Opened 12 years ago

Closed 7 years ago

#3961 closed defect (invalid)

[meta] base size reductions

Reported by: alex Owned by: alex
Priority: high Milestone: future
Component: Core Version: 0.9
Keywords: Cc: alex, Eugene Lazutkin
Blocked By: Blocking:

Description

Base is getting pretty piggy (70K post ShrinkSafe?, 23K gzipped). It may be unreasonable to hit the 50K ceiling without packing, but we should clearly be much closer to it. This is a catch-all but for size-reduction checkins for 0.9

Attachments (4)

lang-fx-reduction.patch (6.5 KB) - added by Adam Peller 12 years ago.
Remove fx defaultEasing hack for old Konqueror (fixed in 3.5.5, according to mumme) Also remove unneeded instance defaults on prototype, other style and size changes
xhr-rpc-reduction.patch (11.3 KB) - added by Adam Peller 12 years ago.
another ~200 bytes. Throw if json filtered comment fails.
cleanup.patch (4.3 KB) - added by dante 11 years ago.
parser.patch (1.5 KB) - added by Adam Peller 11 years ago.
alright, it's not really base. reductions on parser.js

Download all attachments as: .zip

Change History (49)

comment:1 Changed 12 years ago by alex

(In [9936]) slight variable factoring for size. Refs #3961

comment:2 Changed 12 years ago by alex

(In [10146]) size reductions for NodeList?. Refs #3961

comment:3 Changed 12 years ago by alex

(In [10147]) clobber cruft function. Refs #3961

comment:4 Changed 12 years ago by alex

(In [10148]) no one's using "makeCalled()". Removing it. Refs #3961

comment:5 Changed 12 years ago by alex

(In [10149]) size optimizations for loader. Refs #3961

comment:6 Changed 12 years ago by alex

(In [10150]) make sure that the doc system picks up _getText but that we don't pay for it otherwise. Refs #3961

comment:7 Changed 12 years ago by alex

(In [10151]) some size-related changes and a bit of crusty function removal. Refs #3961

comment:8 Changed 12 years ago by alex

(In [10199]) trimming cruft from dojo.Deferred. The toString() was especially useless. Refs #3961

comment:9 Changed 12 years ago by alex

(In [10200]) eliminates the native branch for dojo.* array methods. There wasn't a perceptable or testable speed difference and it cost us in bytes. Refs #3961

comment:10 Changed 12 years ago by Adam Peller

(In [10203]) pick up a couple more bytes from array. Use xor bitwise logic on booleans. Refs #3961

comment:11 Changed 12 years ago by Adam Peller

(In [10206]) Use new docparser comment feature for 'faux declaration'. Refs #3961

comment:12 Changed 12 years ago by Adam Peller

(In [10207]) Fix typo I made in 'faux declaration' of getComputedStyle. Refs #3961

comment:13 Changed 12 years ago by Adam Peller

Resolution: fixed
Status: newclosed

(In [10475]) Add comments for isObject plus style mods. Fixes #4424, #3961

comment:14 Changed 12 years ago by Adam Peller

Resolution: fixed
Status: closedreopened

oops. didn't mean to close.

comment:15 Changed 12 years ago by Adam Peller

(In [10589]) Style and space fix for dojo.isArrayLike. Refs #3961

Changed 12 years ago by Adam Peller

Attachment: lang-fx-reduction.patch added

Remove fx defaultEasing hack for old Konqueror (fixed in 3.5.5, according to mumme) Also remove unneeded instance defaults on prototype, other style and size changes

comment:16 Changed 12 years ago by Adam Peller

(In [10659]) Reductions for _base, including:

Refs #3961

Changed 12 years ago by Adam Peller

Attachment: xhr-rpc-reduction.patch added

another ~200 bytes. Throw if json filtered comment fails.

comment:17 Changed 12 years ago by Adam Peller

(In [10834]) some code reductions. Refs #3961. Throw if json filtered comment fails.

comment:18 Changed 12 years ago by Adam Peller

(In [11157]) I give up. Put back the indexOf searches for comments in place of regexp attempt. Refs #3961, #4829, #4888 Reverts part of change in [10834]

comment:19 Changed 12 years ago by alex

(In [11205]) make the bootstrap smaller through scope protection closures now that the doctool is picking everything up. Big props to Neil for making this possible (and safe). Refs #3961

comment:20 Changed 12 years ago by alex

(In [11212]) updating for removed "!=" selector in dojo.query. Also, size reductions. Refs #3961

comment:21 Changed 12 years ago by alex

(In [11214]) ugg. I keep forgetting to re-enable the query optimizer! Refs #3961

comment:22 Changed 12 years ago by alex

(In [11219]) minor nits. Refs #3961. I'm still very unhappy about how much space the gears detection code takes up.

