While the article is very well written and the author raises a few very good points, I disagree with part of it. Perhaps, I have a slightly different point of view about the subject.

Let me quote a couple sentences from it to analyze them in context:

“Five great programmers can completely outperform 1,000 mediocre programmers”

I completely agree with that sentence (told by Marc Andreessen to the article’s author for his book).

Mediocre programmers don’t add value, they usually subtract from it. They’re usually also mediocre Team players, making things worse.

Later, he contrasts that sentence with this one:

“If you were launching a technology or developing a product, would you rather have five great engineers rather than 1,000 average engineers”

While contextually they seem to be mutually exclusive, I happen to agree with that sentence as well, because I think there’s a not-so-subtle difference between “average” and “mediocre” that the author is simplifying. In my opinion average and mediocre are not the same.

Average engineers add value. If they’re also good team players, they can add a lot of value and they can surely outperform a smaller group of “Rock Stars”.

There’s an important concept that the article also doesn’t quite grasp or at least it’s not presented with the relevance it deserves: Teamwork.

Non-team-players are a bad idea, no matter how great as coders they are. The only way to overcome it would be to get them doing one-man projects, and those are an even bigger very bad idea company-wise.

However, the article links to another article that analyzes Barça’s (the Spanish soccer team) case. As a Barça fan and Barcelona resident I fell particularly identified with the example. Let me add a little bit of my view to the article:

Barça has the best soccer player in the world (Lionel Messi, I’d call him a “Rock Star”) and a lot other players that are way above great (Xavi, Iniesta, Pedro, Villa, to name a few). Most of them were raised on the Club, not bought from other Clubs. Barça also have a few more “average” players in comparison, but most of them are really, really good ones. I wouldn’t say that there’s any mediocre players on the Team.

Curiously enough, Messi played his share of matches with the Argentinian selection (also with a good deal of the best soccer players around, most of them currently playing on the best European Clubs and being extremely successful at it).

The outcome? Completely unremarkable. Messi didn’t score a single goal on the world cup and the team didn’t make it past quarter finals. Why? Because they weren’t a Team, just a bunch of -extremely- talented players doing a mediocre job as a Team.

You can assemble a team of great people if they’re good team players, there’s nothing wrong with it and it will outperform much bigger teams. Barça is a proof of it, but I’m sure there’s a good deal of other examples around.

I consider myself extremely lucky of working head-to-head with a lot of great people (both as Engineers and Team players) at 3Cinteractive and I can’t imagine a better working environment (BTW, we’re hiring). It proves to me that you can assemble a team of great people and accomplish a lot, as long as they’ve got what it takes to work together as a team.

In contrast, mediocre Team players will ruin any team as much as mediocre coders would slow it down.

I think one should aim for hiring at least average programmers that are above average (no concessions there) team members, or average team members that are above average programmers, but I’d ban mediocracy from one’s working environment completely. Oh, and average as in decent, not less

Note: the later article got it all wrong about the “Catalonia is not Spain” part. Those banners are a display of extreme nationalism/separatism with which I’m completely in disagreement.

This is what I’d call my “Programmer’s Grid”, depicting in an image what I’d tried to describe with words above.

And this is my perhaps biased opinion, based on my own experience. What do you think?

Alex

Also, I’ve added a multi-lingual plugin and started translating my latest post to Spanish. It will take some time to get all the older posts translated, but I’ll eventually get that done.

I hope this will help the Kannel spanish-speaking community of users. I know a few of them

It was about time to get my blog up to speed again.

Regards,

Alex

• Failover routes that you manually enable when needed.
• New binds being worked out, not ready for production.
• Test binds you use only from time to time.
• Etc.

This patch allows you to configure in which state a bind will start when booting (or when adding it with the add-smsc command). This is done by a new parameter on the smsc group called start-at-boot, which defaults to true if not present.

In short: just add the following line to an smsc group:

start-at-boot = false

and that smsc will start as “dead” and will require a start-smsc command to activate it.

Download the patch from here.

As usual, this is fairly untested code and should be used with caution.

Please try it and let me know if you encounter any issues.

Alex

This patch addresses that, adding to each smsc the same three counters Kannel uses for global traffic.

NOTE: The text and XML schema on the status page changed to accommodate the new counters, so if you’re parsing it from your code you’ll probably need to make a few changes to be able to parse it correctly.

Please download the patch from here.

As with most new patches, proceed with caution and test it thoroughly before using it on live systems.

Regards,

Alex

It uses two configurable parameters:

To use it, you need to add syslog-level and syslog-facility to the core, smsbox and/or wapbox group. You might also need to tweak your /etc/syslog.conf to suit your needs (I won’t teach you how to configure syslog here, you can Google it yourself isn’t it? ).

Please download the patch from here.

NOTE: This is fairly untested and it’s not part of Kannel’s main tree, so use it with caution.

Give it a spin and let me know if it works for you.

Alex

There were many things that can go wrong when storing data as text on a DB, specially if it involves non-ascii characters.

To fix that, Stipe Tolj recently submitted a patch that modified SqlBox to url-encode the message before writing it to the DB, and url-decodes it when reading from it.

That fixed the issue with special characters, but when you need to send binary messages you also need to deal with the udh field, which can be even more problematic (try writing a binary zero and you’ll know what I mean ).

To fix that, I’ve extended his patch to url-encode/decode the udhdata column as well. With this two fixes you shouldn’t have any issues sending special characters, wap-push and other binary messages either.

Both patches are now part of SqlBox and you’ll get them if you download the latest snapshot, or the SVN head.

IMPORTANT NOTE: If you’re relying on inserting messages on the “send_sms” table, you might need to change your code to url-encode the messages first.

Regards,

Alex

With the release of Kannel 1.5.0 Development Version on Oct. 20th., Sqlbox is now part of Kannel’s distribution.

This is really great news, because you’ll get SqlBox along with Kannel.

You’ll also get OpenSmppBox as well (an Open Source SMPP server implementation for Kannel).

You can get the new release on the Downloads Page as usual. Browse down to the Development Release section.

Hope you enjoy it!

Alex

I hope is a good time for Kannel developers and users to assemble and meet having a good time together as well.

If you happen to be at the Fira this year and want to meet with me just drop me a note in private.

We needed that functionality on one of our servers, so I decided to give it a try. However, we were using OpenSSH 5.3p1 instead of 4.6 (the version the patch was made for) and the patch didn’t apply completely.

After a couple of tweaks, I’ve managed to make it apply on 5.3p1 and, in the same Open Source spirit shown by the original patch author’s , here’s my modified version that applies against 5.3p1:

openssh_scp_sftp_atomic-5.3p1

Check the original patch post on PatchLog for instructions about how to build it and configure. There’s a lot more nice stuff on the site as well.

Nice job Mihai!

Alex

If you use dlrs (Delivery Receipts) with Kannel, you’ve noticed that they’re counted along with MO traffic as incoming messages.

This leads to misunderstandings on how much incoming traffic you’re really sustaining: For example on SMPP, if you request dlrs, depending on the carrier having it enabled or not, you’ll get one or two more “incoming” messages counted (dlrs are very similar to mo messages in structure).

After applying this patch, the status page will display incoming and dlr traffic on separate counter, either for the total amouns as in each and every defined smsc connection. It will also display separate load indicators, so you’ll know at all times if your inbound traffic is actually traffic or just old/delayed dlrs.

Here’s the patch:

kannel-dlr-status

As usual, please try it and let me know what you think.

Alex