop up a dialog warning when they run out of recs. Registering doesn't even require email addresses. TPM doesn't collect them. I'm not sure what the email gets you.
May 2, 2008 5:36 PM | Reply | Permalink
Your cookie suggestion could very well be yummy enough to do the trick. To me it just doesn't seem forceful enough, but I am admittedly scarred by many battles with Black Hats.
In my experience, the only thing e-mail verification buys you is a slightly higher technical bar to leap before abuse can occur. It is one thing for a user to know how to delete a cookie. Typically harder though for the same user to create multiple e-mail boxes and a bot to reply to each one.
I do like this idea, though:
Just make people register and pop up a dialog warning when they run out of recs. Registering doesn't even require email addresses.
Perhaps, if you are not completely turned-off by the idea of required registration (though I would understand if you are), e-mail would NOT in fact be necessary - the number of "recommendations" from each user could simply be stored server-side; wouldn't need cookies at all. Still would not protect against outright fraudulent abuse, but might serve to implement a cap for legitimate readers, yet with a minimum of user headache.
Now, for the $64 question: How do we get this suggestion (and others) to TPM? Wondering if maybe I should write a corollary "Please Recommend THIS post." with no other goal than to establish a ready avenue to provide general feedback to TPM. Am I blind? I remain unable to find a good way to submit requests.
May 2, 2008 6:03 PM | Reply | Permalink
Genghis is right.
Rule #1 of software design: first try the simplest thing that could possibly work.
Rule #2 - Fix the bugs from your attempt at rule #1.
Only then, if you still have a problem, should you consider doing something more complex.
May 2, 2008 6:05 PM | Reply | Permalink
I disagree. Rule #1 is don't write code that has no purpose. Putting in the cookies changes nothing.
May 2, 2008 10:14 PM | Reply | Permalink
I think -- though I haven't fully vetted this -- that the current recommend link functionality is restricted by IP address rather than cookies or account. (It may also be some sort of combination.) Don't see why that couldn't be extended for a total number of "recommends" rather than just one per post.
May 3, 2008 2:02 AM | Reply | Permalink
Not only will I not recommend it, I will not admit to having read it. If you read a comment here purporting to be by me, I will deny it. I'm actually in Bosnia at this minute dodging sniper fire while trying to fill a gas tank with cheap Bosnian gas to bring back to North Carloina and Indiana. I'm sleep-deprived too.
But, serious question, if you want people to read your suggestion, doesn't it have to stay on the page long enough for people to read it? Isn't that what recommend does?
Would Genghis be amenable to this post replacing his "Obama get a SuperDel" post (which seems antithetical to his other post waxing poetic on better reader posts, where if I not mistaken he mentions Josh getting and posting news like Obama's speech or Obama get a SuperDel)?
sorry... INCOMING!!!!! Where's my Magnaquench magnet? Ah, drat... in China....
May 2, 2008 5:15 PM | Reply | Permalink
Momentary dyslexia and grammar failure TWICE: should have been "Obama gets a SuperDel."
spell check would be nice.
May 2, 2008 5:23 PM | Reply | Permalink
Thank you for your non-recommendation.
Though I see now that at least one person has broken the rules and recommended the post anyway. Alas.
Frankly, it would be inaccurate to say I don't care about the topic (I DID post about it after all). But I most definitely don't care about it enough to potentially push other, much more newsworthy and well-researched posts off the Most-Recommended list. Further, I have about zero expectation for TPM to do anything about it anyway. And lastly, recommending this post would serve as prime example of just the sort of garbage to which Genghis referred.
If it really means that much to me, I'll go and implement my own left-leaning blog site, and blow TPM off the internet with my superior interface (don't think the thought hasn't crossed my mind).
May 2, 2008 5:30 PM | Reply | Permalink
Wasn't me
May 2, 2008 5:37 PM | Reply | Permalink
Jade. I encourage you to click the super del link. All will be made clear.
May 2, 2008 5:38 PM | Reply | Permalink
No mails, please... I'm already getting more than I want to handle. I really don't need more.
May 2, 2008 6:06 PM | Reply | Permalink
BTW is this an experiment in reverse psychology? Tell people not to do something and watch how many will, just because you told them not to?
May 2, 2008 6:11 PM | Reply | Permalink
Codegen. Did you hit the recommend button? Coooddegennn?!
Goddammit. Hands off the mouse. Hands off, I say!
May 2, 2008 6:17 PM | Reply | Permalink
Post a Comment