Language:
switch to room list switch to menu My folders
Go to page: First ... 48 49 50 51 [52] 53 54 55 56 ... Last
[#] Thu Feb 04 2010 15:12:02 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Guy Marion, Codesion's executive vice president of marketing, told InternetNews.com that his company recommends that developers not use CVS anymore, as SVN has better support for merging and branching.

???

Am I the only one who get's it? Normally when everybody disgrees with me, I accept that I am wrong (eventually) but in this case, I think everybody's stoned.

I think the whole version control riot bias is an ego thing deep down.

I never played with git, but a friend explained it to me, and I listened to linus' rant about it, and I see what it does, and I just don't see how it's all that incredibly valuable.

Maybe I haven't worked in a large enough enterprise where I have to work with lots of devlopers, but even in so doing, I don't think I get what all the hubub is about.

[#] Thu Feb 04 2010 16:05:53 EST from fleeb @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


Have you had to do a lot of programming at a customer's site?

Git is valuable in a situation like that.

It's awkward to not be able to check out code (because the repository is behind a firewall), make my changes, then have to remember which files changed so when I go to check code back in, I get the right bits.

Git makes it so I don't have to deal with that kind of insanity. I just work as normal, and when I've finished all my code changes, I post what I've done (locally) to the main repository when I get back in.

I can still take advantage of source control's benefits locally, without tainting or otherwise screwing up the main repository.

I don't know about merging and all that, though; I have not had to merge code much, thanks to having to use SourceSafe all the time ('thanks' being relative; I think I might prefer merging).

[#] Thu Feb 04 2010 23:49:18 EST from IGnatius T Foobar @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

 

orkut,
but I guess that's not what I was thinking of.

I think I heard that Orkut is actually the single most popular social network in India.



[#] Fri Feb 05 2010 14:31:20 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Fleeb wins a point for git.

[#] Mon Feb 08 2010 10:43:02 EST from dothebart @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

me hates

for (rpos = (char*) Input; *rpos != '\0'; rpos++)
loops.

and if there are

foo(&rpos);

es inside even more.

Buffer overrun welcome.



[#] Mon Feb 08 2010 13:44:56 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

whasswrong with that?
It's somebody else's problem if they didn't terminate the string.

[#] Mon Feb 08 2010 14:09:38 EST from dothebart @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

its error prone and seems to lead to irretations regulary.

simplifying loop counters / positions and abort conditions seems to be vital to get an error free, maintainable program.



[#] Thu Feb 11 2010 19:37:12 EST from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


latest pet peeve: dealing with a legacy schema. this one has 5 columns on one side of a foreign key and only 2 on the other. Grr!

[#] Fri Feb 12 2010 08:22:52 EST from dothebart @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

static code analysis with gcc:

http://lwn.net/Articles/370717/

some nice pointers.



[#] Fri Feb 12 2010 09:58:13 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

latest pet peeve: dealing with a legacy schema. this one has 5
columns on one side of a foreign key and only 2 on the other. Grr!

That's funny.

[#] Fri Feb 12 2010 11:33:42 EST from Spell Binder @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

For those of us who aren't very DB saavy, what are the issues with having a schema like that?

If I were designing a schema, I'd probably want to place the foreign key in the last column of a record, but, other than aesthetics, I couldn't tell you if there was any real advantage to doing that.
DB Binder

[#] Fri Feb 12 2010 17:25:22 EST from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


Spell, the foreign key is defined sideways.

This is from our invoicing system. The primary key on one of the detail records is MID,SID,INVOICE_DATE. The primary key on the owner record should be (MID,INVOICE_DATE) but is actually (MID,INVOICE_DATE, {3 other columns that aren't required for uniqueness})

So the keys are sideways.
The {3 others} are value columns in the detail record. So I must use those {3 others} to retreive the matching record from the parent record, if I want to be guaranteed of uniqueness via the schema constraints. This also implies that if I query the parent record based on only a (MID,INVOICE_DATE) tuple, then I'm *not* guaranteed of uniqueness, but I should be.

[#] Fri Feb 12 2010 17:28:21 EST from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


The unique columns in the parent table should be a strict subset of the unique colums in the child table, which leads to a cleanly defined one-to-many relationship. What I have is something that's kinda undefined and halfway between one-to-many and many-to-many. Most likely explanation is that somebody generated this schema from a tool, and clicked the wrong number of checkboxes.

[#] Sat Feb 13 2010 16:02:23 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

This isn't your scenario, but one of my prouder moments was when I figurd out something that our dba did not,
We had a primary key on a table that was 3-4 fields longer than what was neccesary for uniqueness.
And we stared at it and stared at it and stared at it. And then it dawned on me. Actually now that I think about it, there's no way he could hav efigured it out. I knew the application, he did not.
Whoever set this table up, put the extra fields in the index, because they were the most heavily hit fields, so the database wouldn't have to go after the actualy data table most of the time.

Which reminds me of another flaw inherent to databases. In code you cna put a comment // this goes after the fields in the index for efficiency.

But in the database you can't. You just have to stare at it and try and figure out why it is the way it is.

[#] Sun Feb 14 2010 16:45:34 EST from dothebart @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Funky shit:

http://freshmeat.net/projects/keyboard-upgrade

some software to make your oldschool ibm keyboard USB enabled... with instant dvorak/colemak.... .conversion



[#] Mon Feb 15 2010 20:17:56 EST from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

But in the database you can't. You just have to stare at it and try

and figure out why it is the way it is.

Oh but you can. See COMMENT ON COLUMN ddl.

[#] Tue Feb 16 2010 14:50:49 EST from Ford II @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

Brilliant. Let's not do it the right way, let's do it the wrong way... BETTER!

[#] Tue Feb 16 2010 14:56:34 EST from LoanShark @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]


Right, the columns become a first class syntax object. Yeah. Lame. But it shows up in those nifty GUI schema browsers, which everyone will ignore, so it must be right.

[#] Fri Mar 05 2010 09:30:49 EST from dothebart @ Uncensored

[Reply] [ReplyQuoted] [Headers] [Print]

being fed up with

handle SIG_HUP nostop

typing on each citadel debugging action, I found out about ~/.gdbinit:

http://tinderbox.dev.gentoo.org/misc/dot.gdbinit

one probably should use that and add it there.



Go to page: First ... 48 49 50 51 [52] 53 54 55 56 ... Last