The way I remember it is ten years ago was the tail end of an era in which hosting was expensive and bandwidth was even worse and small sites that got a little more popular than expected would regularly go dark at the ends of months because they hit a resource cap.
Yeah I miss the days when I used to run a web site for a niche crowd of about ten thousand users on my bedroom pc. Now I'm pushing buttons on heroku and getting grief for dyno overages.
I remember those times. They were beautiful. I could spent hours trying to search for the best free webhosting that includes PHP, MySQL and Phpmyadmin, and didnt fill my website with popups or banners at the top/bottom of the webpage.
Slicehost was great back then. I remember jumping on the early waiting list to get my first server and how much cheaper it was than anything else at the time. The service was top-notch - I'd jump on IRC if something bad happened and there would always be someone there handing out info.
I'm still on Slicehost for a few servers and it's still good, but it feels like it's frozen in time (which it is AFAICT). Can't blame them for cashing out and they didn't really leave us hanging - things are still running fine.
I do remember. I've been doing web stuff for a little over 20 years. It was not a good time.
1. If a file upload failed your site was broken until you managed to upload it again.
2. Deploying something that required a database migration had to be done out of hours, or the site had to be taken offline to do it.
3. Recovering from a disaster was hard, and pretty much always resulted in significant data loss.
4. Moving a site from one server to another was awful, because it meant SSH'ing (or telnet'ing...) on to the box and compiling things from source. If you got anything in the environment wrong the site didn't work and you had to start again.
5. If you worked on anything that ran on multiple servers you pretty much had to script it all anyway. Essentially there were thousands of different (much worse) build-and-deploy processes people had stuck together with various different tools. Going to a new company meant learning a completely new set of processes. Today's processes are just an industry-standard best-in-class approach to doing the same thing.
I'll take Webpack to compile a site and terraform/compose/puppet to deploy the infrastructure over old way of doing things every time.
It was exactly like that, man. Dirt cheap hardware of these days could view web easily. The same is not true for today. Maybe you worked through specific areas or simply was tired to death and now deromanticizing it, because I clearly remember when things went downhill in general.
I've saved a startup once from going bankrupt more or less like this.
They were paying a lot for ASP and a shitty product(cant remember the name..it was something like WebObjects) that was taking too long to load, the newly born company i was hired for would cease to exist, so even thought i was hired as a sysadmin to manage the network, linux and databases, i've used my low skills as a programmer, without nobody asking me, and designed the thing in secret, in about two or three weeks with PHP, Linux and MySql instead.
The (news based, with a lot of personalization page) loaded in about 400 milliseconds instead of what once was a 10 seconds load.
My bosses were a little skeptic at first, because the ASP folks was a team formed with university professors and skilled pros, and i was a teen that was not even hired as a coder. Also it was so much faster that they thought i could only be cheating somehow (like static instead of personalized with data from a backend db)..
Linux and open source was also faced with skepticism, specially when confronted with Microsoft stuff that everybody was used to back then.
The LAMP stack was a wonder back in those days, and it helped a lot to popularize Linux, MySQL, Apache and PHP.
I know today PHP doesn't look so cool giving all the other players, but in the context it was launched it was so much better than all the status quo of the nineties where the options were quite limited.
Although it was about five years ago, yes, my potato server (still running the same one today) couldn't handle that by a long shot. Same with Wordpress, Owncloud, and other web-based things that became a little oversized.
Great summary. Yeah, TextDrive was presented as an alternative to Dreamhost that, although seemingly much more pricey, was not oversold (you can use all the space/bandwidth they give you), had better support, ran more smoothly, etc etc. By and for people who love the web, or something like that.
But there were tons of reliability problems from the start. It didn't help that Jason Hoffman handled criticism very poorly. I vaguely remember one incident where someone complained on the TextDrive forum and, taking it as a personal insult, he deleted the person's (hosting, not forum) account on the spot. There seemed to be always an excuse or a promise that some exciting better thing was around the corner to solve all the problems.
ZFS was one such exciting thing as you mention. There was also supposed to be a new administration interface called TextPanel, and it was constantly being talked up on the forums, how great it was going to be. Complete vaporware. In 2012, the crusty super-slow Webmin interface that was claimed to be temporary is still all that's available.
The shared hosting eventually became reasonably stable and reliable, but I suspect that's only because it wasn't being touched at all, and anyone doing anything nontrivial there had probably moved it to another host. Last fall I tried to set up a WordPress blog. Couldn't, because the PHP version was too old, and when I checked, it turned out (in 2011) the version of PHP being used dated to 2006.
My server just had a multi-day outage starting August 13th due to hardware failure, and is not fully restored even now (a bunch of emails are missing). I suspect this is what precipitated Hoffman's announcement. Apparently lifetime hosting was supposed to mean zero maintenance on Joyent's part.
Honestly, they should've done what Google Fiber recently did for its free internet deal: just say it's "guaranteed for at least 7 years". Many would still have signed up and they'd be within their rights to shut down most of those accounts by now. But what happened instead was a lot of us paid for "lifetime" hosting that lasted only 7 years or less, which was actively maintained for only 1 or 2 years, and during that time never lived up to the quality advertised.
But who knows, maybe Joyent and Hoffman have learned some lessons since 2005. I do expect they'll be good enough to give me a full refund, obviating the need to join any lawsuits.
I was IBM’s “corporate webmaster” in the 1990s. Pretty much the only way I could get support to solve problems was for the web site to crash (and then it would be all hands and executives on deck, which was almost just as bad). e.g. we were mandated to use IBM’s rebadged CERN webserver (variously Lotus Go, Domino Go, Domino, etc) in 1995, but it couldn't handle the load and didn't have any of the fixes/enhancements I’d added to our build of NCSA. We finally cut over to it some time in early 1998. Two months later IBM killed it for an IBM build of Apache (which also didn't have any of the fixes or enhancements we'd made to NCSA and I was prohibited by corporate politics from adding them).
I couldn't use DB2 for www.ibm.com for years because of internal accounting rules requiring commercial licenses, while other accounting rules prohibited giving our group the budget for said licenses.
I'm sure everything is much better now that the cloud has descended.
1996 was my first shared hosting. 96-97 was OK, but I had a project for a client that got shut down after a few days. It was a small ecommerce project, and we got shut down for 'resource abuse' - CPU was spiking and network was going crazy. Had a week of back and forth emails and I think phone with their 'support.
They were poking through my code, bitching that I was doing "select *" queries ("those are inefficient" they kept arguing). After a week we were reinstated - turns out there was someone else running spambot on another hosting account on the same server. But... we were "ecommerce" so obviously must have been the root cause of a spiked CPU. I vowed at that point to not use shared hosting again, and haven't since 1998.
It's been nicer with cheap VPSs over the last few years - easier to get clients set up with their 'own' servers. I know that VPSs are shared as well, and I've seen issues where one VPS abused resources to the point where it affected mine, but that's been pretty rare (2x in the past 5 years I can think of), and it's always easier for people to track down, isolate and resolve.
Also, it was one anecdote point, and was 15 years ago, and was just 'bad service' but life is too short to try to be putting projects as risk to save a few bucks. Projects/sites generally have enough troubles - dealing with shared hosting is just one more thing that can go wrong, imo.
reply