L******t
发帖数: 1985 | 1 From comment of http://www.bradreese.com/blog/11-24-2014.htm:
I don't often praise Cisco, but taking an impartial point of view, Cisco's
action against Arista actually does have strong merit. Case law precedent is
on Cisco's side on this one. And Arista's stock taking a severe beating is
indicative of how things will pan out. Arista is worried big time even if
they are putting up a brave face and defiance in the media.
Is Cisco being a bully? You betcha! Is Cisco being opportunistic? Perhaps.
Do... 阅读全帖 |
|
l***y
发帖数: 791 | 2 using routing to provide security, IMHO, is like having a wisely
structured castle built, before putting in well-armed and well-trained
soldiers. simple firewalling often builds around a DMZ, which is a
network segment. using private IP for that segment saves the money to
buy more public ip addresses. there're more ways than one to do NAT,
but the gist for it is that network engineering should be the first step
in development. having entry and exit points in a large, well segmented
network, vs a |
|
c*****t
发帖数: 1879 | 3 In your case, all the intervals are disjoint, a simple binary earch
mechanism is enough to solve this issue (i.e. just check the even/odd
of array index)
For overlapping ones, the best solution is the interval tree.
Some other algorithms that have slightly longer constant are R
tree, or search trees built on top of GiST (which would be a nice
index scheme for database). |
|
N********n
发帖数: 8363 | 4
Yeah let's wake up, shall we? What is it you do not understand that
"Google has given up on Javascript" - https://gist.github.com/1208618
For years I've been steadily and precisely pointing out Javascript is
neither efficient nor productive enough to be a robust language for
real world development and yet you constantly come out with all sorts
of spins to paint HTML/JS as if some kinda sound solution.
Guess what now? Google has come clean declaring "Neverlearn was right.
We were wrong." Let's... 阅读全帖 |
|
a***n
发帖数: 538 | 5 完整的code
https://gist.github.com/3343011
#include
int main()
{
struct F {
int x;
} foo, *foo_ptr = &foo;
int F::*x = &F::x;
foo.*x = 42;
printf("%d\n", foo_ptr->*x);
}
怎么也看不明白第8行啊 |
|
M****z
发帖数: 1058 | 6 这东西印证了俺对google策略的猜测，Executive Summary已经说得巨清楚了
Google Dash/Dart: Leaked Internal Email (2010)
https://gist.github.com/paulmillr/1208618
Executive Summary
Javascript has fundamental flaws that cannot be fixed merely by evolving the
language. We'll adopt a two-pronged strategy for the future of Javascript:
- Harmony (low risk/low reward): continue working in conjunction with
TC39 (the EcmaScript standards body) to evolve Javascript
- Dash (high risk/high reward): Develop a new language (c... 阅读全帖 |
|
|
|
|
|
|
|
c*******0
发帖数: 5247 | 13
是你在讨论并发，不是我在讨论并发。你的定义是什么？
另外你找的例子要不要太老？两年前的也拿出来说。这种测试要找一大堆，Go 和
nginx的比较
https://gist.github.com/hgfischer/7965620
你要是没有什么直接的reference，还是不要随便下结论为好。我没兴趣claim Go的并
发比Haskell强，不过我倒是很有兴趣看你怎么支撑你的claim。 |
|
w***g
发帖数: 5958 | 14 image的相似度有很多不同的定义，对于不同的应用首先要选一个最符合需要的定义。
然后才是怎么算的问题。常见的相似度定义有
1. 视觉上/一眼看过去相似。主要用color histogram算。传统的CBIR一般都用这个，
但是这个单独自己用的话没啥实际用处。
2. 是否是同一个图像的不同版本。如果不需要检测诸如大图包含小图这种情况的话
就用global feature，比如GIST。如果要检测大图中的小图的话就是local feature，
最牛的算法都有专利保护，比如SIFT和SURF。别的免费可以用的我刚刚都试过一圈，
效果都比较差。
3. 照片是否是同一个rigid object，也用SIFT。
4. 照片是不是同一个人，人脸特征抽取是一大块，基本上是要先register眼耳鼻口
的位置，然后基于这些位置抽取特征。
5. 非rigid object的话得各个击破，比如人脸应为比较重要所以有很多算法可以用。
别的基本上没法做。
6. 图片是不是同一类对象。这个就是图像识别问题了。几年前最牛的是dense sampled
local feature + sparse encodin... 阅读全帖 |
|
|
|
|
|
|
|
|
|
|
|
z*******h
发帖数: 346 | 25 教娃编程趣味性很重要。那本Python for Kids我也有，这本书用Pygame，其实很难用
。所以我用iPad上的Pythonista教娃Python. 这个Pythonista可以调用iOS功能，比
pygame强大多了。比如可以做用大炮喂鱼吃蛋糕的游戏。像八皇后问题可以做成用手点
屏幕放棋子，然后可以加个auto solver button，用DFS实现。
我们还做了个像超级玛丽那样的游戏。碰撞检测都是自己写的。
这是我们家5年级娃用Pythonista编的一些东西：
https://gist.github.com/flowerspower |
|
|
d****n
发帖数: 1637 | 27 latency numbers every programmer should know.
