Older blog entries for dmerrill (starting at number 102)

I woke up this morning to unrecoverable ECC errors on my primary server, and ordered some new RAM. It took a bit of hunting to find just the right kind, since AlphaServer 800s have been out of production for awhile. Trying to navigate Compaq's website gave me a headache.

Taking advantage of the kick in the butt, I'm installing the LDP services that were running on my machine onto the iBiblio machine which was set up for that purpose awhile ago. I have data backups, so that's not a problem.

I quickly installed Debian on my old firewall machine which has just been sitting unused for two months, so I can get my mail again. Since I should have RAM for the Alpha in one week, I'm just going to let my personal website stay down until then.

Oh, I almost forgot. I have a JOB, coding free software. For several year's I've been hunting it, and now I have it. I start Monday! Woo hoo!

Lampadas development is proceeding quickly and lots of loose ends are coming together. I still expect to go to beta at the end of the month.

At this point, it is pretty much ready to manage meta-data and coordinate work, just like the old LDPDB. It has all the same meta-data support, and more. It can generate a dynamic or static site, with a simple CMS included, so theoretically we can generate the LDP website directly from the meta-data in Lampadas. If configured properly, it should look almost the same as the existing LDP site, but will be completely localized. (Right now we have translations in German, French and English, and need translators to add support for others.) The Lampadas dynamic site seems to be pretty solid, but will need fine tuning and optimization before it is ready to run a site as popular as the LDP.

The Gnome Documentation Project is looking at Lampadas for managing their work, particularly translations. I need to make some changes to fill their needs, but that's no problem. I am quite happy to do it, and as they are pretty sharp folks, their input will make it a better system. I hoped from the very beginning that it would be adopted by some other projects. Now that the day is approaching, I am very excited about it.

I'll spend the rest of the month tweaking things to satisfy the LDP and GDP volunteers so it will have whatever features they consider essential, and fixing bugs of course.

Future releases will add CVS support and interactive features oriented more towards readers. Those were the features I originally started Lampadas for, but I think it is good for Lampadas and good for the doc projects to start with this base set of features and get them tweaked until we're all running smoothly on it. Then we can worry about the more whizz-bang stuff.

I'm still looking for a job. I had a great interview on Friday with a company who wants a programmer to hack on SquirrelMail and some other Free Software code that they run their business on. I think they will make an offer Monday. I sure hope so; the job would be really fun. The pay wouldn't be great, but that is not so important to me as enjoying the work.

I'm interested in your Dublin Core editor. I support OMF, which is based on it, and I have a web-based editor for the meta-data in Python -- see demo.lampadas.org. It isn't exactly an OMF editor, but I've been thinking about making it more compliant. At a minimum it would be interesting to see what you're doing.

The LDP will soon include five languages: English, French, Brazilian (Portuguese), Spanish and Korean. We decided awhile ago to actively work on bringing the translation projects, which traditionally have been separate efforts, into our fold as our equals.

Within the next few weeks and months, these sites will gradually be integrated into the LDP. This is really exciting for all of us, but there are a lot of technical hurdles to work through.

This is an extra incentive for me to have Lampadas ready soon. And also, the Gnome Documentation Project folks are seeking a solution to their management issues, because their DocTable has been broken for some time. Lampadas will do what they need, but it isn't ready yet and they need something soon. So, another incentive for me to get a 1.0 out the door soonish.

This weekend and today I've gotten huge amounts of Lampadas code banged out. I feel like I'm starting to get the hang of Python. The timing couldn't be better.

I posted a demo running the current cvs code at http://demo.lampadas.org. It is populated only with default data right now, so you can't really see much. I'll update it occasionally as I finish significant features.

Estimated beta date is still on schedule: end of July.

Still plugging away on Lampadas. I created an entirely avoidable problem on the mailing list by acting like a fool. Big surprise, huh? So here's the gist of what happened...

What I did was complain about people adding copyright claims in CREDITS when they had submitted very little code. Of course the claims are entirely valid and I had no right to make the complaint. So I made a fool out of myself and had to apologize.

I probably shouldn't be advertising my own foolishness, but I figure maybe someone else can learn from my mistake. I also think one should admit to one's mistakes, learn from them, and then move on. So that's what I'm going to do.

Bite me.

I mean really. Your sarcastic attitude and whining are not going to make our documentation better. I know as well as anyone that we have a lot of work to do if we are going to get our documentation up to the standard we all want it to meet. There are many problems with our documentation infrastructure as you say, I'll not deny that, although you completely misunderstand the nature of the problems. Hint: it's not the license. But it pisses me off when smart asses like you criticize in such completely unconstructive ways.

