This is the current home of the “open design studio” that I’ve been yammering about over on davidseah.com. I’m planning on creating all the collateral necessary to restart my own small design business from scratch, and make it available while blogging and documenting the process. I think it will be interesting and fun while giving me a sense of purpose, which is important to me because without that I’m unlikely to feel quite as motivated.
The progress so far…
Although I have not yet put them online yet, I’ve created two starting templates:
- A very simple single-page website template that is using some very cheesy techniques. I’m almost ashamed to put it up, which is why I am totally going to :-)
- A slightly-more complex multi-page website, which is based on the code I made for my friend Angela a few months ago.
Behind the scenes, I’ve been pulling together the support technology:
- Upgrading my web hosting server, which means moving 15 sites from my old server to the new. The reason for this is that Media Temple, my hosting company, has a new server architecture that has PHP5 and MySQL5; my current server is still running the older versions, which are increasingly being phased out. Plus, the new server (dv) 3.5 has twice as much memory as the old 3.0 version, and the upgrade is free. However, it’s a multi-day process. This has occupied about 3 days of my time.
- Consolidating my digital media server organization and choosing a backup strategy. This has taken another 3 days of my time, consolidating all my disk backups and creating one easy-to-backup uber system while purging and recategorizing.
- Choosing a source control system. I’m planning on storing the web templates and other media using a public repository of some kind. I actually lost my old Subversion repository on the old (dv) 3.0 because I forgot about it. I still have sources on local backup, but I’ve lost all the history. It’s not a huge blow as there are no projects in active development there, but it does make me think I need a better repository backup strategy (Subversion doesn’t make it particularly easy). After evaluating the current hot source control solutions, I narrowed it down to either Mercurial and Git. And Git wins, at least in theory, because of its distributed nature (everyone has a full copy of the repo), its ease of creating speculative branches (which encourages experimentation), and its patch-distribution features. Mercurial has many of these same features, but apparently does not support named tags and because of that, loses my interest. Mercurial has the reputation, though, of being much clearer in its documentation and ease-of-use. This ordinarily would have been the big win, but after reading up on Git’s underlying design philosophy I’d have to say that I just want to be a part of it. We’ll see. I’ll be setting up the Git repository after getting it configured…I’m still wrapping my head around how I’ll organize that repository.
- Flip-flopping on how to document and store the materials online. The reason you’re reading this post now is that I’ve decided to go back to WordPress for this blog for two reasons. First, WordPress is a great blogging CMS, and I’ve really missed it. My main site runs off of Expression Engine, and while I like it now that I’ve lived with it for a couple of years, WordPress is much more useful as an inexpensive dynamic platform (read: free) for small business clients.
I’ve also registered the domain name agenceum.com, riffing off the word “lyceum”. This design business niche thing is kind of like the new media school I once imagined starting; perhaps at some point in the future, I’ll move some content to that separate domain. Just thinking ahead.
5 Comments
This is great, Dave. I’d love to see more of this sort of thing from you.
I can’t wait to see how this progresses. Follow Scott Jordon , CEO of ScotteVest clothing. He believes in this open policy of sharing and just launched a new website with tons of his Tribe followers giving input along the way- b/c he’s smart enought to ask.
I write this not to take sides, or even to advise you on a source control system (in my opinion, Git is the more powerful of the two), but your reasons for choosing Git are kind of strange:
Both Git and Mercurial are distributed, and with both systems, everyone has a full copy of the repo.
Git has “real support” for speculative branches, but Mercurial users just clone their repositories when they want to do speculative work – it’s slightly more onerous, but completely do-able. (And since cloned repositories use hard links, they don’t take up [any significant] additional space on your hard disk.)
“Patch distribution features” – I can’t comment on that since I’m not sure which features we’re talking about (admittedly, in neither system do I do much with patches).
Named tags – I wonder where you heard that Mercurial doesn’t support them? http://hgbook.red-bean.com/read/managing-releases-and-branchy-development.html
Mark: thanks for the words of encouragement!
Chris: neat, it’s good to know about other people who are doing something similar. Thanks!
Erik: Oh, that’s good to know…thanks for correcting my understanding. The real speculative branches thing is something that I like philosophically…especially deleting them permanently, though admittedly I haven’t experienced this in practice yet. My reasons are strange, yes…I’m essentially just going with the one that piques my curiosity more, and for now it’s git. I probably will try Mercurial at some point as well…we’ll see.
Sensible choice, then – by starting with Git, you’ll jump the learning hurdle earlier as opposed to later – it will be easier to learn Mercurial someday if you have the need, and in the meantime I think Git seems to be getting more mind share so you can’t really go wrong with learning how to use it.
I started with Git and then tried Mercurial. Overall, I like Mercurial a bit more as it has less conceptual overhead. But they’re both great, and both a huge improvement over Subversion.