How does the suggestions work?


#1

Hi,

I’ve used some credits voting for a couple of suggestions but can’t seem to find out how the systems works. For example how is it decided which suggestions to implement? Is it when it reaches a certain number of credits?

Is it simply a case of the DH guys deciding what they want, in which case, why bother with the different amount of credits?

I think the least they could do is explain the process. If it’s made clear somewhere else on the site and I’ve missed it I apologise now!

TIA

Phillip
http://www.planetphillip.com/
Commentary and Analysis of Sci-Fi Single Player First Person Shooters


#2

I think it’s a case of how many votes the thing gets. Ideally, the ones with most votes will get implemented. But currently there isn’t any way of knowing how many each thing has got… unless you vote for the suggestion that asks that and it gets implemented <_< (Personally I think this should be done anyway so we know we aren’t the only person voting for it and wasting time waiting for it to be implemented).

A reason I stayed with DH is that it seemed they implemented suggestions every month or two. But they haven’t done any for a long time which is kind of disillusioning for me, although they never said they did this anywhere.


#3

It would be nice iif they commited to implementing a set number of suggestions per month but that would be difficult to get working in reality.

They should at least implement one suggestion a month.

Phillip
http://www.planetphillip.com/
Commentary and Analysis of Sci-Fi Single Player First Person Shooters


#4

We do try to add at least one thing new every month. We also try to mostly only work on the suggestions that have gotten the most votes. At this point several of the top suggestions are things that are relatively difficult to implement, though. They are on the long-term to-do list, but we have only so many resources so we have to budget our time out.

If you are curious about the likelihood of a specific suggestion being implemented soon, I can take a look at the votes for you.

  • Dallas
  • DreamHost Honcho

#5

That would be very kind of you =)

Isn’t the whole idea of the points system that the most difficult need more points? Even if they are difficult and worth a lot, if it has the votes shouldn’t it be implemented?


#6

Yeah, I guess some of the suggestions should cost a lot more votes than they do! The people assigning the relative point costs aren’t always the ones implementing the suggestion itself.

An example of a difficult to implement suggestion is the username@domain logins for email. We actually already once tried to roll out a solution for that several months back but were forced to roll it back when it didn’t work as advertised. Something like that has the potential to break email for everybody so we have to be extremely careful about it, and ‘careful’ usually means ‘slow’ in our case.

We think we may have to upgrade all of our email servers to the next version of Debian to properly implement the username@domain email logins and that’s a whole other major project in itself. It’s all on the to-do list, but we need to set aside the developer and admin time to really do it well.

Lately, some of our services have been taking up more time than was expected so that has been eating into our time to implement suggestions. We’re not ignoring the suggestions list! :sunglasses:

  • Dallas
  • DreamHost Honcho

#7

A related discussion on the suggestions system, here:

http://discussion.dreamhost.com/showflat.pl?Cat=&Board=forum_troubleshooting&Number=21289&page=0&view=expanded&sb=5&o=7&part=

Basically, I’d recommend against viewing the suggestions system as a democratic system, wherein high votes ‘win’ versus low votes.

Suggestions serve as a guide for our Marketing and Dev/Admin teams to consider what to offer, and usually we focus on things that are either A) super easy to implement ASAP, or B) have a lot of demand. There are so many other criteria, though, that suggestions are only one variable - albeit a very important one - we use to make plans.

  • Jeff @ DreamHost
  • DH Discussion Forum Admin

#8

I think the “vote” system is brilliant, personally. It allows the customer to participate in the growth of DreamHost, at least on some level. It sometimes conjures up an analogy of little chicks chirping frantically in an attempt to get their parent to give them the choicest morsels, but ultimately it makes DreamHost seem more like a family, with parents and children, than simply a business entity trying to make a buck or two.

DreamHost peeps can now get valuable feedback from the voting system, this forum, and the new wiki. The company has essentially turned its customer base into a fact and fault-finding team, with the whole organism dedicated to the improvement of the service - that is good for DreamHost, and it is good for the customers.

This philosophy is one of the reasons I am so comfortable being hosted by DreamHost. I’m already getting great value for my money, but at the same time I feel like I’m part of the team that is helping to improve the service - something that ultimately makes it better for me too. That is why I’m happy to contribute to the new DreamHost wiki.


Simon Jessey
Keystone Websites | si-blog


#9

Well said.

And I understand that some things are major and affect more than one system. I think you should remove what seems to be a 5 point cap on your suggestions. Maybe start putting 7 votes in show really difficult ones.

As said before, I think certain ones should be implemented anyway because they are just too useful to not be implemented. But that’s just my opinion ^_^;


#10

(like better statistics reporting…)


#11

My apologies for replying so late. Very long weekend away!

I fully understand your posistion about implementing the suggestions and all the problems related it them.
I don’t think the issue lies with the implementation more with the presentation. As somebody said earlier “You shouldn’t view the votes system as a democratic…” Which is fair enough it’s your/their business and can be run exactly as you see fit. However, the wording, both “vote” and the intro, give the impression of a much more democratic system.

I have some more “Suggestions”!

  1. Rename the “Votes” to “Suggestion” points. This may make things clearer for new uses.
  2. Re-word the introduction to exactly explain what happens. You explain the “voting” proceedure but not the underlying process behind it.
  3. You are honest enough to say that not every suggestion will be added. If somebody takes the time and effort to make a suggestion the least you could do is respond. I accept you don’t want to get into long winded discussions (like this one?) but a simple polite “No, because… Thank you” Seems fair.