FYI:
你那个数据库io最好的情况是disk io 再加上300ms delay.
几百个数据比较不出来啥，一百万个就会很大不同了。
Latency Comparison Numbers
--------------------------
L1 cache reference 0.5 ns
Branch mispredict 5 ns
L2 cache reference 7 ns 14x
L1 cache
Mutex lock/unlock 25 ns
Main memory reference 100 ns 20x
L2 cache, 200... 阅读全帖 |
|
M**a
发帖数: 4816 | 28 肯定不是抗体。应该抑制剂。
+++++
问：什么是ERLONAT?
答：ERLONAT是由印度NATCO公司生产的治疗晚期非小细胞肺癌的耙向抗肿瘤用药，为500mg的棕红色片剂,其中每片含埃洛替尼250mg。Natco药业有限公司（NPL），是位于海德拉堡市，资产达9亿5千卢比的一家药业公司，以其研究实力和及时发布技术信息而闻名。
+++++
秘鲁生产的Gleevec——依玛体(EMATI)
依玛体 EMATI ® (甲磺酸伊马替尼片)
依玛体是一种对血小板源性生长因数（PDGF）受体蛋白激酶具有抑制作用的新一代靶向性抗癌药。三期临床试验的结果表明它对新诊断慢性粒细胞性白血病（CML）治疗的有效率超过94%，并且有76%的病人可以得到细胞学的缓解，对加速期和急变期的有效率也分别达到了71%和31%。伊马替尼对胃肠道间质肿瘤的有效率也高达67%。
【适应症】：
慢性粒细胞性白血病（CML）
1. 依玛体用于治疗新诊断的费城染色体阳性（Ph＋）的处于慢性期的成人慢性粒细胞性白血病。
2. 依玛体用于治疗费城染色体阳性的处于急变期和加速期的慢性粒细胞性白血病，依玛体也用于治疗已经用... 阅读全帖 |
|
a****d
发帖数: 1919 | 29 the best protein-tyrosine kinase inhibitor so far being developed to treat
CML and GIST |
|
a*******a
发帖数: 4233 | 30 他为人还是不错的，很多学生真心纪念他。
另外他是间质瘤 gist |
|
w*g
发帖数: 270 | 31 Do not mislead the peers. You didn't get the gist.
In NV, you can take the PE exam after you pass the EIT. However, you cannot
get the NV PE license without four year experieces in Civil Engineering
after B.S. |
|
w*g
发帖数: 270 | 32 Do not mislead the peers. You didn't get the gist.
In NV, you can take the PE exam after you pass the EIT. However, you cannot
get the NV PE license without four year experieces in Civil Engineering
after B.S. |
|
d****y
发帖数: 1671 | 33 Hehe,watering even in the first week?
I guess the gist of the government job is retire with pension after 25-30
years service. |
|
W*********r
发帖数: 98 | 34 There are no securities liabilities in China. Practically, investors made
their business decisions based on government's certification of 蚁力神's
medical effect. That trust is the assumption of their conclusion -- because
government cannot certify something that's unfounded.
It is beyond our knowledge whether 蚁力神 tricked
government or simply conspired with government in this scheme. THe gist
is, government shall be aware of the fact that people WILL base their
business decisions on governm |
|
d******a
发帖数: 127 | 35 Agree! Iressa 没什么用.
