mottershead

Actually, the lifetime subscription for $200 is a very good deal if you can afford it. You are ahead of the game after only eight or nine years, since an annual subscription is $25 and a three-year subscription is $68. You don't have to be planning on living for a 100 years. We just made it that long because our system requires that there be some expiration date, and also (obviously) we don't know how long people's lifetimes will be. But it is really supposed to be more of a "lifetime" subscription.

 

We have in fact had quite a few people sign up for lifetime subscriptions. I guess they want to support photo.net, and maybe there is some interest in just uploading photos and not having to think about renewing a subscription every year to keep them online indefinitely.

It is probably going to be a while for me to fix this problem, because I can't reproduce it. It doesn't affect all visitors with IE, and those who are affected see it only on certain forum threads, and even on those threads, only intermittently.

 

My hypothesis is that it is caused by an interaction between the Javascript code on those pages. Javascript code on photo.net pages comes not only from photo.net but also many of the ads involve Javascript code from third-party ad servers. So, the problem might only appear if there is a particular ad, or combination of ads, on the page. Until I figure out which ads are involved (assuming this theory is true), I can't do much about this. If my theory is right, there is a good chance that the problem will just go away by itself when the offending ad comes to the end of its run.

 

It may be "ridiculous" as Vladimir persists in saying in his comments on every thread where this problem has been reported, but if you have IE and can't live with this problem, then the only real workaround at this point is to use a different browser, such as Firefox, Safari, or Opera. I have the impression so far that the problem is "livable-with", since people report that a page refresh or two generally clears things up. Its easy for me to say that, though, since I haven't seen the problem myself. If you can't live with this problem, I'd like to hear from you.

 

I'm not pointing people towards these other browsers because I think they are better than IE, but because they don't seem to have the problem. Having said that, I do happen to think that these are better browsers, and photo.net is designed for them first, especially Firefox, and then the code is tweaked to workaround IE 6 issues. IE 6 has a low reputation among Web developers for standards compliance, and as more developers (including the people at ad agencies) write code assuming "standards compliant" browsers, we can expect to see more problems involving IE. Previously, people wrote first for IE, and many still adopt that approach. But things seems to have tipped in the last year or two, even though IE is still the dominant browser. That is just my opinion, though, and you are perfectly entitled to think that I don't know anything about browsers.

In a way, the Top Rated Photo pages are a virtual exposition. The rating system is the mechanism by which a group of photo.net members select photos in the Gallery for exhibition. The group happens to be all members of photo.net.

No, Carl. A vote is a system for a group ("voters") to make a choice. The outcome of a vote is that one or more of the options presented are "chosen". There are a large number of voting systems, not all of which are familiar to voters in American elections for public office. It is quite correct to regard the photo.net photo rating system as a vote.

 

As for your "ratings system at work" example, those raters are telling you that they aren't interested in an abstract composition involving masses of color -- the way you are. They are telling you that, for them, a blurry picture with a pink cast of some reflective balls on pedestals is weird and uninteresting. You might have some reason in your head why you believe this photo should be regarded as interesting and/or beautiful, but it isn't getting through to these raters. So what?

"Bandwidth" is an ambiguous term. ISP's and web-hosting concerns use it to refer to how much data you are sending and receiving per period of time. You need to look at your agreement. There are several different schemes for charging for bandwidth. The most common one in entry-level situations is to specify the total amount of data that you can send and receive per time period. Usually this is specified as Megabytes or Gigabytes per month. Both input and output are included, usually, but you might have two rates, one for input and one for output.

 

If you reach the cap, various things can happen. It sounds like in your case, they are just letting you go over and charging you for "overage". This isn't good for you, because as you learned it means that you can be surprised by a large bill at the end of the month, before you even realized how much "bandwidth" you were actually using. This is a dangerous situation. One robot being released on your site sucking down a lot of files, or your site being "slashdotted" (or the equivalent) can suddenly result in your using a lot of "bandwidth" and being hit with serious charges, on the order of thousands of dollars. People have been bankrupted by this.

