Category: Gentoo

I’m not talking here about the second half of Golan-Globus productions (who brought you notable gems like Ninja 3: The Domination and Cobra, of course). I’m talking about the Grid Computing framework. So I’m taking a class at Boston University this summer about Grid Computing, and the prof. is going to have us work with the Globus toolkit. Now, since you can install the entire toolkit on linux (but only the java client on winders), I thought I’d try and ebuild it. Turns out, some people already have. Tried, that is.

OMG WTF. I mean seriously. How do you, as a package maintainer, deal with this. I’m talking about the upstream developers — how do they deal with this convoluted build system? Why package all the perl module dependencies with the actual package itself? People have heard of CPAN (gentoo users have it easy), you know. And poor Hanni (on the bug) — this is their first ebuild. My first ebuild was for yahoo messenger, and as simple as that was, matthewbk had to help me with it. So, shout out to Hanni for bravery. A big thumbs down to the globus upstream for an excruciating package (not the contents, just the way to install/configure the contents).

Someone on the bug suggested breaking the ebuilds up into separate components. I would urge the globus people themselves to release it as a suite of components and have mercy on distro maintainers who want to package the thing. The bug request has been in our bugzilla for two years now. Mainly because it’s a b*tch to package.

Update: I just found this article about someone’s exploration into grid computing. Good read.

So between Henrik, Daniel and mainly Diego, I’ve learnt so much about the autotools this last week, that I’ve actually come to like them quite a bit.

At work, we had a messy build system for the longest time. I un-messed it slightly by adding Makefile wrappers around it. But lately, the combination build system has been getting more out of hand. So I put together a protoype of an autotooled build system, and it’s working fantastically. I especially like make dist to create the internal releases.

Before I found Diego to bother him endlessly in irc, I was working of both, the autotut and (of course) the autobook. The autobook seems little out of date. I got furthest with autotut, and then with Diego, it was just like magic. So now we have a fully functional prototype of the new build system in place, and I’m looking to deploy it in a few weeks. I like the added benefit of sanity checking, especially.

At the very least, I have a sincere appreciation of the autotools now 🙂

Edit: I always appreciated them, but in the past it was more of an “at least it’s not Imake” type of appreciation.

So our counsel got back to us with their review of our application. We do have some changes and clarifications to make to the document (I need to find a PDF editor thingy, because the original form is a PDF). So far, Donnie and Lars have given some feedback. By the way, this all happened yesterday, but I did not want to crowd everyone’s feed readers with 4 of me. You can only take so much!

Anyway, I’m hoping to get all the trustees’ feedback by next Tuesday, so that I can make the amendments necessary to the documents and send it back to counsel.

So, this is now the 18th month that I’m still without my old devbox from los angeles. I really need to fix this soon. Because I can’t stand mutt’s interface (I prefer evo), I tend not to read -core, -dev or bugzilla email. I know that’s bad, but those are all too high traffic for me to handle with mutt. Now mutt-fans, don’t flame me, it’s a personal choice thing. Anyway, if you think I might be associated with a bug (and even if I’m in the cc), please send me a personal email or ping me on IRC. That way I’m assured of seeing it.

As some of you might have seen, I did finally do some updates to mit-krb5 and heimdal this weekend. The bugs assigned to kerberos team had been languishing for a couple of months, much to the dismay of quite a few users. At the moment the kerberos team consists of ryan and myself basically. Emanuele works on krb4 stuff mainly, and I have no idea where aliz has disappeared to. Between my short time for developing these days, and ryan’s divided attention to vim and others, kerberos is suffering. So if someone wants to step to the plate and help us out, we’d be appreciative.

I’ve been meaning to write this update now for a couple of days. So, the specialist at our counsel’s legal firm who deals with corporate law is now looking at our 501(c)6 paperwork. He has now all the paperwork together (including the Foundation’s articles of incorporation) to make sure it all meshes. As I say, this happened two days ago, so we’re back in the holding pattern, but we’re definitely still flying 🙂

Just a quick update to let you know that our application is still being reviewed by the legal team. I’m hoping to hear from them soon, so that we can begin our next steps. I wish I had more to say on the subject, but nothing’s really happened lately.

..is a question on quite a few minds lately. Overall the trustees list has been quiet forever. Seriously, we’re not talking about anything. Now, that’s not to say that things are not being done, because they are. As you know, Corey left the project. He was a trustee as well, and he was both the Treasurer and group contact for the foundation’s legal team. Two months ago or so, he finished up the 501(c)6 application, and mailed it to trustees for comment. A month ago he handed off the group contact to me, and so I’m in the driver’s seat for the application. There were not many comments on the application, so we decided to send it off to the legal team for their review.

