There are some advantages to moving GameTZ to new/different computer/server hosting. I can move to a new major version of the operating system that has more advanced components. Also, I should be able to reduce the monthly costs.
It is a fair amount of work to do this. My first step is to upgrade the server I use at home for development and testing (which is like a clone of the live server). Then, I get the site and all its parts setup and working there. The new OS has some changes that may make this challenging. This can take a week or more, depending.
The next setup is setting up a new server for the site to move to. I'm currently thinking I'll use Amazon AWS which is "cloud computing" based. I think this will be cost effective and allow me to optimize the resources needed to keep GameTZ going as we continue a slow decline. The current server is basically just a PC co-located in a data center somewhere. Cloud services are more abstract or virtual. They are priced based on usage and can be adjusted over time.
Generally, I'm also looking for ways to simplify what the server does and reduce its overall footprint (e.g. use less CPU, disk space, network). This also works well with cloud-based hosting since doing this should also lower monthly costs.
When the new server is ready, I'll swap the IP address for the site and update the DNS records (which maps IP to hostname like gametz.com -> 188.8.131.52 as it is currently) to move the live site over to the new server. That can cause disruption, but with some luck it should be seamless for most users. I've done it at least twice in the past.
A concern with all this is that it's really hard to know how much resources GameTZ needs. The current server seems to easily cope with the load. So, I'm going to try to switch to a new server with less. If I go too low, the site may get slow, unresponsive, or not work consistently. So, that could get tricky. But, I think with cloud-based technology, I should be able to adjust this.
tl;dr - I'm going to mess around with some GameTZ internals and you may not even notice, but it's a good thing.
Given that I'm using Linux, Microsoft doesn't seem like a great choice. If I used Microsoft OS/software, then it would be the obvious choice. Azure offers the Linux flavor I use (CentOS), but it can't be a focus/priority for them. AWS dominates the market and has for over a decade. So, it's an easy choice to use them. CentOS has a presence in AWS (offers the OS free via their marketplace). Google may not be a bad option for me. I'm still looking at it. I haven't used any of them, so it's hard for me to feel confident. I'm just reading articles and using pricing calculators mostly so far.
GameTZ started on a PC at a friend's house, then eventually 3 PCs in my basement. Then, co-location (big step for me) into 1 stronger server, but that also is basically just another PC (maybe in a rack). The cloud is pretty different and hard for me to wrap my head around. I'm old school. Aspects of it are worrisome (e.g. separation of disk storage and the options there). Performance of the disk likely matters (big effect on database speed), but it's hard to know in advance how it will be, especially compared to what I'm using now. Also, the hardware I'm on is somewhat obsolete, so it could be that almost anything I move to will be better. And, the site is used less now too, so I'm sure I need less.
Initially, I was just going to move to a different dedicated server, but my current hosting service has been pushing cloud. So, I checked that out. But, they are limited in what the offer. So, I started looking at AWS and started to see how it might work and be even cheaper...
Using the respective price calculators, Google seems to be cheapest for what I need. AWS is $99/mo., Azure $90, Google $77. I'm currently paying $227, so these are all appealing prices and really not a huge difference between them. Though, I may also just be off in my estimates too.
I like Google's console UI and command line tool too. I'm currently leaning toward Google after reading more about them. I'm sure they would all work. I am biased against Microsoft, so it's hard to seriously consider them. It has felt to me that they worked against Linux for a long time, so it's hard to believe they are on board with it now or up-to-speed compared to other providers.
I'm likely to use snapshots for backups instead of what I've been doing as it would help me simplify some stuff. I'd probably have to test a restore to be sure it works. Maybe I should put the db and other key data files on a separate disk resource too, just to isolate it and snapshot it more easily.
I think GameTZ is just too weird to mold what I have into the structures they offer. For example, I could break off the database into a separate db instance, but I think it would make me redo some stuff to fit their schemes. Also, I suspect it would cost more. If I was starting over, I'm sure I'd just go cloud from the start.
Thanks for the advice.
Who has the most redundancies for the price?
Depending on overall size, cloud providers make redundant backups so if one cloud hosting node goes offline, another kicks in without the user seeing any interruption. I'd assume it would be a tie between Amazon and Google.