 

A better scheme, unless you know what you are doing and are controlling your bandwidth usage, is to have some level up to which you will have a flat charge, and another level up to which you can go with "overage" changes, but above that the ISP just drops the packets and you are capped, like a circuitbreaker. This means that you know the minimum and maximum that you can be charged per month.

 

The slight saving feature here is that your connection probably doesn't have that much capacity. On a residential circuit, the capacity is generally fairly low. Also, usually the capacity of the uplink is much lower than the capacity of the downlink in residential circuits, at least in the U.S. You can only physically pump out so much data. And the computer you are using as a server also constrains how much data you can pump. That said, you'd be surprised how much data a consumer PC can pump out over a month on a residential circuit if you don't watch it, and in the worst case you could get hit with a shocking bill.

 

You're currently in a fairly exposed situation. If I were you, I'd turn my server off until I had figured this out and setup some kind of "circuitbreaker" so you don't get hit with a charge for hundreds or thousands of dollars, out of the blue, at which point it will be too late.

I have done most of the coding for the last 4 years, although Rajeev Surati and Patrick Hudepohl have helped with certain things. Before me, beteen 1999 and 2001, there were several people involved in coding. Also, photo.net is based on a large body of code called ArsDigita Community System (ACS), developed originally by Philip Greenspun, and later by a company he founded, ArsDigita. photo.net forked from ACS (version 3.2) about seven years ago and a lot of the ACS stuff in photo.net has since been changed or removed. But there is still a lot of it left, too, especially at the foundations.

What's hard to understand? This person recognizes that rating has value to the site and enjoys doing it; so he rates many photos. He knows from experience that he doesn't enjoy receiving ratings, and the benefits of submitting photos for rating don't equal the costs, for him. So he submits photos "critique only", and just uses the site to exhibit his photos and receive comments. Seems perfectly logical.

 

As for there being 2 and 3 ratings in his distribution. Of course there are. Those are valid rating values, and we want people to use the full range. Out of 6600 images he is saying that less than 2% of them are very much below average and that about 15% of them are below average. Actually, he sounds a little overgenerous to me, since it doesn't make sense that only 14% of the photos he is looking at are below average. Either he is making a point of rating mostly the better photos, or he could stand to increase the number of 2's and 3's that he is giving out.

 

You seem to have this idea that people shouldn't "judge" others unless they are prepared to be "judged" by others. But that isn't how this site is set up. Judging isn't the proper metaphor. The better metaphor is voting, you can vote here, but nothing compels you to be a candidate. Everybody gets to vote, not just the candidates.

Yeah, but the counter doesn't get reset. There is no counter. The code just looks back 24 hours and counts the number of critique requests you have made. If there are 4 critique requests already, you don't get to make another one.

Advertisers and subscribers are about equally important. Subscribers account for about 40% of the revenue of the site. In fact, the site revenue is about 40% subscribers, about 40% Google, and 20% other advertisers. The Gallery, TRP, portfolios, photo pages, etc, only account for about one-third of the advertising revenue. Most of the advertising revenue comes from the forums and the articles (Equipment/Learn/Travel). The Gallery does account for most of the subscriptions.

In other words, while advertising is a bit more important than subscriptions, overall, subscriptions are a lot more important than advertising in the Gallery. The only reason we make any revenue from Gallery advertising is that there is a lot of it, since people don't very often click on the ads in the Gallery. It isn't completely accurate, but if you said that the Gallery produces subscription revenue, and the Forums and Articles produce advertising revenue, you wouldn't be far off. Another thing you could say without being very wrong is that most subscribers pay for themselves and contribute heavily towards the costs for non-subscribers, and the advertising pays the rest for Trial/Guest members and the people who are just passing through.