With regard to your kind offer to look at specific suggestions, I’m very curious about:

  1. The phpmyadmin upgrade
  2. Installing AWStats on all domains.

TIA
Phillip

Phillip
http://www.planetphillip.com/
Commentary and Analysis of Sci-Fi Single Player First Person Shooters


#12

We appreciate the input, but general comments like this are not that helpful! Analog is definitely not the prettiest web stats program around (I’ve used most of the others available at some point), but I’ve generally found the reports themselves to be fine. What is Analog not providing that you’d like to see? What stats programs have you used that you preferred?

  • Dallas
  • DreamHost Honcho

#13

Changing some of the wording in the suggestions area is a good idea. I did not realize there was so much general confusion about it!

I’d like to say again that sometimes a particular suggestion that is very popular has not yet been implemented due to the technical complexity of the implementation and not because we are deciding to not do it at all. That is true for 3 out of the top 6 voted suggestions at the moment. Other suggestions that may get fewer votes are implemented sooner because they are simple or quick to implement.

The phpmyadmin upgrade is currently the 15th most voted on suggestion. However, I believe we are currently working on it and it should be rolled out in the next week or two. We make some of our own modifications to the source to suit our needs so it takes a little longer than a simple upgrade would normally. This is not an official announcement of it but I heard some rumbling about it the other day. Please don’t hold us to it!

The Awstats suggestion (the suggestion says to completely replace Analog with it) is currently the 12th most voted on suggestion. I have been personally using Awstats for a couple of sites hosted by us (including dreamhost.com) for a few months now to get a feel for how well it works and how much server resources it uses up. I can say it is very likely that we will offer Awstats or another stats program as an additional alternative to Analog at some point this year. It is unlikely that we would start running it on all domains all at once as we would not replace Analog with it. I have been using Awatsts for my own website and I do find the reports more visually appealing than Analog’s, but I still find myself turning back to Analog for specific information I have become accustomed to. They each provide different information in different ways.

There are also technical reasons why we stick with Analog as our default stats mechanism, the main one being the amount of cpu resources it takes to generate reports. Based on information I have seen Analog is around 9 times faster than any other program out there. When you’re running it on over 100,000 domains 3 times a day (once for each report we keep) that adds up fast! However, if the stats are not providing our customer base with information they find to be valuable that may not be so important.

You should note that it is quite easy to set up Awstats on a domain yourself. It does take a little know-how to set up the cron job, Awstats itself is easy to configure. We like to leave our system flexible for anyone who wants to take things into their own hands. I have been meaning to write some documentation about how to set up and configure the most common web stats tools on our servers for anyone interested in doing that, but other things have kept me too busy to do so.

  • Dallas
  • DreamHost Honcho

#14

Dallas, I’m failrly new to the game (and brand new to DH) so please understand that you don’t always have to take everything I say as an extremly serious comment. :slight_smile:

The only thing I have against the analog statistics reporting is that it is a bland, ugly and ‘slow’ (by slow I mean that it takes a lot of screen reading to find out some simple info). It works fine, and displays lots of usable info, the interface is just… difficult.

Suggestions… use Webalizer. Has to be the most popular stats reporting available. The interface is clean, has great info (daily, hourly, entry pages, exit pages, country, etc) and displays in both numerical data, as well as graphs for quick analysis.
With a previous host I could even enter in domains or IP addresses I did not want displayed in the reporting (Loved this! I’m always hitting the site from home and at work, playing with little things here and there, testing some crazy scripts from hotscripts.com, etc.) which saves a lot of time fumbling through the ‘numbers’.

That’s just my take on things. Would be nice to see a new stats report come along. But, they’re your servers, you’re resources. I’m not going to leave just because I don’t like the stats reports :slight_smile:


#15

Sorry if I came off a little grumpy! I’m pretty much the one who makes our stats software decisions so maybe it’s a little close to heart. I agree whole-heartedly that Analog’s reports are ugly and difficult to navigate. It is able to ignore specific hosts like your own IP addresses and that is something that is planned to be integrated into the web panel. Analog is extremely flexible and fast and that’s why we have stuck with it for so long.

A lot of people do seem to like Webalizer, but a lot also like Awstats. Awstats currently has more votes in our suggestions area so I’ve been trying it out on some domains. Awstats seems to be under more active development as well. I’ll try out webalizer on some domains for awhile to get a feel for how it works.

This sort of comment is very helpful! Thanks!

  • Dallas
  • DreamHost Honcho

#16

No worries! It’s all good.


#17

WOW, 100,000 domains, three times a day! That puts things into perspective.

I started the process of installing AWStats on my domain but kept getting interupted and felt it was better to wait and give it my undivided attention. That was a few weeks ago and of course I haven’t got round to it yet!!

Would it be possible to make it a one-click install? That way only the users who really want it would get it.

Anyway, keep up the good work.

Regards
PlanetPhillip

Phillip
http://www.planetphillip.com/
Commentary and Analysis of Sci-Fi Single Player First Person Shooters


#18

I would like to add something to let people more easily use an alternative stats program, yeah. The exact form it takes is still open at the moment. Most likely we would just install Awstats and allow people to toggle it on and off for a domain. It would be in addition to the Analog stats we provide for all domains automatically.

  • Dallas
  • DreamHost Honcho

#19

Thanks for updating us, that’s really good news.

Phillip
http://www.planetphillip.com/
Commentary and Analysis of Sci-Fi Single Player First Person Shooters


#20

That’s a great idea, Dallas. I know of several web hosting companies that offer a choice of statistics packages, but having the option of disabling one or the other seems like a wonderful solution.

I would recommend making the default setting of AWStats to be OFF. That way, only people who feel they really need the package will enable it, and less resources will be consumed. Personally, I’d much rather use AWStats. I particularly like the way it recognizes more user agents, and delivers a better summary.


Simon Jessey
Keystone Websites | si-blog