Most outages you see in cloud hosting are not the fault of the site but one of the hosting provider servers going dark for whatever reason (regional issues, weather, etc)
Sorry @Bill - HA = high availability, DR = disaster recovery. "9's" equates to the level of uptime guarantee via SLA's (service level agreements) ... as in "99.99" (which is 52 mins of downtime annually) vs "99.999" (which is 5 minutes of downtime annually). I think by default, you actually get 2 9's with Azure/AWS/Google.
They all have amazing SLA's with NO additional HA/DR. Should you want/need more, they're super easy to simply add on. Pending on the sophistication of your HA requirements, you'd need to recode your application, say something like Always-On (MSFT SQL Technology) - which I doubt you need. This would simply be some backend work + changing your application connection string to communicate with an AG (availability group) instead of a specific server\instance\database. BUT, what you can do, is enable (via a shopping option basically) is to make your entire (or pieces of it) tenant geo-redundant. Meaning, if data center X fails, within Y threshold, you'd be in data center Z - no involvement on your end.
The same old principal exists ... backup your crap. As long as you have backups for DR, you're more than likely good enough with that. We're not dealing with mission critical data here on GTZ. I know Azure offers a SUPER cheap backup option via Azure backups, not sure what Google has. You can obviously do your own, to cloud or on premises, really up to you. The out of the box stuff that clouds offer though, is usually good enough for the masses.
> hahaha @John
> Sorry @Bill - HA = high availability, DR = disaster recovery. "9's" equates to the
> level of uptime guarantee via SLA's (service level agreements) ... as in "99.99"
> (which is 52 mins of downtime annually) vs "99.999" (which is 5 minutes of downtime
> annually). I think by default, you actually get 2 9's with Azure/AWS/Google.
> They all have amazing SLA's with NO additional HA/DR. Should you want/need more,
> they're super easy to simply add on. Pending on the sophistication of your HA requirements,
> you'd need to recode your application, say something like Always-On (MSFT SQL Technology)
> - which I doubt you need. This would simply be some backend work + changing your
> application connection string to communicate with an AG (availability group) instead
> of a specific server\instance\database. BUT, what you can do, is enable (via a shopping
> option basically) is to make your entire (or pieces of it) tenant geo-redundant.
> Meaning, if data center X fails, within Y threshold, you'd be in data center Z -
> no involvement on your end.
> The same old principal exists ... backup your crap. As long as you have backups for
> DR, you're more than likely good enough with that. We're not dealing with mission
> critical data here on GTZ. I know Azure offers a SUPER cheap backup option via Azure
> backups, not sure what Google has. You can obviously do your own, to cloud or on
> premises, really up to you. The out of the box stuff that clouds offer though, is
> usually good enough for the masses.
for lending your insight.
I'm thinking I may be able to move us to the cloud on Thursday.
I got fairly far setting up a new server, made some mistakes along the way, learned a few things, etc. So, I'm starting over to do it right for the server I hope we'll be able to move to. It takes a while to install and setup and configure all the software it uses. Some of it is hard to test without going live. There will likely be some issues on the day I do the move, so I'll try to leave extra time to keep an eye on things.
Google Cloud is fairly stingy with IP addresses, but I think I can get away with just using 1 IP address for everything. I'm also going to keep disk space to 64gb or less and use snapshots for backups. Generally, I'm looking for ways to simplify things and use less resources (save money).
Playing around with it the last few days, I'm feeling more confident about using cloud-based hosting. A lot of stuff that is harder to do with my current setup is abstracted out and possible to do myself in most cases. So, swapping IPs and messing with disks (add/remove/backup/clone/etc) and even the server hardware itself all feels doable to me now. Though, I'm not planning to change it much once I get it the way I want it. If I need to I should be able to fairly easily. For example, there's some chance the new server will be too slow and I may need to scramble to upgrade components, but that should be doable.
The new server I was working on mysteriously went off the network last week and I'm still unsure why. But, I was messing with some odd network settings that day (maybe that caused it somehow). I was able to stop and restart it to get it back. Hopefully, that won't be an issue in the future. It hasn't happened since and I'm not planning to mess with those settings again.
