| My stuff is out there to be used... |
[Jun. 23rd, 2009|10:59 pm] |
Want to see what I've been slaving on over at the YouTube? There are a few reasons why it hasn't been announced yet, and I'll not say why, but I'm all about building communities at YouTube. My work may not change the world, but maybe the communities that are built with it could make it a better place.
See the future of YouTube Groups... er... I mean Collaborative Channels... oh wait, that branding hasn't happened yet :P . It's a completely re-done back and front end for a video-centric forum. You can discuss videos, post videos, rate videos, and even watch as those ratings alter a video's location in group playlists automatically.
There are really only a couple things about the implementation that are really innovative, but there are some nifty ways to interact with discussions and videos that I think makes them far easier to use than the old groups implementation.
See them now at: http://www.youtube.com/channel/
Also, welcome back Spinal Tap: http://www.youtube.com/channel/spinaltapcontest
If you find a bug, I probably already know about it, but go ahead and comment here (they are all screened).
This is totally not an official release statement, but my main project at YouTube is switching in the next week or two, and I didn't want to *not* tell people.
Also, my team lead is leaving for a new startup, so I've got a bit of a mental block this week while I sort out not working with a great engineer and great guy every day. We'll miss you, Nick. |
|
|
| (no subject) |
[Jun. 27th, 2008|11:52 am] |
So, this is the end of my second week working for Google in Santa Monica. More specifically, I work on the YouTube team.
I get access to a lot of really interesting statistical information, like what videos are watched primarily by females, males, etc. It's all really interesting, and if I was a sociologist, I would do a comparative study about what people like to watch (backed up by a huge amount of data). Then again, I can't really share the information I have discovered, even if some of it is pretty obvious if you pause to think for about 5 minutes.
Back to work!
Update: Also, Stan Lee is going to be giving a talk here next week. |
|
|
| (no subject) |
[Jul. 14th, 2007|08:18 am] |
I know, it's been a while since I posted, but like Annie, I've been busy. I started "part time" with my job at Networks in Motion before we moved, then went full time this week. Everything is going pretty well so far. I've been reading huge amounts of source code, getting my hooks into it, documenting whatever I can, etc. The people I work with are great, and I've been able to help a few people with my expertise already.
While I would still like to teach at a University or College, my publication record and inability to relocate for the next at least couple years severely limits my mobility. And since I don't see any ICS Theory faculty retiring, nor do I see ICS hiring an ICS PhD student to be in their theory group (without far more significant research than I have), I suppose I will stick with industry for the time being. There are worse things than getting paid to write software ;).
I think I'm pretty much done with my dissertation (still need a good abstract :/, may want to offer shorter captions for the table of contents), though I need to send it to David for review/comments/approval. For future dissertation writers: it helps to look at the writing style and organization of other dissertations if not in your precise field, at least in the same area. I was going around in circles before I read the first dozen pages of five different dissertations from a parallel processing group; it gave me a variety of things that were obvious what to do and what not to do.
I hope things are going well for everyone. |
|
|
| travel, moving, etcetera |
[Jun. 24th, 2007|05:53 pm] |
After a 5 day trip to Michigan, we are back. We took pictures (Annie's has just started a flickr account, we'll see about posting some of them at some point), spent some time with Annie's family, ate some good food, and got some sun.
The cats seem to be ok, Annie's lamenting her need to be at a coworker's apartment at 7 tomorrow morning to do some work, and I'm looking forward to my first day at work*.
We are both looking forward to moving into our new place this coming weekend, at least in the "it will be nice to be moved in" sense, the actual moving will not be pleasant (it never is). We do have a 17' U-Haul for the better part of two days, and will likely be offering beer, soda, pizza and sushi to helpers. Annie and I will likely post later this week about where and when you can show up to help (and claim your reward).
Related to my first day of work, the most direct route there is on the 73, a toll road ($4 each way without a fastpass). I will likely end up just getting a fastpass and using the toll road (it seems as though it will save 10 minutes each way, which is an overall profit), but for variety and to get a better feel for the area, I'm going to head in without using the 73 tomorrow. Using Google and Yahoo maps, neither offer the ability to avoid toll roads (Yahoo offers the ability to avoid highways). Only Mapquest allows you to say no to toll roads, highways, or seasonally closed roads. Though to be fair, Yahoo allows you to define a multi-hop route (if in a slightly clunky manner to bypass tolls), and Mapquest has the worst interface for seeing directions, entering addresses, and the worst search for addresses. Come on Google, Yahoo, and Mapquest, fix your maps!