 

So the notion that subscribers don't count isn't true, even from a financial point of view. It is the opposite of true. The site wouldn't survive without the subscribers. If we did manage to survive without them, it would only be by having a lot more ads than we currently do, and probably the site would be quite different than it is. The site probably couldn't survive without the advertisting, either, at least not without making subscriptions mandatory. That would make the site tiny in comparison to the current site, so small that I'm not sure anybody would be interested in subscribing to it. So the site needs both subscribers and advertising and we didn't get on the more or less sound financial footing that we have at present, until we were able to attain reasonably significant numbers of both.

 

However, the number of people who are submitting "critique only" photos is pretty small, and the number of people who use the critique only list to decide which photos they will comment upon is even smaller. And the group who think of the Critique Only list as the basis for a community is even smaller than that. So, it is far from correct to think of this group as "the subscribers".

Actually, there is a Board of Directors, and I'm not the CEO. The CEO is Rajeev Surati. However, if you look at the About Us page you will see that I am the Publisher of photo.net and the Editor-in-Chief. In any case, I get to decide things like this.

How could you get 5 in one calendar day? Suppose you already have 4 in a calendar day, and you are trying to post number 5 on that same calendar day. By definition, the calendar day started less than 24 hours before. You look back 24 hours, how many do you find? By assumption, at least 4. Therefore, number 5 wouldn't be allowed.

Being able to rearrange photos in a folder is one of the most common requests, and it is one of the higher priority features for us to implement.

It is 4 in a 24 hour period, not 4 per calendar day. That said, I don't see how someone can post 5 on the same calendar day without violating the 4-per-24-hours constraint. If you send me more information by email, I will look into it.

I removed the comment with its photo attachment. Comments on photos should be on-topic. The topic is the photo. The comment in this case was marginally on-topic, but the only connection between the photo attached to the comment and the photo being discussed was that it also showed some kids at the beach. The comment didn't even make mention of the attached photo, and the attached photo wasn't particularly illustrative of the point in the comment, other than the comment said the beach girl in the subject photo was cute, and the beach kids in the attached photo were cute. Indeed, it is hard to say what the point of the comment was, other than to give the commenter a pretext for posting a cute-kids photo of his own.

 

It's human nature for people to want to share their photos, but photos posted in the Gallery are supposed to be invitations for critique and commentary on that photo. That doesn't make them like No Words threads where people can post any old photo they want to post that is related in their minds to the subject photo. If somebody wants to start a thread where everybody shows their photos of kids at the beach, start a thread in the No Words forum. Or start a NW thread in the Leica forum if the photo is black and white, looks kind of Leica-ish, and has nice bokeh. Don't turn somebody's Gallery photo discussion into a No Words thread where everybody posts his own photos of similar subjects.

Its a bit like a No Words thread, only you have to submit somebody else's Gallery photo. Kind of an interesting idea.

Yann, as I said, I recognize that the Critique Forum is large and needs sub-communities. It is just that I don't favor the development of a sub-community of people whose common point is opting out of the rating system.

 

Critique-only is not a category. In fact, every critique only photo is also in a category. The "critique only" checkbox was intended as a way for people to post a photo in the Critique Forum without being compelled to expose the photo to rating. It was never intended to be a badge around which a "community" would coalesce, and I've given reasons why it is not a good thing for the forum for there to be such a community. Since there is (a relatively small) community that has developed around it, I don't plan to destroy it. But I don't want it to help it get bigger. I am planning other sub-community features that I think will work better and be more consistent with the overall Critique Forum.

John, there are many images that haven't been submitted to the Critique Forum, and they are just as visible as any other images. It is just that if they aren't rated and aren't in the TRP, there has to be a way for people to find them. Otherwise they are effectively invisible, although not literally so. For example, there may be unrated images in the portfolio of someone who by virtue of doing a lot of commenting, is on a lot of "Interesting" lists, etc.

Did enough of the ad get displayed for you to notice which ad it was?