Is there anything you can totally expunge from the database to save on resources? Old listings with zero pictures, zero people want it, zero people have it. Worst case scenario someone requests we add it as a new item later.
Whatever makes things leaner/faster/easier to manage. A lot of things can go without anyone needing them, I'm sure someone that's a site statistics hound will request some obscure information just because they can, but it's not mission critical to keep it around forever.
Or take any unneeded data, put it on a external drive, take it outside for a round of stress release skeet shooting. Whichever works.
I was just looking at the size of the forum archive. It's 3.2gb or 43% of the GameTZ's total database size (7.5gb on disk). The offer/trade messages archive is about 1GB for another 13%.
Database size can impact a few things. Performance, depending (hard to measure, but if that table gets used and it's huge, so likely strains memory and slows other db usage down). Also, I'm planning daily snapshots, so anything on the disk has some small impact on cost that adds up over time. If I move the data off the server, there's network bandwidth costs (also small, but would add up).
It's definitely something to think about. Since I'm now paying for every little thing and I'm cheap, I suspect I will experiment and optimize and do whatever costs the least.
I put a full backup on dvd every 3 months, so that's kind of like cold storage. I may stop doing that though as it's also extra work and I've almost never used it.
I'm not sure I agree about saving data. In the age of FB's recent reputation, I could see making a case for just deleting stuff for the sake of user privacy. I'm not sure who that old/saved data is for. It could be used for evil. More than one person has asked me to delete their old posts (which they made when they were 14 or whatever). Also, simplicity says just delete it unless there's a good reason to keep it.
All data should have a expiration date. Anything held longer then it needs to adds bloat and liability. Unless of course there's mining value to that data and those who produced that data have consented. Most of that doesn't apply here though. You're not holding sensitive data of financial nature.
> Everything is pretty well normalized I'm assuming?
Not especially. I'm not big on normalizing. I'm mostly self-taught with respect to databases. I do have good indexing (as far as I know). That was something I learned early on is absolutely essential (the site was super slow when I first started using a db).
> Although I guess you're not storing any blob or unstructured data...
> Just varchars, bits, ints and datetime yeah?
And TEXT (e.g. for forum posts) which is the same size as a blob for the database I use (64k).
> Do you use any type of date/time dimensions are store for every entry?
Most records have a datetime column in some form.
> If the latter, that can be optimized to dramatically cut down size.
I'm not sure if I follow what you're saying, but if you just mean I can archive (put into another table that used less) older posts then yeah, that's what the forum archive is. It's a little more complicated because there are post records, topic records and a few other related records being used that are interconnected.
> 5 cents per gig/month for hdd storage.
That seems different from what I've seen. I've been using each cloud services "calculator" tool to estimate costs: https://azure.microsoft.com/en-us/pricing/calculat... https://calculator.s3.amazonaws.com/index.html
There seems to be other costs associated with disk space, especially snapshots. It's hard to know in advance how much space I'd need for snapshots, though. But, roughly, I've been thinking I'd keep 60 days of snaps and It may be 2+gb/day from that (maybe 120 to 200gb in snapshot storage, guessing). Also, disk performance (iops) seems to often be tied to size. So, going too small on disk could cause issues with performance (but that's not likely and issue from what I can tell).
> stupid cheap, like less than $10 for 10gb.
That sounds closer to the numbers I was seeing. Multiply that by 12 months and it's $120. That's a good week for amazon commissions. I'm trying to scale this to actual revenue for the site. I'm sure for a real business these numbers are tiny and a joke. Like I said, I'm being cheap here. The issue is somewhat emotional or philosophical. If I feel like the site is costing me, it makes it harder for me to keep doing it. So, the cheaper it is, the easier it is for me to keep it going (not really about money exactly, more the impression of that... Personally, I agree $10 is a joke for most stuff I spend money on).
