samuel

When I ran OpenBSD I used to block on repeated failed attempts(see: http://home.nuug.no/~peter/pf/en/long-firewall.html#BRUTEFOR...).

I used to change ssh's port, but not anymore. If a bruteforce attack is a threat to you, then you are better being compromised by a random script kiddie or spammer than by a targeted attacker looking after your data. Take that scans as free, low skilled, penetration tests.

I don't understand why stock Python doesn't include some sort of optional type annotation, as SBCL and other Lisps. Some years ago there was a lot of fuss in the Python community about including optional type checking in Python 3... what happened? At the end, I found most changes from version 2 cosmetic and not really exciting.

(I'm not the OP)

It's a free ebook about Prolog. I started reading it some time ago, and looked very promising, with a balanced combination between theory and practice.

But, as with any other book about Prolog I've tried to read I hit a wall, and got stuck. Maybe I'm too dumb to grasp it, or just don't have enough time. I took some classes in college, and remember that never really "got it", just the enough pass the exams. I have tried to overcome that a couple times since then but it's harder when you have a job and a family.

I forgot to add, in this case, check the DB. It could be some sort of deadlock...

That's the last resort: almost never you need to get that far(stracing/trussing usually do the trick), and it doesn't have to take "hours and hours" if you have done it before, specially if it's a locking issue. I don't remember if I have had to gdb apache(I think I have), but sendmail + milters for sure.

I don't doubt that your solution has been "battle-tested", but I assure you that apache + mod_php also has. Now if he has a different problem with that new setup what you will recommend then? Change again to cherokee or lighttpd? Or do you really think that nginx + fpm is a silver bullet?

I accept that spending some time cheap shooting randomly(raising os limits, changing threading model, upgrading to new version...) may be worth the time, though. But changing completely your middleware isn't a cheap shot in my book.

I wouldn't recommend to blindly change his setup. This is the sysadmin equivalent of changing lines of code almost randomly to fix a bug.

When there is a problem the very first step is to find out what's happening. Apache is rock solid software so there must be something in his setup or software that isn't right.

My line of action would be:

  - check the logs(apache and system ones), looking for errors or limits reached.
  - if no clue, try to reproduce it, same os, software versions. Make scripts to simulate the load.
  - strace the processes when they lockup (dtrace 'em in Solaris).
  - get a coredump(kill -SIGABRT), compile apache and their module with symbols (-g CFLAG), an gdb it.

I'm not a sysadmin anymore, but that's what I would have done when I was one.

Oracle DDL doesn't respect transactions, but at least you can wrap it into an autonomous transaction so it won't commit your main transaction.

Said that, it's a great feature. It allows you to deploy your stored procs transactionally, instead of the locking nightmare that Oracle's model may result into.

His book "Head First Data Analysis" seems a fairly gentle introduction to data analysis. I've read it and it's enough to learn a couple of tricks. Where to go from here it's harder to figure out. I'm refreshing my stats classes with the Pearson, and looking for a good introductory R book.

This stats course from Berkeley isn't bad, too: http://www.stat.berkeley.edu/classes/s133/schedule.html

What means "personal"? Do you do it mainly for the fun or for the money? If it's the second, it's a no brainer. Use the tools you know and focus on you problem domain.

If its an experiment or an exploratory project, well, use whatever appeals to you.

What about Nokia? They smartphones also sucks, and they have already sued Apple for IP infringement.

You can't be serious.

That guy is a sadic. It's OK to be scared. It's OK to confuse a camera with a gun, and shoot before discovering that it really was a weapon the hard way. That shit happens at war.

But aiming a severely wounded guy, begging him to take a weapon so he can shoot without permission it's sadistic. That guy is a sadic. Now we can discuss if he was born sadic or just the result of careful training, but it's hard to deny, watching the video and hearing the conversations that he wanted to shoot, and was having fun, the kind of fun a hunter feels.

Of course the US Goverment, the Army, whoever, is to blame because covering the thing up. But that soldier is a murderer, no question about that.

No, it isn't. The first { encloses the class and the second encloses the "anonymous constructor" or whatever is called. It's code executed at object creation, even before the "named" constructor.

I can't agree. What's revolutionary about the iPad? Only that isn't running a desktop OS? There have been Windows CE tablets for ages. The difference is realization, which is everything, of course, it's the difference between Altavista and Google. The iPad will be successful But revolutionary?

On other side, you have the first device that it's at the same time, a tablet and a eink reader. I don't know if it will be successful, probably not at the same scale, but certainly it's much more revolutionary.

Anyway, if anything, I think this device could hurt more Kindle sales than iPad ones.

If they would be serious about that, they would distribute IE8 for Win2k.

I don't know if C is the best answer for him. Ada has most of C advantages(unless you are doing "dirty" C tricks), while retaining type safety, exceptions, bound checking... (Haven't used it in years, though).

I'm in my nth effort to learn (Common) Lisp, and this is the implementation I've chosen, mostly because its portability (read it as I could run it in Windows, despite not being my preferred environment).

The only thing that may worry me a little at the improbable case of using it for production code is: strings are encoded in UTF-32!! That's 4 times more memory than necessary for my usual use cases...

I wouldn't call that "thing" a netbook, but a small laptop. The iBook was about the same size and nobody called it a "netbook".

