@Rory The "line" between DBA.se and SO questions is pretty blurry. Basically, if it's an expert question, or you want feedback from actual database professionals, then put it here. Otherwise, just use your judgement.
Yes. It's not as good as it used to be, but it still pays disproportionately well. I don't think there are many contract markets in the world that pay better than the city - I think it even pays better tnan New York.
@SimonRigharts Change your name - eventually googlebot will forget about your old SO nick. That's what I did when I saw a certain SO posting rising to the top of google search results on my name.
Unfortunately said posting wasn't terribly complimentary about job interview processes, so I thought it wasn't a good idea for it to feature prominently in my search results.
I would like to see chat rooms indexed by search engines. To make this possible, the 'load older messages' button at the top of chat rooms would need to work without JavaScript, or else the chat room page should show all messages if JavaScript is not enabled.
@JackDouglas 13 hours ahead of the UK if they're on daylight saving, 11 hrs ahead if we're on daylight saving here. Occasionallt 12 hrs ahead as our DST and their DST aren't quite in sync.
Merriman keeps the NoSQL faith, though. He believes the New Aggregation Framework can be even simpler than using SQL as it's implemented in databases such as Oracle. "It's cleaner to build a query," he said. "If you want to build a query for Oracle, for example, you have to do string concatenation to do the SQL statement. We are writing a query generator."
And the thing with changing my name is that I want my SO/DBA.se answers (which are generally decent, I hope) to be one of the things that people see when they search for me
@JackDouglas No. He's full of shit. Stonebreaker had a fairly cogent critique of MapReduce. Essentially it doesn't have any concept analogous to indexes, so you have to process your entire data set to filter anything out.
Ask any Teradata weenie of you want a great diatribe about the failings of MapReduce as a database architecture.
@JackDouglas @SimonRigharts has the classic kiwi cultural cringe, worrying that he's good enough to do the work. I wasted several years in NZ for the same reason.
Once he gets here he will realise just how few IQ points are actually necessary to hold down an I.T. job in London.
But "you have to do string concatenation to do the SQL statement" - is he talking about Java/PHP/whatever? I don't do much string concatenation in PL/SQL or T-SQL
@ConcernedOfTunbridgeWells @SimonRigharts is coming to London?
@ConcernedOfTunbridgeWells Holding down a job once you've got it is a bear-chase situation (you don't have to be the fastest, you just have to be not-the-slowest)
@JackDouglas Absolutely - there's a chronic skill shortage in London. Combine that with the amount of cash in the square mile and you get a really strong contract market.
@ConcernedOfTunbridgeWells not really my point - I was referring more to the many thousands of completely inappropriate candidates pushed by agents who know next to nothing about the job they are putting them forward for
I have a table which fills out the rows I have put in there with spaces if the thing i put in there isn't long enough.
For example:
I have a string called
'ABC'
but the column is a nchar(10)
so the value which is put in there becomes
'ABC '
Does anyone know what the problem could b...
@SimonRigharts My working theory is that the contract rate reflects the market rate for someone who's any good at that skill, and if you're any good and working as perm then you're selling yourself short.
Possible Duplicate:
Encouraging people to explain down-votes
Allow users to leave an anonymous comment when voting
I have had four or five down-votes in the past few days on answers that I thought were helpful (and so did others, including the OP, in most cases). Of course none of th...
The system will pick it up.
I had someone blow their daily 30 votes on downvoting every one of my last 30 answers. A day later, I had my 60 rep back and account was suspended.
And my stalker got badges too...
No, no, I'm fine. People don't agree with me, think I'm being childish, whatever. There was one question yesterday with a bunch of answers and mine was the only one down-voted even though a few were less useful and mine was no less correct than any of them. Just seemed like sour grapes of one kind or another. Four other down-votes in about 24 hours, with no explanation and nothing obvious (except the one that was just a link to Erland's article that has been up for years)
The reputation points are absolutely meaningless to me. I am concerned though if (a) people are just being jerks to be jerks or (b) there was a legitimate issue with my answer.
(a) I realize I can do nothing about, but I don't know why anonymous down-voting seems to be encouraged, or at least why they have absolutely no interest in motivating people to provide reasons. (b) if there was an issue with my answer, I want to know what it is. I don't fling turd out there and if someone doesn't understand or if I've made some kind of mistake, I want to help resolve the issue wherever it is. Drive-by down-voting doesn't help anyone (including the googlers).
@AaronBertrand downvotes bug me for that reason too. Downvotes on UX bug me the most even though I have like...4. Most of the time someone comments how something could be improved (and I usually get the DV removed after fixing).
However will you promote yourself to system test then?
Sorry, one too many gigs where the locals manage to turn environments into a total melodrama.