Also, I don't think I really need the disk space and using less is simpler too (less to potentially backup, etc.) Part of simpler is reducing my workload which is another factor (also kind of emotional/philosophical). If GTZ is cheap/easy to keep going, I'm more likely to do it. It has been 20+ years, I'm fairly burned out and it's easy for me to question why I keep doing it (especially when I get some belligerent user to deal with or whatever). I can't do much about that, but I can optimize the hosting.
> What's your database back end?
MariaDB (mysql). MyISAM tables (which are older-style and not especially relational) Is that what you mean? It runs on the same machine as everything else (web server, helper servers, mail, cron jobs, backups, etc.)
> Can you simply do some partitioning? Have you revisited
> your indexing strategy as the site has grown? Assuming you're maintaining them via
> daily maintenance?
I'm not following you here. Partition what exactly? the database? Generally, I find that breaking what my server does up into separate parts/instances adds to cost (given my small scale overall). I looked into it a little, but I couldn't find a way to make it cheaper. And 1 server, 1 disk is simpler to manage too. Maybe you meant something else, though.
I certainly use indexes on my database tables and I think most queries are fairly fast (e.g. pages load quick enough here). I don't do daily maintenance of indexes (if that's what you mean). I'm not even sure what that might be. I sometimes (like once a year) use "optimize table" on them, but it doesn't make a huge difference from what I've noticed. Come to think of it, there is something I do for item searches daily which helps build a better index for that. Matching also does some daily work on the site too (though that rarely gets used by people from what I can tell). ...
> All data should have a expiration date. Anything held longer then it needs to adds
> bloat and liability. Unless of course there's mining value to that data and those
> who produced that data have consented. Most of that doesn't apply here though. You're
> not holding sensitive data of financial nature.
I don't think it's especially sensitive. It's mostly just old forum posts or stuff the user had done publicly (e.g. bios). I have done things to keep some of that stuff out of search engine indexes. But, in most cases, I do still have the data.
It is interesting to think about. I could do more to remove it. Abandoned accounts get their listings removed after a couple years. I could remove bios too, maybe other stuff. Trades/rating are connected to other active accounts, though. So, deleting those would impact others reputations and generally I never remove those. But, I do have a "purge" thing I can do which removes name/email/bio and I can change the username if that helps too.
Deleting stuff can impact security of active users too, though. e.g. If I deleted some famously evil 10 BTRs user data, they can much more easily come back and rip more people off. It happens even if I don't delete the data, but keeping the data has helped me stop them in many cases. Generally, if they make a new account, my system can detect it by comparing the info they provided to past accounts. But, if I delete the past banned accounts, I'll miss it. Actually, more commonly, the same system finds someone making a new account that had a past account without trouble and I just merge the old account into the new one and it helps them.
> How many users does GameTZ have in the UK? GDPR Act starts next month in the UK.
> Even though you're a US business, there may be some legalese you need to show new/current
> UK users to be compliant.
*sigh* That article seemed to be asking more questions than it answered. If there was some simple list of what I'd need to do to comply, it would help. I can certainly delete anyone's data that asks. Is that enough, I wonder. I have to do it by hand to some degree though. I don't have a way for users to do it themselves.
And, it's worrisome for the same reason I gave before. If a user can delete all their data here, then scammers would always do that and we'd all be more at risk...
nice that was fast., its acting a lot different on my computer at work but functions fine, just a lot of things I cant see, the menu bars at the top and bottom, icons and avatars, pos and neg buttons, and other things, but it could just be temporary, will see if it continues, but likely nothing on your end our company uses this horrible invincea freespace browser that blocks a lot of content. like I always couldn't see most pictures on it and stuff, like signatures never displayed before and the like.
> just a lot of things I cant see, the menu bars at the top and bottom, icons
> and avatars, pos and neg buttons, and other things,
that is strange. it may be the image server ( https://images.kenyonhill.com
) -- does that work for you? It should say "Nothing to see here, move along."
Your computer may still be looking at the old server for that. It just needs to update its mapping of the name ( images.kenyonhill.com ) to the new ip address ( 184.108.40.206 ).
Rebooting may help.
> tonymack21 wrote:
>> just a lot of things I cant see, the menu bars at the top and bottom, icons
>> and avatars, pos and neg buttons, and other things,
> does that work for you? It should say "Nothing to see here, move along."
