Category: Uncategorized

Ken Thompson

Ένας καλός φίλος μου έστειλε χτες βράδυ με email ένα attachment:

ken thompson's nametag
ken thompson's nametag from a unix symposium in Greece ~ November 1987

Επειδή ο φίλος δεν μπορούσε να βρει κάποιον να του εξηγήσει πως βρέθηκε αυτό το name tag, ρώτησα τον ίδιο τον Ken Thompson και ιδού η απάντηση:

i dont remember exactly, but i think it was in the mid 80’s. i was invited by the government – department of science or technology or something like that. the occasion was at a time when the government was trying to require unix for sub contractors for standardization.

Ταυτόχρονα ρώτησα και τον Γιάννη Κοροβέση, έναν από τους πρωτεργάτες του Internet στην Ελλάδα:

Λοιπόν, ναι ήμουνα και παρών στο Ξενοδοχείο (Intercontinetal ?) γύρω στα 1989 μαζί με τον ΚΕΝ ήταν και ο B.Gopinath και ένας τρίτος που δεν θυμάμαι προς το παρόν.

Οργανωτής ήταν ο Γιάννης Κοζατσας του ΥΚ της ΓΓΕΤ του τότε (ΥΒΕΤ, Ε=Ερευνα)

Το πιο απλό γεγονός, μία απορία μπορεί να οδηγήσει σε κομάτια της Ελληνικής τεχνολογικής ιστορίας που είναι σχετικά άγνωστα.

Όποιος ξέρει / θυμάται περισσότερα για το γεγονός, παρακαλείται να σχολιάσει.

ΥΓ 2024/09/11 Στο twitter ποσταρίστηκε το σχετικό ρεπορτάζ από τον τότε τύπο:

Image

sendmail: Should I use $*, $+ or $- ?

When writing a sendmail.mc rule, you can use some operators on the left hand side, like $- (match exactly one token), $+ (match one or more tokens) and $* (match zero or more tokens). You may find yourself in a situation where for example you want to use a certain delivery agent for some of your users. Normally you would write something like the following:

LOCAL_CONFIG
Kmitsos btree -m -a.mitsos /etc/mail/mitsosusers

LOCAL_RULE_0
:
R$-  < @ $=w . > $*              $: $(mitsos $1 $)  $3
R$- . mitsos  $*     $#mitsos $: $1

The above example seems to be correct, right? But what if you have a user in mitsosusers that contains a dot (.) in the user name (for example yiorgos.adamopoulos)? Because the . is a token separator (see the OperatorChars definition in your sendmail.cf), $- will not match the name. So the correct ruleset in this case is:

R$+  < @ $=w . > $*              $: $(mitsos $1 $)  $3
R$+ . mitsos  $*     $#mitsos $: $1

I got bitten by this sometime ago, and that is why I am sharing it.

What I like about the NoSQL crowd

Although I am not a big fan of the NoSQL movement (mostly because many of its advocates use arguments I do not agree with) there are a few things that I like about the NoSQL crowd and I want to write them down*. Most of what follows stems from discussions through the years with our DBA and some friends who are members of the “Greek Database Mafia“.

For more than two decades the dominance of the relational model (even though no commercial system fully implemented it†) was undisputed. Nobody ever got fired for choosing a commercial RDBMS for an application, where instead one would look suspicious if one dared to propose something different. This situation is no different than what Rob Pike described in “Systems Software Research is Irrelevant” for Operating Systems:

  • For example it took 10+ years for the R-Tree to enter the commercial systems, although it was solving a real problem. In the meantime if you were lucky and your system offered extensibility you could write it on your own.
  • No matter how novel the system, for it not to be marginalized it had to have an “SQL layer”. No SQL queries, no sales. Provide an SQL layer and all innovation of the product stays unused.
  • Ones proposal for an RDBMS purchase had to be among three or four commercial products. Anything else would likely be considered “a hacker’s choice” because “We make money! We cannot go that way!”

You could say that databases outside academic research had come to a halt. You don’t believe me? Just ask Yannis Ioannidis who uses to say that “Databases are dead“‡ in the most emphatic way when he wants to stir things up in a conversation.

And this is what I like about the NoSQL crowd (== implementers, advocates and integrators) . They do not care about established standards. They are not afraid to experiment in a “real environment”. Some of them may focus on a single problem and solve it well. Others may aim at a wider range of problems. But no system is stopped from being developed and deployed because it not “SQL compilant” or not relational. And even though some of these solutions resemble CODASYLo, once again there is action in the field.