* I forgot to mention publicly that I will be starting a job on June 25th (part time for a week, then a week off to move, then full time starting July 9). I don't know how much I should or should not mention about it, but the pay is good, and it seems like I will be doing full-time development in Python and/or C. Don't get me wrong, I would still very much like to teach, but I just don't have the paper publications to make me qualified right now (I've got a journal version of a conference paper that needs to be submitted, a paper I'm going to submit directly to journal, and Andreas seems to think that a linear time variant of a solution to register moves I offered is enough for a paper). |
|
|
| Working... |
[Jun. 6th, 2005|10:28 pm] |
Hey everyone, so I'm up in San Jose, doing work, and I would be reading and commenting on people's journals when I have free time, but unfortunately, my boss' cablemodem service provider (Comcast I believe), has been having serious service issues. Doing a traceroute tells us that two hops up their routing jumps from 12ms to 1200ms (that is 1.2 seconds) roundtrip.
Apparently SBC in this area has started to offer $15 DSL, which is a screaming deal. My boss is talking about getting it, keeping his cablemodem so that when the current service is acting up, he can swap the cable and call it good.
Anyways, it comes down to me not being able to surf the net much at all because it is so slow. I can get email all right, but that is about it. Surfing on a modem would literally be faster. |
|
|
| Good news... |
[Oct. 15th, 2004|05:59 pm] |
I am finally able to tell people about what I have been working on in the last 5+ months.
I built the back end of Affini, A Network Of Trust.
Aside from incoming and outgoing SMTP, and a small Perl daemon, I built everything that handles email at Affini. The "Inbox" and "Queue" on the site both query my backend. The Butlers that hit your mailboxes and do the right thing, are mine. If you are seeing email information, or manipulating actual emails, that is a query to my search engine. I've written all but one of the protocol handlers by hand in an asynchronous style (hundreds of connections without a problem), and I estimate that it can easily handle 25,000 users/processor (I've simulated 50,000 on my laptop). It can be scaled to multi-processor machines, and multi-machine clusters.
Oh what less than 10,000 lines of Python can do.
How sweet it is.
Check it out, take a look. Tell us what you think. |
|
|
| Brain working good. |
[Aug. 26th, 2004|07:11 am] |
After arguing with my boss about various aspects of the project, and being basically wrong on all fronts, I was a little frustrated last evening (but not much so, the guy's smarter than I am). I worked for around an hour until around midnight implementing and testing a piece of the infrastructure. He came out and I showed him it, and he had always implemented the thing another way. He pulled out "Data Structure Techniques" by Thomas A. Standish, and suggested I give it a read, and that he may have been doing it wrong all along, at which point he went to bed (I'm staying in his son's bedroom).
I gave the relevant portions of the book a read, and discovered that while my way was 'better', it had no guarantees, and that a third method beat us both. "Good enough." I say, and go to bed around 12:30AM.
At 6:45 this morning, I wake up from my slumber for no apparent reason, and get to thinking about the project. Nothing sparks my interest to keep me up further, so I decide that I should probably think a bit about the "research" I've "been doing" for my advisor this summer, as I'm in San Jose working, skipping a meeting with him. My brain walks down two reasonable paths of thought simultaneously, and I have a eureka moment (the instant in time where you know you are seconds away from a real discovery). One minute later, and I have the crucial insight to solving a problem that has been just outside of reach for months.
Now, there's something you should know about my advisor, and if you have a moment, I urge you to take a visit to his website. Take a peek at his CV. Yeah, those papers are heady, I don't care who you are.
About two weeks ago I met with him and said that we couldn't prove anything substantial about the approach I was taking. He started suggesting a few options, and I asked if he had any link to papers on the 5 or 6 things he suggested, he said they all boiled down to a single thing, penned it on the whiteboard in 30 seconds, at which point I had said, "Yeah, I may be able to do something with that." So I spent the proceeding week and a half not doing anything with it, because I didn't really have time, and the direction he offered was the critical bit of information that I needed to solve the problem.
The real question I pose to those who know David, is whether he had already solved the problem, but he wanted me to (because he's such a righteous guy). I may have to ask him that one day.
So here I am at 7:40, in my PJ's at my employer's house in San Jose, CA, typing up an email about my simultaneous not being smart enough, and being smart enough. I should email my advisor and change into real clothes.
Good day to everyone. |
|
|
| Stuff. |
[Jul. 18th, 2004|10:55 pm] |
Annie and I are going to be in MN from July 28 - August 1. Or really, we are landing the evening of July 28, and flying out the morning of August 1.