Hi, Christopher. Thanks for your kind words. You are coming in on a pretty old thread from the end of Dec 2003. At that point we were in bad shape, and things got worse for the following six months. The traffic had grown a lot over the previous two years, and was really straining the hardware that had been purchased three or four years before. There hadn't been any new hardware, apart from a Load Balancer, for quite a long time.

 

At the time of this thread, we had just received very generous support from "Sandy" in the Leica forum, and Sandy had also motivated a lot of people to suscribe; so we finally had about $25K in our bank account with which to buy some new hardware.

 

A couple of months after this thread, we did purchase a new database server, upgraded our Oracle license, and bought two Network Attached Storage boxes for the photo database. However, that wasn't enough, and the site slowed down even more during the first half of 2004. Traffic started to fall off, and the site was still slow. During those six months we lost a lot of ground in traffic and membership that we had gained during 2002 and 2003.

 

But by mid 2004, we had figured out how to squeeze the best performance out of the new hardware, and I also identified and fixed a mistake that I had made when I deployed the NAS boxes that was contributing to the poor performance. With those problems fixed, plus a bit more new hardware, things turned around during the summer of 2004, and it has been all progress since mid-2004.

 

One thing I learned from this experience that will be helpful if I ever manage any other web sites is that if you have a choice between more and better content and features, and faster more scalable performance, performance is probably the right decision. Of course, you can't go to extremes: if you don't have any interesting content and features, you won't have any visitors no matter how fast your boxes and pipes may be. But if you already have visitors, and your site can be made faster, you'll probably get a bigger increase in the number of visitors from better performance than from more features and content. People don't like slow web sites, and every time we are able to make photo.net faster, we immediately get more visitors, without any additional content. photo.net is much faster now than it was two years ago, but we still have plenty of room for improvement.

 

Two years down the road, we now have a lot more subscribers and more advertisers (especially Google), and thanks to them, things are finally on a pretty sound footing financially. There have been quite a few improvements in the graphical design and features of the site, without losing too much of the clean look that people said they liked in 2003. We do have a lot more advertising, and like all advertising on the web, it has crept towards being more obtrusive than I ever would have thought in 2003 would be acceptable to the audience. But subscribers don't see most of it, and I think the advertising is still pretty reasonable compared to other sites that depend as much as we do on advertising revenue. That is thanks to the fact that we also have the subscribers, and we try to keep pretty tight control on what advertisers and advertising we let on the site. Also, Google deserves a lot of credit here. Because of them, we can generate a fair amount of advertising revenue with ads that are quite acceptable on the whole to the audience. Without the Google text link ads, we'd be a lot more dependent on the more in-your-face ads, and there would be a real dilemma about what to do. Advertising on the web would be a lot worse for people if it weren't for Google.

 

We still aren't sitting around eating bonbons, but as long as people keep enjoying the site, signing up for subscriptions, and glancing at (or even clicking on) the ads once in a while, we're in pretty reasonable shape. Thanks for your concern, though, and I hope you continue to enjoy the site.

Vladimir, as I said (twice already in this very thread, and also a few times in the other threads on this topic), if someone can provide me with a URL to a broken thread, I will try to reproduce the problem. There is a URL in the other thread, but it has a typo in it, apparently, in precisely the place where I can't guess what the correct URL should be, namely the forum thread "msg_id".

 

I need someone to point me to a specific thread that shows this behaviour in order to have any chance of finding and fixing this. Even then, I suspect that the problem is an interaction with something else on the page that is transitory, such as an attachment to a post, or Javascript associated with an ad, rather than something that is a persistent feature of forum threads. If it was a persistent feature, you would expect to see it on every thread, or a sizeable number of them. Leaving aside attachments and ads, there aren't that many thread variations, and all of the variations have been around for years.

 

If the problem is something like an attachment, there is some hope of finding the problem, since once a thread is poisoned by a weird attachment (etc) it probably stays that way. But if it is an ad that is causing a problem, that is going to be very hard to find, because which ads appear on pages is very much randomized. Also if the problem is an ad, even if it is found, we have to either block the ad (which could be costly if it is from a big campaign) or get the advertiser to fix it, which may not be very quickly.