It doesn't seem to want to connect. I'll try a reboot later after I finish a couple projects I'm working on.
Works fine on my phone so it's local to my end. Probably this freaky browser they force on us
> I was just looking at the size of the forum archive. It's 3.2gb or 43% of the GameTZ's
> total database size (7.5gb on disk). The offer/trade messages archive is about 1GB
> for another 13%.
> Database size can impact a few things. Performance, depending (hard to measure,
> but if that table gets used and it's huge, so likely strains memory and slows other
> db usage down). Also, I'm planning daily snapshots, so anything on the disk has some
> small impact on cost that adds up over time. If I move the data off the server,
> there's network bandwidth costs (also small, but would add up).
> It's definitely something to think about. Since I'm now paying for every little
> thing and I'm cheap, I suspect I will experiment and optimize and do whatever costs
> the least.
> I put a full backup on dvd every 3 months, so that's kind of like cold storage. I
> may stop doing that though as it's also extra work and I've almost never used it.
> I'm not sure I agree about saving data. In the age of FB's recent reputation, I
> could see making a case for just deleting stuff for the sake of user privacy. I'm
> not sure who that old/saved data is for. It could be used for evil. More than one
> person has asked me to delete their old posts (which they made when they were 14
> or whatever). Also, simplicity says just delete it unless there's a good reason
> to keep it.
If you ever decide to get rid of the archive, do you think you could give us somewhat of an advance notice? I actually search archived posts pretty regularly. Sometimes it's just fun to read through old threads with posts from past users. Sometimes I remember something being discussed on here and I want to find it. But, the biggest thing I would miss is being able to access all of my own old posts. Over the years, GameTZ has inadvertently turned into an archive of my video game thoughts/opinions/reviews/etc. I've participated in so many discussions like "Favorite games of all time", "Top NES games", "Favorite RPGs", etc etc. I actually refer to these pretty often. Any time a new/similar discussion pops up here I refer to what I posted the last time as a starting point. Or, sometimes it's a non-GTZ discussion so I just come here to grab my answer that I already spent the time putting a lot of thought into. If you decide to remove the archive, I would love to have some advance notice so that I can scour my old posts and try to salvage anything I can think of, to save in OneNote or something.
> ...t the size of the forum archive. It's 3.2gb or 43% of the GameTZ's
> total database size (7.5gb on disk). The offer/trade messages archive is about 1GB
> for another 13%.
So, roughly 56% of the database's size (on disk) is for rarely-used archive tables.
The daily snapshots I'm doing use about 1.5gb of space. On top of about the first snap used 20gb. I'm keeping 60 days of snapshots (may lower that). So, 20 + (60*1.5) or 110gb of snapshots per month. That costs: $2.86/mo. OK, not much, I'll admit. But, um, $34.32/year. I suspect if I could get rid of the archives, it would save like $10/year at least, maybe $20/year.
The other "cost" is harder to measure. Querying the archives tends to be slow and memory-intensive (not sure, but I assume). That impacts other people using the site, but it's hard to measure. Also, it means gtz needs a beefier server to handle such things. Potentially, we could save more like $25/mo if we downgrade from the current 2CPU 7.5gb server to a 1CPU/4gb server. I'm not sure if that would just be too slow for other stuff, but given that we don't seem to be stressing the current server, it may be worth trying at some point (especially later as usage continues to go down, like in a year).
I use the archive too, so I understand. It just seems crazy that more than half the database is being used for archives that are rarely used by a few people. Maybe we can live without it? Do other websites keep everything around like that? Do we trust the websites that do keep stuff (fb *ahem*)?
There's a vague privacy or security issue related to keeping every forum post ever made by every user for 20+ years. That can be used against people in some way, maybe stuff that got said when they were young and foolish and not looking for a job, etc. I already hide the archive a fair bit to help protect against that, why not get rid of it entirely?
Other thoughts: we could consider "unarchiving" certain key topics if they are worthy of saving somehow. I don't actually love this idea and likely there's a needle in a haystack aspect to this. But, it's a possibility.