There may be some sort of 'get together' in or around St. Paul on Friday, July 30 or Saturday, July 31. Worst-case scenario: we loiter somewhere with a cooler of sodas, chips, etc. Heh.
I'm off to San Jose tomorrow morning until Wednesday evening to do some last-minute integration before our db guy heads off to the San Diego comiccon with his kids, then when he returns, Annie and I head off to MN. Apparently vacations are why "nothing happens during the summer" in many businesses
My plane takes off in 9 hours, so I should be getting to sleep soon. |
|
|
| The 4th. |
[Jul. 5th, 2004|09:40 am] |
Annie, I, my former roommate (who invites us) Michael Shafae, and a handful of other people go to a local park to have some BBQ'd meat with various other good things while we relax and take in the 4th of July afternoon.
While my roommate's plan was to show up early, relax in the shade with a newspaper, expecting everyone else to start showing up at noon, Annie and I were the first ones there (after him) at 1PM.
Michael invited a very nice postdoc researcher in compter graphics (what Michael studies), and he showed up 15 minutes after us.
Other people trickle in at 2:30, 3, 5. Among those that came reltaively early were Michael's uncle, his wife, and their two daughters (one was 22 or so, the other perhaps 13). I would offer names, but I suck at remembering names. They moved to the LA area from abroad around 9 months ago, and aside from being relatively new to speaking the language, are doing relatively well.
While I spent the early part of the afternoon cooking meat on the grill (15 brats, 4 chunks of frozen chicken...from which a funny anecdote spawns from), Annie talked with Michael's relatives, each teaching the other a bit of their language.
A bit later, Kevin showed up with a fellow I've met a few times, named Chris. Now, Since school has let out, I've not seen or heard anything from Kevin, and on Friday night, I only heard a 3rd party report saying "I got some email from him a few days ago". It turns out that for the past few weeks, Kevin has been busting ass on a paper for a conference that is due today. Heh, I was afraid he didn't like me or something, and here he's just busy as hell. I hope the paper gets in, that'd be fucking cool.
Even a bit more time later, Michael's brother Matthew shows up with his girlfriend (again no name here, I suck).
At this time, Michael's uncle gets hot to play some soccer. They toss a frisbee around for a bit, and we wait until around 5, when a couple other people show up (whose names I can't remember, but they live in LA, and were very cool). While we were waiting, a few of us kick a soccer ball around (perhaps a bit too much).
When the last two of us (me and the postdoc) decide we should take a rest (we had been going for like 45 minutes), 5 minutes later the game was on.
I'll save you from a drawn out discussion of the game, suffice it to say, my skills were perhaps of slightly higher quality than they were when I was 9 (when I last played), if only because I played all too much hacki-sac during the summers of '98-00. Those that played include: Michael's uncle, brother, brother's girlfriend, uncle, youngest cousin and the postdoc. Everyone was a at least a bit better than me, and Michael's uncle and brother were quite good. Michael's cousin that played was a bit shy (so didn't try to make many goals), but was good for passing.
In any case, after perhaps 1-2 hours of playing soccer, my body is now feeling it. Calves, thighs, yowser. I'm going to be sitting as much as possible today.
After soccer, we bsed for another hour and a half or so, then we all went our ways. Most to Newport Beach to see the fireworks. Speaking of which, we sat on the roof of our building, and caught glimpses of 8 displays of fireworks. The closest one was Newport Beach, but was on the other side of a hill, so we didn't get to actually see much of it. The next closest was pretty good, and we could actually watch two other displays in that same direction. Cool.
After that...not much else.
I did get some time to do some contract work, and this morning I'm going to be testing the various pieces together. 6 Pieces of software, interacting via various methods, yikes. There's really only one piece I haven't really tested, but it is a fairly major piece (using some very nifty data structures), and is the brains for two other pieces.
Enough from me for now. |
|
|
| The goings on |
[May. 25th, 2004|11:11 pm] |
I'm still waiting for the results of the Phase II exam ("you pass, congratulations" or "you are teh suck"), and in the mean time, I'm keeping myself busy...with homework (computational complexity...my brain hurts), projects (modifying a processor simulator is "fun"), work (I signed a NDA, so can't say anything, but it is quite fun; algorithms, data structures, implementation - oh my), reading papers (gah, so many, brain hurts), and trying to stay sane (crazy going slowly am I?).
The project looks like it will take another 6 hours or so, which isn't that bad, and considering it is not due for another week or two, I'm in good shape.