It is because I never intended "Critique Only" to be a forum, and I basically don't want things to work that way.

 

It encourages people to submit photos for "critique only" in order to gain access to an alternative forum that they view as more desirable and attractive than the regular Critique Forum. But in that alternative forum the photos basically become invisible except to the other participants in that forum.

 

If people are not comfortable with subjecting themselves or their photos to the vagaries of the rating system, I don't mind if the photos are submitted to the Critique Forum as "critique only" photos. But I don't particularly want to encourage that by making it seem desirable. I view the photo ratings as a integral part of the systems of the site for giving visibility to the most interesting photos. I don't want to present people with an either/or choice: where they can participate in a critique sub-group but only if they foreswear their participation in the regular systems of the Critique Forum.

 

Photos submitted as "critique only" shouldn't have any special visibility for getting critiques, compared to other photos. I think critique-only photos should just be in the Critique forum, the same as any other photos. If this means that they receive fewer critiques because they have opted out of one of the main vehicles for visibility on the site, that is the member's choice.

 

I also didn't like how the Gallery was turning into two camps, with some in the "Critique Only" camp being very aggressive about denouncing the rating system in the Site Feedback forum, etc, questioning the intelligence, honesty, and motivation of people who participate in the rating system, and trying to encourage people to stop submitting photos to be rated. Even though most participants in the critique-forum were not aggressive, I didn't like how whenever someone complained about the rating system in the Site Feedback forum, which was often, someone else would chime in that the person complaining should just submit the photos "critique only", thereby gaining access to something called the "Critique Only Forum", which was depicted as much more calm and desirable than the smelly Critique Forum with its obnoxious rating system. The impression was always left that if you *really* were interested in critiques, you would go all the way and submit photos "Critique Only", that anybody who submitted photos to the regular Critique Forum, couldn't really be interested in critique and could only be some kind of ratings-obsessed rube.

 

Almost everybody here wants more and better critiques, including the people who submit their photos to be rated. Why should "Critique Only" photos be the only ones to be entered into a special form where photos are "really" critiqued? The *entire* Photo Critique forum is supposed to be for people who want to critique photos and have their photos critiqued. I understand the need for subgroups. The Critique Forum is currently too large, with many participants. People feel lost. This is why you see a steady stream of requests for additional "Categories", which is another mechanism for creating subgroups. Years ago we experimented briefly with "Critique Circles", small groups of people who would give each other critiques. The pilot was very popular, but we abandoned Critique Circles after creating only about ten of them, because they were too labor-intensive to set up. I am planning to bring back something like the "Critique Circles", with some new ideas about how to make them less labor-intensive to establish and more scalable.

 

Meanwhile, the Critique Only "forum" won't be returning in that form. I haven't removed the code or made the link not work, in deference to the people who were frequenting the Critique Only photo list as a type of forum. But I'm not planning to make that link visible on the site, and eventually the Critique Only "forum" will be superceded by something else that will give all participants in the Critique Forum access to a critique circle of their choosing without being compelled to submit a photo as "Critique Only" .

It is at position 198 in the All/Avg/Week view, which is where it is supposed to be. It shouldn't be in any 3-day views, because it was submitted for critique on 11 May, and today is 15 May.

 

The server doesn't automatically refresh stale TRP views; it only refreshes them after someone asks for a view and it sees that the view is stale. This refresh happens in the background, because some of the views take a long time to compute. So, unless the refresh is very fast, the person whose request for a view triggered a refresh will see the stale view. The refreshed view will be available a few minutes later, but by then the person who triggered it may have moved on, and probably the first person to see the fresh information will be someone else.