Potentially, the archive could be made into a static html tree (not database, just linked static files), even put on a CD or something (though it may be bigger than a DVD-R already). It would be some work to generate that and it could be buggy and not especially searchable, but I'm just brainstorming here. I think mostly it's just not that valuable and kind of waste to keep all that noise around.
I could just leave it as is, I know. But, I feel like on some level, I should make some effort here to smooth the decline of the site out, take on looming issues before they become problems. Money does matter, but it's complicated (related to how much I get paid for keeping this place going [never been much, but how low am I willing to go... and I'm feeling fairly burned out these days and not that into this place or the Internet in general, so it seems to make sense to get costs way down so that's not something that bothers me]). Moving the hosting will save a lot, other stuff will be smaller in comparison. Part of me just loves to over-optimize too, so maybe I'm just nuts and need to be stopped.
I think it's fair to say it's better just to keep everything as-is. But, I think it's always worth considering stripping the site down to the most important stuff that matters to most users (trading/community features). The is a lot of stuff we have been doing that is superfluous (off the top of my head: reviews, mentors, archives, top traders lists, etc.) that may not seem like much but would make things easier to just drop, especially if we nix most of it. We'd have a leaner website that's easy to keep going, has very low costs and low maintenance, etc. That's where I'm coming from on this more or less. I can understand if people don't love this idea, but I've been carrying this burden for a while (with help, surely) and things really are in decline (have been for years and seem to be continuing slowly to go down). Something has to give. We can just let it happen or we can try to work with it, focus what we have left, etc.
> @Scott -- Consider this your advance notice. I'm seriously considering removing
> the forum archive.image
lol thanks for the notice.
Maybe delete sale/trade threads? It'd free up alot of space
> The daily snapshots I'm doing use about 1.5gb of space. On top of about the first
> snap used 20gb. I'm keeping 60 days of snapshots (may lower that). So, 20 + (60*1.5)
> or 110gb of snapshots per month.
It seems like this is excessive. Do you really need that many daily snapshots? I'd usually keep maybe a weeks worth -- then go to just once a week or something. Seems like you'd save a ton by just keeping less daily snapshots that you are very unlikely to need, right?
> That costs: $2.86/mo. OK, not much, I'll admit.
> But, um, $34.32/year. I suspect if I could get rid of the archives, it would save
> like $10/year at least, maybe $20/year.
Yeah, so, I'd just donate $20/year to be able to keep the older forum archives if that is what we're really talking about here?
> The other "cost" is harder to measure. Querying the archives tends to be slow and
> memory-intensive (not sure, but I assume). That impacts other people using the site,
> but it's hard to measure.
But we admit that they are rarely queried, right? I mean, only one of these two are probably true:
1. We query the archives very rarely -- so the impact when it comes to slowness or memory intensiveness on the site seems minimal.
2. We query the archives fairly often -- which means to me that they are certainly more useful than when we discussed it above -- so it seems worth keeping around maybe?
> I use the archive too, so I understand. It just seems crazy that more than half the
> database is being used for archives that are rarely used by a few people.
I guess, to me, the percentage comparison of the archive to the current doesn't mean anything. I mean, who cares the percentage of the DB that is archive versus current? It doesn't make a big difference in my eyes. The site has been around a long time. If the archive was 90% of the total DB size, I still wouldn't think that was a big deal.
I feel like we're still not talking about a ton of actual storage space -- regardless of the percentage of the overall DB it happens to be.
> Maybe we can live without it? Do other websites keep everything around like that?
I think most do. Reddit does. Facebook does. Most of the forums that I use do.
> Do we trust the websites that do keep stuff (fb *ahem*)?
I think there is a difference between trusting them with personal data -- regardless of how old -- and trusting them with things that I posted publicly. I don't worry about trusting sites too much with a public post I made 10 years ago.
> Part of me just loves to over-optimize too, so maybe I'm just nuts and need to be stopped.
> I can understand if people don't love this idea,
> but I've been carrying this burden for a while (with help, surely) and things really
> are in decline (have been for years and seem to be continuing slowly to go down).
> Something has to give. We can just let it happen or we can try to work with it,
> focus what we have left, etc.