I'm almost done with the current homework assignment, which just so happened that once you stared at each problem and looked a little deeper, didn't seem so hard. So, either I'm understanding what is going on (possible), or I'm so completely clueless I've oversimplified the problems (also possible). When I get the graded homework back 1-2 weeks from Thursday, I'll let everyone know.
Paper reading is going slow. I had chosen a certain topic to read, research, and write a survey paper on (the disjoint set-union problem), but it got to the point where I wasn't sure if I could add anything to the body of knowledge, and really which portion of the thousands of papers I should be reading. Then I generalized the topic into the various set algorithms and data structures (building a pseudo-heirarchy), and tried to decide on that. Honestly, I can't remember which topic I told my advisor that I was going to do a bit more reading on to decide, because the week following was last-minute preparation for taking the Phase II. Gah. I should write more things down. Lately, I've just been trying to finish those papers I started previously. Brain hurts.
Well, it is getting late, and I need to give a midterm tomorrow morning at 9AM. Goodnight everyone. Try to sleep well.
P.S. My new laptop is fucking sweet. I can't wait for the memory upgrade (256 ->768) to come in. After the memory: either an extra battery (swap the cd drive for it) to increase the battery running time from 2.5 hours -> 4+ hours, or a DVD-ROM. I'll probably go for the battery, get a DVD-ROM for my (or Annie's) computer, and rip any movies that I want to watch on the road.
P.P.S. After staring at a laptop display for 3-4 hours, I notice the curvature on standard CRTs. How's that for crazy? |
|
|
| As one door opens, another is closed (but not locked) |
[May. 19th, 2004|08:40 am] |
I talked to Akhilesh last night and formally terminated my contract with his company. I thanked him, because they offered me a great opportunity, but this new contract I signed could not be turned down.
He took it very well, and seemed as thankful as I was to have worked together, and said that if I ever needed a job to give him a call. Crazy.
Now, they are going to need another Python developer soon (now). If you know Python amazingly well, having used asynchronous sockets (asyncore), wxPython, can get a handle on Open Cascade and Abaqus (that has macros that are saved and run in pure Python), then they could sure use you. Post here that you think you have the necessary skills, we can talk (not here), and if you look like you can do the job, I'll get you in contact with Akhilesh. |
|
|
| Happenings |
[May. 17th, 2004|11:08 am] |
I know I've been kind of quiet lately, but it is because I'm busy.
I've been working the contract I mentioned recently, a Processor Architecture project (I get to modify a processor simulator), Computational Complexity homework that is hard every week, and I'm reading papers in order to write a survey paper on some sort of set topic (I need to decide which of the sub-sub groups of set algorithms and data structures I want to survey, that I think I may be able to add to). Gah, much work.
Oh, and this Friday (May 21), I am taking theory version of what is known around here (ICS at UCI) as the "Phase 2", which is known in other places as a "PhD qualifying exam", "Quals", etc. I'm not worried (because I don't worry), but I am concerned and have been studying with Kevin, who is also taking the exam. Of the previous exams that we have on hand (they were given to us by the faculty), we've managed to solve a little over half of all problems, and have partial solutions to another 10-20%. If we do as well on our exam, the word from other people is that we should pass.
Here's hoping we do well. I'll probably post what happens Friday afternoon at like 5 or 6 or something. Oh, party at my place at 7. Hopefully we'll have something to celebrate. |
|
|
| My weekend... |
[May. 2nd, 2004|10:35 pm] |
Hey people, Annie and I got back a couple hours ago from being in San Jose, CA for the weekend, and here's why;
William Chang, former chief architect of Infoseek has a good idea (really a pile of good ideas, but that is beside the point), and I have a certain expertise that makes me able to help him and his company implement this idea. Which is why he flew Annie and I up there, had us stay in a very nice hotel, kept Annie entertained by seeing the sights with his wife and children (who are awesome), kept me busy with talks and coding, and yeah.
I met Wesley Chun of "Core Python Programming" fame, his wife, as well as a few other people (but I'm poor with names), who were all quite knowledgeable and entertaining to talk with (both shop and otherwise).
So, what is the final result of spending the weekend in San Jose? I've got a contract offer that pleases me (I'm waiting for the paperwork, we worked all weekend), and the contract will definitely last for the summer months, likely even for the next couple years. Depending on how much work there is to do, and what kind of work there is, I may have the opportunity to write some papers on it, but we'll see.
So yeah, I don't believe I'll be TAing this summer.