Hundreds of volunteers give freely of their time for the LDP and many other projects, all in an effort to make things actually better. What a concept - see a problem, fix it! It seems you'd rather see a problem, bitch about it!

Instead of whining like a spoiled little child, why don't you actually *do* something about it? And no, the Anti User Hostility Documentation License is not "doing something about it".

Oh yeah, that's right. It would require you to work instead of complain. Never mind.

Today was my last day at work. Now I'm officially unemployed. The market does really seem to be picking up here in DC though. I have about 3 or 4 active prospects, so hopefully one will come through. Two are very interesting. Two would pay the bills.

Certed njohnso: who posted a good summary of his Free Sofware work. Would that all Advogato newbies did that!

Announced Lampadas: in this week's LDP Weekly News.

dyork: To what do you attribute your success in getting contributed docs? Whatever you're doing, you're doing it right.

ndw: Added the DocBook Wiki to the LDP's resources list, and added my own DocBook Processor to it. I have some other stuff to add as well. It looks like it will be a great resource, thanks!

It looks like Greg Ferguson is interested in ASPSeek.

I'm only surprised nobody asked before, and of course your speculation is pretty much spot-on. Josh holds the domain, and there's been a bit of a falling-out. Here's what happened.

He told us awhile ago that he had been contacted by many different people about buying the domain, and he had met with them about it. And had signed NDAs with them, so he couldn't tell us who they were. He said he was simply interested in the dollar value of the domain because he is a businessman.

It made all (I think all, maybe just most) of us very uncomfortable that this had gone on without any of us knowing about it. We feel that any volunteer only holds such a thing in trust for the group. At one point Joshua said that he owned the domain, which made us even more uncomfortable. So we asked him to transfer the domain to iBiblio, our sponsor, a project of UNC's Library Sciences and Computer Science departments.

He said he doesn't intend to sell it, but he refuses to release the domain, despite the rest of the staff asking him to do it. So we are concerned about the situation, and figure it is better to move to a new domain now rather than have the domain pulled out from beneath us if he decides to make a few bucks off of it.

If nothing happens with the linuxdoc.org domain, it will continue to work forever. I don't know where you got the 5 years from. Our other services, such as mailing lists and cvs, are off the linuxdoc.org domain already, and we hope others will use www.tldp.org from now on.

The whole situation sucks rocks. Big craggy boulders, in fact, along with the proverbially corresponding hard place. And us in between.

I've a demo up at www.lampadas.org. It is development code. No bug reports, please, I probably know about it already. Comments and suggestions are most welcome, however. Oh, and don't expect too much yet.

Some guys from the French LDP are working on a Python port, which I'm all for. What's online now is the Perl code I did myself. Current work is primarily in i18n, cvs integration, and loading of meta-data into the database from source files that provide it (DocBook, and LinuxDoc to a lesser degree).

It has also been suggested to use Zope, but I am not convinced of the benefit.

Plug, plug, plugging away at Perl module-ization. The LDP Database was client-server -- simple scripts that read from the database and generated HTML. The current design is much, much more modular. There's an underlying library that implements an object hierarchy, then an HTML module that writes HTML primitives or "widgets" (combo boxes, etc.), then a layer that generates tables (such as a table listing the users, a table listing the documents, a table for editing a document, etc.), and finally a topmost module that generates complete pages from the lower level elements. It sounds like overkill, but I am loving the flexibility it gives me.

And the library module should be useful in doing other things with the data in the future (like writing out XML).

I'm being very careful to keep the system working at all times, as I gradually push bit by bit of the code down deeper into the library structure. This is a very good exercise, and is much better than doing a complete rewrite from scratch.

I'm making lots of progress. I'm perhaps half way through with this part of the project, after 3 weeks. There is much more left to do. But once I have the code in a more modular state, implementing the additional features should be much easier. It was starting to feel like a house of cards. Now it feels much more solid, and the web interface itself is hugely better. Big wins all around.

Meeting a few Debian developers here in DC tonight for a mini-keysigning. (I just started the process of becoming a Debian developer officially). It will be nice to meet some more local hackers.

I understood. I was just ranting along with you!

Finished merging user/editor/maintainer tables into a single username table, and updating the web interface and other scripts. Next: dropping mod_auth and using cookies instead.

93 older entries...

New Advogato Features

New HTML Parser: The long-awaited libxml2 based HTML parser code is live. It needs further work but already handles most markup better than the original parser.

Keep up with the latest Advogato features by reading the Advogato status blog.

If you're a C programmer with some spare time, take a look at the mod_virgule project page and help us with one of the tasks on the ToDo list!