One too many conversations where I had to resist the temptation to say: "Sonny, when I was doing your job I had 14 environments on the go and scripted the rollout so it all ran of a config file."
@ConcernedOfTunbridgeWells I don't doubt that. We're in the process of making that happen for our devs. We have two new hires making it so. Because I need to be able to deploy to a new site with minimum fuss, and if I can deploy with minimum fuss, we can upgrade with minimum fuss. And if we can automate deployment to a site, then I can automate deployment to my desktop.
All I really want is a easy to use local repo of the database that I sync to Hg
@jcolebrand That's one of the key benefits - no dev server, so devs don't trip over each other, and the whole system has to have a clean deployment and version control process.
Plus, you don't have to buy DB licensing for the dev server, so dev edition databases or OTN downloads can be used on develpers' personal workstations.
Even if you buy high-end kit like HP Z800s, a few 15K disks makes a dev workstation pretty zippy on database work.
In case you hadn't heard some of my previous rants about SANs, my old HP XW9300s will outperform some production server platforms I've deployed to on ETL jobs.
well my thought was to have a cheap server that i label as 'dev', and have local boxes running as replication slaves . our dataset is small, maybe 5GB useable data. Then have changesets apply to the dev machine using liquibase or something, which would trickle down to the local slaves.
@jcolebrand probyl database mail then if SQL Server 2005+. sp_OA% use was a hack to enable SMTP email in SQL Server 2000. Otherwise, you needed Outlook on your servers to have MAPI
@ConcernedOfTunbridgeWells I was trying to come up with a scenario a couple of days ago where the central server => slave local machines would be disruptive but failed. Probably a lack of experience.
@DTest If you roll out a change to the database schema while someone's in the middle of developing against it without warning then you can break their code unexpectedly. The database schema has dependent items. Also, if you have multiple workstreams you may have people working on code bases that go with different versions of the database that haven't been released yet.
A reference schema for testing or continuous integration is useful, but forcing arbitrary changes into a development environment without warning will create a mess.
Plus, you want the ability to have different development databases at different versions.
I think the place to have a centralised schema is on a continuous integration or test machine.
The way I do this is to always build test environments from your repository model in whatever form you keep it. This could be a set of table creation scripts or it could be done through a modelling tool like Powerdesigner.
Then you have a mechanism where a database of that controlled, versioned schema can be deployed to an environment, so you can work against a known version of the database.
This could be a set of DDL scripts, a restorable backup or whatever.
If the devs have a database server on their workstation they can take responsibility for local deployments - this also has the benefit of forcing them to learn some basic database administration skills.
If you build test environments from your controlled repository model or central schema and give the devs the ability to run up a test environment easily then they have the wherewithall to test against what people are expecting to release. Building test environments from the repositiry model means that testing is always done against a known database schema that's tied to what you're actually intending to release.
Background
I am working on creating a new development process for a small web team of about 4 programmers and 4 designers, with the obvious potential to grow the team in the future.
Our product is a central application that powers client websites that we also design and host.
Previously, we all...
@AaronBertrand If happen to be within shouting distance of Peter Shire, please apologise for my rudeness not replying to his email. FUE on my part when I moved my mail to Office365 and I've only just found it!
It's not specific on tools, but you can plugin your choice of tooling into the basic approach. I've used simple table creation scripts more often than not.
If you have a decent data modelling tool you can generate DDL from the repository model and store that in your source control system. I've done this with sourcesafe (gak), subversion, TFS and an obscure MKS tool based on rcs. About to try it with CA SCM, so we'll see how well it works on that.
@ConcernedOfTunbridgeWells alright, through it. It's a great overview of the process without specific tools used.
You've pretty much convinced me to drop the replication slaves for a environment deployment script. Also, never had any intention of automatically running patch scripts to production. That kind of thing needs a presence in case something goes wrong, imo.
@ConcernedOfTunbridgeWells I've got the sample data I mentioned. Struggling to see how it could be "visualised" as such. I'll prod it over next couple of days before I throw it your way to see if you think you could do anything with it
@MarkStoreySmith I don't have a car here in Bournemouth and it's a couple of hours into London by train. Weekends is probably the best time unless you fancy a trip down here.
If you can get out to Sunningdale on Sat with the data we can run it up there and take a look at it.
@JNK I do all the big shows every year. I was always afraid to speak at the big ones. Last year I o.d.'d on SQL Sats. And it was the first time I spoke at PASS. Will be my first time speaking at Connections in Vegas too.
@ConcernedOfTunbridgeWells Cheers fella, I'll get back to you tomorrow when I know what the weekend looks like. There is literally feck all data, 25MB sat in a spreadsheet currently. Its really the "how on earth can we make this look snazzy" kind of PIA that I'm looking at.
@AaronBertrand We met at the sqlskills event in london last year. Give him my regards.