For me "a netbook" it's something similar to a product really called Netbook, the Psion Teklogix's "Netbook Pro". It was a little laptop with touchscreen, flash storage, and an Xscale processor running Windows CE.

I used to work for a warehousing company, and warehouse managers loved it. They could browse the Intranet, run a TN5250 terminal, answer email, etc... at any point of the warehouse, carrying almost no weight and, the most important thing, with near zero waiting time. It just booted in seconds. It was outrageusly expensive, may be something like 2000eur.

I understand that's the device Google has envisioned for its Google OS (but much cheaper obviously). A non-x86 CPU, flash storage, may be a touchscreen... something just different from the laptop you would use to run Windows, not just smaller.

One more: mkdir(2)

Points 28, 38, 39 may apply here. I haven't studied this directive but its spanish implementation(IANAL, but if you work with personal data you need to know how to comply), and under that law, users should be asked for explicit permission unless Facebook had stated to them beforehand that they could transfer their data to a third party for a particular purpose at the time of the collection of the data. In that case, there should be a signed contract between both parties (Facebook and you), which states the purpose and duration of the treatment of the data. After that treatment, the third party should destroy that since it's not authorized to use it anymore.

Anyway, as everyone here knows, its pretty hard to enforce this kind of law.

Probably, it's illegal at every EU country which implements the 95/46/EC directive(http://eur-lex.europa.eu/LexUriServ/LexUriServ.do?uri=CELEX:...)

IMHO, the point is not that the site isn't flashy, but that it doesn't explain why would you want off-site and crypted backups in the first time. The site I would expect for a backup tool is one which scares the visitors to the hell, full of well known horror stories, and showing how tarsnap would have avoided them.

For example, Jeff Attwood, a well known and "smart" guy isn't he? Well, he also thought his hosting provider was doing adequate backups...

Astalavista? They had backups, but the cracker logged into the backup servers using the credentials stored in backup scripts and deleted them. Did you know that tarsnap may use a key for writing, another for reading and another for deleting?

It's true that being an UNIX oriented tool makes it a "niche" product since most companies use a combination of UNIX and Windows servers. But there is still a big chance of success among web owners running small to medium webs for a living on cheap VPS or shared hosting accounts on foreign countries. They should get very nervous every time they hear an horror story detailing how some site dissapeared because of a bad backup strategy...

My definition for ipv4 adresses exhaustion is: IANA and RIR's have no more unallocated blocks, so everything is left to the free market. If its consequences will be catastrophic or imperceptible is something I can't imagine. I would love to see Google's and Facebook's risk models for the next years, since they are those who could be more severely hit by this.

IPv4 address space will be exhausted(I know, I know, a cheap shot).

It's worse than that. There is a standard way of assesing programmers productivity and is the worst it could be: kLOC's.

LDAP is "Lightweight" only when compared with X.500..

I just would have used "b" to go backwards by words instead of characters. It's not optimal, but much more natural.

The regular expression example makes no sense for me. A regexp is a very specialized way of express what a pattern must match, opossed to a java code that express how to perform the matching.

You don't regexp's compact syntax. That's fine. You would have a point if you say that its more readable, let's say:

   1: EXACT <a>(3)
   3: PLUS(15)
   4:   ANYOF[ab](0)
  15: OPEN1(17)
  17:   EXACT <cd>(19)
  19:   CURLY {0,1}(23)
  21:     EXACT <s>(0)
  23: CLOSE1(25)
  25: END(0)

than

/a[ab]+(cds?)/

but it's much easier to introduce a bug in 20 lines of Java code that deals with an ad-hoc parsing than in a regexp, specially if you have to change the pattern.

Wasn't it obvious? I mean, this is applicable for almost any language, let it be C++, ASP, TCL or whatever isn't considered trendy anymore but it's proven and stable, and isn't completely unsuited for the task(COBOL, 4GL, RPG...)

Takes at the very least a year to became proficient in any language, and spending that time learning about your tools, that's is a luxury a startup doesn't have. Wasn't it pure common sense?

I don't think nobody is saying that is migration is easy or risk free. Some of your complaints I think are more related to how has been done the transition in a particular facility than generally applicable.

About culture, it's true that most doctors just don't type (lot of them refuse to do it). That's why they use voice recognition software or handwrite. But the love to see lab results as soon as they are ready, or prescribing drugs trough a GUI instead of handwriting prescriptions(this saves lives).

Disasters, you are completely right. That's why its needed to develop policies for working "offline". It's not easy, but can be done.

Propietary data formats. That's why Hl7 and Dicom standards exist. They're a f*cking mess, but everyone sort of support them.

I work in IT for my country's National Health System. I made the switch only six months ago so my domain knowledge is still very limited. Anyway, I have already seen how automation may greatly improve doctors and nurses efficiency.

For example, RIS/PACS(Radiology Information System/Picture archiving and communication system) systems may (and do) save millions euros a year only in radiographic film. I've seen figures stimating the ROI and its only 3-4 years. That without mention the advantages of having radiological data instantly accesible from anywhere, which opens the door to telediagnostics and telemedicine.

That's a very clear example. In other cases it may be harder to calculate real savings(if any), but not the improved care level.