But please people, stop marketing them as a “one solution fits all”. For we will again end up in a stagnation era, just like when everyone was storing stuff in an RDBMS for lack and fear of better suited solutions. They do not invalidate relational systems. They fill in the gap left by them.


[*] – As I had promised.

[†] – By the way, did you know that GROUP BY works outside of the “relational box”?

[‡] – I have heard him say this while giving a speech.

[o] – “one usually gets a low-level record-at-a-time DBMS interface“, says Mike Stonebraker.

Open Systems Security – an Architectural Framework

“In the old days” when security information was scarce and many of us began shaping our security mentality (be it white, gray or black) by reading “Improving the Security of Your Site by Breaking Into it” and the Computer Security FAQ and running tools like iss and Crack. I think it was there that I first read about Arto Karila’s PhD thesis. Even though it is an OSI based document, it helped understand basic concepts. However there were two problems with the document:

  1. It was hard to find, and
  2. It was in a weird PostScript format that even modern versions of ghostscript refuse to display.

With the help of a friend I managed to transform it to PDF and upload it to Scribd: Open Systems Security – an Architectural Framework

Of historical value mostly.

Update 2013/04/13: Now available at https://github.com/a-yiorgos/karila

Device drivers in Java?

The following tweet by @DSpinellis:

#USENIX @AnnualTech M. Renzelmann Decaf: Moving Device Drivers to a Modern Language (Java). He says performance impact is 1%

talks about “Decaf: Moving Device Drivers to a Modern Language” which describes a system where large parts of a driver can be written in a better language than C, the example here being Java.

I was certain that this was not the first time I had read about such an idea. This weekend I was able to go through my archive and find out the reference. Back in January 1997 in the NT Insider (Volume 4, Issue 1) Peter Viscarola, while criticizing the multitude of startups founded by anyone who could code a Java applet (this was a pre-dot-boom era remember) wrote:

It’s obvious that we are missing a real opportunity here to capitalize on the convergence of these trends. We need to immediately fund a start-up company to develop a package for writing Windows NT drivers in Java. THINK of it! We could have processor architecture independent device drivers that don’t even need to be recompiled in order to support X86, PPC, and Alpha machines! Amazing! We could create a visual driver development environment, complete with cute animated assistants. And, the drivers could probably have a visual component to them, so you could actually see your toaster-oven driver doing its work. Cool! THEN we could all be challenged, and have fun, and get rich at the same time. Wow! Why didn’t I think of this before?

It would be nice if we could see Peter’s views on the subject 12 years later.

re: The Humbling Power of P v NP

Some engineer out there has solved P=NP and it's locked up in an electric eggbeater calibration routine. For every 0x5f375a86 we learn about, there are thousands we never see.

In “The Humbling Power of P v NP“, Lance Fortnow urges theorists to try and solve P v NP “not because you will succeed but because you will fail” . This is the Kobayashi Maru character test for theorists it seems.

So what about non-theorists?

My answer is: So what if a problem is NP-complete? Does this mean that we are going to use that fact as an excuse to not solve it, or present a lousy hack as a solution? Or do people think that such problems do not come along the way of “a real professional”? They do, but theorists are trained to recognize them when they see them.

Just like theorists then, “practical computer people” must try to solve (using whatever tool they see fit) an NP-complete problem (like the TSP for example). Not because they will solve it optimally, but because there will always be a better solution. And by seeking it and understanding that “computation is a nasty beast” they will become better programmers professionals.

Update: You should also read:

Asking Questions…

We have seen that nobody should be afraid to ask a question. One of the first lessons I got from the USENET is that “Silly questions are the ones never asked”.

Today’s snail mail included the latest issue of NT Insider where Peter Viscarola in his column (Peter Pontificates) deals with the whole “asking a question” issue again:

Being a noob excuses stupidity. In fact, being a noob totaly means asking stupid questions. However, being a noob does not excuse lack of engineering discipline. As an engineer, I simply cannot understand how a fellow engineer can ask a question without at least attempting to put their question in its proper context.

How rightfully said.

I have a dream…

A Simple Dream

Everyone working in configuration management would like to find a way to simplify and perhaps obsolete this dreary job. The ‘impossible dream’ is that everyone writing configuration scripts can use the work of the whole community instead of re-inventing the wheel for each site and purpose. But so far, while configuration tools proliferate, it has been difficult to convince people to distribute and maintain reusable scripts for configuring systems and implementing common services. There seem to be “too many options,” “too many system dependencies,” and “too many site-specific assumptions.”

This was written in 1999. Then the XML holy grail crap became popular and managing organized complexity became even more painful.