Unfortunately, to be able to have enough time for this, I'm going to have to end my contract with Next Gen Aeronautics very soon (a stipulation in the new contract), which makes me sad (I'm going to have to let down a great guy and his company), but considering this new opportunity, I would be a blooming idiot to not take the offer. So, there we have it. There is some documentation that I need to write for Next Gen, so I'll finish that before I end the contract, but I'll probably be nice and not charge them for the 4 hours I spend on the project on Thursday, nor for the 2-3 hours of documentation that I need to do. Yeah, I know I'm too nice, but I don't like to let people down, even when I'm letting them down. Akhilesh, if you are reading this, I'm sorry!
Now it is time to catch up with everything that happened this weekend. I hope everything is going well.
EDIT:
Oh, and in order to get work done, I'm going to need to acquire a laptop. I'm thinking an IBM, I always liked their styling, keyboard and trackpoint. We'll see if it is at or around my price point.
EDIT 2:
Oh, after Annie and I hopped out of the cab, we were walking to the mailbox, and I was going to check the time on my cell phone...but...my cell phone was gone?!? I called it 10+ times from Annie's phone, but no answer. We got back to the apartment, and called the cab company (they gave us the phone number on the card, along with the cab number). We called the company, the cab wasn't answering. We got the number for the Airport cab manager, who had his guys check the cab, and they found my phone. Goodness, I almost lost my cell phone, which would have sucked. |
|
|
| Wha? |
[Apr. 27th, 2004|04:17 pm] |
So, Annie and I are flying to San Jose this weekend. Why? Because there's a company that wants to hire me to do some relatively long-term contract work, and they want me to meet some people, do a little work, etc.
The flight is booked, my contact is booking us a hotel room across the street from the company, and yeah. I'm pretty pumped.
I should probably get some work done. |
|
|
| Yikes... |
[Mar. 1st, 2004|06:10 pm] |
So, I'm installing Open CASCADE and Python CASCADE for a bit of for-contract research, and during the long-ass installation, I notice something...
"The full installation package contains about 70,000 files..."
No wonder it took an hour to install. |
|
|
| Blah... |
[Feb. 28th, 2004|05:05 pm] |
It's been a few days since I updated, so yeah, I suppose it is about time that I did.
I keep getting busier. Thankfully, most of my major time commitments are already done for this weekend, so perhaps I can get some relaxation time in. I still have a few things to do before Sunday evening:
1. Come up with a deliverable report for the contract I'm working on.
2. Construct a GUI for processing TA evaluations (a side-project I volunteered for).
3. Do some timing research for David Eppstein. It shouldn't take too long, but I've been putting it off since early January.
4. Check out Open Cascade and PythonCascade for contract-related research.
1 should be done within the hour, 2 will likely be pushed to tomorrow, 3 should be done by the time I need to pick Annie up from work (3-4 hours), and 4 will also likely be pushed to tomorrow.
Once I'm done with that, I should probably read the 550 articles in comp.lang.py that I haven't yet. But before comp.lang.py, I'll find the time to read everyone's posts here at livejournal. Damn, so much stuff to do. I better get to it.
Oh. Other than being busy, I'm doing pretty good. I'm going to be the TA for ICS 161 next quarter. That will be the 6th time I've been a TA for ICS 161. It was also confirmed that the Theory Phase 2 exam will be given at the end of next quarter, so guess who will be busy studying for 2.5 months? Yep, me and Kevin.
Back to work. |
|
|
| Swamped... |
[Feb. 11th, 2004|11:14 pm] |
After working on Randomized Algorithms homework this weekend and most of Monday, then working on the potential contract work monday, reading the chapters for Computer Architecture yesterday, reviewing more Randomized algorithms yesterday, and working more on the potential contract work today...I'm exhausted.
On the plate for Thursday:
0. Drop off Annie at work.
1. Go to class.
2. Study randomized algorithms, for the randomized algorithms test tomorrow afternoon.
3. Call up potential contract contact to say, "let us meet Friday afternoon, I have something to show you"
4. Pick up Annie from work.
5. Take randomized algorithms midterm.
6. Pick up Valentine's Day present for Annie.
7. Go out for a 2-day-early VD dinner.
8. Chill.
Friday:
0. Pick up Sam at the train station at 11AM.
1. Chill.
2. Go to class.
3. Chill.
4. Meet with contract contact.
5. Eat dinner with Annie and Sam.
6. Chill.
I can't wait for Friday. That'll be a good day. |
|
|
| navigation |
| [ |
viewing |
| |
most recent entries |
] |
| |
|
|