 

So, you probably saw a stale 3-Days Pet Category view. You can see when the views were generated by the date up at the top. If you notice that it was a long time ago, you can ask for the view again, and you may well see the newly-generated version that was triggered by your previous request. Depending on how heavy the calculation is, it might take a few minutes and a few tries before the refreshed view replaces the stale information. For your information, the Day view is computed hourly; the 3-day view is recomputed every 4 hours; the Week and Month views are recomputed daily. The longer views are recomputed every 14 days.

The categories aren't supposed to be keywords precisely defining the subject matter of the photo. They are supposed to be just broad genres of photography. We maybe have headed a bit down the path of making the categories too specific, with categories like "Cars", and "Birds". It would have perhaps been better if we had stuck with higher level categories like "Landscapes", "Nature", etc. Adding the more specific ones just leads to requests for more and more specific categories. I'm not going to remove the existing categories that are too specific, but I don't think we'll be adding more like that, and certainly not any that are even more specific. There are a more than a thousand photos uploaded per day, and currently about 30 categories. A category that isn't going to get at least 10 photos per day in it isn't worth adding, and ideally it should be more than that. Some of the existing categories don't achieve this, but that isn't a precedent for a lot of small categories.

 

It wouldn't be bad to have a tagging or keyword feature, but the way categories are currently implemented in the database and presented in the user interface doesn't lend itself to a long list of very specific categories. And I don't think that people's critiquing/rating interests are that specific, or if they are, I'm not sure they should be. Indeed, as far as rating is concerned, the only category that gets much rating activity as a category is Nudes. I don't know how much people are using the categories to find photos for critiquing.

Mark (Ocean Physics), I noted your report about the keywords. I haven't been able to reproduce that problem.

 

There is no problem with the generation of the keywords, as far as I can see. The keyword generation is done on the server side. If there were a problem with the keyword generation, it wouldn't just affect one browser. Anyway, the end result of the keyword generation is very simple HTML, nothing fancy, and IE couldn't possibly be having a problem rendering it. It is just text inside a DIV. Even IE can format that.

 

Where the keyword problem most likely lies is that the keywords are made hot on the browser side by Javascript code that is fetched via a <SCRIPT> tag by the browser from IntelliTXT. photo.net doesn't actually control that code. I know the IntelliTXT people have been changing it a lot over the past few months. You can see this from the fact that the IntelliTXT bubbles have been getting more and more elaborate. There may be some problem with how this script interacts with IE. It must be a fairly subtle problem, because it seems to relate only to specific pages -- perhaps specific combinations of keywords, or perhaps there is some interaction between the IntelliTXT Javascript and other features of specific pages, possibly including ephemeral features such as what other ads are on the pages. Who knows? In order to file a bug report with IntelliTXT I need the URL's of the threads that are causing the problems, so that I can try to reproduce the problem. (As I've said now several times.)

 

At this point, I couldn't say that the problem people are talking about in this thread is even the same issue, although it is a reasonable possibility, and I thank you for reminding me of that previous post of yours. Also, let me give you the thanks I omitted the last time for your report and troubleshooting work. I intended to acknowledge it at the time, but I got distracted and forgot. I didn't mean to seem unappreciative or rude.

 

Finally, the last post before this one about photo uploading refers to a completely different issue that resulted from a change made earlier today, and that is now fixed.

I'm not walking away from the problem. If people report the specific threads that are causing problems, and I can reproduce them, I will see what I can do. As I said, sometimes the problems are due to things that people attach to their posts. There may also be problems sometimes with the ads, which often include Javascript. If people are seeing problems with some forum threads but not all, the problem is likely to be something that photo.net doesn't fully control. And, as I said, IE is a flaky browser. A lot of problems will go away just by clearing the browser cache.

 

Incidentally, IE accounts for "only" about 80% of the visitors, not 95%. But 80% is still a lot. And this is another reason why I suspect that the problems are caused by something localized, like the browser cache being full or a problem with one of the ads. If we had done something that generally broke IE, I think there would be more reports of it in this forum.