I don't like the trials. Compared to Gleevec for CML and GIST, Iressa is a
garbage. Target Tx for lung cancer? There is a looooooooooooooooong way to
go. |
|
a*****o
发帖数: 174 | 36 2. Harbor Hosp, MD
6 candidates, early Nov.
Morning report around 0800, 2 presentations by an int. and a res. PD, APD
and a hard core Chinese faulty attended.
Intro by PD for 20 mins. Half prelim half categorical, prelims all AMG,
quite a few from Hopkins, cat mainly FMG, 1/3 fellowship placement, note PD
helps a lot to arrange away-rotation and strong support upon fellowship
application.
Very short interview by PD, <5 mins each candidate: a very nice gentleman,
GI-gist. Asked me to show my... 阅读全帖 |
|
R*****n
发帖数: 8658 | 37 别乱介绍药
要仔细看label
sutent approved indication应该只有renal, GIST
avastin approved indication: colorectal, non-samll cell lung |
|
x*****o
发帖数: 135 | 38 谢谢大家.我也注意到这两个药了,SUTENT和GLEEVE合用对付GIST,估计治不了
adenocarcinoma(腺癌).AVASTIN好像毒性很大,导致GI perforation的,我实在不敢介绍
呀.看来现在新的ONCOLOGY的药也就是拣拣好治的病治治,真正nasty的case也没什么办
法. |
|
|
c********g
发帖数: 1106 | 40 math modeling, such as physiologically based pharmacokinetic /dynamic mo
deling, biologically based modeling. the gist is to quantify adverse eff
ect and quantify risk. |
|
j****e
发帖数: 245 | 41 A psychologist was walking along a Hawaiian beach when he kicked a bottle poki
ng up through the sand. Opening it, he was astonished to see a cloud of smoke
and a genie smiling at him.
"For your kindness," the genie said, "I will grant you one wish!" The psycholo
gist paused, laughed, and replied, "I have always wanted a road from Hawaii to
California."
The genie grimaced, thought for a few minutes and said, "Listen, I'm sorry, bu
t I can't do that! Think of all the pilings needed to hold up the |
|
k**e
发帖数: 2728 | 42 All the cases mostly are very extreme in this book. However, life is not
with a clean cut on everything. Trivial cases in life could find its
resembling psychological thinking / mindset in these cases as well. The gist
in this book is to tell people, "being happy is your own thing", if you
cannot find happiness in your yourselves, you have problem and you need to
work on that. It's impossible to find happiness else where but only in
yourself. When you are happy, people around you are happy. If t |
|
w*******9
发帖数: 1433 | 43 I think what you get from the stochastic book is gist the conditional mean E
(Xt|X0) --- it does depend on t. Continue to take expectation you will get E
(Xt)=E(E(Xt|X0)) --- it's now a constant assuming E(X0)=a/(1-b). Hope it
helps.
b) |
|
R*******t
发帖数: 367 | 44 是腺癌吗？如果是GIST，现在有Gleevec，很快就能让肿瘤在PET上失去代谢活性，人称
魔术子弹。 |
|
R*******t
发帖数: 367 | 45 是腺癌吗？如果是GIST，现在有Gleevec，很快就能让肿瘤在PET上失去代谢活性，人称
魔术子弹。 |
|
f*********1
发帖数: 189 | 46 Your symptoms sound like neuroendocrine tumor.
Differentials based on limited CT images are broad: GIST; prior granuloma
disease such as TB or sarcoidosis; pheo especially your symtoms fit well
with it and sounds
like you also have some family history which makes it fit even better,
however pheo usually enhance well with contrast while yours do not enhance
much with contrast, although necrotized pheo may not enhance; adrenal
cortical adenoma especially hyper functioning adenoma, with adrenal cor... 阅读全帖 |
|
b********p
发帖数: 45 | 47 我靠。。。。
bcr-abl in CML
her+ breast cancer
ER/PR+ breast cancer
EGFR mutation in NSCLC
EGFR overexpression in H&N
PDGFR in GIST
ALK in NSCLC (你可以查一下这个从找到mutation到fda approve有多快）
to name few
你的MD是在国内那的吧？
你不要妄想我会妒忌你这种人了。 |
|
C*****D
发帖数: 1299 | 48 什么叫间质瘤？ 是gist 还是leimyoma？ |
|
|