So, as things currently stand, the Foundation’s counsel has us in a holding pattern while they consult experts and review our application. As soon as I hear back, I will report back to the trustees. I’m hoping that the process will now move along with haste.

Please stay tuned, and many many many thanks to Joshua for reminding me to blog about it. I must confess that this info as at least two weeks old, and thus I give my sincerest apologies for the late reporting.

Please do feel free to email me, snail mail me, call me or /msg me if you have concerns and/or further questions.

…the more of an asshole it becomes. Alternatively, the more things change, etc. What I will say is not intended as a remark against Daniel or anyone. It is merely observation. Mistakes were made, and they continue to be made. The sad thing is that they are the same mistakes.

In the days of drobbins (no link, you can google it), there was once an incident where a problem developer was acting up. Well, this problem developer was not malicious to gentoo. He made statements as did drobbins. The incident I speak of was sparked by an indirect quote that the problem developer made to someone else. When drobbins got wind of it, he ordered that the problem developer’s access be revoked. On a few occasions, I did disagree with drobbins (I know the perception about where I stood in those days, but disagreements were kept private). This occasion was one of them. I argued in semi-public (it was avenj’s presence) that the access be restored immediately, because the problem developer’s intentions were not known, and could have been misunderstood (misunderestimated, if you will). Thankfully, drobbins had the presence of mind to recover from his initial reaction and proceed calmly.

Proceed to today. We have a problem developer (I won’t link to him, but it’s enough of a circus and enough of a divisive issue, that if you’ve paid any attention in the last 5 minutes, you know who it is) of sorts now. He has an ongoing case. His case is under investigation by Developer Relations. In fact, the fact-finding portion has ended, and now the decision-making process starts. Yet, a small group of people have taken it upon themselves to revoke his cvs access. And that’s it. Nothing more can be done or said. No decision by the representative community (devrel, the council, etc). Just a unilateral decision within a vacuum, based on supposition of security. Granted, problem developer is a little immature, and tends to like stirring the shit, and indeed has done some stupid things that have been used as the basis for infra’s decision. The fact, however, is that this is now a community distribution. There are proper channels for this sort of thing. There is at least the assumption of civility, wherein people are informed (devrel/the council) beforehand. And the civility of informing the person with an email, not by having their commit failed. That is simply malicious (whether it was meant as such I do not know, but the perception of maliciousness is definitely there).

The thing that’s common in both cases is that the problem developer found out about it not by a nice email or irc chat, but because their cvs commit failed.

This is not a rant against infra, per se. It’s rather a rant against the general direction Gentoo has been taking in the last few months — let’s say 18 months or so. It’s the assumption of mistrust and presumption of disrespect. We used to be the polite ones (and the user community thankfully has remained so). Now, most of us are just assholes to each other and to the users.

Maybe it’s the size of the group. None of this seems prevaricated. It all actually seems like an inevitable consequence of growth. Have we become what we hated in the first place?

So, apparently, the banking industry has made new rules to dick customers out of their money. It’s not enough that the bank holds all your money already. Now they want to keep some of it for good. So if you post-date a check by 2 days, your bank can still cash it and then charge you oodles and ooodles of money for “overdraft protection.” The date is simply unimportant to the honouring the check part. No matter that you have a history of clockwork deposits with them, and that you have an impeccable banking history eith them. No, they would rather risk losing a customer (yes, Bank of America you have lost 2 customers) because they make a couple of hundred bucks by *not paying attention* to the date. That’s just a racket, ffs.

Because, even though their””software” doesn’t look at the date, it sure red-flags if you try to cash a check more than 6 months old. This from the “supervisor” on the other end of the line. So which is it? Does the “software” look at the date or not?

Honestly, this is stupid business practise, and I’m definitely taking my business elsewhere. Bank of America isn’t all the fluffy good neighbourly bullshit they try to sell themselves as. Bank of America is actually acting more like a petty thief. Goodwill actually does have a meaning, you know. You, Bank of America, could’ve let the thing to rest by not trying to tell me you’re doing me a service by returning me 4 out of 10 “overdraft” charges.

As far as I know, a legal signed and dated document is valid on and from the date of the signature (presumably with a reasonable limitation on its lifetime). Same for a check, if common sense would prevail instead of thievery. You sign and date a check, indicating its validity. I’m pretty pissed off with them at the moment, and I’m happy to talk about it. Had they done the honourable thing of apologising and returning the money to my account, this blog entry would have been full of praise.

Certainly the bank that does get my money will be. So far no complaints about the Bank of Oklahoma who has treated us with nothing but respect. We need local, unfortunately, for the everyday banking stuff, so we can’t be exclusive to them. Ah well. I’ll report back on the bank of choice in a few days. But for now, Bank of America is not my bank, they’re just a set of hoodlums.

mood: peeved at corporate greed, which apparently knows no bounds. flame away, people.