I get that. So, you can do whatever you want to do, of course. I just think that the forum archive might make more sense as a "it works, so let's leave it as it is" thing. But, if you go otherwise with it, I'll live.
Thanks for the feedback.
re: how many snapshots to keep -- yeah, hard to say. I certainly have used backups to restore stuff for people many times in the past, even going back months. But, it doesn't happen that often and part of my new plan is just to say "no, sorry the data is lost." I think daily backups are essential to restore the site in case of some catastrophic problem where the disk is lost. But, I agree it's debatable how many days to keep. Keeping more days is generally helpful with partial restores of some kind, but there can be insidious cases where something is deleted and not noticed for a week. I am no longer copying the disk offsite at all since the network costs tend to pile up for doing that. So, I don't want to skimp on the snapshots too much as it's the only backup now. 60 days was a safe number. I could easily drop that to 45 or 30, maybe less. We'll see.
re: archive usage/effect -- I'd have to put some code in to be sure, but I think the archive isn't used much. I'm not sure it follows that this means there's little or no impact. Databases do a lot of stuff to try to make things faster (e.g. in memory caching, etc) and those resources might linger even with minimal usage. I know I've run into a query cache that causes the same query to be faster the next time you run it, for example. Databases are complicated and can be optimized in a lot of different ways (I can only speculate based on limited experience). It's certainly true that querying the forum archive is notably slower than the regular forum. And, I have gotten the impression it "warms up" (gets faster as subsequent searches are done).
re: privacy -- The privacy issue may actually be the biggest concern for me. When GameTZ was just starting out and the Internet was young, I and others were naive in many ways. Privacy was barely a concern back then. GameTZ used to put all kinds of stuff out in public like real name, email address and mailing address. Things changed in the last decade or so. I have since taken much more care to hide info like that, but mostly in places where the data type was know (e.g. we had an address field, so I just stopped showing the field, done).
The forum is different, all the posts are untyped and I have no idea what's in them from a programmatic standpoint. There have certainly been many cases of personal info being posted both by the person themselves, their friends or their enemies. I've deleted many posts after being asked by people wanting privacy. I think going through that with various folks has shifted my thinking more and more. I take it very seriously and will make a big effort to do all I can to remove info like that. But, it's hard to be sure I got it all. All I can really do is search for their name or whatever, but a slight spelling difference or wording can cause me to miss it.
I've had some experience with bad people online. My naivete has been burned out of me. There are people out there who want to do others harm. All those old posts are a gold mine for them, potentially. Stuff that may seem innocent like talking about your family, dog, hobbies, sex life, etc can be used against you. We've had it happen here (Sheesh got crank calls, someone else maybe got fired after someone called their boss) There has been someone over the years targeting various gtz users (with anonymous harassment) that clearly used what was written in the forums.
I don't mean to over make the case here. But, I think there are good reasons to just delete old stuff as it makes that sort of thing impossible. And, given that there is real risk/exposure, I have to question what the benefit of keeping old posts is exactly.
I think not keeping data is becoming more common now and even a feature in some cases. For example, Snapchat deletes their posts very soon after they are sent. Gmail is adding a self destruct email feature soon. I think this is part of a trend where people are becoming more concerned over privacy. There's also the new European GDPR law that is about the same kind of issues, e.g. "the right to be forgotten".
> If you do eliminate archive, would we get a post count on our bio pages? >
> If you do eliminate archive, would we get a post count on our bio pages? >
Any of the still active participants of this thread may have an advantage. http://gametz.com/?S=4690866&A=Forum&area=General&...
> Sid_Ceaser wrote:
>> If you do eliminate archive, would we get a post count on our bio pages? >
>> If you do eliminate archive, would we get a post count on our bio pages? >
> Any of the still active participants of this thread may have an advantage.
a LOT of abandoned and frozen accounts there :/ see a few names that are still around
Let's count to 10,000 in the early 2000s turned into lets use up all of bill's server space in 2018.
Jokes aside, I'm for removing the archive if it improves site performance. I use it from time to time to find old images.image
Buy the domain, redirect it to a Facebook memorial page.
We should chip in to help