comment:23 Changed 12 years ago by alex

(In [11225]) clean up doc formatting and a bit of structure. Also some code size reductions. Refs #4820. Refs #3961

comment:24 Changed 12 years ago by James Burke

(In [11245]) Refs #3961. requireInto is not ready for prime time yet. In particular, it will not work with xd loading. Removing for now, to help reduce base.

comment:25 Changed 12 years ago by Adam Peller

Milestone: 1.01.1

so we're locking down 1.0. moving to 1.1 for continued work, or do we close this and just use it as a checkin reference?

comment:26 Changed 11 years ago by dante

(In [12108]) refs #3961 - tiny tiny cleanup to parser, and note on usage of dojo/method dojo/event

comment:27 Changed 11 years ago by Adam Peller

(In [12224]) remove deprecated argument juggling in declare.js. style fixes. !strict Refs #3961

comment:28 Changed 11 years ago by Adam Peller

(In [12376]) Reduce toPixelValue with a regexp, remove redundant return statements. Refs #3961 !strict

comment:29 Changed 11 years ago by Adam Peller

(In [12377]) Remove wildcards in toStyleValue regexp. Still just as accurate as it was before. Refs #3961 !strict

comment:30 Changed 11 years ago by Adam Peller

(In [12378]) One more time... don't assume pos=0 Refs #3961 !strict

comment:31 Changed 11 years ago by Adam Peller

(In [12379]) Don't need to seed cache with width/height. Refs #3961 !strict

comment:32 Changed 11 years ago by alex

(In [12418]) drop base size a bit. Refs #3961

comment:33 Changed 11 years ago by alex

(In [12798]) lastIndexOf() was totally borked. This fixes, rolling back changes from r10203. Refs #3961. !strict

comment:34 Changed 11 years ago by alex

(In [12818]) shorten NodeList? code, add toggleClass(), and update documentation. Refs #4205. Refs #3961. !strict

comment:35 Changed 11 years ago by alex

Summary: Base size reductions[meta] base size reductions

comment:36 Changed 11 years ago by alex

(In [12828]) comment cleanup and code shortening. Also expanding test cases to handle default param packing. Refs #3961. Refs #5962. Refs #3121. !strict

comment:37 Changed 11 years ago by bill

Milestone: 1.11.2

Move all milestone 1.1 tickets to 1.2, except for reopened tickets and tickets opened after 1.1RC1 was released.

comment:38 Changed 11 years ago by alex

(In [14051]) getting even more miserly w/ the bytes. Refs #3961 !strict

comment:39 Changed 11 years ago by alex

(In [14052]) some more byte scrimping. Refs #3961 !strict

Changed 11 years ago by dante

Attachment: cleanup.patch added

comment:40 Changed 11 years ago by dante

i've added cleanup.patch [cla/me] -- was looking for sanity check / approval before committing. moves dojo css-class functions into anon function, re-uses the defined "d" variable, and "obfuscates" the use of the .className property by making it a string (once) and refing the variable in each of the three functions. shaves some bytes but only after shrinkSafe makes a pass and renames it to _01 or something.

please review. unit tests for html pass FF3, ie6 + 7 ... is a trivial patch ultimately. (quite a few ie6 DOH failures in other places though I notice ... )

comment:41 Changed 11 years ago by bill

Milestone: 1.2future

Changed 11 years ago by Adam Peller

Attachment: parser.patch added

alright, it's not really base. reductions on parser.js

comment:42 Changed 10 years ago by Eugene Lazutkin

Cc: Eugene Lazutkin added
Owner: changed from anonymous to Eugene Lazutkin
Status: reopenednew

comment:43 Changed 9 years ago by Eugene Lazutkin

(In [21057]) Tiny style change, !strict, refs #3961.

comment:44 Changed 7 years ago by ben hockey

Owner: changed from Eugene Lazutkin to alex
Status: newpending

Does this ticket really need to be kept open?

If there's no response within 14 days it will close automatically.

comment:45 Changed 7 years ago by trac-o-bot

Resolution: invalid
Status: pendingclosed

Because we get so many tickets, we often need to return them to the initial reporter for more information. If that person does not reply within 14 days, the ticket will automatically be closed, and that has happened in this case. If you still are interested in pursuing this issue, feel free to add a comment with the requested information and we will be happy to reopen the ticket if it is still valid. Thanks!

Note: See TracTickets for help on using tickets.